Design the T-S Fuzzy Controller for Magnetic Levitation System via LMIs

• Main Authors: Cheng-Xi Pan, 潘承熙
• Other Authors: Ying-De Zhang
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/91490675592175735326

碩士 === 國立海洋大學 === 電機工程學系 === 91 === This thesis relies on energy viewpoint (Lyapunov Asymptotic Stability Theorem) to design Takagi-Sugeno (T-S) fuzzy controller for large-scale and complex control Systems via Linear Matrix Inequalities (LMIs). The main feature is approximately model in non-linear system by fuzzy rule base, where the consequent parts become type of linear function. With regard to the construct between controller and plant by using the concept of parallel distributed compensation (PDC) to carry out these designs. Final transforms stability analysis question into a system of linear matrix inequalities (LMIs). Moreover, through the use of Matlab engineering application software to solve feedback vector F. The main design purpose is to decrease the number of fuzzy rules compare to traditional fuzzy controller (example: Mamdani Fuzzy Controller). The fuzzy rules depend on whether their associated membership functions are triggered parallel and interactive. The way it in increases follow the arithmetic progression less than traditional fuzzy controller increasing to exponentially. Magnetic levitation (Maglev) system uses non-mechanical contact electromagnetic force to resist gravity so that the body be balanced and suspended on the air, it is one of the typical important mechatronic systems to be studied. In this thesis, mathematic model of magnetic levitation system is analysis and discussed first. Secondly, we design a Mamdani fuzzy controller and a T-S fuzzy controller for this system and compare their resulting response by computer simulations. The simulation result shows that the T-S fuzzy controller can work with less fuzzy rules. Keywords：Linear Matrix Inequalities、T-S fuzzy controller、Parallel Distributed Compensation、Magnetic Levitation。