A short-term marginal price forecasting model based on ensemble learning

Abstract

This The system marginal price reflects the short-term supply and demand of electricity goods in the electricity market, which is an important economic link to the participating members of the market. The traditional prediction model has a large error and low generalization ability to forecast the short-term marginal price. Therefore, this paper proposes an ensemble learning algorithm for short-term marginal price forecasting based on AdaBoost. In this paper, the main factors influencing the short-term marginal electricity price are analyzed. Based on the AdaBoost algorithm, the short-term marginal electricity price forecasting problem is modeled. Four prediction models (C4.5, CART, Linear neural network, BP) are compared, and a short — term marginal price forecasting algorithm is proposed. By comparing the actual values with the predicted values, our proposed algorithm is superior to SVM and BP algorithm, which has high application values in power plant engineering.