Fig. 2From: Power flow adjustment for smart microgrid based on edge computing and multi-agent deep reinforcement learningThe concept of the DRL modelBack to article page