基于无线输电的电动汽车动态充电负荷预测  

Electric Vehicle Dynamic Charging Load Forecasting Based on Wireless Power Transmission

作　　者：刘杨 王维庆[2] 王海云[2] LIU Yang;WANG Weiqing;WANG Haiyun(State Grid Urumqi Power Supply Company, Urumqi 830011, Xinjiang, China;College of Electric Engineering, Xinjiang University, Urumqi 830047, Xinjiang, China)

机构地区：[1]国网乌鲁木齐供电公司,新疆乌鲁木齐830011 [2]新疆大学电气工程学院,新疆乌鲁木齐830047

出　　处：《电网与清洁能源》2018年第1期128-136,共9页Power System and Clean Energy

基　　金：自治区重点实验室项目(2016D03021);国家高技术研究发展计划(863计划)(2013AA050604);教育部创新团队项目(IRT1285)~~

摘　　要：无线充电式电动汽车因动态充电方式的便捷性,在未来规模化应用中,其充电负荷将给电力系统带来重大影响。提出了基于无线输电的电动汽车充电负荷预测方法,根据动态无线充电频率高、速度快的特点,建立了以超级电容为辅助电源的系统模型,得到了动态充电负荷的计算因子。考虑到影响电动汽车充电行为的因素众多,提出了以大数据为基础的混合分类模型,利用TAN分类器和粗糙集的互补作用,提高分类精度和效率,达到准确预测充电负荷的目的。仿真结果表明,动态充电负荷克服了静态充电负荷在时间和空间尺度下聚集的缺点,能与电网良好互动,从而为电网负荷削峰填谷。Because of its convenience, the electric vehicle with wireless charging is expected to be widely used, thus the dynamic charging load resultant from this mode on a large-scale will bring significant impacts on the power system in the future. This paper proposes a load forecasting method based on wireless power transmission. According to the characteristics of the high speed and high frequency the dynamic wireless charging, a system model based on super capacitor for auxiliary power supply is established and the calculation factor of the dynamic charging load is thereby obtained. Taking into account numerous factors that affect the charging behavior of electric vehicles, a hybrid classification model based on large data is proposed, using the complementary effect of TAN classifier and rough set to improve the classification accuracy and efficiency so as to achieve the accurate predicting of the charging load. The simulation results show that dynamic charging load has overcome the shortcomings of the load aggregation in time and space as in the static charging, and its good interaction with the power grid helps to shave peak loads and fill valley loads for power grids.

关 键 词：电动汽车 无线充电 超级电容 负荷预测 混合分类器 

分 类 号：TM714[电气工程—电力系统及自动化]