Artificial Control For Complex Nonlinear Systems

With the development of science and technology, the plants to be controlled are becoming more and more large and complex. As a result, it is difficult to get the accurate mathematic model. Even we can build mathematic model for complex systems, it may be too complicated to design a controller with conventional means which are based on accurate mathematic models. It is the artificial intelligent control based on knowledge that takes bright future for the above problems. Furthermore, the accurate mathematic models are not needed by artificial control. Fuzzy and neural network controllers are two kinds method in the artificial intelligent control. In this paper, we concentrate on the designing of intelligent controllers for complex nonlinear systems.Stability is one of the most important indexs for systems. It is difficult to analyze the stability of fuzzy systems which are nonlinear in essence. Now, Based on T-S fuzzy model, Lyapunov stability theory and linear uncertainty theory are often used to analyze the system's stability. In the following, There are some work.First, based on fuzzy dynamic state space model, the fuzzy controller and observer are designed for the nonlinear discrete systems. According to fuzzy membership functions for each rule, the system is divided into some subspaces. On every subspace, the system is equal to linear certain system with uncertainty. For the certainty, fuzzy state feedback controller with pole placing is designed. Correspondingly, for the uncertainty, the state feedback linear supervising controller is designed.Second, the fuzzy control is applied to Chaos systems. After building the local dynamic model for Chaos systems, the local controller and observer are designed. With the piecewise Lyapunov theory, the stability of the closed system is analyzed. The common positive P is not needed .So it is less conservative than the old methods and it bring about the new idea and methods for the Chaos systems' control.Third, based on the T-S fuzzy dynamic model, using the concept of CDF (compensation and division for fuzzy model) and the approach of linear matrixillinequality , a new kind of fuzzy controller and the stability analysis of the closed-loop system are done. It shows that the controller with concept of CDF is less conservative than that with PDC.In the end, Based on the normal BP algorithm, Chaotic mechanism is introduced to train the neural network model reference adaptive controller and identificationer. So the complex plants with uncertainty and alternation with time are controlled. The simulation shows the Choatic BP algorithm is more effective than the normal BP algorithm. Furthermore, by using identification and control online, the system is more robust and adaptive than ever.