基于神经网络及证据理论的CGF搜潜策略改进

Neural Network and Evidence Theory Based Improvements on Submarine Search Strategy of CGF

针对目前多计算机生成兵力(Computer Generated Force,CGF)协同反潜中无法充分利用战场信息实现CGF搜潜策略优化的问题,在CGF间可有效实现信息交互的前提下,将信息融合技术引入到反潜作战仿真中;利用BP(Back Propagation)人工神经网络获取证据信息的基本置信分配,通过改进的D-S(Dempster-Shafer)证据理论对反潜CGF获取的信息进行综合,以此改进反潜CGF的移动策略。仿真实验表明,在反潜CGF搜潜过程中引入人工神经网络和信息融合技术,可有效提高CGF搜潜的成功率。

Since the Computer Generated Force （CGF） can not take full advantage of battlefield information to optimize CGF submarine search strategy in the cooperative anti-submarine con^bat, the information fusion technology was introduced into the anti-submarine warfare simulation based on the effective interaction of the information between CGF. The basic confidence assignment of the evidence information was obtained using Back Propagation （BP） artificial neural network, and the information from antisubmarine CGF was synthesized by using the improved DS evidence theory. And thus the anti-submarine CGF mobile strategy was improved. Simulation results show that the introduction of artificial neural network and information fusion technology in antisubmarine CGF submarine search process can improve the anti-submarine efficiency.