Publication

Abstract : The key aspect of power-utility firms is load forecasting. Utility companies may save millions of dollars by using load predictions with lower error rates. In this study, forecasting was implemented by applying a seasonal method and a time series method, namely SARIMA and LSTM, and model was evaluated using a 32-year dataset. Electricity consumption is a seasonal dataset with non-stationary nature and SARIMA model was successfully used to forecast the future values of the dataset. In comparison with LSTM, SARIMA produces the best results with the minimum mean square error and root mean square error. Also, a comparative study on wind energy forecasting using LSTM and SARIMA is carried out and the latter one gave better performance.