Modeling And Control For Ball&Plate System Based On Fuzzy Inference

Ball&Plate System is a nonlinear, multivariable and underactuated system, which has high theoretical and practical value. In this thesis, the Ball&Plate system is treated as experimental platform and the control objective is to make balls track reference signals. Three tasks are discussed in this thesis:the mathematical model of Ball&Plate System is built firstly; secondly, based on this fuzzy model and H2/H∞theory, the robust algorithm and optimum algorithm are designed to control the Ball&Plate system in real time. To verify the effectiveness of the control scheme, the Ball&Plate system is simulated by Matlab tools; thirdly, a Kalman filter is also added into the controller to enhance the anti-jamming capability.In the modeling section, based on mechanism modeling method, the model of Ball&Plate System is built and simplified, whose controllability is also proved. The system model is also discussed further by fuzzy modeling method, which is different from mechanism modeling method and system identification method.In the section of controller design, according to the optimal control theory and robust control theory, the H2and H∞controllers are derived. The H2control system has better performance and the H∞control system has better robustness. Considering the uncertain factors in the system, such as disturbances and measurement noises, a Kalman filter is utilized in the control system.Experimental results illustrate that the H2/H∞control method can obtain excellent control effect, and the Kalman filter can eliminate the impacts of uncertain factors on the system performance.