Research On Transformer District User Identification Based On Power Meter Reading Data

Among many assessment indexes of power grid management,the line loss rate of low-voltage distribution network has become an important reference standard.Due to the transformation of low-voltage distribution network and other reasons,there is often a real situation that the actual relationship between the customer and transformer and the customer-transformer connectivity relationship recorded in the system files is not consistent.At present,the power grid company mostly uses manual on-site line inspection to check the relationship between the customer and transformer.In the case of complex lines or hidden lines,it will use the solution of applying for power failure troubleshooting.In order to solve the problem that it is difficult to identify the relationship between customer and transformer in low-voltage distribution network,this paper proposes a method based on constrained least square and support vector machine to identify the relationship,which is applied to intelligent transformer district management.On the basis of no auxiliary equipment installed on site,data mining is carried out by using the power consumption data of the power consumption information collection system to complete the real-time detection of the line loss rate in the transformer district and effectively identify the customer-transformer relationship in the transformer district with relatively less effective data when necessary.The main contents of this paper are as follows:The mathematical model and the network model of the relationship between the customer and the transformer are established;through the study of the relationship between the user power consumption and the line loss in the transformer district,the method of reasonably distributing the fixed loss energy of the transformer district to the power consumption of each user is proposed;the correlation between the user power consumption data and the energy consumed by users in the same transformer district are analyzed.A customer-transformer connectivity relationship identification method based on constrained least square and support vector machine is proposed.According to the mathematical model and data characteristics of the customer-transformer connectivity relationship in the low-voltage distribution network,from the perspective of "0-1 linear integer programming",the applicability of enumeration,implicit enumeration,simulated annealing,genetic algorithm,particle swarm optimization and mixed integer linear programming in solving the problem of relationship identification in the transformer district is discussed through experiments;According to the network model and data characteristics,by relaxing the constraints of 0-1 integer programming,considering the interaction between the energy consumed by users in the same district,a relationship identification method based on constrained least square and support vector machine is proposed.In Linux environment,the customer-transformer connectivity relationship identification method based on constrained least square and support vector machine is implemented by C++ language.Aiming at the problem of data missing,the paper explains the reason and advantage of using average interpolation method through experiment comparison;uses effective set method to realize the application of constrained least square in transformer district user identification;uses libsvm toolkit to realize the training classification of support vector machine;uses simulation data experiment to illustrate the selection skill of support vector machine kernel function;the field data shows the availability of the method.