No abstract

Forecasting hourly spot prices for real-time electricity markets is a key activity in economic and energy trading operations. This paper proposes a novel two-stage approach that uses a combination of Auto-Regressive Integrated Moving Average (ARIMA) with other forecasting models to improve residual errors in predicting the hourly spot prices. In Stage-1, the day-ahead price is forecasted using ARIMA and then the resulting residuals are fed to another forecasting method in Stage-2. This approach was successfully tested using datasets from the Iberian electricity market with duration periods ranging from one-week to ninety days for variables such as price, load and temperature. A comprehensive set of 17 variables were included in the proposed model to predict the day-ahead electricity price. The Mean Absolute Percentage Error (MAPE) results indicate that ARIMA-GLM combination performs better for longer duration periods, while ARIMA-SVM combination performs better for shorter duration periods.

Accurate forecast of the hourly spot price of electricity plays a vital role in energy trading decisions. However, due to the complex nature of the power system, coupled with the involvement of multi-variable, the spot prices are volatile and often difficult to forecast. Traditional statistical models have limitations in improving forecasting accuracies and reliably quantifying the spot electricity price under uncertain market conditions. This paper presents a hybrid model that combines the results from multiple linear regression (MLR) model with an auto-regressive integrated moving average (ARIMA) and Holt-Winters models for better forecasts. The proposed method is tested for the Iberian electricity market data set by forecasting the hourly day-ahead spot price with dataset duration of 7, 14, 30, 90, and 180 days. The results indicate that the hybrid model outperforms the benchmark models and offers promising results under most of the testing scenarios.INDEX TERMS ARIMA, energy price, forecasting, Holt-Winters, hybrid model and regression.

No abstract

No abstract