基于分层的SVDDBN的空间作战决策研究

Decision-Making Method for Space-Based Combat Based on Hierarchical Structure Varied Discrete Dynamic Bayesian Network

为了实现空间作战平台在不确定知识基础上的快速自主智能攻击决策,以天基反卫星作战为背景,提出利用分层的变结构离散动态贝叶斯网络(SVDDBN)来建立攻击决策的图形模型,给出了该模型的推理算法公式和用于攻击决策的变结构网络参数的自适应生成算法。在此基础上设定初始条件进行模型算法的仿真验证,试验结果证明了本文提出的算法的可行性,它可以使空间作战平台利用打击链前端传来的经过模糊处理数据进行快速的推理决策,从而使空间作战平台具有有效的自主化决策能力。

A Decision-making graphical model for Space-Based Combat is presented based on Hierarchical Structure Varied Discrete Dynamic Bayesian Network（SVDDBN） to achieve fast unmanned tactical decision-making for Space Combat Vechicle（SCV） with uncertain information.A Calculational Inference Algorithm for SVDDBN is used to ratiocinate conclusion of battle field situation assessment and decision-making of Space-Based Combat,and an adaptive parameter generation algorithm is proposed to generate Bayesian network parameters when its structure varied.The simulation results show that our method is efficacious.Fuzzy processed data which given by sensor can be used by our method to make quick inferential decision,so that it enhances the capablity of self decision-making of SCV effectively.