In recent years, the loads needing electrical power in military aircraft and civil jetkeep increasing, this put huge pressure on the electrical power system (EPS).As EPS becomes more powerful and complex, its reliability and maintenancebecomes difficult problems to designers, manufacturers and customers. Toimprove the mission reliability and reduce life cycle cost, the EPS needs healthmanagement.This thesis developed a set of generic health management methods for the EPS,which can monitor system status; diagnose faults/failures in component levelcorrectly and predict impending faults/failures exactly and predict remaininguseful life of critical components precisely. The writer compared a fewdiagnostic and prognostic approaches in detail, and then found suitable ones forEPS. Then the major components and key parameters needed to be monitoredare obtained, after function hazard analysis and failure modes effects analysisof EPS. A diagnostic process is applied to EPS using Dynamic Case-basedReasoning approach, whilst hybrid prognostic methods are suggested to thesystem. After that, Diagnostic, Prognostic and Health Management architectureof 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 GDPreport is arranged in the Appendix A.
展开▼