Open AccessAn Incremental Deep Model For Computing Electrical Power Load Forecasting Based Social Factors
Author(s) -
Waseem Alromema,
Ahmed Alahmadi,
Salman Zakariay
Publication year - 2020
Publication title -
international journal of information systems and computer sciences
Language(s) - English
Resource type - Journals
ISSN - 2319-7595
DOI - 10.30534/ijiscs/2020/02962020
Subject(s) - computer science , electricity , deep learning , electrical load , term (time) , load shedding , electric power system , demand forecasting , electric power , industrial engineering , heap (data structure) , architecture , artificial intelligence , operations research , power (physics) , engineering , voltage , electrical engineering , algorithm , art , physics , quantum mechanics , visual arts
Load forecasting (LF) is critical for guaranteeing adequate limit and controlling of the power business in numerous nations, which theeconomies dependingon electricity. Its production (load) and consumption (demand) have to be in equilibrium at all times since storing electricity, in a considerable quantity, results in high costs. Therefore, the forecasting of the electrical load problem in many countries become crucial and critical in the recent years. In this paper, a novel deep model architecture for LFintroduced, which integrates the features of dataset in discovering the most influent factors affecting electrical load usage. In addition, different LF strategies introduced and their interrelations just asthe intensity of neural organizations to rough the heap estimating. The deep model is based on in three terms time: Long-term (yearly), Mid-term (Monthly), and Mid-term (Weekly), which can possibly provide interrelated deep learning models. Moreover, to generating more accurate predictions based the hierarchal learning architecture. The dataset used is introduced in the case study, which is power load in Giga-watt from years 2006 to 2015. The load forecasted for the year 2016 and is validated to check its accuracy