Fuzzy Neural Network Based Controller Design And Analysis For The Stochastic Nonlinear System

The practical engineering systems always are nonlinear systems.There inevitably exist a lot of uncertainties and stochastic disturbance in systems,which will have bad influence on stability and performance of the systems.Therefore,the theoretical research of the control problems for stochastic nonlinear uncertain systems is of great theoretical significance.In the meanwhile,as far as flight control is concerned,how to overcome these negative effects and assure the flight safety has practical application value.This thesis focuses on the analysis and design of stabilization controllers for stochastic nonlinear uncertain systems,and applies these theoretical results into the stability control of flight systems.The main research contents are as follows:The current studies of stochastic nonlinear control problem at home and abroad are introduced,and the current research focuses and problems to be resolved are analyzed;Under the framework of stochastic Lyapunov stability theory,combining backstepping technique and FNN(fuzzy neural network)method,the adaptive controller is designed for a class of pure feedback stochastic nonlinear uncertain systems,which renders the responses of the closed-loop systems are bounded in probability.The number of adjustable parameters can be reduced effectively;In order to overcome the shortcoming of backstepping,the low-pass first order filter is applied to avoid “parameter explosion”.By using of the dynamic surface control method,the SGUUB(semi-globally uniformly ultimately bounded)of the closed-loop systems is obtained.Moreover,the filter time can be adjusted in order to decrease the filter error and accelerate the speed of convergence;Meanwhile,the theoretical method is applied to hypersonic flight vehicle longitudinal model with stochastic perturbance.An adaptive fuzzy neural network dynamic surface controller is designed,to make the signals of the closed-loop system semi-globally uniformly ultimately bounded.And the simulation verifies the effectiveness of the proposed method;Finally,the summary and prospect are proposed in the thesis.