Sliding-mode-aided fuzzy control system design

Includes bibliographical references (pages [73]-75)Instead of using the trial-and-error approach, this thesis proposes a systematic procedure of a fuzzy controller design. In this procedure, a symmetric decision table is decided by the sliding mode approach first Then, two scaling factors, Kj and K2, are adjusted with the aid of the phase plane portraits. Since the problem which always comes with the sliding mode control is the control discontinuity, or the chattering, a more precise symmetric decision table is used in this paper to smooth out the chattering in a thin boundary layer neighboring the switching surface. The simulation results indicate that the suggested design procedure improves the stability of a system with nonlinearity or with disturbance. In digital control systems, a proper sampling time is essential to the desired time response. A sampling time too small or a sampling time too big will affect the system seriously. A proper range of the sampling time for a fuzzy control system can be chosen. In the sliding-mode-aided fuzzy control system design, a proper sampling time is deeply related to the scaling factor, K2. In addition to introducing the relation between the sampling time and K2, this paper offers a method to choose the values of the sampling time and K2 which is easily implemented and also makes the system stable.M.S. (Master of Science