Multi-step ahead forecasting for electric power load using an ensemble model

作者：

Highlights：

• The proposed model is used for multistep prediction of univariate electric power.

• The proposed model has a two-layer structure.

• The experiments are conducted with different prediction horizons.

• The best forecasting performance is obtained by the proposed model.

• The proposed model has high robustness compared to other models.

摘要

•The proposed model is used for multistep prediction of univariate electric power.•The proposed model has a two-layer structure.•The experiments are conducted with different prediction horizons.•The best forecasting performance is obtained by the proposed model.•The proposed model has high robustness compared to other models.

论文关键词：Multi-step prediction,Electric power load,Ensemble models,NeuralProphet,Lightgbm

论文评审过程：Received 27 June 2022, Revised 17 August 2022, Accepted 19 August 2022, Available online 24 August 2022, Version of Record 5 September 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118649