Power Consumption Prediction via Improved Gaussian Mixture Clustering Model for Automatically Clustering

Abstract

Power consumption is a very important factor in smart grids for load management process. Forecasting electricity consumption is the first step in dealing with load management. Aiming at the energy consumption prediction of power time series, this paper proposed an energy consumption prediction approach based on improved Gaussian mixture clustering. In this study, Gaussian mixture model clustering is used to group the characteristics of power time series, and the proposed clustering fitness evaluation is used to determine the number of clusters automatically. The results showed the accuracy and efficiency of the model and the ability to compete with current techniques for forecasting electricity consumption based on the use of the mean absolute error (MAE) to measure the accuracy of the prediction, as the model was able to achieve better predicting results.