A Comparative Study of the Performance of Machine Learning based Load Forecasting Methods

Abstract

A significant aspect of power system control is short-term power load forecasting. It is the basis for the preparation of network systems, the exchange of energy and load scheduling. The accuracy of power load forecasting is directly linked to the security, stability and economic activity of the power system. In this paper, short-term load forecasting methods based on machine learning are studied and evaluated, and the efficiency of the load forecasting methods based on machine learning is compared to a conventional load forecasting approach widely used.