基于DBN-TOPSIS法的空中目标融合威胁评估

Threat assessment of air target fusion based on DBN-TOPSIS

针对已有空中目标威胁评估方法主观性强、稳定性弱、评估过程不连续的问题,综合考虑目标运动特性与目标属性,提出了一种基于动态贝叶斯(dynamic bayesian network,DBN)和逼近理想解法(TOPSIS)的融合威胁评估方法DBN-TOPSIS。通过分析来袭目标特征指标间的依赖关系,建立空中目标威胁评估指标体系。采用模糊理论处理连续型特征指标,统一指标形态,利用DBN进行动态威胁度等级概率推理。构造模糊DBN推理结果与TOPSIS评估矩阵之间的映射关系,采用TOPSIS法将威胁评估概率一维向量转换为确定数值,进行空中多目标威胁度准确排序。实验结果表明,融合威胁评估方法具有较好的合理性和稳定性。

Aiming at the problems that the current air target threat assessment has strong subjectivity,weak stability and discontinuous assessment process,this paper proposes a fusion threat assessment model DBN-TOPSIS based on dynamic Bayesian network(DBN)and TOPSIS by considering the target motion characteristics and the target attributes.An air target threat evaluation index system is established through the analysis of the dependent relationship between the indicators of the incoming targets.The fuzzy theory is used to process the continuous characteristic indexes and unify the shape of indexes,and DBN is applied to carry out probabilistic reasoning of the dynamic threat level.The mapping relationship between fuzzy DBN evaluation results and TOPSIS evaluation matrix is constructed.TOPSIS method is applied to convert the one-dimensional vector of threat assessment probability into a definite value.The accurate ranking of the air multi-target threat is obtained.The experimental results show that the fusion threat assessment method is reasonable and stable.