Fuzzy Identification Of Nonlinear Systems With Control

This thesis takes the complex nonlinear systems as object and mainly studies their identification and control methods.First, it offers an overview of the methods of identification and control for nonlinear systems and their current situation.Second, it introduces the main theories of fuzzy logic system, including the technology and approach for nonlinear system identification and their shortcomings. Based on these, the author proposes a new method of fuzzy modeling for nonlinear system called fuzzy multi-model identification method, in which, the conclusion parameters of each rule are identified separately by using fuzzy weight cursive least square algorithm. The simulative result demonstrates the effectiveness of this method.Third, it discusses the data topological homeomorphism transform control method of nonlinear systems, which has better application in practice, but its theory analysis is not enough. The author gives answers to some extent about whether the dynamic characteristics of the original system are changed after data transform and analyzes the convergence of affine nonlinear system under the simple control law.Finally, the author gives an example of the neutralization process of ammonium nitrate in chemical industry to prove the validity of the identification and control methods proposed in this thesis. A model is build by using the multi-model identification method, and an effective control result is achieved by using the data change method. The simulation result shows the better precision.