Software-Intensive Equipment Fault Diagnosis Research Based on D-S Evidential Theory

Aiming at limitations of the current fault diagnosis of the software-intensive equipment (SIE), considering the advantages of D-S evidential theory at dealing with multi-information, this paper presents a method of fault diagnosis at the decision level based on D-S evidential theory. The method establishes system structure of fault diagnosis, constructs the reasonable basic probability assignment algorithm, carries out multi-criteria fusion using D-S fusion model and method. The method is proved to be effective for fault location by instance. It makes diagnosed information more definite and improves the accuracy of diagnosis.