Research On Characteristic Analysis Of Electricity Consumption Of Special Transformer Users And Identification Of Electricity Theft

The accuracy of electric energy measurement and quality reliability directly affect the economic benefits of users and the utilization of social energy.Theft of electricity refers to the illegal use of electric power resources.This practice has seriously affected the accuracy of measurement.It not only caused great losses to electric power companies,but also seriously threatened the safe operation of the power grid.With the continuous escalation of anti-tampering methods,the means of stealing electricity are also changing,and they are developing in a hidden and high-tech direction.Ordinary electricity inspection personnel are difficult to identify by the factors such as personal ability.It is an urgent technical problem to accurately resolve the measurement anomalies caused by the illegal use of electricity by users.In order to solve the above problems,this paper proposes an intelligent algorithm based stealing detection model,which fully considers the user's power load and power consumption characteristics.Firstly,analyze the user load data based on clustering algorithm,according to the load data,the users are divided into different categories and different daily load characteristic curves are calculated,match the user's daily load curve with the user's category characteristic curve to find the suspected abnormal power customer.Finally,the abnormal user is further identified based on the user's power consumption parameter to determine the final suspected power stealing user.The main research content of this article is as follows:(1)Based on the related technology of cluster analysis and BP neural network,a steal detection model based on load,voltage,current,and other measurement data is proposed.The model was verified by using the special user measurement data of a city in Yunnan Power Grid.(2)Detailed analysis of the user's classification process based on the user load curve,using the fuzzy C-means algorithm to get the user's load characteristic curve,The status of the user is analyzed by the degree of deviation of the user's load curve from its load characteristic curve,which provides a basis for power stealing discrimination.(3)For the traditional BP neural network algorithm is easy to fall into the local minimum value of the deficiencies,this paper uses the mental evolution algorithm to optimize it.The optimized BP neural network algorithm is applied to the power stealing detection model,and a power stealing identification model based on power parameters is established.Experiments show that compared with the traditional BP neural network,the accuracy of the optimized neural network is higher.This model also provides a new idea for the detection of electricity theft.