基于RS-DBN的电子对抗目标清单生成方法

由于电子对抗作战目标类型和工作方式多样, 变化速度快, 有效信息难以充分获得, 且在不同作战阶段呈现出不同特点, 使用传统评估方法难以对其等级排序实施精确评估。对此, 提出一种基于随机集的动态贝叶斯网络电子对抗目标等级评估方法。首先，对电子对抗作战目标清单生成方式进行梳理, 确定了评价指标体系, 并根据作战阶段的变化特点，结合动态贝叶斯网络完善了评价体系。然后, 充分考虑作战过程中数据获取不完整的特点, 通过引入随机集方法将传统贝叶斯网络的节点参数求解方法进行拓展, 使用区间数学的思想得到了较为准确的动态贝叶斯网络节点参数。最后，进行了案例仿真计算和结果分析, 并对节点概率确定方法进行算法复杂度讨论。结果表明，所提方法更加适合样本不完整的军事评估问题, 评估结果与实际作战基本一致, 使用的算法具有高效性、适用性和推广性。

Method of electronic countermeasure targets' list generation based on RS-DBN

Because the types and working methods of electronic countermeasures are diverse and change rapidly, effective information is difficult to obtain completely, and different operational characteristics are presented in different operational stages, so it is difficult to accurately evaluate their ranking by using traditional evaluation methods. Therefore, a target level evaluation method of electronic countermeasures based on random set dynamic Bayesian network is proposed. Firstly, the generation method of electronic countermeasures combat target list is combed, the evaluation index system is determined, and the evaluation system is improved based on the changing characteristics of the combat stage combined with the dynamic Bayesian network. Then, after fully consider the characteristics of incomplete data acquisition in combat, the traditional method of solving node parameters of Bayesian network is extended through the introduction of the random set method, and the more accurate node parameters of dynamic Bayesian network are obtained by using the idea of interval mathematics. Finally, the simulation calculation and result analysis of an example are carried out, and the complexity of the algorithm is discussed by the method of determining the node probability. The results show that the method proposed in this paper is more suitable for military evaluation with incomplete samples, and the evaluation results are basically consistent with actual combat situation. The algorithm used are efficient, applicable and popular.