Analysis of power consumption monitoring data and identification of excess power consumption based on the proposed algorithm

Abstract

In the framework of this work, an algorithm for detecting overconsumption of electric energy in low-voltage networks is proposed. The research is based on a systematic approach based on a set of methods for the intelligent collection and processing of information. Data Mining technology is used to extract, compress, sample, analyze and present data. The method of expert assessments determines the fundamental criteria for influencing the process under study. The data array is processed by dividing it into clusters. An analysis of the process under study using the decision tree method based on the IBM SPSS Statistics statistical package is proposed. In the course of the work done, elements that meet the established selection criteria were extracted from a large array of data. As a result, an algorithm is obtained that is applicable to the analysis of behavioral electricity consumption, allowing you to track quantitative and qualitative indicators of overconsumption of electricity during the production activity of the object under study, at a selected time interval. The use of this algorithm makes it possible to optimize the process of solving the problem of uncontrolled use of energy resources by identifying illegal business activities, in particular cryptocurrency mining.