多变量非线性飞行控制系统的神经网络动态逆控制方法  

Control Method of Dynamic Inversion with Neural Network Used for Multi-Variable Nonlinear Flight Control System

作　　者：钱克昌[1] 陈自力[1] 李建[1] 

机构地区：[1]军械工程学院光学与电子工程系,河北石家庄050003

出　　处：《计算机系统应用》2010年第11期56-59,共4页Computer Systems & Applications

基　　金：军队十一五装备预研基金(9140A25050106JB3412)

摘　　要：针对多变量非线性飞行控制系统,从理论上对其逆系统的解析形式进行了详细推导,根据神经网络逼近逆系统的原理分析,提出了一种由静态神经网络和积分器组成的动态神经网络,构造了多变量非线性飞行控制系统的神经网络动态逆控制系统,并利用动力伞飞行控制系统进行了仿真验证,结果表明这种控制方法完全满足控制要求,具有较好的抗干扰能力和鲁棒性能。Aiming at multi-variable nonlinear flight control systemt,his paper analyzes the analytical inverse system in theory detailedly.By the tracking principle of neural network,a dynamic neural network is presented,which consist of static neural network and integrators.The paper designs a control method of dynamic inversion with neural network used for multi-variable nonlinear flight control system,and makes simulation of flight control system of power parafoil.Simulation results demonstrate that the control method has strong ability of control and robustness.

关 键 词：飞行控制系统 动态逆 神经网络 多变量 

分 类 号：TP29-AD[自动化与计算机技术—检测技术与自动化装置]