Statistical approaches for cleaning BCTC load data

负载数据是指在交付点级别上，电表在一定时间间隔内记录的电力消耗。负荷数据是电力系统运行、规划和维护中最关键的数据之一，它包含了有关日常状况、评估系统可靠性性能指标、节能以及系统规划和扩展的重要信息。 然而，以细粒度准确地收集所有负载数据是一项具有挑战性且成本高昂的任务。多年来，不列颠哥伦比亚省输电公司（BCTC）一直在努力解决这一问题的负荷数据质量。检测损坏或其他无效数据并不是一件容易的任务，因为由于电力消耗行为、断电事件和客户动态的不确定性，负载数据自然是随机的。由于需要在去除无效数据和噪声的同时保留负载数据中的原始模式，而不是人为地弥补丢失或损坏的数据，因此难度更大。

Load data refers to the power consumption recorded by meters at certain time intervals at the delivery point level. Load data is one of the most critical data for operation, planning and maintenance power systems, and contains vital information for day-to-day conditions, for evaluating system reliability performance indices, for energy saving, and for system planning and expansion. However, collecting all load data accurately in fine granularity is a challenging and costly task. For many years, British Columbia Transmission Corporation (BCTC) has been trying to address this issue of load data quality. Detecting corrupted or other invalid data is not an easy task because load data is naturally random due to uncertainty in the power consumption behaviors, outage events and dynamism of customers. The difficulty is compounded by the requirement of preserving the original patterns in the load data while removing invalid data and noises, instead of artificially making up the missed or corrupted data.