Research On Fuzzy Internal Model Control Algorithm Of Nonlinear System Based On C-R Fuzzy Model

Most systems are nonlinear in industrial process, whose characteristic are complicated, such as nonlinear, time varied and multi-mode. So that the application of the conventional linearity theory is limited, and the model and control of nonlinear system is one of main research in control field. For nonlinear modeling, the approximation ability of Fuzzy system modeling is favorable. For nonlinear control, IMC is so simple, effective and feasible that IMC is paid more attention to. The thesis combined Fuzzy method and IMC to study model and control of nonlinear system based on C-R Fuzzy model. The main study is as followed.(1) Firstly, C-R fuzzy model structure of the complex nonlinear system is researched. The C-R fuzzy model structure includes fuzzy inference rules and local linear models which is liable to show the system knowledge. It is a effective modeling method. In the model process, the anti-fuzzy modeling is needed that simplified the model process.(2) Secondly, the identification methods of C-R fuzzy model are studied. The local models parameters and membership function are obtained by the identification methods. Because of the blindness of choosing the initial factor in identification algorithm 1, the author changes the initial membership and the condition of final judgment, the identification algorithm 2 is studied. For the shortcomings of the assumption of the inference values in the identification algorithm 1 and 2, the identification algorithm about determining the optimum inference values is researched. According to the problems of calculating the volume, a dynamic identification method based on the clustering method by relational grades is proposed. The number of judgment is reduced and the computing time is saved by the identification algorithm improved.(3) Thirdly, considering the difficulty of get accurate model and inverse model of object on controlling the nonlinear system with nonlinear IMC. Using the C-R fuzzy model to construct the nonlinear system. And using the advantage of the local linear analytical to construct the fuzzy inverse model by the desired output. On that basis, Nonlinear Fuzzy Internal Model Control Algorithm Based on C-R Fuzzy Model is proposed: CR-FIMC.(4) forthly, on the basis of CR-FIMC, we change the C-R fuzzy model to a C-R fuzzy step model, whose structure is similar to step response model, then using the C-R fuzzy step model to construct the nonlinear system. On that basis, the model and the inverse model are introduced into the FIMC and PID. Further, the CR-PID controller was added to the CR-FIMC structure in order to speed up tracking of error convergence and improve CR-IMC controller performance, So a dual controller with a fuzzy internal model control and PID in conjunction based C-R fuzzy step model is proposed: CR-FIMC-PID.