Research on improved evidence theory based on multi-sensor information fusion

Abstract In view of the lack effective information fusion model for heterogeneous multi-sensor, an improved Dempster/Shafer (DS) evidence theory algorithm is designed to fuse multi-sensor information. The first introduces compatibility coefficient characterize between evidence, obtains weight matrix each proposition, and then redistributes basic probability distribution focal element obtain a new source. Then concept credibility introduced, average support used improve synthesis rule, so as result. Compared with other algorithms, proposed can solve problems existing in DS when dealing highly conflicting certain extent, results are more reasonable convergence speed faster.