Fault Detection for Air Conditioning System of Civil Aircraft Based on Multivariate State Estimation Technique

Civil aircraft air conditioning system (ACS) is a key system in aircraft, the system conditions monitoring and health management technology is the breakthrough key technical of the domestic civil aircraft operators support. With civil aircraft ACS as the object, combined with ACS existing monitoring parameters system and the route common failure mode, this paper proposes a method of civil aircraft ACS subregion fault detection. Using multivariate state estimation technique (MSET), a region characteristic parameters healthy baseline model of ACS is constructed under normal operation conditions, and carry out online monitoring of the baseline model to achieve the ACS fault detection. The method is validated by using actual quick access recorder (QAR) data provide by airline. Using the proposed method, by examining changes in the baseline model, ACS fault can be effectively detecting before the failure occurs, realize the purpose of online early fault warning of ACS, which provides support for the airline to arrange aircraft maintenance programs reasonably and reduces aircraft downtime.