How to improve PHEV electric mileage ratios? Factor decomposition with Explainable AI
利用日本大型数据集和可解释增强机模型，分解影响插电式混合动力汽车电动里程比的关键因素，发现充电习惯和驾驶者性格比交通条件等影响更大，为政策制定提供依据。
Plug-in hybrid electric vehicles (PHEVs) are yet another practical solution for reducing emissions, although the effect depends on the electric mileage ratio (EMR). This study explored the relationship between EMRs and charging habits, traffic conditions, driver disposition, socioeconomic attributes, and area characteristics using an Explainable Boosting Machine (EBM). Utilizing an extensive dataset from Japan , the fifth-largest emitter worldwide, the EBM offers higher prediction accuracy and clearer interpretations than traditional methods. This approach allowed us to conduct a comprehensive analysis, thereby filling an important gap in the literature. Charging habits and driver disposition had a greater impact on the EMR compared with other factors. Encouraging home charging could be more effective in improving the EMR than developing public charging stations by promoting policies based on the learned EBM. Our results provide useful information for developing policies to improve the on-road emission reduction benefits of PHEVs.