In this paper we consider techniques for reducing the generated number of rules in learning fuzzy controllers of the state-space action-reinforcement type that can be simply implemented and that behave well in the presence of process noise. Fewer rules lead to better performance, less contradiction in controller action estimation, smaller required execution-time and make it easier for a human to comprehend the generated rules and possibly intervene.

Adaptive Control is used in order to improve close loop dynamics with a fuzzy controller when process parameters are unknown or fluctuate form an initial value. The way in which the adaptive control environment may be applied is the following. First we obtain a linear fuzzy controller. Second, we apply the adaptive rules by means of actuating directly over fuzzy variables which change their value. The techniques are based on Lyapunov functions. Third, we comment about extending this method to non-piecewise linear controllers using the contrast definition for a fuzzy controller.

This paper presents and adaptive fuzzy controller using fuzzy neural networks(FNNs). The adaptive controller uses two FNNs. One FNN is used to identify a fuzzy model of controlled object. The other FNN is used as a fuzzy controller. The fuzzy controller is designed with the linguistic rules of the fuzzy model. The response of the designed control system is checked with a linguistic response analysis proposed by the authors. An adaptive tuning of the control rules of the FNN controller is made possible utilizing the fuzzy model. Simulations using nonlinear controlled objects were done to verify the proposed control system.

A design method for linear combiner type filters, based on a fuzzy variant of the usual design method, is introduced and analyzed. Design results are exemplified.

Three optical fuzzy calculating modules, MAX/MIN, NOT/THROUGH, and SUP/THROUGH operating modules, are proposed. The MAX/MIN operating on inputted 2 membership functions. The NOT/THROUGH operating module calculates the complement of the membership function. The SUP/THROUGH operating module outputs an image representing the supremum (least upper bound) of the membership function. The THROUGH operation passes the image of the inputted membership function from the entrance to the exit. This paper demonstrates that these modules can output the image into which the modules transform inputted images on the basis of operation on fuzzy logic.

Most linguistic models of processes or plants known are essentially static, that is, time is not a parameter in describing the behavior of the object's model. In this paper we show two models for synchronous finite state machines (FSM) based on fuzzy logic, namely the Crisp-State-Fuzzy-Output (CSFO FSM) and Fuzzy-State-Fuzzy Output (FSFO FSM). As a result of the introduction of the FSM models, the improved architectures for fuzzy logic controller have been defined. These architectures featuring pipelined intelligent fuzzy controller are discussed in terms of dimensionality of the model. VLSI integrated circuit implementation issues of the fuzzy logic controller are also considered. The presented approach can be utilized for fuzzy controller hardware accelerators intended to work in the real-time environment.

This paper presents the prototype framework for automated integration of CMOS current-mode fuzzy logic circuits using an intelligent module approach. The library of modules representing the standard fuzzy logic operators was built. These modules were finally used to synthesized sophisticated fuzzy logic units. Fuzzy unit designs were made based upon the results of a newel methodology of the current mirror-based fuzzy logic function synthesis. This methodology is actually incorporated into the presented framework. As an example, the membership function unit was synthesized, simulated, and the final layout was generated using the presented framework. Finally, the fuzzy logic controller unit (FLC) was generated using the proposed framework. Simulation as well as measurement results show unquestionable advantages of the proposed fuzzy logic function integration system over the classical design methodology with respect to the area, relative error and performance.

We have proposed a fuzzy model for behavior of vehicles in the road traffic simulation system with microscopic model for analyzing traffic jam in the broad areas. It can exactly simulate each vehicle's behavior. We propose a new hardware processor to simulate fuzzy decision-making mechanism for its model. This paper describes the functions, performance and structure of the hardware processor.

This paper presents the decision support system using fuzzy knowledge to adapt the cutting conditions chosen by a conventional expert system to a particular machine tool, workpiece and clamping system. These preliminary results demonstrate the capability of fuzzy logic to adjust cutting parameters taking into account parameters difficult to quantify.

A method of automatic learning of fuzzy if-then rules with certainty factors from the given input-output data is developed. A certainty factor expresses the degree to which a fuzzy if-then rule is fitting to the given data. Fuzzy if-then rules with certainty factors are generated without optimization techniques. The obtained fuzzy if-then rules can be regarded as an approximator of a non-linear function. This method is applied to GMDH (Group Method of Data Handling) to cope with difficulty in approximating multi-input functions with fuzzy if-then rules.

This paper focuses on the usage of the fuzzy set theory in decision making systems. The approach to calculation of generalized membership function, based on application of method of principal components is proposed. For solving of the problem of fuzzy forecasting the development of Bayes procedure is used. The structure of decision making system, in which following procedures are fulfilled, is discussed.

In multi-attribute decision making, human beings influenced with various factors often change their decisions. This paper presents a new approach to express the changes in the decision makings when they got new information. The new approach uses the fuzzy neural network (FNN) which has been proposed by the authors. The FNN identifies the weights to the attributes with the back propagation learning. Through experiments, it is shown that the changes of subjects' decision can be described by the changes of their weights to the attributes.

Fuzzy controllers still remain ill-accepted in the control community. As a matter of fact, their design relies on a new relation between the material world and the scientists. Whereas some theoritical studies are carried out on this subject, experimentations on processes show what fuzzy techniques can bring to the control theory.

In the last years fuzzy control has grown up to an important methodology of control engineering. In spite of the successful realizations of the underlying concepts in industrial products there has only been little effort regarding a semantical foundation of the prevailing heuristics that are used in fuzzy control. For this reason we want to outline promising approaches to an interpretation and better mathematical justification of fuzzy control, where the fundamental ideas of using equality relations to specify fuzzy environments for crisp data are presented. It turns out that Mamdani's classical max-min-inference is a consequence of our model.

A fuzzy logic controller derived from the variable structure control (VSC) theory is designed. Unlike the conventional design of the fuzzy controller, we do not fuzzify the error and the rate of error, but fuzzify the sliding surface. After the fuzzy sliding surface is introduced, the fuzzy rules are defined based on the sliding control theory. It will be shown this sliding mode fuzzy controller is a kind of VSC that introduces the boundary layer in the switching surface and that the control input is continuously approximated in the layer. As a result we can guarantee the stability and the robustness by the help of VSC, which were difficult to insure in the past fuzzy controllers. Simulation results for the inverted pendulum will show the validity.

Existing fuzzy control methods do not perform well when applied to systems containing nonlinearities arising from unknown deadzones. We propose a novel two-layered fuzzy logic controller for controlling systems with deadzones. The two-layered control structure consists of a fuzzy logic-based pre-compensator followed by a conventional fuzzy logic controller. Our proposed controller exhibits superior transient and steady-state performance compared to conventional fuzzy controllers. We illustrate the effectiveness of our scheme using computer simulation examples.

In this paper, an optimal tuning Algorithms is presented to automatically improve the performance of fuzzy controller, using the simplified reasoning method and the proposed complex method. The method estimates automatically the optimal values of the parameters of fuzzy controller, according to the change rate and limitation condition of output. The controller is applied to plants with dead time. Then, computer simulations are conducted at step input and the performances are evaluated in the ITAE.

This paper describes the implementation of an autonomous fuzzy logic controller. The controller is endowed with basic control principles and learning constructs which enable it to autonomously modify its control policy based on system performance. The controller lies dormant when system response is satisfactory but if rapidly initiates adaptation in real time when adverse performance is observed. The autonomous fuzzy controller is implemented on an Intel MCS-51 series micro-controller board using an inexpensive 8-bit Intel 8031 processor. The 11.06 MHz micro-controller operates at a rate exceeding 200 "global" look-up table reinforcements per second. This is important when developing practical on-line adaptive controllers for fast systems. It is also significant because an initial controller look-up table could be incorrect or non-existent. The relatively high learning rate enables the controller to learn to control a system even while it is controlling it.

This paper describes a fuzzy network circuit of analogue and digital mixed operation. The circuits are suggested for membership function, MIN function and normalization function using either linear voltage-controlled MOSFET resistance or pulse stream operation. The analogue-digital hybrid fuzzy hardware is extensible to the fuzzy-neural network as its basic configurations are already used in URAN-I of 135,424 synaptic connections.

We present a F.L.C. (Fuzzy Logic Controller) to control of the oscillation frequency of a V.C.X.O. (Voltage Controled Crystal Oscillator). This F.C.X.O. maintains stable its oscillation frequency inside a range of 1 ppm (one part per millon), with temperature between -55$^{\circ}C$ to＋75$^{\circ}C$.

This paper proposes a design method of fuzzy phase-lead compensator and its self-learning by neural network. The main feature of the fuzzy phase-lead compensator is to have parameters for effectively compensating phase characteristics of control systems. An important theorem which is related to phase-lead compensation is derived by introducing concept of frequency characteristics. We propose a design procedure of fuzzy phase-lead compensators for linear controlled objects. Furthermore, we realize a neuro-fuzzy compensator for unknown or nonlinear controlled objects by using Widrow-Hoff learning rule.

This paper proposes a novel chaos generator using the model of Josephson junction. Constructing an equivalent circuit of Josephson element by using an operational amplifier, we have made a chaos generator. The feature of this generator is to generate several kinds of oscillations as well as chaotic oscillations by only changing a DC bias current to the junction. In this paper, it is described in detail how to construct the circuit and what kind of oscillations is realized in the circuit. The experimental results of oscillation modes are compared with simulation results with a satisfactory agreement.

In this paper, we describe a chaos simulator as a developing tool for applications of chaos engineering. This simulator is composed of three modules, such as generation module of chaotic signals by deterministic rules, determination module whether observed time series is chaos or not, and nonlinear system identification module by self generating Neuro Fuzzy Model.

In this paper, we apply the self generating neuro fuzzy model (SGNFM) to the dimension analysis of the chaotic time series. Firstly, we formulate a nonlinear time series identification problem with nonlinear autoregressive (NARMAX) model. Secondly, we propose an identification algorithm using SGNFM. We apply this method to the estimation of embedding dimension for chaotic time series, since the embedding dimension plays an essential role for the identification and the prediction of chaotic time series. In this estimation method, identification problems with gradually increasing embedding dimension are solved, and the identified result is used for computing correlation coefficients between the predicted time series and the observed one. We apply this method to the dimension estimation of a chaotic pulsation in a finger's capillary vessels.

This research provides the results of a trial to generate the chaos by using nonlinear function constructed by fuzzy inference rules. The chaos generation function or chaotic behavior can be obtained by using Takagi-Sugeno fuzzy model with some constraint of the relationship of its parameters. Two examples are shown in this research. The first is simple example that construct of logistic image by fuzzy model. The second is more complicated one that provide the chaotic time series by non-linear autoregression based on fuzzy model. Simulated results are shown in these examples.

This paper describes a simple random signal generator employing by CMOS analog technology in current mode. The system is a nonlinear dynamical system described by a difference equation, such as x(t＋1) = f(x(t)) , t = 0,1,2, ... , where f($.$) is a nonlinear function of x(f). The tent map is used as a nonlinear function to produce the random signals with the uniform distribution. The prototype is implemented by using transistor array devices fabricated in a mass product line. It can be easily realized on a chip. Uniform randomness of the signal is examined by the serial correlation test and the $\chi$2 test.

The chaotic fuzzy logic systems behave in a more complex way than crisp chaotic systems, and they can show some advantages in complex applications. Such an application is introduced in this paper, namely in scrambling and ciphering the signals.

A fuzzy controller designed by mini-max-gravity(MMG) method is essentially nonlinear with respect to the controller's input and output relationship, and stability analysis is thus needed to construct a stable control system. This paper deals with a design method of a position-type MMG fuzzy controller stable in a sense of Lyapunov when considered is a single-input-single-output linear, stable plant. We first introduce a method to construct a Laypunov function by using an eigen-value of A matrix of the linear, stable plant dynamics and then we derive an asymtotic stability condition in terms of scale factors for fuzzy state variables and controller gain. The stability condition is found reasonably practical through comparing the theoretical stability region with that obtained from simulations.

Fuzzy control algorithms are developed based on fuzzy models of systems. The control issues are posed as multiobjective optimization problems involving goals and constraints imposed on system's variables. Two basic design modes embrace on-and off-line control development. The first type of design deals with the time and state-dependent objectives and pertains to control determination based upon the current state of the system. The second design mode gives rise to explicit forms of fuzzy controller that is learned based on a given list of state-control associations. Both the fuzzy models as well as fuzzy controllers are realized as logic processors.

Fuzzy logic rule-based controller has many desirable advantages, which are simple to implement on the real time and need not the information of structure and dynamic characteristics of the system. Thus, nowadays, the scope of the application of the fuzzy logic controller becomes enlarged. But, if the controlled plant is a time-varying and nonlinear system, it is not easy to construct the fuzzy logic rules which usually need the knowledge of an expert. In this paper, an approach in which the logic control rules can be self-organized using genetic algorithm will be proposed and the effectiveness of the proposed method will be verified by computer simulation of the 2 d.o.f. planar robot manipulator.

By introducing the notion of constraint-oriented fuzzy inference, we will show that it provides us ways of fuzzy control methods that has abilities of adaptation, learning and self-organization. The basic supporting techniques behind these abilities are“hard”processing by Artificial Intelligence or traditional computational framework and“soft”processing by Neural Network or Genetic Algorithm techniques. The reason that these techniques can be incorporated to fuzzy control systems is that the notion of“constraint”itself has two fundamental properties, that is, the“modularity”property due to its declarativeness and the“logicality”property due to its two-valuedness. From the former property, the modularity property, decomposing and integrating constraints can be done easily and efficiently, which enables us to carry out the above“soft”processing. From the latter property, the logicality property, Qualitative Reasoning and Instance Generalization by Symbolic Reasoning an be carried out, thus enabling the“hard”processing.

To improve limitations of fuzzy PI controller especially when applied to high order systems, we propose two types of fuzzy logic controllers that take out appropriate amounts of accumulated control input according to fuzzily described situations in addition to the incremental control input calculated by conventional fuzzy PI controllers. The structures of the proposed controller were motivated by the problems of fuzzy PI controllers that they generally give inevitable overshoot when one tries to reduce rise time of response especially when one tries to reduce rise time of response especially when a system of order higher than one is under consideration. Since the undesirable characteristics of the fuzzy PI controller are caused by integrating operation of the controller, even though the integrator itself is introduced to overcome steady state error in response, we propose two fuzzy controllers that fuzzily clear out integrated quantities according to situation. The first contr ller determines the fuzzy resetting rate by situations described fuzzily by error and error rate, and the second one by error and control input. The two structures both give reduced rise time as well as small overshoot. To show the usefulness of the proposed controller, that are applied to systems that are difficult to get satisfactory response by conventional fuzzy PI controllers.

Instead of Cartesian product for in combining multiple inputs for fuzzy logic controllers, a method using fuzzy relation in inference is proposed. Moreover, fuzzy control rule described by fuzzy relations is derived from given conventional fuzzy control rule by fitting concept. It will be shown through several examples that the proposed technique gives smoother interpolation than conventional ones.

A neural model predictive control strategy combining a neural network for plant identification and a nonlinear programming algorithm for solving nonlinear control problems is proposed. A constrained nonlinear optimization approach using successive quadratic programming combined with neural identification network is used to generate the optimum control law for complex continuous chemical reactor systems that have inherent nonlinear dynamics. The neural model predictive controller (MNPC) shows good performances and robustness. To whom all correspondence should be addressed.

Lipreading through visual processing techniques help provide some useful systems for the hearing impaired to learn communication assistance. This paper proposes a method to understand spoken words by using visual images taken by a camera with a video-digitizer. The image is processed to obtain the contours of lip, which is approximated into a hexagon. The pattern lists, consisting of lengths and angles of hexagon, are compared and computed to get the fuzzy similarity between two lists. By similarity matching, the mouth shape is recognized as the one which has the pronounced voice. Some experiments, exemplified by recognition of the Japanese vowels, are given to show feasibilities of this method.

This paper describes the use of fuzzy control and decision making to simulate the control of traffic flow at an intersection. To show the value of fuzzy logic as an alternative method for control of traffic environments. A traffic environment includes the lanes to and from an intersection, the intersection, vehicle traffic, and signal lights in the intersection. To test the fuzzy logic controller, a computer simulation was constructed to model a traffic environment. A typical cross intersection was chosen for the traffic environment, and the performance of the fuzzy logic controller was compared with the performance of two different types of conventional control. In the hardware verifications, fuzzy logic was used to control acceleration of a model train on a circular path. For the software experiment, the fuzzy logic controller proved better than conventional control methods, especially in the case of highly uneven traffic flow between different directions. On the hardware si e of the research, the fuzzy acceleration control system showed a marked improvement in smoothness of ride over crisp control.

This paper makes a trial to build the model of car-following in the state of starting to stable driving on the basic of driver's knowledge that is easily characterized by linguistical cognition. There are three main steps in building the model. Firstly, each driver's rule of three testees is studied in linguistical experssion by the interview and questionary surveys that are repeated once a day for ten days. Secondly, quantification of the linguistical expression is investigated by driving experiments that includes the questionary survey to the testee in the test vehicle, and the membership functions of variables of rule are obtained. Thirdly, implicaton and composition of fuzzy inference is made by Max-Min Methods and defuzzification by gravity method. It can be said that the proposed model of car-following based on driver's knowledge is practically allpicable to the estimation of drivering of car-following on trunk roads in urban area.

In this paper, the fuzzy ISODATA algorithm is applied to forecasting liquefaction of sand in the antiseismic structures. According to the data of the earthquake taken place in Tang Shan in 1976, we construct a model of mathematics, on which we forecast 32 samples in the earthquake of Tang Shan. The correct rate of forecast is 90.7 %.

Various forms of hardware alternatives exist for the implementation of fuzzy logic controllers. In this paper, we describe a systematic framework for realizing fuzzy heuristics on programmable-logic-devices. Our approach is suitable for the automated development of fuzzy logic controllers.

For the purpose of applying to a high-speed control system, such as engine control for automobile application, we propose an architecture of a fuzzy inference processor, which can realize high-speed inference, high-resolution, and can be implemented with small chip area. We have designed a single chip based on the architecture, and confirmed the performance, such as 140 kFLIPS with 8-bit resolution.

In this paper the caracteristics of the fuzzy flip-flop which was proposed as a fuzzy sequential circuit is firstly mentioned. Secondly the circuit construction of typical fuzzy flip-flip circuits using VHDL (Very high speed integrated circuit Hardware Description Language) compiler and simulator is presented. Finally the possibility of the application of the fuzzy sequential circuit will be mentioned.

A fuzzy microcontroller is presented implementing a simplified inference mechanism. Fuzzification, rule composition and defuzzification are carried out by means of (basically) analog current-mode CMOS circuits operating in strong inversion. Also a voltage interface is provided with the external world. Combining analog and digital techniques allow a programming capability.

A high speed fuzzy processor using bipolar technology is proposed in this paper. The hardware system uses a high-speed current-mode membership function circuit and normalization technique. The new membership function circuit generates an ideal membership function of the fuzzy set and its circuit is also simple and available for VLSI implementation. Several techniques have been implemented to speed up response of the processor. The fuzzy processor has been designed and implemented in bipolar circuit technology. The experiments and simulations show that the response speed is below 100ms. It can also be expected that the fuzzy processor can be integrated on one chip and its response time is only about the order of nanoseconds.

A new hardware architecture achieves high speed, high precision fuzzy inference capabilities while maintaining Flexibility on par with software approaches. This flexibility allows unmodified, uncompromised porting of fuzzy system designs into hardware. The architecture is also scalable and offers data resolutions from 8 bits to 32 bits.

In this paper, we present an international benchmark used in the adaptive control specialist community in order to evaluated the fuzzy control performances. Before solving the corresponding problems, we introduce some improvements on a classic fuzzy controller in order to consider high order systems and time delays. At the end of this paper, the simulation results obtained with the extended Fuzzy Controller will be compared with those obtained with a Supervised Adaptive Controller.

As well-known, fuzzy control has been recognized to be of great usefulness in many engineering fields. However, the present design methods of fuzzy control systems depend on trial and error the thing that limits its usefulness. Therefore, an effective and convenient support tools for design and evaluation are greatly needed as well as the establishment of the design methods and guidling. From these backgrounds, we have developed a fuzzy control simulator[1, 2] which has various fuzzy control methods such as "direct method", "indirect method" and "fuzzy-PID method". This paper deals especially with the "direct method" function of the simulator. The simulator was developed for personal computers and programed in C language.

In this paper we propose a new stability theorem and a robust stability condition for linguistic fuzzy model systems in state space. First we define a stability in linear sense. After representing the fuzzy model by a system with disturbances, A necessary and sufficient condition for the stability is derived. This condition is proved to be a sufficient condition of the fuzzy model. The Q in the Lyapunov equation is iteratively adjusted by an gradient-based algorithm to improve its stability test. Finally, stability robustness bounds of a system having modeling error is derived. An example is also included to show that the stability test is powerful.

We propose a new PI FLC which utilizes the error(e), change of error( e), and previous control(u(k-1)). It is shown by simulation that the proposed scheme gives better performance in steady state than the conventional PI FLC.

We focus our attention on grading of table meat in accordance with the standard of Japan Meat Grading Association, and construct a beef grading system by image processing. For image processing of beef grading, it needs some techniques such as a shading correction, separation of color image data, and classification of color image data into some grades, for the system construction. However, there are various kinds of weak points in usually used methods for these techniques. Then the authors propose and introduce new approaches using Neural networks and fuzzy inference for the techniques above mentioned, which is very convenient and ensure the high precision.

This paper presents an auto tuning method of fuzzy inference using Genetic Algorithm. The determination of membership functions by human experts is a difficult problem. Therefore, some auto-tuning methods have been proposed to reduce the time-consuming operations. However, the convergence of the tuning by the previous methods depends on the initial conditions of the fuzzy model. So, we proposes an auto tuning method for the fuzzy neural network by Genetic Algorithm (ATF system). This paper shows effectiveness of the ATF system by simulations.

Fuzzy inference system which inferences and processes the Fuzzy information is designed using digital voltage mode neural circuits. The digital fuzzification circuit is designed to MIN,MAX circuit using CMOS neural comparator. A new defuzzification method which uses the center of area of the resultant fuzzy set as a defuzzified output is suggested. The method of the center of area(C. O. A) search for a crisp value which is correspond to a half of the area enclosed with inferenced membership function. The center of area defuzzification circuit is proposed. It is a simple circuit without divider and multiflier. The proposed circuits are verified by implementing with conventional digital chips.

This paper presents a new method to automatically design fuzzy logic controller(FLC). The main problems of designing FLC are how to optimally and automatically select the control rules and the parameters of membership function (MF). Cell state space algorithms (CSS), differential competitive learning (DCL) and multialyer neural network are combined in this paper to solve the problems. When the dynamical model of a control process is known. CSS can be used to generate a group of optimal input output pairs(X, Y) used by a controller. The(X, Y) then can be used to determine the FLC rules by DCL and to determine the optimal parameters of MF by DCL and to determine the optimal parameters of MF by multilayer neural network trained by BP algorithm.

This talk presents the overview of the author's research and development activities on fuzzy inference hardware. We involved it with two distinct approaches. The first approach is to use application specific integrated circuits (ASIC) technology. The fuzzy inference method is directly implemented in silicon. The second approach, which is in its preliminary stage, is to use more conventional microprocessor architecture. Here, we use a quantitative technique used by designer of reduced instruction set computer (RISC) to modify an architecture of a microprocessor. In the ASIC approach, we implemented the most widely used fuzzy inference mechanism directly on silicon. The mechanism is beaded on a max-min compositional rule of inference, and Mandami's method of fuzzy implication. The two VLSI fuzzy inference chips are designed, fabricated, and fully tested. Both used a full-custom CMOS technology. The second and more claborate chip was designed at the University of North Carolina(U C) in cooperation with MCNC. Both VLSI chips had muliple datapaths for rule digital fuzzy inference chips had multiple datapaths for rule evaluation, and they executed multiple fuzzy if-then rules in parallel. The AT & T chip is the first digital fuzzy inference chip in the world. It ran with a 20 MHz clock cycle and achieved an approximately 80.000 Fuzzy Logical inferences Per Second (FLIPS). It stored and executed 16 fuzzy if-then rules. Since it was designed as a proof of concept prototype chip, it had minimal amount of peripheral logic for system integration. UNC/MCNC chip consists of 688,131 transistors of which 476,160 are used for RAM memory. It ran with a 10 MHz clock cycle. The chip has a 3-staged pipeline and initiates a computation of new inference every 64 cycle. This chip achieved an approximately 160,000 FLIPS. The new architecture have the following important improvements from the AT & T chip: Programmable rule set memory (RAM). On-chip fuzzification operation by a table lookup method. On-chip defuzzification operation by a centroid method. Reconfigurable architecture for processing two rule formats. RAM/datapath redundancy for higher yield It can store and execute 51 if-then rule of the following format: IF A and B and C and D Then Do E, and Then Do F. With this format, the chip takes four inputs and produces two outputs. By software reconfiguration, it can store and execute 102 if-then rules of the following simpler format using the same datapath: IF A and B Then Do E. With this format the chip takes two inputs and produces one outputs. We have built two VME-bus board systems based on this chip for Oak Ridge National Laboratory (ORNL). The board is now installed in a robot at ORNL. Researchers uses this board for experiment in autonomous robot navigation. The Fuzzy Logic system board places the Fuzzy chip into a VMEbus environment. High level C language functions hide the operational details of the board from the applications programme . The programmer treats rule memories and fuzzification function memories as local structures passed as parameters to the C functions. ASIC fuzzy inference hardware is extremely fast, but they are limited in generality. Many aspects of the design are limited or fixed. We have proposed to designing a are limited or fixed. We have proposed to designing a fuzzy information processor as an application specific processor using a quantitative approach. The quantitative approach was developed by RISC designers. In effect, we are interested in evaluating the effectiveness of a specialized RISC processor for fuzzy information processing. As the first step, we measured the possible speed-up of a fuzzy inference program based on if-then rules by an introduction of specialized instructions, i.e., min and max instructions. The minimum and maximum operations are heavily used in fuzzy logic applications as fuzzy intersection and union. We performed measurements using a MIPS R3000 as a base micropro essor. The initial result is encouraging. We can achieve as high as a 2.5 increase in inference speed if the R3000 had min and max instructions. Also, they are useful for speeding up other fuzzy operations such as bounded product and bounded sum. The embedded processor's main task is to control some device or process. It usually runs a single or a embedded processer to create an embedded processor for fuzzy control is very effective. Table I shows the measured speed of the inference by a MIPS R3000 microprocessor, a fictitious MIPS R3000 microprocessor with min and max instructions, and a UNC/MCNC ASIC fuzzy inference chip. The software that used on microprocessors is a simulator of the ASIC chip. The first row is the computation time in seconds of 6000 inferences using 51 rules where each fuzzy set is represented by an array of 64 elements. The second row is the time required to perform a single inference. The last row is the fuzzy logical inferences per second (FLIPS) measured for ach device. There is a large gap in run time between the ASIC and software approaches even if we resort to a specialized fuzzy microprocessor. As for design time and cost, these two approaches represent two extremes. An ASIC approach is extremely expensive. It is, therefore, an important research topic to design a specialized computing architecture for fuzzy applications that falls between these two extremes both in run time and design time/cost. TABLEI INFERENCE TIME BY 51 RULES {{{{Time }}{{MIPS R3000 }}{{ASIC }}{{Regular }}{{With min/mix }}{{6000 inference 1 inference FLIPS }}{{125s 20.8ms 48 }}{{49s 8.2ms 122 }}{{0.0038s 6.4㎲ 156,250 }} }}

During the past decade, several specific hardwares for fast fuzzy inference have been developed. Most of them are dedicated to a specific inference method and thus cannot support other inference methods. In this paper, we present a hardware architecture called KAFA(KAist Fuzzy Accelerator) which provides various fuzzy inference methods and fuzzy set operators. The architecture has SIMD structure, which consists of two parts; system control/interface unit(Main Controller) and arithmetic units(FPEs). Using the parallel processing technology, the KAFA has the high performance for fuzzy information processing. The speed of the KAFA holds promise for the development of the new fuzzy application systems.

An alternative approach to the design of application-driven fuzzy systems is proposed. A broad class of fuzzy systems applications requires a certain fuzzy partition of the input space while it demands for simple numerical quantities. For this class, a dedicated fuzzy system archictecture is presented and a design strategy is proposed. Both the single-input/single-output and multi-input/multi-output cases are considered. Numerical analysis are complete illustrating several aspects of the proposed framework.

In this paper, we describe hardware architecture of fuzzy processors for reasoning involving fuzzy control“Heuristics”. This we believe will lead to fuzzy systems that are closer to the way humans process domain knowledge for decision making. One noticeable beneficial effect based on our notion of fuzzy heuristics is the significantly reduced number of rules required.

This paper addresses the structure and its associated learning algorithms of a feedforward multi-layered connectionist network, which has distributed learning abilities, for realizing the basic elements and functions of a traditional fuzzy logic controller. The proposed neural-network-based fuzzy logic control system (NN-FLCS) can be contrasted with the traditional fuzzy logic control system in their network structure and learning ability. An on-line supervised structure/parameter learning algorithm dynamic learning algorithm can find proper fuzzy logic rules, membership functions, and the size of output fuzzy partitions simultaneously. Next, a Reinforcement Neural-Network-Based Fuzzy Logic Control System (RNN-FLCS) is proposed which consists of two closely integrated Neural-Network-Based Fuzzy Logic Controllers (NN-FLCS) for solving various reinforcement learning problems in fuzzy logic systems. One NN-FLC functions as a fuzzy predictor and the other as a fuzzy controller. As ociated with the proposed RNN-FLCS is the reinforcement structure/parameter learning algorithm which dynamically determines the proper network size, connections, and parameters of the RNN-FLCS through an external reinforcement signal. Furthermore, learning can proceed even in the period without any external reinforcement feedback.

The Multisensor Fusion Problem (MFP) deals with the methodologies involved in effectively combining together homogeneous or non-homegeneous information obtained from multiple redundant or disparate sensors in order to perform a task more accurately, efficiently, and reliably. The inherent uncertainties in the sensory information are represented using Fuzzy Numbers, -numbers, and the Uncertainty-Reductive Fusion Technique (URFT) is introduced to combine the multiple sensory information into one consensus -number. The MFP is formulated from the Information Theory perspective where sensors are viewed as information sources with a fixed output alphabet and systems are modeled as a network of information processing and processing and propagating channels. The performance of the URFT is compared with other fusion techniques in solving the 3-Sensor Problem.

Supervised learnmg 01 recurrent neural networks (RNNs) is discussed. First, we review the present state of art, featuring their major properties in contrast of those of the multilayer neural networks. Then, we concisely describe one of the most practical learning algorithms, i.e. backpropagation through time. Revising the basic formulation of the learning algorithms, we derive a general formula to solve for the exact solution(s) of the whole connection weights w of RNNs. On this basis we introduce a novel interpretation of the supervised learning. Namely, we define a multidimensional Euclidean space, by assigning the cost function E(w) and every component of w to each coordinate axis. Since E=E(w) turns up as a hyper surface in this space, we refer to the surface as learning surface. We see that topological features of the learning surface are valleys and hills. Finally, after explicating that the numerical procedures of learning are equivalent to descending slopes of the learning surface along the steepest gradient, we show that a minimal value of E(w) is the intersection of curved valleys.

An association memory, solving an optimization problem, a Boltzmann machine scheme learning and a back propagation learning in our chaos neuron models are reviewed and some new results are presented. In each model its microscopicrule (a parameter of a chaos system in a neuron) is subject to its macroscopic state. This feedback and chaos dynamics are essential mechanisms of our model and their roles are briefly discussed.

This paper describes a new type of neuron model, the inputs of which are interfered with one another. It has a high mapping ability with only single unit. The learning speed is considerably improved compared with the conventional linear type neural networks. The proposed neuron model was successfully applied to the prediction problem of chaotic time series signal.

This paper describes a neo fuzzy neuron which was produced by a fusion of fuzzy logic and neuroscience. Some learning algorithms are presented. The guarantee for the global minimum on the error-weight space is proved by a reduction to absurdity. Enhanced is that the learning speed of the neo fuzzy neuron exceeds 100,000 times of that of conventional multi-layer neural networks.

System identification is a major component for a control system. In biosystems, which is nonlinear and dynamic, precise identification would be very helpful for implementing a control system. It is difficult to precisely identify such non-linear systems. The measurable data on products from 2,3-butanediol fermentation could not be included in a process model based on kinetic approach. Meanwhile, a predictive capability is required in developing a control system. A neural network (NN) dynamic identifier with a by/(1＋ t ) transfer function was therefore designed being able to predict this fermentation. This modified inverse NN identifier differs from traditional models in which it is not only able to see but also able to predict the system. A moving window, with a dimension of 11 and a fixed data size of seven, was properly designed. One-step ahead identification/prediction by an 11-3-1 BPNN is demonstrated. Even under process fault, this neural network is still able to perform several-step ahead prediction.

In this paper, an intelligent control scheme with multi-stage fuzzy inference is developed for a myoelectric prosthesis to achieve natural control with tactile feedback based on fuzzy control strategies. Strain gauges and a potentiometer are added to the prosthesis for tactile feedback with a PWM motor driver developed to save the battery power. According to the multi-stage fuzzy inference, the prosthesis can determine the stiffness of the object and hold an object without injuring it, meanwhile, the hysteresis phenomenon is an 80196KC single-chip microcontroller to replace the original controller.

The construction of the rulebase of a fuzzy controller is usually difficult because experts' knowledge is often hard to derive. To remedy such a problem, a number of self-learning schemes for rulebase formulations were proposed. One of the popular approaches is the reinforcement learning. Many successful examples employing such an idea were proposed and claimed to be with good results in the literature. The purpose of this paper is to discuss and make comparisons between some of the related work in order to provide a better picture regarding their performances. A numerical algorithm for the analysis of nonlinear as well as fuzzy dynamic systems, the Cell-to-Cell Mapping, is used. The analytical results reveals the true behavior of the learning schemes.

Most of the today realized fuzzy logic control applications has been designed using different heuristic approaches for synthesis and implemented with conventional programming languages on general purpose microcontrollers. This paper aims to present a new methodology to design a fuzzy controller. The methodology is based on the Cell-to-Cell approach to extract the control law. A set of fuzzy rules is then found by using a FAM (Fuzzy associative memories) approach. The proposed procedure was implemented to control the rotor position of a DC motor.

In this paper a fuzzy controller for a flexible arm with one degree of freedom is presented. Goal of the control is to drive the manipulator to the position $\theta$0 avoiding the oscillations due the elasticity of the arm. The performances of the fuzzy controller are evaluated through a series of simulations that shows appreciable results both for the transient and the steady behaviour.

Fuzzy logic based Control Theory has gained much interest in the industrial world, thanks to its ability to formalize and solve in a very natural way many problems that are very difficult to quantify at an analytical level. This paper shows a solution for treating membership function inside hardware circuits. The proposed hardware structure optimizes the memoried size by using particular form of the vectorial representation. The process of memorizing fuzzy sets, i.e. their membership function, has always been one of the more problematic issues for the hardware implementation, due to the quite large memory space that is needed. To simplify such an implementation, it is commonly [1,2,8,9,10,11] used to limit the membership functions either to those having triangular or trapezoidal shape, or pre-definite shape. These kinds of functions are able to cover a large spectrum of applications with a limited usage of memory, since they can be memorized by specifying very few parameters ( ight, base, critical points, etc.). This however results in a loss of computational power due to computation on the medium points. A solution to this problem is obtained by discretizing the universe of discourse U, i.e. by fixing a finite number of points and memorizing the value of the membership functions on such points [3,10,14,15]. Such a solution provides a satisfying computational speed, a very high precision of definitions and gives the users the opportunity to choose membership functions of any shape. However, a significant memory waste can as well be registered. It is indeed possible that for each of the given fuzzy sets many elements of the universe of discourse have a membership value equal to zero. It has also been noticed that almost in all cases common points among fuzzy sets, i.e. points with non null membership values are very few. More specifically, in many applications, for each element u of U, there exists at most three fuzzy sets for which the membership value is ot null [3,5,6,7,12,13]. Our proposal is based on such hypotheses. Moreover, we use a technique that even though it does not restrict the shapes of membership functions, it reduces strongly the computational time for the membership values and optimizes the function memorization. In figure 1 it is represented a term set whose characteristics are common for fuzzy controllers and to which we will refer in the following. The above term set has a universe of discourse with 128 elements (so to have a good resolution), 8 fuzzy sets that describe the term set, 32 levels of discretization for the membership values. Clearly, the number of bits necessary for the given specifications are 5 for 32 truth levels, 3 for 8 membership functions and 7 for 128 levels of resolution. The memory depth is given by the dimension of the universe of the discourse (128 in our case) and it will be represented by the memory rows. The length of a world of memory is defined by: Length = nem (dm(m)＋dm(fm) Where: fm is the maximum number of non null values in every element of the universe of the discourse, dm(m) is the dimension of the values of the membership function m, dm(fm) is the dimension of the word to represent the index of the highest membership function. In our case then Length=24. The memory dimension is therefore 128*24 bits. If we had chosen to memorize all values of the membership functions we would have needed to memorize on each memory row the membership value of each element. Fuzzy sets word dimension is 8*5 bits. Therefore, the dimension of the memory would have been 128*40 bits. Coherently with our hypothesis, in fig. 1 each element of universe of the discourse has a non null membership value on at most three fuzzy sets. Focusing on the elements 32,64,96 of the universe of discourse, they will be memorized as follows: The computation of the rule weights is done by comparing those bits that represent the index of the membership function, with the word of the program memor . The output bus of the Program Memory (μCOD), is given as input a comparator (Combinatory Net). If the index is equal to the bus value then one of the non null weight derives from the rule and it is produced as output, otherwise the output is zero (fig. 2). It is clear, that the memory dimension of the antecedent is in this way reduced since only non null values are memorized. Moreover, the time performance of the system is equivalent to the performance of a system using vectorial memorization of all weights. The dimensioning of the word is influenced by some parameters of the input variable. The most important parameter is the maximum number membership functions (nfm) having a non null value in each element of the universe of discourse. From our study in the field of fuzzy system, we see that typically nfm 3 and there are at most 16 membership function. At any rate, such a value can be increased up to the physical dimensional limit of the antecedent memory. A less important role n the optimization process of the word dimension is played by the number of membership functions defined for each linguistic term. The table below shows the request word dimension as a function of such parameters and compares our proposed method with the method of vectorial memorization[10]. Summing up, the characteristics of our method are: Users are not restricted to membership functions with specific shapes. The number of the fuzzy sets and the resolution of the vertical axis have a very small influence in increasing memory space. Weight computations are done by combinatorial network and therefore the time performance of the system is equivalent to the one of the vectorial method. The number of non null membership values on any element of the universe of discourse is limited. Such a constraint is usually non very restrictive since many controllers obtain a good precision with only three non null weights. The method here briefly described has been adopted by our group in the design of an optimized version of the coprocessor described in [10].

The notion of behavioural dependence of fuzzy concepts is introduced. Examples are given along with first results concerning classical aggregation operators.

In radiation protection and nuclear safety, there are many uncertainties or fuzziness due to subjective human judgement. It is desirable to have a theory by which both non-probabilistic uncertainties, or fuzziness, of human factors and the probabilistic properties of machines can be treated consistently. Fuzzy set theory seems to be an effective tool for analyzing the risk and safety of complex man-machine systems such as nuclear power plants.

The International Atomic Energy Agency's Statute in Article III.A.5 allows it“to establish and administer safeguards designed to ensure that special fissionable and other materials, services, equipment, facilities and information made available by the Agency or at its request or under its supervision or control are not used in such a way as to further any military purpose; and to apply safeguards, at the request of the parties, to any bilateral or multilateral arrangement, or at the request of a State, to any of that State's activities in the field of atomic energy”. Safeguards are essentially a technical means of verifying the fulfilment of political obligations undertaken by States and given a legal force in international agreements relating to the peaceful uses of nuclear energy. The main political objectives are: to assure the international community that States are complying with their non-proliferation and other peaceful undertakings; and to deter (a) the diversion of afeguarded nuclear materials to the production of nuclear explosives or for military purposes and (b) the misuse of safeguarded facilities with the aim of producing unsafeguarded nuclear material. It is clear that no international safeguards system can physically prevent diversion. The IAEA safeguards system is basically a verification measure designed to provide assurance in those cases in which diversion has not occurred. Verification is accomplished by two basic means: material accountancy and containment and surveillance measures. Nuclear material accountancy is the fundamental IAEA safeguards mechanism, while containment and surveillance serve as important complementary measures. Material accountancy refers to a collection of measurements and other determinations which enable the State and the Agency to maintain a current picture of the location and movement of nuclear material into and out of material balance areas, i. e. areas where all material entering or leaving is measurab e. A containment measure is one that is designed by taking advantage of structural characteristics, such as containers, tanks or pipes, etc. To establish the physical integrity of an area or item by preventing the undetected movement of nuclear material or equipment. Such measures involve the application of tamper-indicating or surveillance devices. Surveillance refers to both human and instrumental observation aimed at indicating the movement of nuclear material. The verification process consists of three over-lapping elements: (a) Provision by the State of information such as - design information describing nuclear installations; - accounting reports listing nuclear material inventories, receipts and shipments; - documents amplifying and clarifying reports, as applicable; - notification of international transfers of nuclear material. (b) Collection by the IAEA of information through inspection activities such as - verification of design information - examination of records and repo ts - measurement of nuclear material - examination of containment and surveillance measures - follow-up activities in case of unusual findings. (c) Evaluation of the information provided by the State and of that collected by inspectors to determine the completeness, accuracy and validity of the information provided by the State and to resolve any anomalies and discrepancies. To design an effective verification system, one must identify possible ways and means by which nuclear material could be diverted from peaceful uses, including means to conceal such diversions. These theoretical ways and means, which have become known as diversion strategies, are used as one of the basic inputs for the development of safeguards procedures, equipment and instrumentation. For analysis of implementation strategy purposes, it is assumed that non-compliance cannot be excluded a priori and that consequently there is a low but non-zero probability that a diversion could be attempted in all safeguards ituations. An important element of diversion strategies is the identification of various possible diversion paths; the amount, type and location of nuclear material involved, the physical route and conversion of the material that may take place, rate of removal and concealment methods, as appropriate. With regard to the physical route and conversion of nuclear material the following main categories may be considered: - unreported removal of nuclear material from an installation or during transit - unreported introduction of nuclear material into an installation - unreported transfer of nuclear material from one material balance area to another - unreported production of nuclear material, e. g. enrichment of uranium or production of plutonium - undeclared uses of the material within the installation. With respect to the amount of nuclear material that might be diverted in a given time (the diversion rate), the continuum between the following two limiting cases is cons dered: - one significant quantity or more in a short time, often known as abrupt diversion; and - one significant quantity or more per year, for example, by accumulation of smaller amounts each time to add up to a significant quantity over a period of one year, often called protracted diversion. Concealment methods may include: - restriction of access of inspectors - falsification of records, reports and other material balance areas - replacement of nuclear material, e. g. use of dummy objects - falsification of measurements or of their evaluation - interference with IAEA installed equipment.As a result of diversion and its concealment or other actions, anomalies will occur. All reasonable diversion routes, scenarios/strategies and concealment methods have to be taken into account in designing safeguards implementation strategies so as to provide sufficient opportunities for the IAEA to observe such anomalies. The safeguards approach for each facility will make a different use of these procedures, equipment and instrumentation according to the various diversion strategies which could be applicable to that facility and according to the detection and inspection goals which are applied. Postulated pathways sets of scenarios comprise those elements of diversion strategies which might be carried out at a facility or across a State's fuel cycle with declared or undeclared activities. All such factors, however, contain a degree of fuzziness that need a human judgment to make the ultimate conclusion that all material is being used for peaceful purposes. Safeguards has been traditionally based on verification of declared material and facilities using material accountancy as a fundamental measure. The strength of material accountancy is based on the fact that it allows to detect any diversion independent of the diversion route taken. Material accountancy detects a diversion after it actually happened and thus is powerless to physically prevent it and can only deter by the risk of early detection any contemplation by State authorities to carry out a diversion. Recently the IAEA has been faced with new challenges. To deal with these, various measures are being reconsidered to strengthen the safeguards system such as enhanced assessment of the completeness of the State's initial declaration of nuclear material and installations under its jurisdiction enhanced monitoring and analysis of open information and analysis of open information that may indicate inconsistencies with the State's safeguards obligations. Precise information vital for such enhanced assessments and analyses is normally not available or, if available, difficult and expensive collection of information would be necessary. Above all, realistic appraisal of truth needs sound human judgment.

A direct method of fuzzy inference and a fuzzy algorithm with learning function are applied to the steam generator level control of nuclear power plant. The fuzzy controller by use of direct inference can control the steam generator in the entire range of power level. There is a little long response time of fuzzy direct inference controller at low power level. The rule base of fuzzy controller with learning function is divided into two parts. One part of the rule base is provided to level control of steam generator at low power level (0%∼30% of full power). Response time of steam generator level control at low power level with this rule base is shown generator level control at low power level with this rule base is shown to be shorter than that of fuzzy controller with direct inference.

Fuzzy set theory has been extensively researched in various fields of engineering. In nuclear science, a significant influence of fuzzy sets can be noticed. However, applications of fuzzy set theory to nuclear engineering is novel. In this paper, we start with a basic statement of the decision-making process based on fuzzy set theory, and then apply it to nuclear science with some practical applications (a fuzzy decision making in an accidental release to the atmosphere as well as in a problem of land suitability classification). We believe that the use of fuzzy set theory in nuclear science has potential advantages.

Nuclear reactor operation is a human intensive task; one of the features of a problem for which fuzzy controllers present the most suitable solution. The performance of the fuzzy controllers can further be improved through tuning. In this work, application of a fuzzy controller in real-time control of a nuclear reactor is presented. The fuzzy controller is tuned on-line using direct gradient search method.

The presence of older installations, in particular nuclear facilities, demands extra studies concerning the safety evaluation of those installations. One of the aspects to deal with is the safety of the several transmission lines in a nuclear installation, for instance the safety of control, safety against fire, etc.. . This paper investigates the use of fuzzy set theory in the inspection of transmission lines of nuclear installations at SCK/CEN, Belgium.

A Fuzzy Logic Controller for handing the swell/shrink problems of nuclear steam generators is designed, implemented and tested on the compact nuclear simulator at Korea Atomic Energy Research Institute. Its performance is found to be better than of the PI controller originally being used. In terms of the total variations for the control actions and for the flow error curve, the ones by the fuzzy controller are found to be less than one third of those by the PI controller.

The objective of this paper is to develop a fuzzy logic based decision-making system to detect low current faults using multiple detection algorithms. This fuzzy system utilizes a fuzzy expert model which executes an operation without complicated mathematical models. This fuzzy system decides the performance weights of the detection algorithms. The weights and the turnouts of the detection algorithms discriminate faults from normal events. This system can also be a generic group decision-making tool for other areas of power system protection.

The Objective of this paper is to provide fuzzy control designers with a design tool for stable fuzzy logic controllers. Given multiple sets of data disturbed by vagueness uncertainty, we generate the implicative rules that guarantee stability and robustness of closed-loop fuzzy dynamic systems. We propose the cell-state transition method which utilizes Hsu's cell-to-cell mapping concept [1]. As a result, a generic and implementable design methodology for obtaining a fuzzy feedback gain K, a fuzzy hypercube [2], is provided and illustrated with simple examples.

In this paper, we notice the fact that a human learning process is characterized by a process under a natural language environment, and discuss an approach of learning based on indirect linguistic instructions. An instruction is interpreted through some meaning elements and each trend. Fuzzy evaluation rule are constructed for the searched meaning elements of the given instruction, and the performance of a system to be learned is improved by the evaluation rules. In this paper, we propose a framework of learning based on indirect linguistic instruction based learning using fuzzy theory: FULLINS(FUzzy-Learning based on Linguistic IN-Struction). The validity of FULLINS is shown by applying it to helicopter flight control.

The paper deal with the differences between a fuzzy logic controller with a complete linguistic description and one with an incomplete linguistic description. The conditions to get a complete crisp controller by using a fuzzy logic controller with incomplete description are analyzed, and an application to the control of an analog PLL circuit is described, [1].

In this paper, the design technique of fuzzy controller using pole placement method and the stability analysis of the system are discussed. The consequent parts of the fuzzy model representing the fuzzy control system are descrived by linear stated equations. It cannot be guaranteed that the total fuzzy system is stable even if all subsystems are stable. The range of the consequent parameters of fuzzy feedback controller which is stable for each fuzzy subspace of the input space are derived, using a rather simplified stability criterion. Then, the consequent parameters of fuzzy controller is determined with the sufficient condition that the fuzzy feedback controller maintain robust stability for the model of other subspace.

In this paper, the absolute stability criterion of nonlinear plants with sector bounded nonlinear feedback is derived. The result obtained is useful for applications, in particular, stability analysis and design of fuzzy logic controllers.

The purpose of this work is to build a fuzzy model of a batch culture for a process control. The process is highly nonlinear system with large delay. This paper presents two methods of modeling the process behavior. One is a method of recognizing them by fuzzy rules that are contracted by the pattern analysis in consideration of skilled operators' way. The other is a method of predicting them by approximate linear models and fuzzy rules by statistic analysis.

A fuzzy expert system for prediction and mitigation of sludge bulking was developed for an activated sludge process which treats waste water from a food industry. The developed system is able not only to infer the degree of progress of sludge bulking but also to generate remedial operation guides which may be sent to the local controllers as remote set points. One of the important consequences through this study is the BI (Bulking Index) inferred by the bulking prediction expert system was found to have a close correlation with the SVI (Sludge Volume Index) which is a practical measure of degree of bulking but needs tedious chores for its measurement.

This paper deals with an industrial application of a fuzzy feedback combined learning control to an industrial batch free radical polymerization reactor. As a result, the plant has reduced the batch reaction time by 50 minute and stabilized both by 40 percent reduction of the standard deviations of product qualities, such as the total solid content and the graft gum, and by 45 percent reduction of the standard deviation of the batch reaction end time.

A refuse incineration plant is a complex process, whose multi-variable control problems can not be solved conventionally by deriving an exact mathematical model of the process. The usage of advanced fuzzy technologies within the suitable development methodology is demonstrated by a controller implemented for the refuse incineration plant in Hamburg-Stapelfeld, Germany.

The paper presents an effective method for finding the propagation structure of the real origin of a system malfunction. It uses a combined system model consisting of Structural Model (SM) in the form of Fuzzy Directed Graph and Behavior Model (BM) as a set of Fuzzy Relational Equations $A\;{\circ}\;R\;=\;B$. Here a specially proposed fuzzy inference technique is checked and investigated. Finally a test example for fault diagnosis of an industrial system is given and analyzed.

Fuzzy logic places a considerable burden on an inference engine for applications such as control or approximate reasoning. Various neural network architectures have been proposed to deal with the computational task, and yet, maintain flexibility in the desired traits of the final system. Recently, we introduced a trainable network architecture whose nodes implement weighted Yager additive hybrid operators for fuzzy logic inference in an approximate reasoning setting. In this paper we examine the utility of such networks for control situations. We show that they are capable of learning control functions which are piece-wise monotonic in each of the variables. The learning ability is demonstrated through an example.

We propose two-degree-of-freedom fuzzy neural network control systems. It has a hierarchical structure of two sets of control knowledge, thus it is easy to extract and refine fuzzy rules before and after the operation has started, and the number of fuzzy rules is reduced. In addition an example application of automatic vehicle operation is reported and its usefulness is shown simulation.

We propose the new method about the neural-based pattern recognition by using Hadamard transform for the improvement of learning speed, stability and flexibility of network. We can obtain the spatial feature of pattern by Hadamard transformed pattern. We carried out an experiment to estimate the effect of Hadamard transform. We tried the learning of numeric patterns, and tried the pattern recognition with noisy pattern. As a result, the learning times of the network for the 'Hadamard' case is smaller than that of usual case. And the recognition rate of the network for the 'Hadamard' case is higher than that of usual case, too.

This paper presents the idea of a neural fuzzy controller with application to the control of an industrial machining process. The structure of such a controller, which links the idea of a fuzzy controller and a neural network, is suggested. Results of comparative simulations indicate that the proposed neural fuzzy controller performs equally well as a fuzzy logic controller; moreover, it is more flexible and allows faster data processing.

Fuzzy algorithm is essentially nondeterministic, but to guarantee the stable control the fuzzy control program should be deterministic in practice. Fuzzy controllers with matrix representation is very simple in construction and very fast in computation. The value of the matrix is not adequate at the first place, but can be modified by using the neural networks. We apply the simple heuristic techniques to modify the matrix successfully.

A dynamic-structure system is one that has the flexibility to change the system configuration automatically so as to operate in an optimal manner. A conceptural model for a dynamic-structure system is presented in this paper. In this model, the interchangeable components of the overall system are grouped together. Their activity levels are evaluated by an intelligent preprocessor that is associated with the group. A knowledge-based task distribution system evaluates the activity levels and makes decisions as to how the components operating below capacity should be shared with workcells that have similar components that are overloaded. Associated decision making can be effected through fuzzy logic and particularly the compositional rule of inference. A simulation example is given to illustrate the application of dynamic structuring.

For the design of multivariable fuzzy control systems the decomposition of control rules is preferable since it alleviates the complexity of the problem. In some systems, however, inference error of the Gupta's decomposition method is inevitable because of its approximate nature. In this paper, we propose a new multivariable fuzzy controller with a coordinator which can reduce the inference error of the decomposition method by using an index of applicability.

In this paper, two methods of fuzzy modeling are prsented to describe the input-output relationship effectively based on relation characteristics utilizing simplified reasoning and neuro-fuzzy reasoning. The methods of modeling by the simplified reasoning and the neuro-fuzzy reasoning are used when the input-output relation of a system is 'crisp' and 'fuzzy', respectively. The structure and the parameter identification in the modeling method by the simplified reasoning are carried out by means of FCM clustering and the proposed GA hybrid scheme, respectively. The structure and the parameter identification in the modeling method by the neuro-fuzzy reasoning are carried out by means of GA and BP algorithm, respectively. The feasibility of the proposed methods are evaluated through simulation.

The purpose of the paper is to explain some heuristic, common sense suppositions of fuzzy control. It is shown that Fuzzy Control is a kind of quasilinear interpolation of prototypes. Control function can be sufficiently exact represented as piecewise-linear function. The best interpolation is connected with normalized intersected fuzzy sets.

Proficiency in creating a knowledge base is required for high accuracy fuzzy control. To overcome this a fuzzy inference method is proposed that take these membership functions from the probability densities showing the distribution of the mesurement values. And a method using a rough fuzzy knowledge base automatically created from the basic measurement data and tuned using the gradient method is proposed. In actual tests, these were applied to automatic high accuracy adjustment devices for magnetic head and for high frequency circuits with good results.

We describe six important defuzzification methods and their respective merits and shortcomings, dependent on the rules, domains, etc. Furthermore, we present an alternative approach, the so called ξ-Quality defuzzification method, for the case that the output fuzzy sets have different shape or are asymmetric.

This paper studies the evaluation of mechanical design plans through fuzzy cluster. Plans are classified into two sets, 'good' and 'bad'. The membership of a plan to the 'good' set is numerically equal to the distance to the 'bad' set. The central parameter of the 'good' set is defined as '1', and that of the 'bad' set '0'. This will greatly simplify calculations. The result of the calculating example proves the method available.

We look at the problem of defuzzification in situations in which in addition to the usual fuzzy output of the controller there exists some ancillary restriction on the allowable defuzzified values. We provide two basic approaches to address this problem. In the first approach we enforce the restriction by selecting the defuzzified value through a random experiment in which the values which have nonzero probabilities are in the allowable region, this method is based on the RAGE defuzzification procedure and makes use of a nonmonotonic conjunction operator. The second approach which in the spirit of the commonly used methods, a kind of expected value, converts the problem to a constraint optimization problem.

A promising approach to get the benefits of neural networks and fuzzy logic is to combine them into an integrated system to merge the computational power of neural networks and the representation and reasoning properties of fuzzy logic. In this context, this paper presents a fuzzy neural network which is able to code fuzzy knowledge in the form of it-then rules in its structure. The network also provides an efficient structure not only to code knowledge, but also to support fuzzy reasoning and information processing. A learning scheme is also derived for a class of membership functions.

This paper introduces a new model for forecasting groundwater level on the basis of analysing defect of finite element method. The new model is built with fuzzy sets and neural networks. It is convenient for use. We computed the groundwater level of one city in P. R. China with it and got a very satisfactory result. It can be popularized to corecast groundwater level of mine.

In this paper, a fuzzy-neural interpolating network is proposed to efficiently approximate a nonlinear function. Specifically, basis functions are first constructed by Fuzzy Membership Function based Neural Networks (FMFNN). And the fuzzy similarity, which is defined as the degree of matching between actual output value and the output of each basis function, is employed to determine initial weighting of the proposed network. Then the weightings are updated in such a way that square of the error is minimized. To show the capability of function approximation of the proposed fuzzy-neural interpolating network, a numerical example is illustrated.

In the paper, a new design method of rule-based fuzzy modeling is proposed for model identification of nonlinear systems. The structure indentification is carried out, utilizing fuzzy c-means clustering. Fuzzy-neural networks composed back-propagation algorithm and linear fuzzy inference method, are used to identify parameters of the premise and consequence parts. To obtain optimal linguistic fuzzy implication rules, the learning rates and momentum coefficients are tuned automatically using a modified complex method.

The problem of the optimization of fuzzy relational models for dealing with (non-fuzzy) numerical data is investigated. In this context, interfaces optimization assumes particular importance, becoming a determinant factor in what concerns the overall model performance. Considering this, several scenarios for building fuzzy relational models are presented. These are: (i) optimizing I/O interfaces in advance (independently from the linguistic part of the model); (ii) optimizing I/O interfaces in advance and allowing that their optimized parameters may change during the learning of the linguistic part of the model; (iii) build simultaneously both interfaces and the linguistic subsystem; and (iv) build simultaneously both linguistic subsystem and interfaces, now subject to semantic integrity constraints. As linguistic subsystems, both a basic type and an extended versions of fuzzy relation equations are exploited in each one of these scenarios. A comparative analysis of the differ nt approaches is summarized.

We consider a fuzzy controller corresponding to PI controller. This controller is applied to a controlled object which is a first order lag system with dead time. An antecedent part is divided into 3, 5, and 7 parts ( membership function of triangle shape ), and a consequent part into 3, 5, and 7 parts ( membership function of singleton ). In each combination of an antecedent part and a consequent one. We compare control efficiency under the performance criteria such that the overshoot is kept 20% and the ITAE index is minimized.

We investigate a systematic design procedure of automated rule generation of fuzzy logic based controller for uncertain dynamic systems such as an engine dynamic model.“Automated Tuning”means autonomous clustering or collection of such meaningful transitional relations in the state-space. Optimal control strategies are included in the design procedures, such as minimum squared error, minimum time, minimum energy or combined performance criteria. Fuzzy feedback control systems designed by the cell-state transition method have the properties of closed-loop stability, robustness under parameter variabtions, and a certain degree of optimality. Most of all, the main advantage of the proposed approach is that reliability can be potentially increased even if a large grain of uncertainty is involved within the control system under consideration. A numerical example is shown in which we apply our strategic fuzzy controller design to a highly nonlinear model of engine idle speed contr l.

An Optimum Fuzzy Controller which can be sued to direct the driver to control a running train in an optimum operation way has been developed. In the development process of the controller, the theory and technology of Optimum Control and Fuzzy Control are applied. Practical field tests have been carried out in P.R. of China. In order to make the function of the controller more perfect, the controller is improved by the advanced fuzzy control technology and tool in Japan. The computer simulation of the improved controller has been finished.

Fuzzy techniques are applied to the positioning of an elastic beam. The advantage is that the system model is not needed. A simple fuzzy friction compensator is also used. The final position is achieved within 3/2 the period of the fundamental mode. A fuzzy set of rules is applied for large-angle positioning, with adaptations that reduce the effects of shock. In this case, the final position is achieved within two fundamental periods. There is typically some final error attributed to the dry friction.

There is an opinion of regarding a simple fuzzy logic controller as a kind of Variable Structure Controller in recent years. The opinion may provide an analytical basis which describes the robustness to uncertainty and the stability of a fuzzy logic controller. So in this paper, a fuzzy logic controller based on the Variable Structure System with is designed for a robot manipulator which is a class of complex, nonlinear system with uncertainty. Fuzzy control rules, membership shape of the I/O variables of the fuzzy logic controller are designed for guaranteeing the stability of an overall control system. From a computer simulation of dynamic control of a two link robot manipulator, the design procedure of the fuzzy logic controller is validated.

In this paper, we present a fuzzy-number-oriented methodology to model uncertain geometric robot environment and to manipulate geometric uncertainty between robot coordinate frames. We describe any geometric primitive of robot environment as a parameter vector in parameter space. Not only ill-known values of the parameterized geometric primitive but the uncertain quantities of coordinate transformation are represented by means of fuzzy numbers restricted to appropriate membership functions. For consistent interpretation about geometric primitives between different coordinate frames, we manipulate these uncertain quantities using fuzzy arithmetic.

In this paper, we reason about the fuzzy dynamic equation based on L-R type fuzzy number, analyze and solve the fuzzy dynamic property and the fuzzy dynamic response of singledegree of freedom system. Specific expressions of fuzzy dynamic response are response.

The elevator group control systems are the control systems that manage systematically three or more elevators in order to efficiently transport the passingers. In the elevator group control system, the area-weight which determines the load biases of elevators is a control parameter closely related to the system performance. This paper proposes a fuzzy model based method to determine the are-weight. The proposed method uses a two-stage fuzzy inference model which is built by the study of area-weight properties and expert knowledge. The proposed method shows the more desirable results than the conventional method in the simulations that use real traffic data.

The purpose of this paper is to construct a prediction system on the chance of rain in a local region using a fuzzy relational model. The prediction system consists of two parts. One is a prediction part on the chance of rain. The compositional law of fuzzy inference, proposed by Zadeh, is applied to predict the chance of rain. The other is a learning part of a fuzzy relational model using input-output data. A simple and fast learning algorithm is used in this part. Simulations are carried out by the actual weather data in our city and their results show the validity of prediction by the fuzzy relational approach.

This paper proposes a new on-line fuzzy model identification(ONFID) algorithm in which the fuzzy model evaluation stage is incorporated. The fuzzy model evaluation is performed by the fuzzy equality index which is known to be a useful tool to evaluate the performance of the identified fuzzy model. Then the fuzzy model is updated according to the result of the evaluation. Proposed ONFID algorithm can sensibly identify to the system changes. To show the usefulness of the proposed algorithm, it is applied to the fuzzy model identification problem of the gas furnace and the output prediction problem of the flexible joint manipulator which is a nonlinear system.

Numerical solution of inverse problem of Takagi-Sugeno fuzzy model is proposed. The method is located on the application of numerical optimization to the fuzzy model. Steepest descent method is used for the numerical optimization. We use the linear approximation of fuzzy model, called pseudo first order approximation, by fixing the membership value on the neighborhood of the corresponding input. It is introduced in order to reduce the difficulty of optimization process. The efficiency of this method is shown by a numerical experiment.

This paper discusses the hull form generation from fuzzy model constructed with actual ship data using fuzzy concept. SAC, which is the most important factor in the hull form generation, is expressed by a fuzzy model describing the relationships among design parameters, which have a great influence on SAC, through model identification process with the actual ship data and design parameters. Then, we can infer the SAC of an aimed ship through the process of fuzzy inference and decide the offset of a front view by making the fuzzy model between SAC and offset as well. In conclusion, this paper makes a step forward from the geometrical definition, which has been used for hull form generation so far, to direct mathematical formulae about the relationship between design parameters and offset. So, if the design parameters are given, we can generate the hull form taking such properties into account.

A fuzzy mathematical model is presented that can be applied to support the inspection of organizations by an internal or external evaluation group. The model offers the opportunity to deal with the situation of common practice in which such evaluations are considered as a time series rather than single events. A hypothetical but realistic example is given to illustrate the computational procedure involved.

This paper describes simulation of navigating a sailboat around obstacles to a goal as quickly and safely as possible. Navigation strategies using concepts from fuzzy control are compared with more conventional ones through application at the levels of choosing an optimal heading and steering the sailboat towards that heading.

In this paper, we propose an algorithm of obstacle avoidance using fuzzy inferences. After the basic idea of the path generation algorithm using piecewise polynomials is described, the obstacle avoidance problem using fuzzy inferences is considered. Main concept of the avoidance algorithm is to modify intermittent point data using fuzzy inferences and to generate the collision free path based on the modified data. Finally, simulation result demonstrate the effectiveness of the proposed algorithm.

The fuzzy control theory is applied to control a container crane, which is a very complicated system and controled manually by experts. As reference velocities of trolley and hoist of the container crane, we use those decided by experts, and express them by fuzzy model. We control the crane to follow the reference velocities by using fuzzy controllers. The fuzzy controllers are designed on the container crane. We made a model container crane and applied the suggested method to it

Presented in this paper is a newly developed motion planning method of an autonomous mobile robot(MAR) which can be applied to flexible manufacturing systems(FMS). The mobile robot is designed for transporting tools and workpieces between a set-up station and machines according to production schedules of the whole FMS. The proposed method is implemented based on an earlier developed real-time obstacle avoidance method which employs Kohonen network for pattern classification of sonar readings and fuzzy logic for local path planning. Particulary, a novel obstacle avoidance method for moving objects using a collision index, collision possibility measure, is described. Our method has been tested on the SNU mobile robot. The experimental results show that the robot successfully navigates to its target while avoiding moving objects.

This paper presents a methodology of path planning and navigation for an autonomous mobile robot. A fast algorithm using decomposition technique, which computes the optimal paths between all pairs of nodes, is proposed for real-time calculation. The robot is controlled by fuzzy approximation reasoning. Our new methodology has been implemented on a mobile robot. The results show that the robot successfully navigates to its destination following the optimal path.

The paper discusses the problem of controlling systems with a very high number of input variables effectively by fuzzy If . . . then rules. The basic idea is the partition of the state space into domains, which step can be done even iteratively several times, and every domain has its own sub rule base referring to a considerably lower number of variables than the original space. In this manner the number of necessary rules is drastically reduced and time complexity of the control algorithm remains acceptable.

The paper introduces a new method for fuzzy processing. The method allows handing a piece of information lost in the classic fuzzification process, and thus neglected by other methods. Processing the result after fuzzification is sustained by the interpretation that the input-output set mapping, specified by the IF-THEN rules, can be regarded as a direct mapping of their corresponding alpha-cuts. Processing involves just singletons as intermediary results, the final result being a combination of singletons obtained from fired rules.

Fuzzy optimal control is considered. An optimal sequence of controls is sought best satisfying fuzzy constraints on the controls and fuzzy goals on the states (outputs), with a fuzzy system under control Control over a fixed and specified, implicitly specified, fuzzy, and infinite termination time is discussed. For computational efficiency a small number of reference fuzzy staters and controls is to be assumed by which fuzzy controls and stated are approximated. Optimal control policies reference fuzzy states are determined as a fuzzy relation used, via the compositional rule of inference, to derive an optimal control. Since this requires a large number of overlapping reference fuzzy controls and states implying a low computational efficiency, a small number of nonoverlapping reference fuzzy states and controls is assumed, and then interpolative reasoning is used to infer an optimal fuzzy control for a current fuzzy state.

We have designed a multiple-valued fuzzy Approximate Analogical-Reseaning system (AARS). The system uses a similarity measure of fuzzy sets and a threshold of similarity ST to determine whether a rule should be fired, with a Modification Function inferred from the Similarity Measure to deduce a consequent. Multiple-valued basic fuzzy blocks are used to construct the system. A description of the system is presented to illustrate the operation of the schema. The results of simulations show that the system can perform about 3.5 x 106 inferences per second. Finally, we compare the system with Yamakawa's chip which is based on the Compositional Rule of Inference (CRI) with Mamdani's implication.

In this short note we show that a number of conclusions unacceptable to our intuitions or commonsense knowledge can be drawn from Zadeh's possiliblity theory.

By the exponential representation form (EF) for fuzzy logic, any fuzzy value a (in fuzzy valued logic or fuzzy linguistic valued logic) can be represented as Bc, where B is called the truth base and C the confidence exponent. This paper will propose the basic concepts of this form and discuss its interesting properties. By using a different truth base, the exponential form can be used to represent the positive and the negative logic in fuzzy valued logic as well as in fuzzy linguistic valued logic. Some Simple application examples of EF for approximate reasoning are also illustrated in this paper.

We tried gas identification by using one semiconductive gas sensor. As a method of gas identification, we used the fuzzy reasoning with fuzzy set of slope of gas pattern which is divided into arbitary interval. As a result, we got a good successful rate for hydrogen 66.6%, propane 79.1%, butane 100%, methane 100%, city gas 79.1% and alcohol 91.6%, respectively.

Though fuzzy control is very popular at present, the application field of fuzzy system will be wider if we design it as a man-machine system. We suggest, in this paper, a man-machine cooperating system which makes easy the manual control of a triple inverted pendulum by simple fuzzy controller, and verify its effectiveness by experiments.

Machine learning in an uncertain or unknown environment is of vital interest to those working with intelligent systems. The ability to garner new information, process it, and increase the understanding/ capability of the machine is crucial to the performance of autonomous systems. The field of artificial intelligence provides two major approaches to the problem of knowledge engineering-expert systems and neural networks. Harnessing the power of these two techniques in a hybrid, cooperating system holds great promise.

The base of proposed decomposing approach is multilevel process of agregation (simplificative transformation) of the description of the project structures. The new classification of fuzzy choice operators is suggested to obtain the decomposing correlations.

A novel digital signal processing technique based on fuzzy rules is proposed for estimating nonstationary signals, such as image signals, contaminated with additive random noises. In this filter, fuzzy rules are utilized to set the filter parameters, taking the local characteristics of the signal into consideration. The introduction of the fuzzy rules is effective, since the rules to set the filter parameters is usually expressed ambiguously. Computer simulations verify its high performance.

We present a flexible retrieval system of face photographs based on their linguistic descriptions in terms of fuzzy perdicates. While natural for describing a face, linguistic expressions are also subjective, which affects the retrieval result. Thus, the capability of a retrieval system to adjust to different users becomes very important. In this research we use fuzzy logic techniques, for describing image data, inference for retrieval and adjustment to a new user. Experimental results of the adjustment are also included.

This paper presents a fuzzy system that estimates the optimal bit allocation matrices for the spatially active subimage classes of adaptive transform image coding in noisy channels. Transform image coding is good for image data compression but it requires a transmission error protection scheme to maintain the performance since the channel noise degrades its performance. The fuzzy system provides a simple way of estimating the bit allocation matrices from the optimal bit map computed by the method of minimizing the mean square error between the transform coefficients of the original and the reconstructed images.

We are concerned with developing a robust method for comprehensive scene analysis. In particular, we address the problem of representing spatial relations between regions in a segmented 2D image. Spatial relations are modeled as fuzzy sets. The method has two aspects, symbolic and quantitative, consisting of assigning labels for spatial relations and numeric degrees to which a relation holds respectively. The procedure of deriving a spatial relation is hierarchical taking into account geometric/physical characteristics of the regions in question.

This paper investigates the use of Fuzzy vector quantizer(FVQ) in speech synthesis. To compress speech data, we employ K-means algorithm to design codebook and then FVQ technique is used to analysize input speech vectors based on the codebook in an analysis part. In FVQ synthesis part, analysis data vectors generated in FVQ analysis is used to synthesize the speech. We have fined that synthesized speech quality depends on Fuzziness values in FVQ, and the optimum fuzziness values maximized synthesized speech SQNR are related with variance values of input speech vectors. This approach is tested on a sentence, and we compare synthesized speech by a convensional VQ with synthesized speech by a FVQ with optimum Fuzziness values.

This work presents a net-based structure to model approximate reasoning using fuzzy production rules, the Fuzzy Petri Net model. The Fuzzy Petri Net model is formally defined as a n-uple of elements. It allows for the representation of simple and complex forms of rules such as rules with conjunction in the antecedent and qualified rules. Parallel rules and conflicting rules can be modeled as well. We also developed an analysis method based on state equations and two fuzzy reasoning algorithms. Finally, the proposed method is applied to an example.

In this paper, first, the fuzzy Petri net inference mechanism with learning function is proposed by using the extended fuzzy Petri nets. Secondly, a control system with this new inference engine is proposed. This system can do automatically and easily the knowledge acquisition from the operator's empirical data and can also be controller adaptively under the big parameter change.

In this study, we newly formulated the link capcity allocation problem and the link capacity allocation and routing problem in an voice/data integrated network by the fuzzy set concept. We developed efficient algorithms for the above fuzzified problems and successfully showed that the fuzzy set theory is the powerful tool for the design problems in communication networks.

We present here two Petri nets formalisms that can deal with uncertainty by the use of necessity-valued logic. The first and basic model, called necessity-valued Petri nets (NPN), can at the same time deal with uncertainty on markings are on transitions. The second model, called necessity-valued Petri nets (TNPN), is an extension of both NPN and timed Petri nets.

In the cellular mobile communications as decreasing the cell radius to increse the reuse factor of frequencies, the handoff requests are increasing so that the efficient handoff decision making becomes a crucial problem. In this simulation study, we evaluate a set of handoff algorithms based on fuzzy-multicriteria decision making. These algorithms uses the parameters including the received signal strength intensity, the bit error rate and the distance between a mobile station and a base station. We compare the fuzzy algorithms in terms of call block ratio and handoff request ratio and call force ratio, and show the applicability of those algorithms in the cellular mobile communication systems.

In this paper, we describe the use of certain optimization techniques, principally dynamic programming and high level computational methods, to enhance the capabilities of a fuzzy adaptive neural network controller which we had developed for on-line control and adaption on complex nonlinear processes. Potential applications to an array of processes from diverse fields are discussed.

A fuzzy control system typically requires“tuning,”or adjuctment of the parameters defining its linguistic variables. Automating this process amounts to applying a second“metacontrol”layer to drive the controller and plant to desired performance levels. Current methods of automated tuning rely on a single crisp numeric functional to evaluate control system performance. A generalization of Box's complex algorithm allows more realistic tuning based on lexicographic aggregation of multiple ordinal scales of performance, such as effectiveness and efficiency. The method is presented and illustrated using a simple inverted pendulum control system.

This paper proposes that the best reasoning(i.e. rule evaluation) method which should be used in a fuzzy system significantly depends on the reasoning environment. It is shown that allowing for dynamic switching of reasoning methods leads to better performance, even when only two different reasoning methods are considered. This paper discusses DSFS (Dynamic Switching Fuzzy System) which dynamically switches and finds the best reasoning method (from among 80 different possible reasoning methods) to use depending on the reasoning situation. To overcome the reasoning speed and memory problem of DSFS due to its computational requirements, the DSFS Switching Reasoning Table method is proposed and its higher performance as compared to a conventional fuzzy system is shown. Finally, efforts to obtain general relationships between the characteristics of different reasoning methods and the actual control surface/environment is discussed.

Fuzzy mathematics is used to elicit and evaluate human psychophysical responses in panel tests. The fundamental instrument used is a bar graph whose data is then converted to a paired comparison matrix. Form this matrix we use the theory of Perron and Froebenius to obtain an eigenvalue and eigenvector which indicates not only the panelist's comparitive responses but also the consistency of the responses from that panelist. Tests were done to evaluate the procedure.

This paper presents a simulation study on two self-learning control systems for a fuzzy prediction model of CO (carbon monoxide) concentration:linear control and fuzzy control. The self-learning control systems are realized by using Widrow-Hoff learning rule which is a basic learning method in neural networks. Simulation results show that the learning efficiency of fuzzy controller is superior to that of linear controller.

In controlling a system having many variables to control and multi objectives to satisfy such as a roof crane system, it is often difficult to obtain fuzzy If-Then rules in usual ways. As an alternative, we can more easely obtain rules in such a manner that we obtain each independent group of rules using partial variables for a partial objective. In this case, obtained rules can be conflicting with each other and conventional inference methods cannot handle such rules effectively. In this paper, we propose a roof crane controller with optimal velocity profile generator and a fuzzy logic controller with an inference method suitable for such conflicting rules.

In this paper, a predictive fuzzy control algorithm to supervise the elevator system with plural elevator cars is developed and its performance is evaluated. Elevator group controller must decide which of the cars is suitable for responding the registered hall call and allocate it to the selected car controller. In most cases, the purpose of group control is to minimize waiting time of passengers and occurrence of long wait as much as possible. The proposed algorithm ensures the efficient operations of the group cars and provides the improved level of service, coping with multiple control objects and uncertainty of system state. The feasibility of the proposed control algorithm is evaluated by graphic simulator on computer.

In order to improve productivity, an intelligent control system is presented in the pater. In this intelligent control system, a feedforward neural network and a fuzzy feedback mechanism are adopted to achieve a constant milling force with an adjustable feedrate under a variety of cutting conditions in milling operations.

We are describing the architecture of a fuzzy logic controller using pulse-width-modulation (PDM) technique and a pipeline structure. Features of this controller are: A new architecture for the inference unit, reduced chip area and less I/O-pins. Additionally we present two different rule-bases: one hardwired with reduced chip-area and the other programmable for prototyping. Also an architecture of a parallel minimum-gate is shown.

An 8b Fuzzy Coprocessor (FC) is presented that has eight programmable fuzzy algorithms and up to 256 inputs, 64 outputs and 16,384 rules. The 6.4mm2 chip fabricated in 1.0$\mu\textrm{m}$ CMOS technology can be used as a stand-alone device or as a macrocell for microcontrollers. Operating at 20MHz crystal frequency, it has a peak performance of 7.9M rules/s. Perspectives of future FC generations are also outlined, including a 12-16b resolution, additional fuzzy set operations, and optimized inference and defuzzification strategies.

A Fuzzy Microprocessor(FMP) is presented, which is suitable for real-time control applications. The features include high speed inference of maximum 114K FLIPS at 20MHz system clocks, capability of up to 128-rule construction, and handing of 8 input variables with 8-bit resolution. In order to realize these features, the fuzzifier circuit and the processing element(PE) are well optimized for LSI implementation. The chip fabricated in 1.2$\mu\textrm{m}$ CMOS technology contains 71K transistors in 82.8 $\textrm{mm}^2$ die size and is packaged in 100-pin plastic QFP.

The computer users who desire the comprehensive use of an advanced network system which permits to link the diversified types of computer from one to others, still tend to increase in number. However, one of the ways to install a highly reliable facilities relative to the network system is to inherit the know-how preserved in the specific technology. Therefore, we will propose the solution of this problem. This paper describes COMNETS, an expert system for constructing distributed networks.

In this paper, a fault evaluation method is proposed, which is to determine whether analog electronic circuits are faulty or not. In our method, evaluation characteristics of an expert test engineer are defined by means of directed graphs. By performing a multi-stage fuzzy inference based on the graphs, novice test engineers can derive a fault evaluation result satisfied by the expert. The effectiveness of our method is checked by some experiments for an amplifier circuit.

This paper describes a model of the subjective reliability analysis, which uses a fuzzy set, natural language expressions and parameterized operations of fuzzy sets, and reflects analysts' subjectivity. The model has the problem of many different analysis results being obtained since the results depend on their subjectivity. As one of the solutions two kinds of mutual agreements based on the analysis results are considered. One is the intersection and the union of the fuzzy sets obtained by the analysis. The other is the weighted average of the fuzzy sets. This paper gives these interpretations from the viewpoint of system reliability analysis. This paper also shows examples of these considerations.

This paper presents the application of fuzzy control to three links of a Rhino robot and compares its performance to traditional PD control. The dynamics of motion of robot links are governed by nonlinear differential equations. The fuzzy controller, being an adaptive technique, gives better performance than the traditional linear PD controller over a typical operational range. The fuzzy controller reaches the desired position with no overshoot, which is unlikely with the PD controller.

Wang and Medel proved (1991) that fuzzy systems with product inference, centroid defuzzification, and everywhere positive membership functions (in particular, Gaussians, Wang, 1992) are capable of approximating any real continuous control function on a compact set to arbitrary accuracy. Kosko (1992) proved that fuzzy systems, in which membership functions have compact support, and combination operation (V-operation) for rules is the sum, are also universal approximators. In this paper, we generalize this result of Kosko and prove that for any &- and V-operations, any defuzzification procedure, and any basic membership function with a compact support, the resulting fuzzy controls are universal approximators. Also, Wang's result is transfered to min-inference.

The main purpose of this paper is to introduce and develop the notion of a fuzzy measure in separable Banach space. This definition of fuzzy measure is a natural generalization of the set-valued measure. Radon-Nikod m theorems for fuzzy measure are established.

This paper presents a fuzzy neural network, called the fuzzy counterpropagation network, that structures its inputs and generates its outputs in a manner based on counterpropagation networks. The fuzzy counterpropagation network is developed by incorporating the concept of fuzzy clustering into the hidden layer responses. Three learning algorithms are introduced for use with the proposed network. Simulations demonstrate that fuzzy counterpropagation networks with the proposed learning algorithms work well on approximating bipolar and continuous functions.

The fuzzifier circuit DPFC 7 is presented. Its features are: programmable membership function, CMOS digital interface, analog and current mode internal processing and integrability without external components. It has been designed to obtain a basic efficient block for fuzzy processing, to be included in a future architecture.