Research On Satallite Power System Diagnosis Based On Qualitative Model

As the satellite system becomes more and more complicated and automated, the danger of satellite misdiagnosis is more serious. So it is very important to diagnose its fault in time. The Satellite Power System, as an important component of the satellite supplies the power to the whole. The system performance will have a direct impact on the work of the state, so as to affect the life of the satellite. Fault diagnosis technology of the satellite ensures safety and reliability. This is essential and model-based diagnosis is a vital technical branch of it. In this thesis a study is made of the satellite power subsystems with the background of fault diagnosis technology, and makes a comprehensive and in-depth study of the fault diagnosis technology based on a qualitative model.On the basis of analysing the current domestic and international model-based diagnostic techniques, this thesis summarizes the advantages and disadvantages of diagnostic method in the satellite power system applications, and introduces the diagnostic reasoning engine Livingstone and diagnostic tool SMV.This paper analyzes the satellite power distribution subsystems for structural features, construction principles, and the working logic of satellites working in orbit and the practical operation under solar illumination. After analysis of the satellite power system fault model, this thesis summarizes the model design of the main components of which there are several categories of failure mode, and also summarizes the causes leading to failure mode. A detailed analysis of the power control equipment principle, and components, then uses Stanley software modeling. The satellite will be divided into four regional operations, and modeling the corresponding state model in four regions. A detailed description is made of each component model and the relation between input and output, some of the normal working states, and the components, constraint relationships.According to the failure mode analysis, this paper builds a model verifier of each scene as well as a description of the scene statement. Under this scenario model logic reasoning is analyzed through SMV model verifier software. This gets fault diagnosis result, and verifies accurately the FMEA failure mode. By selecting different observation points, it designs the scenes getting the different verification results, compares the componets, state values change after execution. It examines and verifiers the engine diagnostic validity and accuracy of the reasoning knowledge.