Maintenance Decision-making Model Based on Partially Observable Markov for Railway Traction Substation Equipment

Traditional planned maintenance mode for railway traction substation equipment may lead to excessive maintenance or inadequate maintenance. To solve this problem, this paper combined the actual railway inspection operations and presented a novel condition-based maintenance decision-making model under partial observation. In consideration of the random failure and deterioration failure, partially observable Markov process(POMDP) was used to describe the state transition process of the device. Furthermore, with considering the uncertainty of equipment repair, the instantaneous availability of equipment is solved. And then, the quantitative expression of the failure risk and maintenance risk were provided, in which the inspection interval and maintenance time was the parameters. Finally, the maintenance decision-making model based on POMDP was built up to minimize the sum of equipment risk and system operation risk. Genetic algorithm was used to solve this model. The example analysis shows that the model is feasible and effective, and has a certain practical application value.