A Method of Updating Load Characteristic Database of Typical Consumers Based on Machine Algorithm

Abstract

The daily load curves of the electric power system can reflect the actual electricity consumption characteristics of consumers, so daily load curves of the electric power system is widely used in load modeling. However, due to the complexity of most consumers' power consumption structures. it is difficult to distinguish the typical characteristics. So it is necessary to establish a more accurate load characteristics database of typical consumers to achieve the typical industry classification of different consumers. In order to solve the problem that the current power consumption characteristics are not obvious and it is difficult to select typical consumers, this paper proposes a method of updating typical consumers load characteristics database based on machine algorithm by using the support vector machine algorithm. First of all, read the historical load characteristic database of typical consumers and train the historical daily load curve data. Secondly, comprehensively use the support vector machine algorithm to identify the industry of the new consumers daily load data and determine its typical degree. Finally, eliminate the data with poor characteristics in the database, so as to realize the update of the typical consumers load characteristic database. The results show that the proposed method can improve the cluster quality, realize the database updating and optimization, and truly reflect the power consumption characteristics of consumers in a certain area.