Artificial Neural Network for short-term load forecasting: An evaluation based on Chhattisgarh State

Load forecasting is a technique of prediction of future occasion, final occasion, final results or trend of load demand based to on historical and probabilistic data. Short term load forecasting has a significant effect at the efficiency of operation. In short term load forecasting, the goal is to estimate the load for the following half an hour up to the next two weeks. In this study artificial neural network (ANN) approach is presented for load forecasting particularly for Chhattisgarh State. A prediction of very next day is shown with different parameters in order to count those parameters which perform well and are able to provide the best results in prediction. The ANN model is trained using an input dataset which consists of previous year data along with a day before data of same date of predicted load as two different features of input dataset. These features further involve both generation and load demand in it It covers the timeframe from 1 JAN 2022 to 30 NOV 2023 for both training and testing method. After training and testing of model, the result gets evaluated through statistics matrices such as MAPE (mean absolute percentage error) and MAE (mean absolute error) in the end, ensuring accuracy and fitness of the model.