Ventilation control strategy using low-dimensional linear ventilation models and artificial neural network

Directly using CFD (Computational Fluid Dynamics) for ventilation prediction and control is time-consuming and data-storage-expensive. Construction of linear system is of great importance for ventilation online control. Following with previous construction of linear ventilation models (LVM), in this work we continued to develop a new reliable model with the methodology of LLVM (Low-dimensional Linear Ventilation Models)-based ANN (artificial neural network) aiming at ventilation online control based on indoor pollutants response. A large database was firstly built using CFD simulations considering different ventilation modes, ACHs (air change rates per hour) and individual pollutant sources. Corresponding experiments were also conducted for validations. For the sake of rapid prediction, CFD-validated LVM (Linear Ventilation Models) method was used. Next, LLVM method was utilized for reconstruction and expansion of CFD database. Then, the well-trained LLVM-based ANN was employed to predict indoor CO2 concentration levels. An evaluation index (Er) was defined to assess performance of different ventilation modes, considering weighting factors for both ventilation rates and CO2 concentration (i.e., 0.45 and 0.55). It is found that a single ventilation mode was not efficient for all scenarios, e.g., different pollutant source strength and positions. With ventilation control using LLVM-based ANN model, both indoor pollutant concentration and energy consumption were largely reduced (up to 30% and 50%) when compared with not using ventilation control. This work can potentially provide an efficient and effective strategy for ventilation online control of indoor environment in the perspectives of both health and energy efficiency.