Forecasting of Daily Energy consumption

The goal

To predict a Daily Peak Load for a Middle Eastern utility company.

What we did

We developed a time series forecasting model. Time series were with multiple seasonality, so appropriate class of forecasting models were used. The final model was chosen based on the performance of the models on the testing dataset. The following statistical methods were used: TBATS, Holts-Winter exponential smoothing, ARIMA models, Fourier decomposition of time series.

• Client: Utilities, Middle East