Smart sliding mode control for magnetic levitation ball system

• Main Authors: Yong-Tang Jhou, 周永堂
• Other Authors: Huann-Keng Chiang
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/35076111535681711831

碩士 === 國立雲林科技大學 === 電機工程系碩士班 === 101 === Precision control and hardware cost are considered in the technological advances. The system precision control and lower cost is important in industry. Magnetic levitation ball system used coil current to generate electromagnetic force. Hence, the ball can suspend in the air. It can avoid the friction of the mechanical components. However, the magnetic levitation ball system is an unstable open loop system. The traditional control theory uses the operating point linearization method to design the controller. Hence, the system is unstable to work at the operating point which is sensitive to the parameter variations and external interferences. In this thesis, we establish mathematical model of the magnetic levitation ball system. We proposed three controllers in the magnetic levitation ball system. The first method is exponential reaching sliding mode controller. The second method is the sliding mode controller with immune fuzzy compensation which used fuzzy control to approximate antigen-antibody binding force nonlinear term in immune theory. The immune fuzzy controller used output feedback compensated when the system has parameter variations and external interferences. The third controller is an adaptive neural sliding mode controller with particle swarm optimization in offline. Particle swarm optimization searches center and width of Gaussian function. Moreover, we use Lyapunov stability to guarantee the magnetic levitation ball system asymptotically stable. Finally, the three controllers can reduce chattering phenomenon, and track reference command accurately in experimental results.