切换随机非线性系统的神经网络控制分析-neural network control analysis of switched stochastic nonlinear systems.docxVIP

AbstractArtificial neural network (ANN) is a information processing network system based on the development of a simulation of the human brain mechanisms in modern neurobiology research, which not only has to deal with numerical data on the general computing power, but also dealing with knowledge of thinking, learning and memory. Artificial neural networks will play a more role in the field of science and technology as well as peoples understanding of the mechanism of brain information processing and the deepening of artificial neural network level intelligence, one of them is the application of neural network control for hybrid systems.As we know, the practical systems are very complex. May be, they have many complex characteristics, such as nonlinear, multi-models, strong coupling, infinite dimension, randomicity, uncertainty, and so on. The dynamic random switching system in this paper is a special kind of hybrid system is described by the family of subsystems and their associated components to switch strategy. Each sub-system with a random item is used to simulate the real life phenomenon of a number of randomly generated, such as noise and so on. Precisely because of this uncertainty, making the system has increased the difficulty of theoretical analysis, however, with the neural network technology continues to mature, in respect of a class of hybrid systems more complicated, people have begun to study progress, and have begun to apply the practical areas, such as radar and image signal processing, including data compression, feature extraction, noise suppressionand signal recognition, etc. The aim of this paper is an attempt toIto differential methods,neural network control and Back-stepping control of the area into a common method of study of this topic to find the subject of a number of key issues in the solution.The principal research work and novelties are listed as follows:A neural network sliding mode control method for a class of switched nonlinear sys