Wholesale Electricity Price Forecasting Using Integrated Long-Term Recurrent Convolutional Network Model

Electricity price forecasts have become a fundamental factor affecting the decision-making of all market participants. Extreme volatility has forced participants to hedge against volume risks and movements. Hence, getting an accurate forecast from few hours days ahead is very important challenging due various factors. This paper proposes integrated long-term recurrent convolutional network (ILRCN) model predict electricity prices considering majority contributing attributes as input. The proposed ILRCN combines functionalities neural long short-term memory (LSTM) algorithm along with novel conditional error correction term. combined can identify linear nonlinear behavior within input data. ERCOT wholesale data load profile, temperature, other factors for Houston region been used illustrate model. performance forecasting verified using performance/evaluation metrics like mean absolute accuracy. Case studies reveal that shows highest accuracy efficiency in compared support vector machine (SVM) model, fully connected LSTM traditional LRCN without stage.