• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1995.10b
• /
• pp.165-173
• /
• 1995

In order to eliminate position errors existing at the steady state in the motion control of robotic maniprlators, a new fuzzy control algorithm is proposed using three variables, position error, velocity error and integral of position errors as input variables of the fuzzy controller, This controller is applied to the tracking control of robotic manipulators in Cartesian space. Three dimensional look-up table is used to reduce the computational time in rel-time control. Simulation and experimental studies are conducted to evaluate the control performance for the two axis direct drive SCARA robot system.

• Journal of Bio-Environment Control
• /
• v.10 no.1
• /
• pp.42-49
• /
• 2001

This study was carried out to develop a fuzzy control technique of ventilation window for controlling a temperature in a greenhouse. To reduce the fuzzy variables, the inside air temperature shop was taken as one of fuzzy variables, because the inside air temperature variation of a greenhouse by ventilation at the same window aperture is affected by difference between inside and outside air temperature, outside wind speed and the wind direction. Therefore, the antecedent variables for fuzzy algorithm were used the control error and its slop, which was same value as the inside air temperature slop during the control period, and the conclusion variable was used the window aperture opening rate. Through the basic and applicative control experiment with the control period of 3 minutes the optimum ranges of fuzzy variables were decided. The control error and its slop were taken as 3 and 1.5 times compared with target error in steady state, and the window opening rate were taken as 30% of full size of the window aperture. To evaluate the developed fuzzy algorithm in which the optimized 19 rules of fuzzy production were used, the performances of fuzzy control and PID control were compared. The temperature control errors by the fuzzy control and PID control were lower than 1.3$^{\circ}C$ and 2.2$^{\circ}C$ respectively. The accumulated operating size of the window, the number of operating and the number of inverse operating for the fuzzy control were 0.4 times, 0.5 times and 0.3 times of those compared with the PID control. Therefore, the fuzzy control can operating the window more smooth and reduce the operating energy by 1/2 times of PID control.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.10a
• /
• pp.207-213
• /
• 1997

In this paper, we generalize Kolmogorov' strong law of large numbers to the the case of independent fuzzy random variables.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.11a
• /
• pp.33-36
• /
• 1997

We prove a uniform strong law of large numbers for sequences of fuzzy random variables.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.11a
• /
• pp.276-279
• /
• 2005

A review of how the functions of two or more independent variables can be approximated by using fuzzy systems is provided in this paper. We start with an exact represention of a linear interpolation function of two independent variables by using a fuzzy system. Next, we describe how this function can be approximated by another fuzzy system with a lesser number or with a desired number of output fuzzy sets. Thus, a reduction of the storage needed is achieved by storing the fuzzy rules or equivalently the output fuzzy set numbers instead of storing the whole discrete function values. A description on how the cubic spl me interpolation function can be represented exactly by using the fuzzy system method is provided, along with a few examples where fuzzy rule tables with a size of 7$\times$7 provide a representation of the functions with relative errors of the order of $10^{2}$ or less.

• Journal of Electrical Engineering and Technology
• /
• v.5 no.3
• /
• pp.503-510
• /
• 2010

This paper proposes the use of fuzzy applications to a 4-machine and 10-bus system to check stability in open conditions. Fuzzy controllers and the excitation of a synchronous generator are added. Power system stabilizers (PSSs) are added to the excitation system to enhance damping during low frequency oscillations. A fuzzy logic power system stabilizer (PSS) for stability enhancement of a multi-machine power system is also presented. To attain stability enhancement, speed deviation ($\Delta\omega$) and acceleration ($\Delta\varpi$) of the Kota Thermal synchronous generator rotor are taken as inputs to the fuzzy logic controller. These variables have significant effects on the damping of generator shaft mechanical oscillations. The stabilizing signals are computed using fuzzy membership functions that are dependent on these variables. The performance of the fuzzy logic PSS is compared with the open power system, after which the simulations are tested under different operating conditions and changes in reference voltage. The simulation results are quite encouraging and satisfactory. Similarly, the system is tested for the different defuzzification methods, and based on the results, the centroid method elicits the best possible system response.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.49 no.7
• /
• pp.378-388
• /
• 2000

In this paper, we propose Fuzzy Polynomial Neural Networks(FPNN) based on Polynomial Neural Networks(PNN) and Fuzzy Neural Networks(FNN) for model identification of complex and nonlinear systems. The proposed FPNN is generated from the mutually combined structure of both FNN and PNN. The one and the other are considered as the premise part and consequence part of FPNN structure respectively. As the consequence part of FPNN, PNN is based on Group Method of Data Handling(GMDH) method and its structure is similar to Neural Networks. But the structure of PNN is not fixed like in conventional Neural Networks and self-organizing networks that can be generated. FPNN is available effectively for multi-input variables and high-order polynomial according to the combination of FNN with PNN. Accordingly it is possible to consider the nonlinearity characteristics of process and to get better output performance with superb predictive ability. As the premise part of FPNN, FNN uses both the simplified fuzzy inference as fuzzy inference method and error back-propagation algorithm as learning rule. The parameters such as parameters of membership functions, learning rates and momentum coefficients are adjusted using genetic algorithms. And we use two kinds of FNN structure according to the division method of fuzzy space of input variables. One is basic FNN structure and uses fuzzy input space divided by each separated input variable, the other is modified FNN structure and uses fuzzy input space divided by mutually combined input variables. In order to evaluate the performance of proposed models, we use the nonlinear function and traffic route choice process. The results show that the proposed FPNN can produce the model with higher accuracy and more robustness than any other method presented previously. And also performance index related to the approximation and prediction capabilities of model is evaluated and discussed.

• Structural Engineering and Mechanics
• /
• v.22 no.6
• /
• pp.665-683
• /
• 2006

The strength behaviors of Fiber Reinforced Plastics (FRP) Composites can be greatly influenced by the properties of constitutive materials, the laminate structures, and load conditions etc, accompanied by many uncertainty factors. So the reliability study on FRP is an important subject of research. Many achievements have been made in reliability studies based on the probability theory, but little has been done on the roles played by fuzzy variables. In this paper, a fuzzy reliability model for FRP laminates is established first, in which the loads are considered as random variables and the strengths as fuzzy variables. Then a numerical model is developed to assess the fuzzy reliability. The Monte Carlo simulation method is utilized to compute the reliability of laminas under the maximum stress criterion. In the second part of this paper, a generalized fuzzy reliability model (GFRM) is proposed. By virtue of the fact that there may exist a series of states between the failure state and the function state, a fuzzy assumption for the structure state together with the probabilistic assumption for strength parameters is adopted to construct the GFRM of composite materials. By defining a generalized limit state function, the problem is converted to the conventional reliability formula that enables the first-order reliability method (FORM) applicable in calculating the reliability index. Several examples are worked out to show the validity of the models and the efficiency of the methods proposed in this paper. The parameter sensitivity analysis shows that some of the mean values of the strength parameters have great influence on the laminated composites' reliability. The differences resulting from the application of different failure criteria and different fuzzy assumptions are also discussed. It is concluded that the GFRM is feasible to use, and can provide an effective and synthetic method to evaluate the reliability of a system with different types of uncertainty factors.

• Journal of the Korean Data and Information Science Society
• /
• v.12 no.1
• /
• pp.99-101
• /
• 2001

In this note, we show that a linear regression model, using entropy and degree of nearness of fuzzy numbers, suggested by Wang and Li[FSS 36, 125-136] seems to be unreasonable by an example.

• Proceedings of the KIEE Conference
• /
• 2002.07d
• /
• pp.2818-2820
• /
• 2002

In this paper, we design neurofuzzy networks architecture by means of linear fuzzy inference. The proposed neurofuzzy networks are equivalent to linear fuzzy rules, and the structure of these networks is composed of two main substructures, namely premise part and consequence part. The premise part of neurofuzzy networks use fuzzy space partitioning in terms of all variables for considering correlation between input variables. The consequence part is networks constituted as first-order linear form. The consequence part of neurofuzzy networks in general structure(for instance ANFIS networks) consists of nodes with a function that is a linear combination of input variables. But that of the proposed neurofuzzy networks consists of not nodes but networks that are constructed by connection weight and itself correspond to a linear combination of input variables functionally. The connection weights in consequence part are learned by back-propagation algorithm. For the evaluation of proposed neurofuzzy networks. The experimental results include a well-known NASA dataset concerning software cost estimation.