基于数据分类重建的10 kV变压器负载损耗超标预警方法

Excessive Loss Warning Method of 10 kV Transformer Load Loss Based on Data Classification Reconstruction

10 kV变压器运行数据中心在存储过程中可能会出现电力系统数据采集与监控(SCADA)数据残缺,传统变压器负载损耗超标预警方法中无法识别缺失数据,影响预警结果。为此,设计一种基于数据分类重建的10 kV变压器损耗超标预警方法。设计适配变压器的数据采集装置总体架构,分别重构采集到的SCADA数据,修复缺失数据,将变压器输入特征量归一化处理,计算负载损耗,结合衰退指标构建超标预警架构,完成10 kV变压器负载损耗超标预警。测试结果表明,提出方法可有效识别负载损耗数据,且负载损耗的计算值与实际情况更接近。

In the storage process of 10 kV transformer operation data center,the data of supervisory control and data acquisition(SCADA)of power system may be incomplete,and the missing data cannot be identified in the traditional early warning method for excessive transformer load loss,which affects the early warning results.An early warning method for excessive loss of 10 kV transformer based on data classification reconstruction is designed.We design the overall architecture of the data acquisition device,reconstruct the collected SCADA data,respectively,and repair the missing data.The input characteristic quantity of transformer is normalized,the load loss is calculated,and the excessive loss warning framework is constructed in combination with the recession index,the excessive loss warning of load loss of 10 kV transformer is completed.The test results show that the proposed method can effectively identify the load loss data,and the calculated value of load loss is closer to the actual situation.