Implication of Diverse Modalities for Electrical Load Forecasting

Abstract

The extensive research and development over aeras have proposed numerous techniques for load forecasting (L.F). The major targeted area was either the grid, commercial or residential consumers but the industrial sector was not widely explored; specially the spaciousness of the grid isolated industries or independent power plants (IPP’s). The paper intent to investigates the L.F methods and highlight their flaws, comprehensive review of their limitations and existing potential challenges to associated with electrical L.F. The study furthers three potential research questions and respective objectives attained after extensive quality assessment criteria for inclusion or exclusion of literature. However, in electrical load forecasting, the choice of parameters and model selection criteria leads to an optimization challenge. The methods proposed with trial and error combined with expert knowledge to optimize the input parameters have been widely used in past. Such custom-made approaches are difficult to be considered in the isolated industries or IPP’s due to their nature of operations. Also, the renewable energy integration with conventional grids has commended towards more precise and accurate L.F. This paper presents a comprehensive review of forecasting techniques for electrical load forecasting and examine the models. Additionally, the research gaps are also discussed.