Improving the performance of machine learning models by integrating partly physical control response models in short-term forecasting of aggregated power system loads

Keyphrases

• Physical Control 100%
• Short-term Forecasting 100%
• Aggregated Power 100%
• Power System Load 100%
• Temperature Variation 100%
• Machine Learning Models 100%
• Control Response 100%
• Response Model 100%
• Machine Learning 50%
• Modeling Approach 50%
• Prediction Accuracy 50%
• Hybrid Method 50%
• Extreme Weather Conditions 50%
• Measure Data 50%
• System Dynamics 50%
• Model Integration 50%
• Forecasting Power 50%
• Electrical Load 50%
• Thermal Storage Capacity 50%
• Thermal Dynamics 50%
• Load Forecast 50%
• Control Actions 50%
• Control Load 50%
• Physically-based 50%
• Data-driven Model 50%
• Integration Approach 50%
• Outdoor Temperature 50%
• Dynamic Control 50%
• Trial Data 50%

INIS

• performance 100%
• control 100%
• machine learning 100%
• power systems 100%
• forecasting 100%
• dynamics 75%
• data 50%
• accuracy 50%
• comparative evaluations 25%
• information 25%
• hybrids 25%
• thermal storage 25%
• power 25%
• modeling 25%
• variations 25%
• saturation 25%
• capacity 25%
• coverings 25%
• weather 25%
• outdoors 25%
• houses 25%

Engineering

• Model Response 100%
• Power Engineering 100%
• Learning System 100%
• Thermal Energy Storage 33%
• Field Trial 33%
• Model Integration 33%
• Storage Capacity 33%
• Electrical Load 33%
• Load Forecast 33%
• Integration Approach 33%
• Outdoor Temperature 33%
• Measurement Data 33%
• Control Action 33%
• Systems Dynamics 33%
• Hybrid Method 33%
• Information Source 33%