Reinforcement Learning Neural Network Control for T-S Fuzzy Nonlinear System

摘要：

In this paper, a new control scheme based on fuzzy model and reinforcement learning neural network is presented for a class of nonlinear system with unknown uncertain nonlinearities. Firstly, the T-S fuzzy model is adopted for modelling of the known nonlinear system approximately; fuzzy state feedback control law is designed to track the desired output signal. Secondly, a reinforcement learning neural network control is used to improve the scheme of the fuzzy state feedback control. The effect of the unknown uncertainties and the error caused by fuzzy modelling is overcome by adaptive tuning of the weights of the MLP neural network on line. Finally, the proposed scheme is applied to a fault fighter control system and the simulation results show it is excellent and effective.

关键词： Reinforcement Learning Neural Network T-S fuzzy model State Feedback Nonlinear System

作者: XIAO Di ZHANG Guang-ming HU Shou-song

作者单位: College of Automation,Nanjing University of Technology,Nanjing 210009,P.R.China College of Automation,Nanjing University of Aeronautics and Astronautics,Nanjing,210016,P.R.China

会议类型: 国际会议

会议名称: 2008高等智能国际会议(2008 International Conference on Advanced Intelligence)

会议地点: 北京

会议语种:英文

在线出版日期: 2008-10-18（万方平台首次上网日期，不代表论文的发表时间）