D-S规则推广及其在飞机目标识别中的应用研究

A study on the extended D-S rule and its application in aircraft target recognition

在多传感器特征信息融合算法中,D-S理论得到了广泛应用,但该理论在处理多特征冲突问题时识别效果不十分理想。针对这种情况,本文推广了D-S规则,得出了一种处理冲突信息的新方法。在分析推广规则受融合顺序影响的基础上,提出了最佳融合顺序,较好地利用了冲突信息,并通过仿真比较了D-S规则和推广规则应用于飞机目标识别的效果。结果表明:本文提出的D-S理论推广规则能有效解决多特征冲突的飞机目标识别难题。

For data fusion algorithms of multi-sensors, the D-S rule is widely used extensively. But it can not resolve the problem of feature collision ideally. An extended D-S rule is proposed. Based on the analysis of the extended rule influenced by the fusion order, the best fusion order is presented, which makes a good use of the collision information. With the comparison between the effects of the D- S rule and the extended rule in target recognition, the simulation results show the extended rule is more efficient in dealing with the collision information for aircraft target recognition.