Development and application of linear ventilation and temperature models for indoor environmental prediction and HVAC systems control

Abstract It has been of great importance to develop better control techniques for heating, ventilation and air conditioning (HVAC) systems to provide occupancy driven energy and comfort management due to the significance of building energy conservation. Following with our previous work, where low-dimensional linear ventilation model (LLVM) and artificial neural network (ANN) were incorporated to realize online control of indoor air quality (IAQ), we continued to expand the work on indoor thermal comfort (ITC) with low-dimensional linear temperature model (LLTM) and contribution ratio of indoor climate (CRI) to provide dependable support for HVAC online control. Two steps were required to be implemented, respectively considering pollutant and temperature responses. As a premise for control, the database was constructed by CFD, which was verified by the corresponding experiments. Linear ventilation model (LVM) and linear temperature model (LTM) could be well employed to expand the CFD database. Then, combined with satisfying ANN and CRI along with low-dimensional linear model (LLM) for database reconstruction, LLVM-based ANN and LLTM-based CRI could rapidly predict the distribution of indoor environmental parameters (e.g., CO2 concentration and temperature) within acceptable errors. Next, the comprehensive evaluation indices were defined to provide weighting factors for indoor environment (IAQ and ITC) and air conditioning energy. By using the current control strategy, the HVAC energy consumption resulting from ventilation and air conditioning loads could be significantly decreased up to 50% and 32% respectively. This study will further promote an intelligent control strategy to improve the deteriorated condition between occupant comfort and HVAC energy conservation.