基于马尔科夫链充电负荷预测的多区域充电桩优化配置研究

发布时间：2018-05-11 03:37

本文选题：电动汽车 + 马尔科夫链　；参考：《工程科学与技术》2017年03期

【摘要】：考虑用户出行习惯的复杂性、多样性,对区域充电桩进行合理的配置以满足充电需求。首先,通过马尔科夫链模型描述电动汽车用户一天出行过程中在行驶、充电、不充电也不行驶3种决策行为下,动力电池荷电状态的变化情况。以此确定该过程中电动汽车用户的实时充电行为,得出不同类型电动汽车的快充、慢充负荷需求。然后,考虑规划区内电动汽车的移动特性及其不同时段不同类型电动汽车辆数,预测各区域各时段充电负荷的需求情况。最后,以投资、运维成本最小为目标建立区域充电桩优化配置模型。该模型计及了电动汽车移动特性均衡等约束条件,并通过粒子群优化算法求解。对33节点4区域系统电动汽车充电负荷需求预测及其充电桩配置进行仿真,仿真结果验证了所提方法的有效性和可行性。
[Abstract]:Considering the complexity and diversity of the user's travel habits, the regional charging piles are reasonably configured to meet the charging demand. Firstly, Markov chain model is used to describe the change of charged state of electric vehicle under three kinds of decision behavior: driving, charging, charging and driving. The real-time charging behavior of electric vehicle users is determined, and the demands of different types of electric vehicles for fast charging and slow charging are obtained. Then, considering the moving characteristics of electric vehicles in the planning area and the number of different types of electric vehicles in different periods of time, the demand for charging load in different periods of time in each region is forecasted. Finally, the optimal configuration model of regional charging pile is established with the aim of minimum investment and operation cost. The model takes into account the equilibrium constraints of the moving characteristics of electric vehicles and is solved by particle swarm optimization (PSO). The demand forecast of charging load and charging pile configuration of 33-bus 4-region system electric vehicle are simulated. The simulation results show that the proposed method is effective and feasible.
【作者单位】：四川大学电气信息学院;国网福建省电力有限公司电力科学研究院;国网福建省电力有限公司厦门供电公司;
【基金】：国家自然科学基金资助项目(51377111) 四川省科技厅应用基础项目资助(2015JY0128) 四川大学引进人才科研启动经费资助项目(20822041A4161) 国家电网总部科技项目资助
【分类号】：TM715;U491.8

【相似文献】