分时电价下的充电站负荷预测

Charging station load prediction under time-of-use pricing

为有效应对分时电价(ToU)策略对充电站的影响,给充电站运营商、电力系统管理者和政策制定者提供科学依据,通过分析历史充电数据、用户行为模式和气象信息,识别定价对用户充电行为的直接影响。针对这一问题,采用一种混合模型,将时间序列模型与用户行为模型相结合。通过设计特征嵌入层等方法引入Transformer模型以更好地捕捉时间序列数据中的复杂模式,而用户行为模型则有助于全面考虑用户充电偏好和响应。实验结果表明,混合模型在充电站负荷预测方面表现了卓越的性能。通过对分时电价的重点关注,成功地预测了未来一段时间内充电站的负荷需求。

In order to effectively deal with the impact of Time-of-Use pricing(ToU)strategy on charging stations and provide scientific basis for charging station operators,power system managers and policy makers,the direct impact of pricing on user charging behavior was identified by analyzing historical charging data,user behavior patterns and meteorological information.To solve this problem,a hybrid model was adopted,which combined the time series model with the user behavior model.The Transformer model was introduced through methods such as feature embedding layers to better capture complex patterns in time series data,while the user behavior model helped to fully consider user charging preferences and responses.Experimental results show that the hybrid model has excellent performance in the load prediction of charging stations.By focusing on ToU,the research successfully predicts the load demand of charging stations in the future period.