Pan-cmac Neural Network Than The Performance Of Its Ivsc Control

Neural network (NN) is a new technology developed at the end of 20th century. A novel thread is provided for modeling and control of uncertain nonlinear systems due to the favorable nonlinear mapping capability and self-adaptive learning capability of neural networks. NN used in control system aims at nonlinearity,uncertainty and complexity of systems. As for high precision control of uncertain nonlinear systems based on NN, the generalization capability of NN is so important that the control precision is affected directly by it . The main content in this paper include generalization performance research of CMAC NN and the application in Integral Variable Structure Control (IVSC) for uncertain nonlinear systems. The first chapter is introduction, where are the main research content of this paper and the significance. In the second chapter, a all-round introduction is presented firstly, which includes the history,principle and application of neural network. Secondly, research on generalization performance of NN is stated. Finally the application in control field and the development situation of NN is introduced. In the third chapter, the structure and principle of CMAC NN are narrated in detail. The learning and generalization capability as well as some improvement of CMAC is talked. When the performance of a single CMAC is difficult to advance a little more, valid information learned by every network can be utilized synthetically by combining a number of trained CMAC neural networks with combination-weights. Because more information is obtained, the combination network usually has better generalization capability than the best one. Multiple-CMAC Combining Method is talked at length in this chapter, which is a contribution of this paper that generalization capability one CMAC can not reach can be arrived at by combining several networks. The fourth chapter gives the statement of the application of CMAC NN to Integral Variable Structure Control (IVSC). Variable Structure Control is a control method in common use for uncertain nonlinear systems. The border dithering and selection of sliding surface are the main problems confronted. Usually Variable Structure Control is combined with some other control methods to make full use of...