IMPROVING THE FUZZY SYSTEM PERFORMANCE BY FUZZY SYSTEM ENSEMBLE

This paper proposes a fuzzy system ensemble (FSE) that can improve the system performance in non-linear and complex problems. Each fuzzy sys tem in the FSE is built from two design stages, each stage of which is performed by different genetic algorithms (GAs). The first stage gene rates a fuzzy rule base that covers as many of the training examples a s possible. The second stage builds fine-tuned membership functions th at make the system error as small as possible. These two stages are re peated independently upon the different partitions of input-output var iables. The system error will be reduced further by invoking the FSE t hat combines multiple fuzzy systems with an equal system error weighti ng method where the weight constant is inversely proportional to the F S's error. Applications of the FSE to both the truck backer-upper cont rol and the Mackey-Glass chaotic time-series prediction are presented. For the truck control problem, control performance of the proposed me thod is compared with the approach of Wang and Mendel [IEEE Transactio ns on System, Man, and Cybernetics 22 (1992) 1414] in terms of either the rate of successful controls reaching to the goal or the average tr aveling distance. For the chaotic time-series prediction problem, pred iction accuracy of the proposed method is compared with that of other fuzzy or neural network predictors in terms of non-dimensional error i ndex (NDEI). (C) 1998 Elsevier Science B.V. All rights reserved.