Fault-tolerant control for outdoor ventilation air flow rate in buildings based on neural network

This paper describes a supervisory control scheme that adapts to the presence of the measurement faults in outdoor air flow rate control using sensor-based demand-controlled ventilation, maintains an adequate indoor air quality and minimizes the resulting increase in energy consumption. A strategy, which is based on neural network models, is employed to diagnose the measurement faults of outdoor and supply flow sensor, and accomplishes the fault-tolerant control of outdoor air flow when faults occur. The neural network models are trained using the data collected under various normal conditions. The residuals between the measurements of flow sensors and the outputs of the neural network models are used to diagnose the faults. When the fault of outdoor or supply air flow sensor occurs, the recovered estimate of outdoor or supply air flow rate obtained on the basis of the neural network models is used in the feedback control loop to regain the control of outdoor air flow. Tests using dynamic system simulation are conducted to validate the strategy. The control, IAQ and energy performances of the system under fault-tolerant control strategy in the presence of the faults in air flow sensor are also presented.