Abstract：Aiming at the difficulty in the intelligent fault diagnosis for the newly developed product because of the insufficiency of the data about its failure probability, a novel fault diagnosis method based on multi-information fusion with the weighted D-S evidence theory was proposed. The multi-information fusion with D-S evidence theory was used to deal with the problem of the lack of the model of the fault probability distribution or the non-effective of the exactly mathematical analysis. In order to regulate the degree of confidence of the present fault diagnosis estimation according to the past accuracy about the fault diagnosis, the weighted D-S evidence theory was applied. Then modified fault diagnosis with the past fault diagnosis accuracy was fulfilled. The results of the fault diagnosis for the late-model marine meteorological instrument show the proposed method is feasible and efficacious while the fault probability and the knowledge are insufficient.
2011, Vol. 28 
