信息融合中传感器可信度的动态确定及应用

Dynamic determination of sensor credibility in data fusion and its application

基于经典冲突系数和证据距离函数给出的修正冲突系数,提出了确定传感器动态可信度的方法.将所提出的方法应用到目标识别系统中,算例表明:所提出的方法可以根据系统收集的传感器报告来确定各个传感器报告的可信度,在证据高度冲突情况下也能够正确识别目标.

Based on the modified conflict coefficient which takes the consideration of classical conflict coefficient and evidence distance compressively,a new model to dynamically determine the sensor credibility through data collected in multisensor systems was constructed in this paper.A numerical example in target recognition was used to illustrate the efficiency of the method.It is shown that correct results can be obtained when the evidences conflict with each other even in a high degree.