From one building to many: transferability of a deep reinforcement learning agent for optimizing pollutant exposure and energy consumption

Key Points

• The research aims to evaluate how well a deep reinforcement learning (DRL) agent can be applied across multiple buildings to improve energy use and health outcomes.
• Assessment of a deep reinforcement learning agent's capabilities across various building scenarios.
• Analysis of energy consumption and pollutant exposure metrics.
• Evaluation of occupant health improvements associated with optimized settings.
• DRL agent improved energy efficiency in modeled buildings.
• Significant reduction in pollutant exposure documented.
• Transferability of the agent demonstrated scalability for broader applications.

Abstract