摘要
针对无人机空战中单机对抗模式态势评估既要综合考虑连续型和离散型变量的影响,又具备不确定性的推理能力的特点,建立了一种基于混合动态贝叶斯网络的无人机空战态势评估模型。分析空战中敌我飞行数据,提取空战中战斗力要素,依据因果关系建立混合动态贝叶斯网络模型的网络拓扑结构。利用主观经验和历史数据,设计了先验概率、条件概率和状态转移概率等离散型网络参数及正态分布式连续型网络参数。采用链式算法和滤波算法进行网络推理。经仿真实验验证,证明了混合动态贝叶斯无人机空战态势评估模型具备快速性、有效性和容错性。
The situation assessment for the one-to-one mode of unmanned aerial vehicles air combat needs to consider the influence of continuous and discrete variables,and the uncertain reasoning as well. According to the above analysis,the situation assessment for unmanned aerial vehicles air combat based on Hybrid Dynamic Bayesian Network has been established.The flight data in an air combat has been analyzed. Some capability elements in an air combat have been extracted. According to the causal relationship,the topological structure of the Hybrid Dynamic Bayesian Network model has been established. By the subjective experience and the historical data,discrete network parameters and continuous network parameters have been designed. Discrete network parameters include prior probabilities,conditional probabilities and state transition probabilities.Continuous network parameters are some normal distributions. The chain algorithm and the filter algorithm are used for network inference. The simulation experiments showed that the situation assessment model for unmanned aerial vehicles air combat based on Hybrid Dynamic Bayesian Network has the advantages of rapidness,effectiveness and fault-tolerance.
出处
《指挥控制与仿真》
2017年第4期1-6,39,共7页
Command Control & Simulation
关键词
无人机
态势评估
混合动态贝叶斯网络
unmanned aerial vehicles
situation assessment
Hybrid Dynamic Bayesian Network
