Electricity Spot Market Price Forecasting by Using LSTM Optimized by ETO Algorithm

As renewable energy resources rapidly develop and electricity markets undergo continuous reforms, the forecasting of electricity spot market price has become increasingly critical. Accurate price forecasting not only aids market participants in formulating more effective trading strategies but also enhances the operational efficiency of power systems. This study proposes a forecasting methodology for electricity spot market price based on Long Short-Term Memory neural networks (LSTM) optimized by Exponential-Trigonometric Optimization (ETO) algorithm to improve prediction accuracy. The LSTM-based forecasting model capable of capturing long-term dependencies within time series data is employed. Meanwhile, the ETO algorithm is employed for the optimization of the hyperparameters of the LSTM model, thereby mitigating overfitting and enhancing the model's adaptability and generalization capabilities. Experimental results demonstrate that the LSTM model optimized by ETO algorithm for electricity spot market price forecasting outperforms compared forecasting methods across performance metrics including MAPE, RMSE and MAE, which verifies the proposed forecasting model is effective and practical in this paper. This research not only provides effective decision-support tools for electricity market participants but also offers new perspectives and methodologies for future studies in electricity price forecasting.