考虑集群辨识的海量用户负荷分层概率预测

随着电力公司等传统能源企业向综合能源服务商的加速转型,原有的粗放式用户用电管理模式逐渐难以满足电力营销管理的需求。针对海量用户场景提出了用电模式分层聚类方法及用户集群辨识模型。基于用户集群辨识结果提出了条件残差模拟负荷概率预测模型,进行负荷分层概率预测,以实现对用户精细化用电管理。通过典型案例验证了所提方法的可行性与优越性。

Load-stratified Probability Forecasting for Massive Users Considering Cluster Identification

With the accelerated transformation of traditional energy companies such as power companies to integrated energy service providers, the original extensive power management model of users has gradually become difficult to meet the needs of power marketing management. A stratified clustering method of electricity consumption patterns and a model of user cluster identification are proposed for the scenarios including massive users. Based on the identification results of user clusters, this paper proposes a probability forecasting model of conditional residual simulation load. The load-stratified probability forecasting is carried out to realize the refined power management of users. A typical case verifies the feasibility and superiority of the proposed method.