快换式公交充电站短期负荷预测方法的研究 research on short-term load forecasting methods of electric buses charging station.pdfVIP

第 41 卷 第 4 期 电力系统保护与控制 Vol.41 No.4 2013 年2 月 16 日 Power System Protection and Control Feb.16, 2013 快换式公交充电站短期负荷预测方法的研究 张维戈，颉飞翔，黄 梅，李 娟，李亚芬 (北京交通大学电气工程学院，北京 100044) 摘要：为了应对电动汽车规模化发展之后庞大的充电站负荷对电力调度配送等环节带来的影响，依据当前北京奥运快换式公 交充电站的相关实测数据，对快换式公交充电站负荷特征及其影响因素进行分析，提出一种基于模糊聚类分析与BP神经网 络相结合的快换站短期负荷预测方法。利用该方法对四组不同的测试日进行预测实例分析，并与单一BP网络模型的预测效 果进行比较。统计结果表明，基于模糊聚类分析与BP神经网络相结合的预测方法具有较高的预测精度，满足一定的应用要 求，适用于快换式公交充电站短期负荷预测工作。 关键词：电动汽车；快换式充电站；短期负荷预测；模糊聚类分析；BP神经网络 Research on short-term load forecasting methods of electric buses charging station ZHANG Wei-ge, XIE Fei-xiang, HUANG Mei, LI Juan, LI Ya-fen (School of Electrical Engineering, Beijing Jiaotong University, Beijing 100044, China) Abstract: In order to deal with the impact of the huge load of charging stations on power scheduling and distribution after large-scale developments of electric vehicles(EV), according to the measured data of Beijing Olympics quick-change bus charging station, this paper analyzes the load characteristics of the quick-change bus charging station and its influencing factors, then puts forward a short-term load forecasting method of quick-change station based on fuzzy clustering analysis and BP neural networks. The paper uses this method to predict four groups of different testing days, and compares it with the predicting answer of a single BP neural network model. The statistical results show that the method of the combination of fuzzy clustering analysis and BP neural network has a higher accuracy and it meets certain application requirements. It reveals the combined method is suitable for quick-change bus charging stations short-term load forecasting. This work is supported by National High-tech R D Program of Chi