The Research On Fuzzy Sliding-mode Control Method

In the 1950s, Sovist researcher-Emelyanov first proposed the conception of variable structure. Furthermore, the theory of variable structure system(VSS) was developed by Utkin and Itkis etc. In the 1970s, the VSS got extensive attention of the western researchers for its distinct features. It has been deeply studied from the different aspects with many mathematic methods. Now the VSS already becomes a relative independence research branch. Variable structure control(VSC) is an effective robust control strategy. In the recent years, with the development of all kinds of control theory, such as robust control, adaptive control and fiizzy control etc, the theory of VSC has also obtained development rapidly.Based on the study status of VSS and the new requirements from the practice for the theory of VSC, some ways are put forward to resolve the problems existed in VSC, also the corresponding results are given. The main contents are as follows:1. A variable structure control method based on T-S fuzzy neural network (FNN) is brought forward. The linear combination of the error and its derivative are selected as switch surface, and then the tendency law method is used to design VSC control law. T-S FNN is imported, so that the actual output can trail the given value. The self-organize and self-regulate ability of T-S FNN is utilized to adjust some parameter of the subjection function online, so that the anticipant control purpose can be satisfied.2. A fuzzy integral VSC method based on integral VSC method is put forward. The characteristic of integral VSC method and the fuzzy control is integrated. Synchronously the uncertainties such as parameter change are estimated by affiliating gray model, control impact is improved by import compensate control.3. Some research on three-order inverted pendulum has been done based on the fuzzy integral VSC method above. The essence thought is to get the linear model near the work pointfrom the nonlinear model, look upon all the items ignored during linear course as uncertainties, the stabilization of the system is guaranteed by adjusting the coefficient of the sign function online according to the uncertainties.The summarization and prospect are given in the end of the paper.