| The analysis of power load characteristics has always been a very important research topic.In the context of the lean management of smart grids,the research on the analysis of power load characteristics is gradually refined and deepened.Through non-intrusive load identification,load decomposition is used to analyze the characteristics of the end of the electric load,which helps to improve the accuracy of load forecasting.At the same time,the overall analysis of the traditional regional grid load is of great significance to the safe and economic operation of the power system.This paper studies the non-intrusive load identification algorithm and the load characteristic analysis of traditional large power grids respectively.The main tasks completed are as follows:1)Aiming at the problem that the normalized V-I trajectory characteristics cannot effectively identify loads with similar V-I trajectory shapes but large differences in current amplitudes,this paper proposes a non-invasive load identification algorithm based on neural networks and the composite feature.The composite feature is composed of pixelated V-I trajectory characteristics inspired by image recognition technology and current harmonic characteristics.The algorithm uses a two-layer structure with a total of 3 neural networks to identify loads with the composite feature.The composite feature makes up for the shortcomings of the V-I trajectory characteristics drawn by the normalized values,and improves the effect of load identification.The two-layer neural network structure makes the two component features of the composite feature comlement each other well.This paper uses instance data to evaluate and verify the non-invasive load identification algorithm based on neural networks and the compound feature.2)Combined with the actual line data collected by the substation in a certain area,the load characteristics are analyzed,and a Web application is created to realize the visualization of the load characteristics analysis.The load characteristics of the transformer stations in the line are compared.To find the similarity of the load between different transformer stations in the line,the load characteristics of similar transformer stations are compared. |