A method for maintenance decision based on condition monitoring

Abstract

Condition based maintenance (CBM) is a kind of realtime efficient maintenance decision, which is desired to solve the question about “lack of maintenance” and “excess of maintenance”. Through CBM, the maintenance cost can be reduced and the risk of catastrophic failure can be minimized to make the equipment running with maximum effectiveness. Based on the data of condition monitoring, combined with the factors such as maintenance times, degradation of equipment, human factors, working conditions and environment, a Weibull Proportional Hazards Model (WPHM) was established. With the goal of average maintenance cost and the constraint of availability, the genetic algorithm is adopted to optimize the threshold to guide the maintenance decision. Finally, maintenance advice is provided under certain confidence level. Combined with the monitoring data of diesel engine, the WPHM was established after estimation of parameters. And the corresponding maintenance decision is analyzed by the proposed method. The proposed method is of great significance for other repairable systems about CBM.