With the acceleration of urbanization and the development of public transportation systems,the operational efficiency and energy utilization efficiency of subway systems have attracted more and more attention.Flywheel energy storage technology provides new solutions to energy utilization problems in rail transit systems with its high-power cycle capabilities.In this paper,Markov decision process is used to describe the energy management problem of a single flywheel energy storage system,and a reinforcement learning algorithm based on deep Q network is used to learn the optimal dynamic adjustment strategy for charge and discharge thresholds.By building a simulation environment on Matlab/Simulink platform,the developed energy management algorithm is tested,and the results are compared with fixed charge and discharge threshold strategies and random charge and discharge threshold strategies,which shows that this strategy has significant effects on improving power utilization efficiency and system operation stability.
flywheel energy storage systemenergy managementMarkov decision processdeep reinforcement learningDeep Q-Network(DQN)
万方数据
维普
