Abstract
The air-conditioning systems of the buildings have been diversified in recent years, and the complexity of the system has been increased. At the same time, safety and stability of the system and the low-running cost are demanded. To respond these demands, various researches have been done. The developments of the fault detection and diagnosis systems have received greater attention to make the air-conditioning system more efficient. Most of those systems require the faults pattern to detect the faults. However, the dynamical systems are not easy to analyze and nor to collect the fault patterns in advance. We propose a model simulation of the air-conditioning system to collect the fault patterns and a real time faults diagnosis system using a recurrent type neural network (RNN).
