一种D-S证据推理的改进方法

An Improved Method of D-S Evidential Reasoning

D-S证据理论是多传感器信息融合常用的算法,但它无法解决证据的可信度问题。本文根据传感器的局部决策先算出局部决策值,构造整个系统的支持矩阵,然后求这个支持矩阵的特征向量,以此作为各个传感器的可信度。这个可信度就可以作为各个传感器的权值,以此修正D-S证据的融合算法。最后,给出了一个对比仿真实验,证明此算法的在处理错误或大偏差证据的有效性。

D-S evidential reasoning is a usual method in multi-sensor fusion, but it can抰 determine the reliability of the evidence. The paper firstly calculates the local decision on the basis of the measurement, on which the support matrix is based. Then the eigenvector is gained from the support matrix, and it is the reliability vector of the system. The paper improves the D-S evidential reasoning by giving a weight to evidence that equals to the evidence抯 reliability. At last, an example is given out. It proves that the method described above is valid when some evidences are false or far from the real value.