Research On Analysis Method Of Distribution Network Line Loss Based On Data Driven

In the current international environment,natural resources are increasingly exhausted,and energy conservation and emission reduction has become the focus of every country.As an indicator of energy loss in the transmission process,line loss is closely related to the line loss management and technical level of power supply enterprises.It is one of the important indicators for the assessment of power supply enterprises.Compared with the high-voltage transmission network,the medium and low-voltage distribution network has the defects of complex structure,incomplete measurement system and unclear topology,which makes the line loss of the distribution network unclear and difficult to reduce the loss reasonably,resulting in a relatively high line loss rate of the system.According to relevant statistics,the line loss of medium and low voltage distribution network in China accounts for about 50% of the line loss of the whole power system,especially the line loss of 10 k V medium voltage feeder.Therefore,it is necessary to focus on the line loss of 10 k V medium voltage feeder.With the arrival of the big data era,the multi-source data platform built by power supply enterprises provides rich data for the line loss analysis and calculation of distribution network.However,due to the wide distribution range of 10 k V feeders,the differences in feeder attributes and operating parameters,and the large regional and environmental impact,some traditional line loss analysis and calculation methods are no longer applicable to the current environment,which hinders the development and promotion of line loss management,It reduces the work efficiency of power supply enterprises.Therefore,based on the multi-source data platform,considering that there are two common types of dirty data in load data: abnormal value and missing value,this thesis establishes a load data cleaning model for abnormal value identification and missing value filling to reduce the impact of dirty data on subsequent line loss analysis and calculation.Then,using the cleaned data and the feeder topology model parsed from the GIS system of distribution network,the forward and backward generation line loss calculation model is constructed to calculate the line loss rate,and the weak position of feeder line loss is located according to the refined calculation results as the theoretical basis for distribution network planning.Finally,in view of the wide distribution range of large-scale distribution networks and the unclear causes of line loss,this thesis designs a set of identification model of line loss causes of feeders based on feeder classification,identifies the causes of line loss of the overall feeders to eliminate irrelevant factors,identifies the causes of line loss of the classified feeders to find out the causes of each type of line loss,and considers the rationality of new feeder planning based on the identification of line loss causes,After determining the category,the line loss cause identification model of the category is used to judge whether the new feeder planning is reasonable.Taking the feeder of a large Prefecture Level Power Supply Bureau in Guangdong Province as an example,the validity and practicability of the data cleaning model,line loss calculation and cause analysis model in this thesis are tested.The data cleaning,line loss calculation and cause analysis model in this thesis can deal with the abnormal and missing values in the sample,refine the line loss calculation of the feeder and mine the cause of the feeder line loss.The processing of outliers and missing values can bring high-quality data for the subsequent line loss calculation and line loss cause mining,and obtain more reliable line loss calculation and cause mining results;By calculating the line loss of sample feeders,we can understand the line loss level and refined loss of each feeder,which is convenient for engineers to make reasonable rectification;Through the scientific classification of sample feeders and the mining of line loss causes,we can clarify the focus of loss reduction in various regions,optimize the allocation of distribution network resources,exchange the minimum investment for the maximum benefits,and actively respond to the call of national energy conservation and emission reduction.