A method for condition evaluation based on DSmT

Condition-based maintenance (CBM) is an optimal maintenance strategy due to improved planning for rehabilitation or replacement based on current evaluation of condition. In this paper, a new method of information fusion-DSmT(Dezert-Smarandache Theory) developed from DST(Dempster-Shafer Theory) and Bayesian theory is introduced to dealing with condition evaluation. With the research and analysis of condition evaluation, the model of condition evaluation based on DSmT is presented in which the generalized basic belief assignment and the rule of information fusion are built. The performance is compared with DSmT and DST for challenging realistic condition evaluation. The result shows that the DSmT enhance the performance of condition evaluation by reducing the time on computing and increasing the quality of fusion result.