Fault-toleration And Diagnose On Multi-sources Information Flow Orient Spaceflight Appilication

Fault diagnosis and tolerance of a system is a comprehensive technique involving in many disciplines and an important approach to the guarantee of system’s security and reliability. The continuous development of fault diagnosis and tolerance technique for different applications leads to the prosperity of relative theories and methods. Aerospace system is a complex of various but relative sub-systems. How to guarantee the normal and reliable work of a system by fault diagnosis is an important problem. Considering that the system is under remote control and in autonomous work, autonomous reconfiguration and recovery are as important as the former. Thus, an exploration to fault diagnosis and tolerance is highly valued in both theoretical research and practical application. The results of the research are listed below.With a further study on the multi-sources information flow of a distributed test system, this paper suggests a correlative fault diagnosis model based on the integrated information platform. The relative information about capability parameter, test requirements, test number and expert knowledge is classified and reorganized, and then, it is followed by the proposal of integrating all the system information in a diagnosis platform. For better adaptation to different types of data of information sources, different packaged database and corresponding data interfaces are designed. The specially-designed fault diagnosis procedure can be a guarantee to the operationalization of the fault diagnosis based on multi-sources information flow.With a further research on the topology structure of a distributed test and control network, based on the exclusiveness of the shortest distance from a node to any other one, the concept of the minimum fault sensor is established. What follows is a needed confirmation to prove that minimum fault sensor finds faults in shortest path. Thus, the fault diagnosis based on minimum fault sensor and the arithmetic of minimum fault sensor are proposed. Last, the effective application of this method is illustrated.The variation of connection when nodes go wrong is analyzed. The reconfiguration method which takes the minimum variation of network average degree as its terminal aim and the application of the above-mentioned method are suggested. The conclusive result in this part is that the suggested method is applicable to aerospace test network for it can effectively recover a network’s former connection even if the original structure of the net can not be recovered.The relationship between the fault rate of a network and its tolerance rate is explored. The concept of functional forwarding index is established. And the characteristics of functional forwarding index are listed in details. Based on this functional forwarding index, a kind of fault diagnosis and tolerance are suggested. Furthermore, its application is analyzed.In a whole, this paper analyzes the application and effectiveness of fault diagnosis and tolerance based on multi-sources information flow. Besides, it suggests the working law of fault diagnosis and tolerance in a network based on the integration of multi-sources information flows. Last, the approach to the realization of fault diagnosis and tolerance with the relative technical approaches is proposed.