Optimization Method of Power Equipment Maintenance Plan Decision-Making Based on Deep Reinforcement Learning

The safe and reliable operation of power grid equipment is the basis for ensuring system. At present, traditional periodical maintenance has exposed abuses such as deficient excess maintenance. Based on a multiagent deep reinforcement learning decision-making optimization algorithm, method plans proposed. In this paper, an model plan that takes into account reliability economics constructed with constraints safety its constraints. distributed recurrent Q-networks adopted to solve model. uses high-dimensional feature extraction capabilities multiobjective problem planning. Through case analysis, comparative results show proposed algorithm better ability, well lower cost. Accordingly, can realize optimal decision plan. expected value shortage cost obtained by $71.75$ $MW·H$ $496000$ $yuan$.