基于BP神经化贝叶斯网络的空袭目标毁伤效果评估

针对空袭目标毁伤因素的复杂多样性,以及贝叶斯网络在评估时对先验知识获取和对网络节点限制方面存在的不足,提出一种BP神经化贝叶斯网络的评估方法,解决空袭目标的毁伤问题。分析了空袭目标毁伤的影响因素,并进行了知识表示;分析了贝叶斯网络结构处理多个父节点问题时存在的几个缺陷,建立了空袭目标毁伤评估的BP神经化网络结构,给出了"BP神经化模块"的决策方法。通过实例分析,结合Netica仿真,验证了该评估方法的准确性,操作简单、有效性好。

Battle Damage Assessment of Air-raid Based on BP Neuralization Bayesian Network

Aiming at the complex and diversity of battle damage factors of air-raid,and the existing shortage of a priori knowledge acquisition and network node restrictions when Bayesian network to assess,the assessment method based on the BP neutralization Bayesian network is brought out,which to solve the problem of the battle damage of air-raid. The influence factors of the battle damage of air-raid are analyzed,and the knowledge is showed,Analyzing the structure of bayesian network existing several defects when is dealing with multiple parent nodes,the network structure of battle damage assessment of air-raid based on BP neutralization Bayesian network is built. The method of decision-making of “BP neutralization module”is given out. The case through to analysis,and combining with the simulation with Netica,the accuracy of the assessment method is verified. There is a simple operation and good effectiveness.