Design of a Dynamic Feedback LSTM Electricity Price Forecast of Smart Grids

In this study, a dynamic feedback long short-term memory (DFLSTM) predictive model has been developed for price prediction in the smart grid market. The model’s performance was rigorously evaluated using key performance indicators (KPIs), resulting in metrics, including MAE, MSE, RMSE, and a notably high R2. These outcomes signify the model’s exceptional accuracy and ability to capture intricate price patterns within the smart grid market. The substantial R2 value, in particular, demonstrates the model’s robustness and capacity to elucidate around 92.29% of the variance in market prices.