D-S证据理论及其改进算法研究

在多传感器数据融合中,D-S证据理论对于处理不确定性问题是一种有效的方法。但是,当证据严重冲突或完全冲突时,利用D-S证据理论法计算容易给出错误的结果。改进的证据理论法虽然解决了上述问题,但通过分析数据试验发现改进方法在处理一般的证据时效果不理想。文章将经典证据理论和改进方法相结合,发挥两种方法的优点,以达到更好的识别效果。

Research on D-S Envidence Theory and its Improvement Algorithm

In multi-sensor information fusion,D-S evidence theory is an effective way to process uncertainty.However,it is easy to get the wrong result when evidence is seriously or completely conflict.Although the improved evidence theory solves these problems,it is found that it is not satisfied when processing general evidence by analyzing the data.Therefore,it will get better recognition result to combine the classical evidence theory and the improved method.