基于DNN的空空导弹不可逃逸区生成算法研究  

The algorithm research of air-to-air missile non-escape zone generation based on DNN

作　　者：张绍良[1] 南英[1] ZHANG Shaoliang;NAN Ying(College of Astronautics,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

机构地区：[1]南京航空航天大学航天学院,江苏南京210016

出　　处：《飞行力学》2021年第3期68-74,共7页Flight Dynamics

基　　金：国防科技重点实验室基金资助(61425010102)。

摘　　要：针对目标采取最优逃逸策略下的空空导弹发射时机决策问题,提出了导弹不可逃逸区模型,并采用半直接配点法计算不可逃逸区边界。然后,设计了一种深度神经网络(DNN)模型,对不可逃逸区边界样本数据进行训练,得到不可逃逸区DNN生成模型。仿真结果表明,不可逃逸区DNN生成模型可以实时高精度生成不可逃逸区边界,与样本值的平均误差率小于0.9%,满足战场环境中实时性和准确性要求,为最优对抗条件下空空导弹发射时机决策提供了技术支持。With the target adopted the optimal escape strategy,the air-to-air missile non-escape zone model was proposed for solving launch timing decision problem of air-to-air missile.The semi-direct collocation method was used to solve the boundary of the non-escape zone.Then,the deep neural networks(DNN)was designed to train the sample data of the non-escape zone boundary,and the corresponding DNN generation model had been obtained.The simulation results show that the DNN model can generate the non-escape zone boundary in real time with high accuracy,and the mean error rate is less than 0.9%,compared with the sample value,which can satisfy the real-time and accuracy requirements in the battlefield environment.This paper provides the technical support for the launch timing decision of air-to-air missile in the optimal confrontation condition.

关 键 词：空空导弹 不可逃逸区 半直接配点法 深度神经网络 

分 类 号：TJ762.23[兵器科学与技术—武器系统与运用工程]