基于神经网络的BTT导弹逆控制方案设计  

Design of Inverse Control of BTT Missile System Based on RBF Neural Network

作　　者：王强 王晓恩[1] 杨志峰[2] 

机构地区：[1]中国人民解放军空军驻二院军事代表室,北京100854 [2]空军工程大学导弹学院,陕西三原713800

出　　处：《上海航天》2010年第2期29-33,共5页Aerospace Shanghai

基　　金：国防科技重点实验室基金(51431040203JB3201)

摘　　要：对基于神经网络的倾斜转弯(BTT)导弹逆控制进行了研究。用径向基函数(RBF)神经网络结构和最近邻聚类算法,对导弹系统逆动力学系统进行动态模型辨识,以辨识模型为控制器与BTT导弹控制系统串联构成动态伪线性系统;用逆系统法设计了一种用于BTT导弹非线性控制的经典控制与神经网络在线自学习综合控制方案,实现了导弹三通道的线性化控制和输出的渐近无差跟踪。仿真结果表明:该方案可根据设计指标要求实现对BTT导弹的非线性控制,且有较强的鲁棒性。The inverse control of BBT missile based on neural network was studied in this paper. A RBF neural network applying nearest neighbor clustering algorithm was used to realize the identification of the inverse dynamic system model of missile. And the system which was made by the identification model and the BTT missile control system in series was to make adynamic pseudolinear. The on-line self-learning control strategy which combined inverse control based on classical control with neural network was proposed for the nonlinear control of the BBT missile, in which the linearization control of the three channels of the missile and the output＇s indistinctive tracting gradually were realized. The simulation results showed that the design of BTT missile control system would not only realize nonlinear control, but also possess excellent robustness and meet the designing requirement.

关 键 词：BTT导弹 RBF神经网络 逆控制 在线自学习 最近邻聚类算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]