Short-term Load Forecasting using Bootstrap Aggregation based Ensemble Method

This paper deals with the study of a prediction algorithm based on bootstrap aggregation technique applied to short term load forecasting problem. Short term Load forecasting is a crucial task in power system studies as it forms the basis for several analysis that include load flow analysis, contingency planning, financial planning, energy management, load scheduling etc. Thus, in this work authors investigate the performance of Bootstrap algorithm in load forecasting studies. Bootstrap aggregation technique works on the principle of creation of several predictors from the available data. These predictors independently generate their responses which are then aggregated to finally obtain the result. The performance of the algorithm is studied using daily load profile data of Bengaluru city in India. The ensemble result derived here indicates that the technique is highly promising and must be further explored for load forecasting studies.