FUZZY SLIDING MODE CONTROLLER DESIGN BASED ON FUZZY UNION REASONING

• Main Authors: Ti-Hung Chen, 陳帝宏
• Other Authors: Chung-Chun Kung
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/93303870249014571153

碩士 === 大同工學院 === 電機工程研究所 === 87 === Recently, fuzzy logic controllers have been applied in many fields, many of which cannot easily be modeled in mathematical processes. Besides, the control action is based on the linguistic control rules, so fuzzy logic controllers are fit for human’s thought. However, there are some drawbacks need to be solved. Firstly, it lacks a systematic and efficient algorithm to design fuzzy logic controllers. Secondly, for a complex plant, the fuzzy rule base become very complicated and the number of fuzzy rules increases exponentially as many as the number of input variables increases linearly. Thirdly, it is so time-consuming to find an optimal fuzzy logic controllers. In this thesis, to overcome the mentioned drawbacks, a fuzzy union reasoning based fuzzy sliding mode controller is presented. In the process of designing the suggested controller, we firstly employ the sliding mode control technique to design the fuzzy rules, so that the fuzzy controller has the characteristics of stability and robustness. Next, we adopt the fuzzy union reasoning to construct the fuzzy control rules and to obtain the fuzzy controller. The fuzzy control rule base of the controller is composed of a series single-input-single-output relations coupled by union. So that, while the additional inputs to the controller, the total number of rules would only increase linearly. Finally, the elitist genetic algorithm is applied to search the optimal fuzzy logic controller. The simulation results demonstrate that this algorithm is an effective systematic formulation to design fuzzy logic controllers. Moreover, the suggested method can control plants with fewer fuzzy rules and being able to efficiently reduce the number of fuzzy rules. It also has the property of robustness to certain parameter variations and external disturbance.