Dynamic Programming Approach to Optimal Maintenance Scheduling of Substation Equipment Considering Life Cycle Cost and Reliability

摘要

Asset management of transmission system is an essential issue for smart grid and condition monitoring system, especially maintenance planning. Maintenance scheduling is the most important part of asset management. The scheduling affects the reliability of equipment, the reliability of substation system and life cycle cost (LCC). Since the equipment in substation hardly fails to operate, the information about failure is not yet systematically recorded. This paper focuses on long term maintenance based on available data of historical planning and operation data. In particular, the Weibull distribution model is used to estimate reliability function of equipment. A maintenance planning may lead to different final condition of effective age of equipment. We propose the optimal maintenance planning in conjunction with the specified final condition of equipment and compare the LCC and reliability. By using dynamic programming (DP), the algorithm gives the minimum LCC while satisfying reliability and final condition. Numerical results demonstrate the effectiveness of the DP approach. Comparing with the results using the genetic algorithm, the DP method provides the maintenance planning with less LCC.