The Ts Model, Its Applied Research, Including Mode Control Method

Because almost every industrial process exhibits significant nonlinear characteristic, the application of the conventional linear theory is limited. So the modeling and control of nonlinear system becomes one of the main researches in the control field. For nonlinear system, fuzzy modeling is one of the promising methods for describing nonlinear behavior. The T-S model consists of fuzzy reasoning and local linear model group. It is suitable to express structural knowledge, and does not require defuzzification processing. Moreover, the T-S model can approximate a nonlinear system with an arbitrary accuracy. Thus, the T-S modeling has attracted much attention. In addition to, considering the nonlinear system control, internal model control (IMC) approach has gained high popularity due to its outstanding features such as simplicity, good tracking performance and robustness. So, the control scheme based on T-S model is studied in this paper, which combined fuzzy system modeling method and IMC approach.Firstly, the structure of T-S model for nonlinear system was studied. The T-S model is identified by fuzzy C-means (FCM) algorithm and least squares. Compared with the system model based on hybrid pi-sigma neural network, it has high modeling precision.Secondly, the identification method based on adaptive genetic algorithm (GA) for T-S model was studied. Adaptive GA is adapted to simultaneously optimize both premise parameters and conclusion parameters in the T-S model. So the disadvantages of easily falling into local minimal by using the FCM algorithm and the least squares method for the parameter identification could be overcome, and the identification accuracy was improved.Thirdly, aiming at the shortcoming of slower convergence rate in the T-S model identification method based on adaptive GA, a T-S model identification method based on improved particle swarm optimization (PSO) algorithm was studied. The premise parameters and conclusion parameters could be simultaneously optimized by using the improved PSO algorithm, the convergence rate could be effectively speeded up, and the modeling accuracy was improved.Finally, according to the IMC principle, an IMC method based on the T-S model for a class of nonlinear systems was given. The T-S model and it's inversion of the nonlinear system were constructed, and the difficulties of getting the accurate model and it's inversion for the IMC of the nonlinear system were overcome. Then, the internal model controller was designed. The simulation results show that the proposed method could provide a better dynamic performance of both the command tracking and disturbance rejection and robustness against parameters perturbation.