With the restructuring of formerly vertically integrated utilities, the energy market behaves like a competitive market, which has resulted in an increased focus on the formulation of forecasting techniques. The contribution of this work is twofold. Firstly, we analyze and evaluate the impact of renewable sources on price forecasts and use them in model training. Next, we propose a bootstrap aggregated-stack generalized architecture for very short-term electricity price forecasting to facilitate market participants in formulating strategies in real time. The stacking phase integrates extreme gradient boosting and random forest, which is then bagged to obtain a computationally efficient model. The final combination of feature engineering and ensemble architecture is observed to outperform the existing techniques.

随着以前垂直整合的公用事业的重组，能源市场的表现就像一个竞争性市场，这导致人们越来越重视预测技术的制定。这项工作的贡献是双重的。首先，我们分析和评估可再生资源对价格预测的影响，并将其用于模型训练中。接下来，我们提出一种用于非常短期电价预测的自举聚合堆栈通用体系结构，以帮助市场参与者实时制定策略。堆叠阶段整合了极端梯度提升和随机森林，然后将其装袋以获得计算效率高的模型。观察到特征工程和集成体系结构的最终组合要优于现有技术。