基于时域特征检测用户用电不确定性行为

针对用户用电不确定性程度难以量化,如何选取合适用户参与电网新业务的问题,提出一种基于时域特征分量提取的用户用电不确定性行为检测方法,首先依次提取用电数据的周期分量、趋势分量和随机分量,并从随机分量中检测用户的不确定行为点,其次分别从短期和长期两个维度提出用户局部和整体不确定性指标,并对用户不确定性进行排序.结果 表明,所提方法可以准确检测用户的不确定行为点,量化用户的不确定性程度,在检测精度,效率和速率等方面均比其他方法具有优势,为电网公司在不同时段选取用户参与调控提供依据.

Uncertainty Behavior of Electricity Consumption is Detected Based on Time Domain Characteristics

In view of the difficulty of quantifying the degree of uncertainty in power consumption by users, how to select suitable users to participate in new grid services, a method for user uncertainty behavior detection based on the extraction of time-domain feature components was proposed. Firstly, the periodic component, trend component and random component of the electricity consumption data were extracted in sequence, and the uncertain behavior points of users were detected from the random components. Secondly, the local and global uncertainty indexes of users were put forward from short-term and long-term dimensions respectively, and the user uncertainty was sorted. The results show that the proposed method can accurately detect the uncertain behavior points of users, quantify the degree of uncertainty of users, and has advantages over other methods in terms of detection accuracy, efficiency and speed, which provides a basis for power grid companies to select users to participate in regulation at different time periods.