动态交通环境下的纯电动车辆多目标出行规划
Multi-objective optimization for traveling plan of fully electric vehicles in dynamic traffic environments

为缓解电动车辆出行过程中包括里程不足、充电时间长、充电站稀少以及电池循环寿命有限等固有问题, 提高电动车辆的行驶性能以及驾驶员的接受程度, 需要为其推荐合理的出行与充电方案。然而, 目前出行方案制定方法没有考虑到交通环境复杂多变的特性, 并且仅能提供在单一目标下的出行方案, 难以为驾驶员提供综合考虑多种因素的出行策略。该文提出了一种在动态随机路网环境下的考虑多目标多约束的电动车辆出行规划策略。该出行规划策略考虑到交通环境的时变随机特性, 利用多目标蚁群优化方法计算求解最优Pareto解集, 为驾驶员推荐包括出行路径、各路径上的行驶速度、充电位置与模式、空调使用等出行要素。研究结果表明: 基于动态随机路网的出行方案相比于基于静态确定性路网的出行方案更为优秀; 相比于单一目标出行方案, 基于多目标优化的出行策略综合性能更好。仿真结果证明了该方法能够协调各优化目标与约束条件, 合理推荐电动车辆的出行方案解集, 提升电动车辆的使用性能。
Abstract：Reliable routing and charging plans are needed for fully electric vehicles (FEVs) to ease the intrinsic drawbacks of FEVs, such as mileage limitations, overly long charging times, limited charging stations, and limited battery lifetimes to enhance the driving performance and drivers' acceptance. However, most research has overlooked the effects of changing traffic system features and has provided travelling strategies for only a single travelling objective rather than strategies that consider multi-factors simultaneously. This paper describes a multi-objective, multi-constraint travelling plan optimization strategy for dynamic time-dependent traffic networks for FEVs. This optimization strategy includes the time-dependent stochastic changes of the traffic environment with the multi-objective ant colony optimization method used to calculate the optimal Pareto travelling plan set. The result provides drivers with information including travelling route, travelling speed on each road, charging locations and modes, and air-conditioner usage. The results show that a travelling plan within a dynamic time-dependent traffic environment is better than that within a stable deterministic traffic environment. The travelling result with multi-objective optimization is better than a single objective strategy. This multi-objective, multi-constraint optimization method provides reasonable travelling plans that enhance the driving performance of FEVs.