Nonlinear multi-model second level generalized predictive control based on switching
A nonlinear multiple models second-level generalized predictive controller design method based on an er-ror switching strategy is proposed for a class of nonlinear discrete-time systems with zero dynamic instability caused by parameter jumps.Initially,the space of unknown parameters is divided into multiple subsets,and multiple nonlinear pre-diction models are established in each subset,with the identification of unknown parameters.Subsequently,a second-level adaptive method with constraints is utilized to obtain the parameter estimates of each subset of the virtual model,and the corresponding generalized predictive control effect is calculated to address the zero dynamic instability issue more effec-tively.To mitigate the impact of parameter jumps on the system,the model output error performance index is employed to select the optimal generalized predictive controller for controlling the nonlinear system at each moment,followed by stability analysis.Ultimately,simulation results,when compared with existing methods,demonstrate that the proposed generalized predictive controller exhibits excellent tracking performance and anti-interference capabilities.
nonlinear systemmultiple modelssecond level adaptivegeneralized predictive control