Electric vehicle charging period planning method based on deep reinforcement learning

The invention discloses an electric vehicle charging period planning method based on deep reinforcement learning. The method comprises the following steps: establishing a Markov process model of an electric vehicle charging period optimization planning process, and performing training optimization on the Markov process model by using a near-end strategy optimization method to obtain a trained Markov process model; and inputting the electric vehicle charging total power optimization target value and the charging demand, and outputting an electric vehicle charging time period planning result. The charging peak of the electric vehicle can be relieved, the current charging time period of the electric vehicle is optimized on the premise that the charging requirement of the electric vehicle is met, and the charging requirement distribution of all the electric vehicles is more uniform. Meanwhile, the power supply pressure can be reduced, and the power demand of the power distribution network is more balanced. The ch