Exploratory Weather Data Analysis for Electricity Load Forecasting Using SVM and GRNN, Case Study in Bali, Indonesia

Accurate forecasting of electricity load is essential for companies, primarily planning generators. Overestimated or underestimated value may lead to inefficiency generator deficiency in the grid system. Parameters that affect demand are weather conditions at location In this paper, we investigate possible parameters load. As a case study, choose an area with isolated system, i.e., Bali Island, Indonesia. We calculate correlations various during period 2018–2019. use two machine learning models design Generalized Regression Neural Network (GRNN) and Support Vector Machine (SVM), using features from parameters. scenarios add one-by-one which The results show parameter highest correlation temperature, then followed by sun radiation wind speed parameter. obtain best prediction GRNN SVR coefficient 0.95 0.965, respectively.