Author(s)

Dolores De Groff, Perambur Neelakanta

Department of Electrical Engineering and Computer Science, Florida Atlantic University, Boca Raton, FL, USA.

This paper proposes a unique approach to load forecasting using a fast convergent artificial neural network (ANN) and is driven by the critical need for power system planning. The Mazoon Electrical Company in Oman provided the real data for the study of monthly load forecasting using ANNs, which are presented in this paper. The link between past, present, and future temperatures, loads, and humidities is learned by the artificial neural network (ANN). The test ANN predicts reasonably accurate results of predicted power loads. The underlying exercise uses a traditional multilayer ANN architecture with feed-forward and backpropagation techniques in addition to a recently proposed fast-convergence algorithm that is deduced in terms of eigenvalues of a Hessian matrix associated with the input data of temperature and humidity changing over time. The anticipated results are cross verified with actual power load data obtained.

Load Forecasting, Artificial Neural Network, Backpropagation Algorithm, Eigenvalues, Fast Learning Rate, Power System

Groff, D. and Neelakanta, P. (2025) Artificial Neural Network-Based Electric Load Forecasting. Journal of Computer and Communications, 13, 150-159. doi: 10.4236/jcc.2025.136010.