电动汽车具有低排放、低噪声、高能效等特点，已俨然成为新能源汽车的中坚力量。随着车路互联的发展，车联网背景下的电车大规模应用已成为一种必然，而随之带来的对电量的巨大需求将成为亟待解决的难题。目前这方面的研究是相对零碎的，特别缺乏研究在车联网背景下，综合考虑车路协同及无线充电技术，来实现对电车能耗优化的绿色生态驾驶及出行。为弥补这不足，本研究将从深度挖掘车联网大数据出发，首先探究电车的能量耗散机理，建立以数据驱动的高精度能耗预测模型；进而分析信号配时与驾驶速度的耦合关系，通过部署点-线-面的车路协同控制策略，并采用强化学习来同步实现电车的绿色生态驾驶；最后考虑无线充电布设与出行路径选择的相互制约，构建诱导绿色出行的双层多目标优化模型，运用在线学习机制实现绿色出行实时诱导并最终建立满足车路协同一体化的电车生态管理系统。本研究对推动电车市场并引导绿色生态出行，具有重要的理论研究意义和实际应用价值。

Electric vehicle (EV) has become a mainstream force in the development of new energy transportation thanks to its overwhelming advantages of energy efficiency, low greenhouse gas and noise emission. Under the context of connected vehicles (V2X), the massive deployment of EVs is becoming inevitable with the promotion of the cooperative vehicle-infrastructure development. Nevertheless, the huge demand of electricity with the wide spreading deployment of EVs will pose great challenges to the nation’s electric power system. At present, systematic study combining vehicle infrastructure cooperation, wireless charging and energy consumption optimization is absent. In order to fill the gap, this project is going to study and eventually establish an eco-friendly driving and traveling system for EVs based on the deep learning of the comprehensive data of connected vehicle network. The study could be decomposed into three stages. At the first stage, data-driven based accurate energy consumption model of electric vehicles will be built by investigating the mechanism of energy consumption and regeneration. With the help of the established energy consumption estimation model, at the second stage, the coupling relationship of vehicle speed and signal timing will be further analyzed, and then a dynamic EV eco-friendly driving control system can be developed by incorporating the vehicle-infrastructure cooperation controlling strategies at intersection-arterial-network levels and applying the deep learning method. At the third stage, a dual-level multiple objective optimization model for EVs’ eco travelling will be developed by considering the optimal deployment of wireless charging facilities and EV drivers’ route choices. Furthermore, by applying on-line learning mechanism, a real-time eco-travelling guiding system can be built; and eventually, an ecological electric vehicles management and control system can be established by integrating the work from all three stages. This project will be of vital academic importance and application value in promoting the electric vehicle market and eco-friendly driving and travelling.