Integrated Building Heating, Cooling and Ventilation Control

Current research studies show that building heating, cooling and ventilation energy consumption account for nearly 40% of the total building energy use in the U.S. The potential for saving energy through building control systems varies from 5% to 20% based on recent market surveys. In addition, building control affects environmental performances such as thermal, visual, air quality, etc., and occupancy such as working productivity and comfort. Building control has been proven to be important both in design and operation stages. Building control design and operation need consistent and reliable static and dynamic information from multiple resources. Static information includes building geometry, construction and HVAC equipment. Dynamic information includes zone environmental performance, occupancy and outside weather information during operation.. At the same time, model-based predicted control can help to optimize energy use while maintaining indoor set-point temperature when occupied. Unfortunately, several issues in the current approach of building control design and operation impede achieving this goal. These issues include: a) dynamic information data such as real-time on-site weather (e.g., temperature, wind speed and solar radiation) and occupancy (number of occupants and occupancy duration in the space) are not readily available; b) a comprehensive building energy model is not fully integrated into advanced control for accuracy and robustness; c) real-time implementation of indoor air temperature control are rare. This dissertation aims to investigate and solve these issues based on an integrated building control approach. This dissertation introduces and illustrates a method for integrated building heating, cooling and ventilation control to reduce energy consumption and maintain indoor temperature set-point, based on the prediction of occupant behavior patterns and weather conditions. Advanced machine learning methods including Adaptive Gaussian Process, Hidden Markov Model, Episode Discovery and Semi-Markov Model are modified and implemented into this dissertation. A nonlinear Model Predictive Control (NMPC) is designed and implemented in real-time based on Dynamic Programming. The experiment test-bed is setup in the Solar Decathlon House (2005), with over 100 sensor points measuring indoor environmental parameters such as temperature, relative humidity, CO2, lighting, motion and acoustics, and power consumption for electrical plugs, HVAC and lighting. The outdoor environmental parameters, such as temperature, relative humidity, CO2, global horizontal solar radiation and wind speed, are measured by the on-site weather station. The designed controller is implemented through LabVIEW. The experiments are carried out for two continuous months in the heating season and for a week in cooling season. The results show that there is a 26% measured energy reduction in the heating season compared with the scheduled temperature set-points, and 17.8% energy reduction in the cooling season. Further simulation-based results show that with tighter building façade, the cooling energy reduction could reach 20%. Overall, the heating, cooling and ventilation energy reduction could reach nearly 50% based on this integrated control approach for the entire heating/cooling testing periods compared to the conventional scheduled temperature set-point.