Abstract:
In recent years, the loads needing electrical power in military aircraft and civil jet
keep increasing, this put huge pressure on the electrical power system (EPS).
As EPS becomes more powerful and complex, its reliability and maintenance
becomes difficult problems to designers, manufacturers and customers. To
improve the mission reliability and reduce life cycle cost, the EPS needs health
management.
This thesis developed a set of generic health management methods for the EPS,
which can monitor system status; diagnose faults/failures in component level
correctly and predict impending faults/failures exactly and predict remaining
useful life of critical components precisely. The writer compared a few
diagnostic and prognostic approaches in detail, and then found suitable ones for
EPS. Then the major components and key parameters needed to be monitored
are obtained, after function hazard analysis and failure modes effects analysis
of EPS. A diagnostic process is applied to EPS using Dynamic Case-based
Reasoning approach, whilst hybrid prognostic methods are suggested to the
system. After that, Diagnostic, Prognostic and Health Management architecture
of EPS is built up in system level based on diagnostic and prognostic process.
Finally, qualitative evaluations of DPHM explain given.
This research is an extension of group design project (GDP) work, the GDP
report is arranged in the Appendix A.
