Neural Network-Based Evaluation of Energy Consumption Across Diverse Zones

Power demand is increasing in India day by day. The growth in power demand typically refers to the increasing need for electricity due to various factors, including population growth, economic development, urbanization, and technological advancements. Optimizing efficiency is the key to lowering the use of energy. In this paper, the authors designed a feed-forward neural network model for the evaluation of energy consumption across three zones employing the Levenberg-Marquardt (LM) algorithm. The mean square error (MSE) and R- R-squared (R2) values were evaluated for training, testing, and validation datasets. For the training set, 427336.57 was observed as MSE and 0.9919 as the R2 value. The Mu plot was plotted. Mu is a damping factor that controls the trade-off between the gradient descent and the Gauss-Newton algorithm. It is crucial for energy planning and infrastructure development to meet future energy needs sustainably.