Stabilizing Controller Design Using Fuzzy T-S Model on Two Wheeled Self-Balancing Vehicle

• Main Authors: Cao, Jia-Rui, 曹家瑞
• Other Authors: Huang, Chih-Peng
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/75376825038740777417

碩士 === 臺北市立大學 === 資訊科學系 === 104 === In this research, we proposed a controller design method of the Two Wheeled Self-Balancing Vehicle (TWSBV) based on Fuzzy T-S Model (T-S Fuzzy) associated with genetic algorithm (GA). To achieve the stable controller of TWSBV, we used GA to properly seek for the feasible state feedback gains for the T-S Fuzzy controller, which is constructed by the heuristic experiment with two fuzzy membership functions of vehicle body angle and vehicle angular velocity to conquer some nonlinear parameters. With the convergent state feedback gains via GA, we can ensure the TWSBV has the better performances. Through analyzing the TWSBV impulse response characteristics produced from the state-space equation of TWSBV dynamic model, the system’s performance of the controllers can be evaluated. The fitness function of genetic algorithm is formulated by some performance indexes so that we can search for better state feedback gains. Furthermore, using GA’s ability of natural selection can reduce the tuning time of the satisfying state feedback gains. By the well-tuned state feedback gains, we thus can achieve a fuzzy model controller of the TWSBV system. The experimental simulation demonstrates that the proposed method has less balancing time and states’ oscillation, which mean that it indeed has better performance than others.