Short-Term Forecasting of Energy Consumption of Small Settlements in the Far North

This paper examines and analyzes various methods for short-term forecasting of electricity and power consumption for small power systems. Various algorithms such as ARIMA, ETS, TBATS and Naive Forecaster were used to forecast the energy consumption. The calculations were performed using data on energy consumption in small settlements in the Far North. The analysis of time series characteristics was performed. Evaluation metrics such as MSE, RMSE, MAE, MAPE and R2 were used to evaluate the forecasting quality. Forecasting algorithms were implemented using different forecasting methods and compared.