基于小波神经网络的自适应逆控制及其应用  

Adaptive Inverse Control Based on Wavelet Neural Networks and Its Applicationin a Flight Simulator

作　　者：刘金琨[1] 刘涛[1] 

机构地区：[1]北京航空航天大学自动控制系,北京100083

出　　处：《系统工程与电子技术》2003年第5期591-594,共4页Systems Engineering and Electronics

基　　金：航空基础科学基金资助课题(00E51022)

摘　　要：神经网络控制特别适用于具有非线性和不确定性因素的系统。采用小波神经网络(WNN)对飞行仿真转台的直流伺服系统进行实时辨识,得到其逆模型。然后将这一训练后的网络作为前馈控制器与常规反馈控制器结合构成并行自适应逆控制器,控制转台跟踪指定的速度和位置轨线。仿真结果表明该方法的有效性。Neural network control is well-suited for the system with nonlinearity and uncertainty. In this paper, wavelet neural network(WNN) is used to identify the inverse model of the flight simulator. After the training, an adaptive inverse controller which is a combination of a neural network feedforward controller and a traditional feedback controller is formed to drive the simulator to track the preselected trajectories in speed and position. Simulation results verify the effectiveness of the proposed method.

关 键 词：小波分析 小波神经网络 自适应逆控制 飞行仿真转台 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]