基于BP神经网络的电动汽车负荷预测

Electric Vehicle Load Forecasting Based on BP Neural Network

摘要: 大规模电动汽车接入，对电网造成严重冲击，影响电网可靠运行。为减少电动汽车接入对电网的平衡造成的破坏，研究基于大数据的电动汽车用电特性成为增强电网的可靠性与安全性的关键。本文通过电动汽车的利用特性及相关参数，分析不同影响因素对充电负荷的影响，研究各类电动汽车的充电负荷模型。在此基础上结合BP神经网络算法对电动汽车的发展趋势进行预测，建立电动汽车充电负荷预测模型，预测未来几年电动汽车用电负荷趋势，为电网优化及电能分配提供参考依据。

Abstract: Large-scale electric vehicles (EVs) access will cause serious impact on the power grid and affect the reliable operation. In order to reduce the damage to the balance of the power grid caused by the simultaneous access of EVs, the research on the electrical characteristics of EVs based on large data becomes the key to enhance the reliability and safety of the power grid. In this paper, through the characteristics and related parameters of EVs’ operation, the impact of different factors on charging load is analyzed, and the charging load models of various types of EVs are studied. On this basis, combined with BP neural network algorithm, the development trend of EV is forecasted, the charging load forecasting model of EV is established, and the power consumption of EV in the next few years is estimated, which provides a reference basis for grid optimization and power distribution.