Fault tree based fault diagnostics methodology for an aircraft fuel system

improved by reducing the time taken to restore systems to the working state when faults occur. The fault identification process can be a significant proportion of the time taken in the repair process. Having diagnosed the problem the restoration of the system back to its fully functioning condition can then take place. This paper describes the development of a fault diagnostic methodology for an aircraft fuel system. The approach takes into account the dynamics of the system. Using sensors installed to provide information about the current status of certain critical parameters. The information produced for these parameters are then categorised into different trend types using a simple pattern recognition technique. Non-coherent fault trees are then used to identify all possible causes of the observed sensor reading trends. By combining the information provided from all sensors the causal faults can be detected. The approach presented has been developed and tested for small demonstration systems – this paper describes how it has been scaled up for a larger, more representative system and the issues that have been overcome in doing this. The system used exists as an experimental facility where the procedure developed can now be fully tested.