Model order reduction of building energy simulation models using a convolutional neural network autoencoder

Keyphrases

• Autoencoder 100%
• Convolutional Neural Network 100%
• Model Order Reduction 100%
• Building Energy Simulation 100%
• Reduced Order Model 60%
• Original Model 40%
• Detailed Model 40%
• Simulation Time 40%
• Low Dimension 40%
• Deep Learning Methods 20%
• Hyperparameters 20%
• Clustering Algorithm 20%
• Complex Geometry 20%
• Model Support 20%
• Time Series Data 20%
• Total Error 20%
• Detailed Simulation 20%
• Energy Performance 20%
• Numerical Optimization 20%
• Energy Demand 20%
• Scenario Development 20%
• Dimension Features 20%
• Energy Behaviour 20%
• Complex Time Series 20%
• Energy Scenarios 20%
• Multi-zone 20%
• Campus Buildings 20%
• Multi-zone Building 20%
• Retrofit Scenarios 20%
• Whole Building 20%
• Model Simulation 20%
• Thermal Zone 20%
• Complex Buildings 20%
• Reduced Simulation 20%
• Unsupervised Deep Neural Network 20%

Engineering

• Autoencoder 100%
• Energy Building 100%
• Simulation Mode 100%
• Convolutional Neural Network 100%
• Reduced Order Model 60%
• Simulation Time 40%
• Detailed Model 40%
• Proof-of-Concept 20%
• Simulation Tool 20%
• Data Series 20%
• Retrofit 20%
• Clustering Algorithm 20%
• Energy Performance 20%
• Numerical Optimization 20%
• Energy Simulation 20%
• Energy Scenario 20%
• Energy Behavior 20%
• Thermal Zone 20%
• Deep Learning Method 20%