A Competitive and Heuristic Error-Feedback Learning for Fuzzy Modeling

• Main Authors: Hsi-Yan Chang, 張晞彥
• Other Authors: Ji-Chang Lo
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/49286730661173909927

碩士 === 國立中央大學 === 機械工程研究所 === 87 === This article describes a new type of fuzzy system with extrapolation capability to extract MISO fuzzy if-then rules from input-output sample data through learning. The proposed model inherits many merits from Sugeno-type models and their variations that can be found in many fuzzy modeling literature. The model is shown to be an universal approximator of a non-linear mapping. A heuristic error-feedback learning algorithm associated with the new model is suggested. Based on which, the estimator is shown to have a self-adjusting step when approaching a minimum.The power of the algorithm is demonstrated by four numerical examples. Comparison shows that the suggested approach can produce a fuzzy model with a simple methodology in a sense that no other complex optimization techniques are required.