Based on the failure mode and effects analysis (FMEA) of the main parts in the air conditioning system, this paper builds the fault model of the air conditioning system by the hierarchy multi-signal flow method. According to the failure analysis, the fault-test dependency matrix for multi-signal flow of air conditioning system is given out. Moreover, fault detection rate (FDR) and fault isolation rate (FIR) of each component are carried out. It’s found that the FDR and FIR, especially the latter parameter, are low. To improve their values, the model is improved by including additional five measurement points. By this way, the FDR is improved from 91.4% to 100%, and the FIR is improved from 32.9% to 83.9%. The results show that the hierarchy multi-signal flow can be used to improve the FDR and FIR, increase the fault diagnosis efficiency, and guide the design of the health management of aircraft air conditioning.