Load Forecasting and Load Prediction Using Neural Network

Electric load forecasting is essential to energy management because it enables effective planning and resource utilization. Future electricity consumption forecasts must be precise. In this research, the effectiveness of ARIMA and neural networks— two widely used load forecasting techniques was evaluated. In order to get better outcomes, a hybrid model is suggested that combines ARIMA with an ANN. Three Statistical parameters—Precision, Recall, and F-Measure— along with the anticipated load served as the basis for the assessment. The results showed that the suggested model performed better in terms of prediction accuracy and error reduction than both ARIMA alone and ARIMA + Naive Bayes. In particular, compared to ARIMA alone and ARIMA plus Naive Bayes, the suggested model significantly reduced predicting error by 95.92% and 96.75%, respectively. These results show that the suggested model is a good option for estimating electric demand and may significantly improve energy management and planning.