多传感器协同探测证据理论分类融合方法

随着智能化、网络化集群作战等理念和技术的兴起，精确制导武器越来越向智能化、协同化方向发展。多传感器协同探测能够针对不同的探测任务背景和作战需求，提升目标探测性能，还可以跨域整合多种探测平台。但是由于信息的不确定性等特点，使得多传感器数据直接融合可能造成决策困难。因此，在证据理论体系下对信息融合的有效性进行合理分析与度量是很有必要的。提出了一种基于Deng熵的证据理论分类融合算法，以熵减为主要思想，将证据进行分类融合。在决策过程中，将含有证据数最多的类别融合结果作为总体融合结果，避免高冲突证据的影响，提升融合结果的信息有效性。采用算例说明了所提方法不仅能够得到合理正确的结果，并且融合可靠性较高，便于决策与后续的信息处理。

Evidence Theory Classification Information Fusion Method for Multi-sensor Collaborative Detection

With the rise of concepts and technologies such as intelligent and networked cluster operations, precision guided weapons are increasingly developing in the direction of intelligence and synergy. Multi-sensor collaborative detection is able to achieve higher detection performance for different detection mission backgrounds and operational requirements, and can integrate multiple detection platforms across domains. However, due to the uncertainty of information and other characteristics, the direct fusion of multi-sensor data may cause decision-making difficulties. Therefore, it is necessary to analyze and measure the effectiveness of information fusion under the evidence theory system. In this paper, a classification theory of evidence theory based on Deng entropy is proposed. The entropy is reduced to the main idea, and the evidence is classified and merged. In the decision-making process, the result of the category fusion with the largest number of evidences is taken as the overall fusion result, avoiding the influence of high conflict evidence and improving the information validity of the fusion result. The example is used to illustrate that the proposed method can not only obtain reasonable and correct results, but also have high fusion reliability, which is convenient for decision-making and subsequent information processing.