基于MATLAB的几类神经网络稳定性分析仿真系统设计通信工程专业论文.docxVIP

Simulation System Design on Stability Analysis for Several Classes of Neural Networks Based on MATLAB Abstract Artificial neural network is a booming emerging discipline in thefield of high science and technology research. Its theory got rapid development in recent years, and it has become the theory of an international frontier.Currently, it has obtained many concrete results in the fieldsof pattern recognition, automatic control, auxiliary decision-making, image processing, artificial intelligence, and many other researches.This paper studied the structure of the neural network model, characteristic function, the basic theoretical knowledge and so on. In addition, compared and analyzed the theoretical achievement of neural network’s qualitative analysis. Provide the basic theoretical foundation for design the simulation system. MATLAB as an essential tool for engaging in intelligent control theory also plays an important role in the design, in this paper, using the method of it, especially the GUI (graphical user interface) design method, M file writing theory and so on, has carried on the detailed research. Stability analysis of LMI, because many of the neural network theory is expressed by LMI, so when writing a simulation program,the analysis of network stability eventually translate into the study of linear matrix inequality (LMI). In this paper used the theory basis of LMI, solving method, correlation function, whichis the basis of writing a simulation program and the basis. On the basic of analyzing of theory of stability analysis, selecting two stability analysis theories more mature which are namely the stability of Hopfield neural network analysis theory and the analysis of the stability of Cellular neural network.With the help of MATLAB software design simulation system’s operator interface, writing the simulation program and function call program.The simulation program links to the operating interface is generated in the simulation system in order to i