基于改进核密度估计的电动出租车快速充电行为研究

Fast Charging Behavior of Electric Taxi Based on Improved Kernel Density Estimation

电动汽车充电行为研究是充电负荷时空分布预测、充电基础设施规划和有序充电管理的基础。本文基于上海市电动出租车的实测数据,划分充电片段并提取快速充电行为特征变量,开展相关性分析以揭示变量之间的内在相关性,从工作日和休息日两个时间维度揭示快速充电行为规律,提出一种基于扩散方程的自适应扩散核密度估计模型应用于快速充电行为特征变量的概率建模并使用拟合优度检验指标验证该模型的有效性。研究结果表明:电动出租车的快速充电行为在工作日和休息日具有明显的差异性,自适应扩散核密度估计模型可使电动汽车充电行为特征变量的概率建模更加准确,且具有更高的拟合精度。

Charging behavior of electric vehicle(EV)is the basis of spatial-temporal distribution prediction of charging load,charging infrastructure planning and vehicle charging management.This paper collected the electric taxis charging data in Shanghai and extracted the characteristic variables of fast charging behavior for the defined charging segments.The correlation analysis was carried out to reveal the intrinsic relationship between variables and the law of fast charging behavior for both weekday and weekend study periods.An adaptive diffusion kernel density estimation model(ADKDE)was proposed on the basis of the diffusion equation and applied to the probability estimation of characteristic variables of fast charging behavior.The goodness of fit test was performed to verify the effectiveness of the ADKDE model.The results indicate that the fast-charging behavior of electric taxis is significantly different between weekday and weekend.The proposed ADKDE model can improve the accuracy of the probability modeling of the EV charging behavior and the accuracy of model fitting.