What question did this study set out to answer?

The aim is to review and synthesize machine learning and deep learning models for forecasting electrical loads in the industrial sector.

January 23, 2026EnergiesOpen Access

Electrical Load Forecasting in the Industrial Sector: A Literature Review of Machine Learning Models and Architectures for Grid Planning

View Full Paper

Ask AI

Bookmark

Authors

Systematic literature review highlights hybrid models optimizing electrical load forecasting, suggesting enhanced planning approaches.

The aim is to review and synthesize machine learning and deep learning models for forecasting electrical loads in the industrial sector.
Conducted a systematic literature review following PRISMA guidelines.
Analyzed hybrid architectures incorporating statistical techniques, ML, and DL strategies.
Evaluated model performance using statistical measures like MAE, RMSE, and MAPE.
Proposed decoupling output length predictions from core signal forecasting.
Hybrid models combining LSTM and optimization algorithms showed superior performance over conventional methods.
Complex architectures effectively captured non-linear relationships and local trends in data.
Identified a need for tailored metrics for comprehensive, application-based performance assessment.

Eckhoff et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

January 23, 2026EnergiesOpen Access

View Full Paper

Ask AI

Bookmark

Authors

Systematic literature review highlights hybrid models optimizing electrical load forecasting, suggesting enhanced planning approaches.

The aim is to review and synthesize machine learning and deep learning models for forecasting electrical loads in the industrial sector.
Conducted a systematic literature review following PRISMA guidelines.
Analyzed hybrid architectures incorporating statistical techniques, ML, and DL strategies.
Evaluated model performance using statistical measures like MAE, RMSE, and MAPE.
Proposed decoupling output length predictions from core signal forecasting.
Hybrid models combining LSTM and optimization algorithms showed superior performance over conventional methods.
Complex architectures effectively captured non-linear relationships and local trends in data.
Identified a need for tailored metrics for comprehensive, application-based performance assessment.

Eckhoff et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark