2024AIMS MathematicsOpen access

Analysis and forecasting of electricity prices using an improved time series ensemble approach: an application to the Peruvian electricity market

Salvatore Mancha Gonzales, Hasnain Iftikhar, Javier Linkolk López‐Gonzales

Open full text 38 citations

Abstract

In today's electricity markets, accurate electricity price forecasting provides valuable insights for decision-making among participants, ensuring reliable operation of the power system. However, the complex characteristics of electricity price time series hinder accessibility to accurate price forecasting. This study addressed this challenge by introducing a novel approach to predicting prices in the Peruvian electricity market. This approach involved preprocessing the monthly electricity price time series by addressing missing values, stabilizing variance, normalizing data, achieving stationarity, and addressing seasonality issues. After this, six standard base models were employed to model the time series, followed by applying three ensemble models to forecast the filtered electricity price time series. Comparisons were conducted between the predicted and observed electricity prices using mean error accuracy measures, graphical evaluation, and an equal forecasting accuracy statistical test. The results showed that the proposed novel ensemble forecasting approach was an efficient and accurate tool for forecasting monthly electricity prices in the Peruvian electricity market. Moreover, the ensemble models outperformed the results of earlier studies. Finally, while numerous global studies have been conducted from various perspectives, no analysis has been undertaken using an ensemble learning approach to forecast electricity prices for the Peruvian electricity market.

Open-access reader

About this research paper

What this paper is about

In today's electricity markets, accurate electricity price forecasting provides valuable insights for decision-making among participants, ensuring reliable operation of the power system. However, the complex characteristics of electricity price time series hinder accessibility to accurate price forecasting. This study addressed this challenge by introducing a novel approach to predicting prices in the Peruvian electricity market. This approach involved preprocessing the monthly electricity price time series by addressing missing values, stabilizing variance, normalizing data, achieving stationarity, and addressing seasonality issues. After this, six standard base models were employed to model the time series, followed by applying three ensemble models to forecast the filtered electricity price time series. Comparisons were conducted between the predicted and observed electricity prices using mean error accuracy measures, graphical evaluation, and an equal forecasting accuracy statistical test. The results showed that the proposed novel ensemble forecasting approach was an efficient and accurate tool for forecasting monthly electricity prices in the Peruvian electricity market. Moreover, the ensemble models outperformed the results of earlier studies. Finally, while numerous global studies have been conducted from various perspectives, no analysis has been undertaken using an ensemble learning approach to forecast electricity prices for the Peruvian electricity market.

Why it matters

OpenAlex reports 38 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

In today's electricity markets, accurate electricity price forecasting provides valuable insights for decision-making among participants, ensuring reliable operation of the power system. However, the complex characteristics of electricity price time series hinder accessibility to accurate price forecasting. This study addressed this challenge by introducing a novel approach to predicting prices in the Peruvian electricity market. This approach involved preprocessing the monthly electricity price time series by addressing missing values, stabilizing variance, normalizing data, achieving stationarity, and addressing seasonality issues. After this, six standard base models were employed to model the time series, followed by applying three ensemble models to forecast the filtered electricity price time series. Comparisons were conducted between the predicted and observed electricity prices using mean error accuracy measures, graphical evaluation, and an equal forecasting accuracy statistical test. The results showed that the proposed novel ensemble forecasting approach was an efficient and accurate tool for forecasting monthly electricity prices in the Peruvian electricity market. Moreover, the ensemble models outperformed the results of earlier studies. Finally, while numerous global studies have been conducted from various perspectives, no analysis has been undertaken using an ensemble learning approach to forecast electricity prices for the Peruvian electricity market.

Key concepts: Electricity price forecasting, Electricity market, Electricity, Econometrics, Time series, Electricity price, Series (stratigraphy), Ensemble forecasting

Related papers

Back to paper searchBrowse research topicsOriginal source
Analysis and forecasting of electricity prices using an improved time series ensemble approach: an application to the Peruvian electricity market — Research Paper | ScholarLens