Research On Fault Samples Selection And Fault Injection Method For Flight Control Systems

The flight control system is one of the most important subsystems of the aircraft.The interconnection and interaction of various subsystems,especially the rapid development of fly-by-wire control system based on the application of digital technology in the aircraft,lead to a comprehensive feature of the aircraft system function.This makes the problem of fault diagnosis and the reliability evaluation of the system more complicated.Thus,fault injection method is necessary to ensure flight safety and quality.This article takes the flight control system as the research object,systematically established the signed directed graph model and analyzed the fault propagation mechanism.A quantitative sign directed graph model is proposed considering the limitation of the qualitative SDG model.We use the sample which is selected by the designed method of the fault sample optimization based on qualitative/quantitative model to carry out the fault injection experiments,evaluating the system through the experimental fault injection.Firstly,the concept and classification of the fault samples optimization and fault injection method are introduced.Then the research status and development trend of the flight control technology are given,which expound the significance of the research.Then a special modeling approach of the mechatronic system under the introduced concepts related to the improved SDG model is put forward.The optimization strategy as well as the double layered model is used to deal with the branch redundancy and indistinct hierarchy of the SDG model.Secondly,three fault selection methods are proposed based on the SDG model in this paper.First of all,the fault optimization algorithm is put forward based on the fault propagation mechanism in which the node importance expresses a significant role.Then,the quantitative information of the nodes and edges is added to the model,based on which a new fault optimization algorithm is proposed to solve the inaccurate model reasoning.At last,a new method associated hierarchical and quantitative SDG is proposed for the large complex systems.Finally,the proposed injection method which uses the software simulation to present the hardware fault is experimentalized in the embedded system.The results of injection experiments validate the efficiency of the method.Through the analysis of the experimental results and the collected data,the reliability and the fault propagation probability of the system are evaluated.