IMPROVING ELECTRICITY DEMAND FORECASTING THROUGH HYBRID NEURAL NETWORKS AND META-HEURISTICS: A CASE STUDY IN IRAN

A stable and reliable electricity supply is indispensable for various sectors, including industries, households, schools, hospitals, and more. Precise electricity demand forecasting plays a pivotal role in enabling energy producers and distributors to make informed decisions regarding energy generation and distribution, ensuring the stability of the electricity grid. This research addressed the challenge of electricity consumption prediction by harnessing the power of artificial intelligence techniques, specifically hybrid neural networks, synergized with advanced meta-heuristic algorithms, including the shuffled frog leaping algorithm (SFLA), gray wolf optimizer (GWO), and genetic algorithm (GA). Over a decade-long period, we conducted a comprehensive study encompassing industrial, household, and agricultural sectors in Iran. The results of our investigation, evaluated using the coefficient of determination (R2), revealed that the hybrid neural networks coupled with the GWO algorithm exhibited superior predictive performance, particularly in the household and industrial sectors. This innovative approach not only provided more accurate electricity consumption predictions but also furnished valuable insights to empower decision-makers in effective energy management. In summary, our research pioneered the fusion of hybrid neural networks with meta-heuristic algorithms, presenting a novel methodology for electricity consumption prediction. This approach not only enhanced forecast accuracy but also contributed to the advancement of intelligent energy resource management systems, which were vital for the sustainable development of energy sectors worldwide.