基于随机森林算法的台区合理线损率估计方法

Line loss rate estimation method of transformer district based on random forest algorithm

线损管理是电力公司的重点管理内容之一,低压电网普遍采用分台区的管理手段。供电侧数据缺失和营销抄表日期冲突导致的线损率缺失是电力公司线损系统中台区线损数据存在的主要问题。为此,提出了一种涉及多源数据的基于随机森林算法的台区合理线损率估计方法。从线损系统、生产管理系统和营销系统中提取台区、变压器和用户相关数据,建立台区特征数据库;对台区进行聚类分析,并在此基础上建立决策树分类模型和随机森林估计模型;利用上述模型估计台区线损率。以上海电力公司实际数据为例,计算结果验证了所提方法的可行性;并将所得结果与线性回归模型和回归树模型的估计结果进行比较,表明所提方法性能优越。

Line loss management is one of the key management contents of electric power company and district-dividing is a common management means in low-voltage power grid.The missing of line loss rate data caused by the data missing at power supply side and the mismatching of record time is the key problem in the line loss system of electric power company.Aiming at the above problems,a line loss rate estimation method of transformer district based on random forest algorithm is proposed involving multi-source data.Related data of transformer districts,transformers and loads are extracted from line loss system,production management system and marketing system to establish the feature database of transformer district.Then,the transformer districts are clustered,based on which the decision tree classification model and random forest estimation model are established.Finally,the line loss rate of the transformer district is estimated by the above models.Calculative results of Shanghai Electric Power Company verify that the proposed method is feasible and has superior performance compared with the estimation results of linear regression model and regression tree model.