中文摘要：

基于68辆日本爱知县电动汽车1a间的实际出行GPS数据，定量分析了电动汽车在城市圈内行驶的电量消耗规律，尤其是较全面的考察了道路坡度对能耗的影响特征．比较了多层混合效应线性回归模型、最小二乘回归模型和随机效用面板模型的参数估计效果．结果表明，多层混合效应线性回归模型更好的拟合观测数据．道路坡度对能耗的影响最大，下坡坡度显著影响电动汽车回收制动能源的效率．研究成果有效提高了目前电动汽车能耗估计的精度，选择最佳能耗路径的同时，有助于减少电动汽车出行者的里程焦虑．

英文摘要：

The influence factors and characteristics of travelling Electric vehicles＇ （EVs） energy con- sumption within metropolitan are quantitatively analyzed, where the experimental GPS data collected for one year from 68 EVs in Aichi Prefecture, Japan was employed. Research focus on the effects of road gradients on the efficiency of EVs＇ energy consumption. The multi-level mixed effects linear regression model is used to estimate EVs＇ energy consumption, with its estimated coefficients compared to the least squares regression model and the random effects panel data model. The former has better fitness than the other two models in general. The road gradients have great effects on EV energy consumption, especially the downward gradients contribute significant on regeneration efficiency of braking energy. The proposed model improves the accuracy of previous EVs＇ energy consumption estimation models, which may help select the best route with less energy consumption and reduce the range anxiety of current EV drivers.