Adaptive fuzzy predictive control for unknown multivariable nonlinear systems
未知多变量非线性系统自适应模糊预测控制

We present a direct adaptive fuzzy predictive control method for a class of unknown multivariable nonlinear systems. In this method, the plant is represented by a predictive model consisting of a linear time-varying submodel and a nonlinear submodel; the vector composed of fuzzy logic systems is used to design the predictive controller directly. The unknown vector in the controller and the unknown matrix in estimates of the generalized error are adjusted based on the time-varying function of dead-zone. It is proved that the proposed method can make the estimates of the generalized error vector converge to a neighborhood of the origin.