Electricity demand forecasting in Ambon using machine learning techniques

Abstract

This study aims to analyze the impact of electrical load forecasting using Artificial Neural Networks (ANN) to improve power supply reliability and efficiency in Ambon’s electric system. The objective is to develop a reliable forecasting model that supports effective energy management, helping to achieve operational excellence in terms of quality, safety, and cost-efficiency. A quantitative approach was utilized, gathering historical electricity load data from 2019 to 2024, alongside relevant environmental and temporal factors. The data were analyzed using ANN within a Python-based framework to predict future electricity demands accurately. The study employs a structured equation modeling to validate the forecasting model and its components. The findings reveal that the ANN model effectively predicts electrical loads with high accuracy, demonstrating substantial improvements in operational efficiency and energy cost reductions. The model’s ability to incorporate multiple input variables allows for nuanced understanding and prediction of load variations, thereby facilitating better resource allocation and strategic planning. This research contributes uniquely by applying ANN for electrical load forecasting in the context of Ambon’s electrical system, underscoring the integration of AI techniques in improving the operational efficiency of power utilities. The study extends the knowledge on the application of machine learning in the power sector by demonstrating how sophisticated forecasting models can significantly enhance energy management strategies