Determination of Forecast Indicators of Electricity Quality in Mode of Synchronized Vector Measurements

Abstract

The work is devoted to the development of software for forecasting the quality of electricity in an automated system for diagnosing the quality of electricity consumers using cloud technologies. The existing domestic and foreign methods for monitoring the quality of electricity using the technology of synchronized vector measurements are analyzed. The structural scheme of the technology of diagnostics of electricity quality as a new direction at the junction of sciences – information technologies and energy is developed. Based on the experimental data of electricity quality indicators obtained from the synchrophasor, an array of data (dataset) was formed for further processing. Two statistical methods were chosen to study the data and forecast the indicators of electricity quality – the nearest neighbors and ridge regression. With the help of standard Phyton programming language libraries, reading and primary data processing, plotting, statistical processing and implementation of forecasting models were performed. The analysis of the obtained forecast graphs is performed and it is concluded that according to the normalized data the accuracy of the Ridge regression model is higher by 10-15%. The WEB-interface of the system for interactive interaction and visualization of indicators with the output of tables and graphs for analysis, graphical representation and display of the results of diagnostics of electricity quality is designed and developed.