A practical deep reinforcement learning framework for multivariate occupant-centric control in buildings

Keyphrases

• Air Conditioning 100%
• Deep Reinforcement Learning (deep RL) 100%
• Heating Ventilation 100%
• Learning-based 100%
• Reinforcement Learning 100%
• Occupancy-based Control 100%
• Reinforcement Learning Framework 100%
• Q-network 100%
• Duel 100%
• Virtual Environment 50%
• Optimal Control 50%
• Modeling Method 50%
• Control Framework 50%
• Heating Control 50%
• Networked Agents 50%
• Control Task 50%
• Human-in-the-loop 50%
• Cooling Energy 50%
• Real-world Implementation 50%
• Building Control 50%
• Reinforcement Learning Algorithm 50%
• Office Space 50%
• Multiple Subsystems 50%
• Action Space 50%
• Learning Agents 50%
• Human Thermal Comfort 50%
• Comfort Modeling 50%
• Personal Comfort 50%
• Multi-dimensional Control 50%
• Occupant Presence 50%
• Thermal Acceptability 50%
• Conventional Reinforcement 50%
• Action Control 50%

Computer Science

• Deep Reinforcement Learning 100%
• Learning Framework 100%
• Reinforcement Learning 100%
• Virtual Environments 33%
• Control Framework 33%
• Learning Agent 33%
• Actual Deployment 33%
• Human-in-the-Loop 33%
• Loop Operation 33%
• Action Control 33%

Engineering

• Reinforcement Learning 100%
• Air Conditioning System 40%
• Energy Engineering 20%
• Tasks 20%
• Optimal Control 20%
• Thermal Comfort 20%
• Virtual Reality 20%
• Action Space 20%
• Dimensional Control 20%