• Energy Research

The increasing volatility in electricity markets has reinforced the need for better trading strategies by both sellers and buyers to limit the exposure to losses. Accordingly, this paper proposes an electricity trading strategy based on a mid-term forecast of the average spot price and a risk premium analysis based on this forecast. This strategy can help traders (buyers and sellers) decide whether to trade in the futures market (of varying monthly maturity) or to wait and trade in the spot market. The forecast model consists of an Artificial Neural Network trained with the Long Short Term Memory architecture to predict the average monthly spot prices, using only market price-related data as input variables. Statistical analysis verified the correlation and dependency between variables. The forecast model was trained, validated and tested with price data from the Iberian Electricity Market (MIBEL), in particular the Spanish zone, between January 2015 and August 2019. The last year of this period was reserved for testing the performance of the proposed forecast model and trading strategy. For comparison purposes, the results of a forecasting model trained with the Extreme Learning Machine over the same period are also presented. In addition, the forecasted value of the average monthly spot price was used to perform a risk premium analysis. The results were promising, as they indicated benefits for traders adopting the proposed trading strategy, proving the potential of the forecast model and the risk premium analysis based on this forecast.

The increasing volatility in electricity markets has reinforced the need for better trading strategies by both sellers and buyers to limit the exposure to losses. Accordingly, this paper proposes an electricity trading strategy based on a mid-term forecast of the average spot price and a risk premium analysis based on this forecast. This strategy can help traders (buyers and sellers) decide whether to trade in the futures market (of varying monthly maturity) or to wait and trade in the spot market. The forecast model consists of an Artificial Neural Network trained with the Long Short Term Memory architecture to predict the average monthly spot prices, using only market price-related data as input variables. Statistical analysis verified the correlation and dependency between variables. The forecast model was trained, validated and tested with price data from the Iberian Electricity Market (MIBEL), in particular the Spanish zone, between January 2015 and August 2019. The last year of this period was reserved for testing the performance of the proposed forecast model and trading strategy. For comparison purposes, the results of a forecasting model trained with the Extreme Learning Machine over the same period are also presented. In addition, the forecasted value of the average monthly spot price was used to perform a risk premium analysis. The results were promising, as they indicated benefits for traders adopting the proposed trading strategy, proving the potential of the forecast model and the risk premium analysis based on this forecast.

Short-term load forecasting (STLF) plays a vital role in ensuring the safe, efficient, and economical operation of power systems. Accurate load forecasting provides numerous benefits for power suppliers, such as cost reduction, increased reliability, and informed decision-making. However, STLF is a complex task due to various factors, including non-linear trends, multiple seasonality, variable variance, and significant random interruptions in electricity demand time series. To address these challenges, advanced techniques and models are required. This study focuses on the development of an efficient short-term power load forecasting model using the random forest (RF) algorithm. RF combines regression trees through bagging and random subspace techniques to improve prediction accuracy and reduce model variability. The algorithm constructs a forest of trees using bootstrap samples and selects random feature subsets at each node to enhance diversity. Hyperparameters such as the number of trees, minimum sample leaf size, and maximum features for each split are tuned to optimize forecasting results. The proposed model was tested using historical hourly load data from four transformer substations supplying different campus areas of the University of Beira Interior, Portugal. The training data were from January 2018 to December 2021, while the data from 2022 were used for testing. The results demonstrate the effectiveness of the RF model in forecasting short-term hourly and one day ahead load and its potential to enhance decision-making processes in smart grid operations.

Short-term load forecasting (STLF) plays a vital role in ensuring the safe, efficient, and economical operation of power systems. Accurate load forecasting provides numerous benefits for power suppliers, such as cost reduction, increased reliability, and informed decision-making. However, STLF is a complex task due to various factors, including non-linear trends, multiple seasonality, variable variance, and significant random interruptions in electricity demand time series. To address these challenges, advanced techniques and models are required. This study focuses on the development of an efficient short-term power load forecasting model using the random forest (RF) algorithm. RF combines regression trees through bagging and random subspace techniques to improve prediction accuracy and reduce model variability. The algorithm constructs a forest of trees using bootstrap samples and selects random feature subsets at each node to enhance diversity. Hyperparameters such as the number of trees, minimum sample leaf size, and maximum features for each split are tuned to optimize forecasting results. The proposed model was tested using historical hourly load data from four transformer substations supplying different campus areas of the University of Beira Interior, Portugal. The training data were from January 2018 to December 2021, while the data from 2022 were used for testing. The results demonstrate the effectiveness of the RF model in forecasting short-term hourly and one day ahead load and its potential to enhance decision-making processes in smart grid operations.