Partner Institution: Research Fund for International Young Scholars (RFIS)

Project Information: Principal Investigator (PI), 2022.01-2023.12

Project Description: “Smart buildings” refer to buildings that can make intelligent decisions about their energy management. A key component of smart buildings is a “smart brain,” which is an intelligent energy management system capable of making decisions regarding renewable energy availability, heating, ventilation, and air conditioning (HVAC) control, appliance and lighting control, etc. When executing these control decisions, this energy management system should consider other factors such as user behavior and preferences, occupancy schedules, and building characteristics. However, the process of seeking a “smart brain” in “smart buildings” faces a series of basically unresolved challenges. Even the most modern energy management systems cannot implement intelligent decisions in terms of control, demand response, or user behavior. This is because current energy management systems cannot recognize and learn different energy flexibilities of different buildings, and they ignore or do not learn from user behavior. This proposal will utilize learning-based control and reinforcement learning tools to research how to design intelligent energy management programs that can self-optimize and co-evolve with user behavior.