基于MDP随机路径模拟的电动汽车充电负荷时空分布预测

针对电动汽车时空转移随机性的问题,计及实时交通与温度,提出了一种基于马尔可夫决策过程随机路径模拟的城市电动汽车充电负荷时空分布预测方法。首先,根据各类车型充电方式与出行特点对各类电动汽车进行分类;其次,根据蒙特卡洛方法建立各类电动汽车的时空转移模型,采用马尔可夫决策理论对出行路径进行实时动态随机模拟;根据电动汽车实测数据建立温度、交通能耗模型,计算得到实时单位里程耗电量。最后,以某典型城区为例,对不同温度、不同交通状况下电动汽车区域充电负荷进行计算。仿真结果表明,区域内快充负荷较大的节点充电波动性较大,环境温度升高或交通拥堵状况恶化会导致充电负荷高峰的持续时间增高。

Spatial-Temporal Distribution Prediction of Charging Load for Electric Vehicle Based on MDP Random Path Simulation

Contrapose the spatial-temporal transfer random of electric vehicles, temperature and traffic conditions are taken into consideration, and a spatial-temporal distribution prediction method of charging load for urban electric vehicle based on Markov decision process(MDP)random path simulation is proposed. Firstly, the types of electric vehicles are classified according to their spatial transfer and charging characteristics. Secondly, the real-time dynamic random path simulation based on Monte Carlo method and Markov decision theory are used to establish the spatial-temporal transfer model of various electric vehicles. The model of temperature and traffic energy is established according to the measured data of electric vehicles, and the energy consumption is obtained accordingly. Finally, taking a typical urban area as an example, the charging load is calculated under different temperatures and different traffic conditions. The simulation results show that the charging load of the nodes with fast charging in the region is higher, and ambient temperature rise or deterioration of traffic congestion can lead to an increase in charging load.