• Journal of the Korean Institute of Intelligent Systems
• /
• v.12 no.3
• /
• pp.231-238
• /
• 2002

As a modeling method where the merits of fuzzy inference system and evolutionary computation are put together, evolutionary fuzzy modeling performs global approximate optimization. The paper proposes fuzzy clustering as fuzzy rule generation process which is one of the most important steps in evolutionary fuzzy modeling. With application of fuzzy clustering into the experiment or simulation results, fuzzy rules which properly describe non-linear and complex design problem can be obtained. The efficiency of evolutionary fuzzy modeling can be improved utilizing the membership degrees of data to clusters from the results of fuzzy clustering. To ensure the validity of the proposed method, the real design problem of an automotive inner trim is applied and the global approximation is achieved. Evolutionary fuzzy modeling is performed for several cases which differ in the number of clusters and the criterion of rule selection and their results are compared to prove that the proposed method can provide proper fuzzy rules for a given system and reduce computation time while maintaining the errors of modeling as a satisfactory level.

• IE interfaces
• /
• v.15 no.4
• /
• pp.338-348
• /
• 2002

In this paper, we present a restricted tabu search(RTS) algorithm that schedules jobs on identical parallel machines in order to minimize the maximum lateness of jobs. Jobs have release times and due dates. Also, sequence-dependent setup times exist between jobs. The RTS algorithm consists of two main parts. The first part is the MATCS(Modified Apparent Tardiness Cost with Setups) rule that provides an efficient initial schedule for the RTS. The second part is a search heuristic that employs a restricted neighborhood generation scheme with the elimination of non-efficient job moves in finding the best neighborhood schedule. The search heuristic reduces the tabu search effort greatly while obtaining the final schedules of good quality. The experimental results show that the proposed algorithm gives better solutions quickly than the existing heuristic algorithms such as the RHP(Rolling Horizon Procedure) heuristic, the basic tabu search, and simulated annealing.

• Journal of information and communication convergence engineering
• /
• v.6 no.2
• /
• pp.222-227
• /
• 2008

The important problem in designing a Fuzzy Logic Controller(FLC) is generation of fuzzy control rules and it is usually the case that they are given by human experts of the problem domain. However, it is difficult to find an well-trained expert to any given problem. In this paper, I describes an application of genetic algorithm, a well-known global search algorithm to automatic generation of fuzzy control rules for FLC design. Fuzzy rules are automatically generated by evolving initially given fuzzy rules and membership functions associated fuzzy linguistic terms. Using genetic algorithm efficient fuzzy rules can be generated without any prior knowledge about the domain problem. In addition expert knowledge can be easily incorporated into rule generation for performance enhancement. We experimented genetic algorithm with a non-trivial vehicle controling problem. Our experimental results showed that genetic algorithm is efficient for designing any complex control system and the resulting system is robust.

• Proceedings of the KIEE Conference
• /
• 2006.11a
• /
• pp.21-23
• /
• 2006

As a general rule for evaluating dependability of a system, reliability is commonly considered which barely rays attention to the system behavior, however the estimation is based on the assumption of a fault-frost system, which may be impracticable and inaccurate especially for complicated system. This paper introduces a security and dependability integrated approach to analyze the availability of a fault-active system both from dependability and security points of view. Two fault modes involved are discussed about the impairment to the system reliance. The approach can be well applied to estimate and quantify the attribute of system robustness with the help of Markov chain process, which is good at solving status related problem. The comparison result between dual system and IEC61850-based almighty backup system is shown to sup-port the suggested approach.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.57 no.4
• /
• pp.551-559
• /
• 2008

Increase in power consumption can cause a serious stability problem of an electric power system without construction of new power plants or transmission lines. Also, it can generate large power loss of the system. In costly and environmentally effective manner to avoid constructing the new infrastructures such as power plants and transmission lines, etc, the distributed generation(DG) has paid great attentions so far as a solution for the above problem. Selection of optimal location and size of the DG is the necessary process to maintain the stability and reliability of existing system effectively. However, the systematic and cardinal rule for this issue is still open question. In this paper, the method to determine optimal location of the DG is proposed by considering power loss when the DG is connected to an electric power grid. Also, optimal size of not only the corresponding single DG but also the multi-DGs is determined with the proposed systematic approach. The IEEE benchmark 30-bus test system is analyzed to evaluate the feasibility and effectiveness of the proposed method.

• Journal of Korea Multimedia Society
• /
• v.22 no.6
• /
• pp.665-675
• /
• 2019

In this paper, we compare and analyze the classification performance of deep learning algorithm Convolutional Neural Network(CNN) ac cording to ensemble generation and combining techniques. We used several CNN models(VGG16, VGG19, DenseNet121, DenseNet169, DenseNet201, ResNet18, ResNet34, ResNet50, ResNet101, ResNet152, GoogLeNet) to create 10 ensemble generation combinations and applied 6 combine techniques(average, weighted average, maximum, minimum, median, product) to the optimal combination. Experimental results, DenseNet169-VGG16-GoogLeNet combination in ensemble generation, and the product rule in ensemble combination showed the best performance. Based on this, it was concluded that ensemble in different models of high benchmarking scores is another way to get good results.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.5
• /
• pp.939-948
• /
• 2012

The most widely used data mining technique is to explore association rules. This technique has been used to find the relationship between items in a massive database based on the interestingness measures such as support, confidence, lift, etc. Association rules are frequently used by retail stores to assist in marketing, advertising, floor placement, and inventory control.In general, association rule technique generates the rule, 'If A, then B.', whereas negative association rule technique generates the rule, 'If A, then not B.', or 'If not A, then B.'. We can determine whether we promote other products in addition to promote its products only if we add negative association rules to existing association rules. In this paper, we proposed the negatively attributable and pure confidence to overcome the problems faced by negative association rule technique, and then we checked three conditions for interestingness measure. The comparative studies with negative confidence, negatively pure confidence, and negatively attributable and pure confidence are shown by numerical examples. The results show that the negatively attributable and pure confidence is better than negative confidence and negatively pure confidence.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.39 no.3
• /
• pp.170-179
• /
• 2016

In this paper we have tackled the outstanding inventory planning problems over new product launching period in a more holistic manner by addressing first the definition of efficient business rules to effectively control and reduce the inventory risks followed by the rigorous explanations on the implementation guide on suggested inventory planning rules. It is not unusual for many companies in the consumer electronics market to make a great effort to reduce the time to launch a new product because the ability to bring out higher performing products in such a short time period greatly increases the probability for them to remain competitive in the high tech market. Among so many newly developed products, those products with new features and technologies appeal to many potential customers while products which fail to win customers by design and prices rapidly disappear in the market. To adapt to this business environment, those companies have been trying to find the answer to minimize the inventory of old products so they can move to next generation products quickly with less obsolete material. In the experimental implementation of our rule-based inventory planning, Company 'S' reduced the inventory cost for the outgoing products as low as 49% of its peak level of its preceding product version in just 5 month after the adoption of rule-based inventory planning process and system. This paper concluded the subject with a suggestion that the best performance of rule-based inventory planning is guaranteed not from one-time campaign of process improvement along with system development but the decision maker's continuing support and attention even without seeing any upcoming business crisis.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.29B no.11
• /
• pp.120-129
• /
• 1992

In this paper we propose a fuzzy-neural network(FNN) which includes both advantages of the fuzzy logic and the neural network. The basic idea of the FNN is to realize the fuzzy rule-base and the process of reasoning by neural network and to make the corresponding parameters be expressed by the connection weights of neural network. After constructing the FNN, a novel controller consisting of a conventional P-controller and a FNN is explained. In this control scheme, the rule-base of a FNN are automatically generated by error back-propagation algorithm. Also the parallel connection of the P-controller and the FNN can guarantee the stability of a plant at initial stage before the rules are completely created. Finally the effectiveness of the proposed strategy will be verified by computer simulations using a 2 degree of freedom robot manipulator.

• International Journal of Control, Automation, and Systems
• /
• v.4 no.3
• /
• pp.372-381
• /
• 2006

The purpose of this study was to develop a process management system to manage ingot fabrication and improve ingot quality. The ingot is the first manufactured material of wafers. Trace parameters were collected on-line but measurement parameters were measured by sampling inspection. The quality parameters were applied to evaluate the quality. Therefore, preprocessing was necessary to extract useful information from the quality data. First, statistical methods were used for data generation. Then, modeling was performed, using the generated data, to improve the performance of the models. The function of the models is to predict the quality corresponding to control parameters. Secondly, rule extraction was performed to find the relation between the production quality and control conditions. The extracted rules can give important information concerning how to handle the process correctly. The dynamic polynomial neural network (DPNN) and decision tree were applied for data modeling and rule extraction, respectively, from the ingot fabrication data.