随机时变路网环境下电动公交充电调度与充电设施规划协同优化研究

Due to high environmental benefits, electric buses are generally recognized as an effective way to deal with the problem of urban traffic pollution. However, the popularization and application of electric buses is limited by the defects of their short driving range and long charging time. Meanwhile, large-scale and disorderly charging services cause excessive load fluctuation for the power grid. Reasonable charging facilities layout and charging schedule can eliminate the public's "range anxiety" over electric buses and reduce the impact on the power grid, which have attracted much attention in recent years. Uncertainties in the road network are not fully considered in the existing research, thus the charging schedule and charging facility planning scheme were designed to be beyond the reality. In this project, the sources and the impact mechanism of uncertainty in the road network are analyzed, and stochastic time-dependent variables are used to represent the uncertainties. Subsequently, the electric bus charging facilities planning, the charging scheduling, and their collaborative relationship are modelled. The expectation method and the chance-constrained method are used to deal with the uncertain variables, and interactive intelligent algorithms are designed to solve models with large-scale, dynamic and nonlinear characteristics. The practicality and validity of the models and algorithms are validated through a series of simulation and comparative analyses. The contributions of this project will enrich and perfect the charging scheduling and charging facility planning theoretical system, as well as provide the theoretical guidance and method support for the large-scale application of electric buses.

电动公交车由于环保效益高被公认为是解决城市交通污染问题的有效途径，但续驶里程短和充电时间长的缺陷限制了其推广应用，同时大规模无序充电会导致配电网负荷波动过大。合理的充电设施布局和充电调度，可以消除人们对电动公交车的“里程忧虑”、降低对配电网的冲击，成为近些年的研究热点。现有研究未能充分考虑路网环境中的不确定性，致使所设计的充电调度与充电设施规划方案过于理想化。项目对路网环境中不确定性的来源和影响机制进行分析，用随机时变变量进行刻画；分别对电动公交充电设施规划、充电调度以及两者之间的协同关系进行优化建模，使用期望值法和机会约束法对不确定变量进行处理，设计能适应大规模、动态性、非线性等特征的交互式智能算法进行求解；通过实际案例的模拟仿真和对比分析，验证模型和算法的实用性和有效性。项目成果将丰富和完善现有的充电调度与充电设施规划理论体系，为电动公交车的大规模应用提供理论指导与方法支撑。

为实现“碳达峰”、“碳中和”的目标，低碳出行将成为我国新时期经济社会可持续发展的重要战略之一。由于环保效益高，大力发展电动公交车被认为是解决日益严峻的交通能源消耗和环境污染问题的有效途径。但电动公交车续驶里程短和充电时间长的缺陷限制了其推广应用，同时大规模无序充电会导致配电网负荷波动过大。本项目从管理的视角出发，研究了以下四个方面的内容：1）电动公交网络能量需求预测；2）电动公交充电设施规划；3）电动公交充电和换电优化；4）电动公交充电调度与充电设施规划协同优化。本项目充分考虑路网环境中的不确定性因素，并用随机时变变量进行刻画；将负载质量、是否开空调和行驶时段考虑在测算电动公交网络年需求能量的过程中，使得计算结果更符合实际；分别对电动公交充电设施规划、充电调度以及两者之间的协同关系建立优化模型，使用期望值法和机会约束法对不确定变量进行处理，设计能适应大规模、动态性、非线性等特征的交互式智能算法进行求解；通过实际案例的模拟仿真和对比分析，验证模型和算法的实用性和有效性。项目的研究结果证明，通过合理地充电设施布局和充电调度，不但可以消除人们对电动公交车的“里程忧虑”，还可以降低对配电网的冲击。本项目成果丰富了现有的充电调度与充电设施规划理论体系，为提高电动公交车的运营管理水平提供了决策支持。