基于神经网络模型的自适应控制系统设计及仿真

基于神经网络模型的自适应控制系统设计及仿真

学生姓名： 班级： 指导老师：

摘要：伴随智能控制技术和远程监控技术的发展，自适应控制算法在工业、

日常生活、科学研究等领域得到广泛运用。由于人们对于智能控制的需求日益复杂多变，便出现了基于模仿神经元工作的神经网络学科。基于神经网络上的自适应控制算法不仅具备传统PID控制算法的自动调整参数功能，更具备复杂情况下参数调整，进而更符合现实需要。

本文探究了基于BP神经网络和RBF神经网络的自适应控制算法，根据仿真效果图可分析出其优势所在。BP神经网络是一种误差反向传播型神经网络，具有对外界刺激和输入信息进行联想记忆的能力；径向基函数(RBF) 神经网络是一种性能良好的前向网络，在simulink模型仿真中可知：RBF神经网络与BF神经网络相比，RBF神经网络的学习速度更快，在参数自适应调整方面具有具有很强的聚类分析能力和较快的收敛速度。

关键词：自适应控制算法 BF神经网络 RBF神经网络 simulink

指导老师签名：

Neural network model-based adaptive control

system design and simulation

Student name: Class: Supervisor:

Abstract :Accompanied by the intelligent control technology and remote

monitoring technology development, and adaptive control algorithms have been widely used in industry, daily life, such as scientific research fields. Increasingly complex due to the demand for intelligent control, giving rise to a neural network based on imitation of neurons disciplines. Adaptive control algorithms based on neural network not only with traditional PID control algorithm automatically adjusts the parameters, parameter adjustment, with complex situations and thus more realistic.

This paper explores the adaptive control algorithm based on BP neural network and RBF neural network, simulation effect diagram to analyze its advantages. BP neural network is an error back propagation neural network, capable of associative memory to external stimuli and input information; radial basis function (RBF) neural network is a good performance to the network in simulink model simulation shows: RBF neural network learning faster, cluster analysis has a strong capacity and faster ,speed of convergence in the parameters of the adaptive RBF neural network is compared with the BF neural network.

Keyword: Adaptive control algorithm BF neural network RBF neural network simulink

Signature of Supervisor: