基于证据理论与DBN的炮兵远程火力毁伤评估

Fire Damage Assessment on Artillery Long-distance Firepower Based on Evidence Theory and Dynamic Bayesian Network

针对炮兵远程火力毁伤评估中获取信息的信度和效度等问题,将证据理论引入DBN(Dynamic Bayesian Network,动态贝叶斯网络)进行毁伤效果的评估。通过构建炮兵远程火力毁伤指标体系,利用模糊集合理论将毁伤指标进行离散划分,建立基于证据理论与DBN相结合的毁伤评估模型,并基于连接树(Junction Tree)算法实现对火力毁伤效果的动态评估。最后,通过仿真验证分析,得出了较为可靠的结论,为远程火力毁伤效果科学评估提供了有效的方法。

Aimming at on the problem of the reliability and certainty of the damage information in the assessment of artillery long-distance firepower damage,the evidence theory is introduced into DBN to estimate the effect of the damage. The index system of artillery long-distance firepower is built,and it is discretely divided through fuzzy-set theory. The model of the artillery long-distance firepower damage assessment which is based on the dynamic Bayesian network combined with the evidence theory is built to implement the dynamic assessment of firepower damage through the junction tree algorithm. And it is qualitatively analyzed,which promotes the scientificalness and accuracy of the long-distance firepower damage assessment.Finally,the model is validated by simulation,and the more reliable conclusion is drawn which provides effective ways to assess the long-distance firepower damage scientifically.