自动驾驶仪神经网络控制的HLA仿真研究  

HLA Simulation Research of Autopilot Based on Neural Network

作　　者：马建伟[1,2] 曾愿晓[1] 

机构地区：[1]河南科技大学电子信息工程学院,河南洛阳471003 [2]中国空空导弹研究院,河南洛阳471009

出　　处：《弹箭与制导学报》2010年第1期27-29,36,共4页Journal of Projectiles,Rockets,Missiles and Guidance

基　　金：总装备部预先研究基金资助

摘　　要：在对空空导弹自动驾驶仪各功能模块需求和能力分析的基础上,以组件的重用和可扩展为目标,运用HLA技术划分各联邦成员,成功组建神经网络自动驾驶仪分布式仿真系统。设计过程中的关键技术如对FOM/SOM设计和Simulink模型代码转换分别做了研究,给出了基于Matlab/Simulink所建立的空空导弹自动驾驶仪神经网络控制算法与HLA集成的一个工程样例。Based on the analysis of autopilot of air-to-air missile function modules＇ demand and ability, aiming at improving components reusability and system extensibility , using HLA distributed technology to carry function partition . Distributed simulation system about autopilot based on neural network was established successfully. The key techniques during designing this system, such as FOM/SOM design and code conversion of Simulink models were studied and presented a sample of autopilot of air-to air-missile control system from Matlab/Simulink to HLA federation. Simulation result shows that the simulation system meets user＇s needs perfectly,, and each component can be customized due to individual needs.

关 键 词：空空导弹 自动驾驶仪 高层体系结构 运行支撑框架 联邦对象模型 仿真对象模型 

分 类 号：TJ765[兵器科学与技术—武器系统与运用工程]