Component-based Failure Mode And Effect Analysis For Aviation Software Systems

With the rapid development of electronic technology, computer technology and software engineering, the aviation software system is becoming more and more important in aviation airborne system, which plays a crucial role in safety and reliability of aviation airborne system. To improve the reliability of the aviation software system, software system designing which can analyze the reliability of the aviation software system attracts more and more emphases. As an important analysis method on security and reliability, failure mode and effects analysis(FMEA) is widely used in aerospace, automation, nuclear, electronic, mechanic, software products and other fields. The method can identify failure mode with high risk in software system, so as to decide the improvement approach.Traditional FMEA determines the risk priorities of failure modes through the risk priority number(RPN), which is a mathematic product of the risk factors occurrence, severity and detection. However, FMEA has its inherent weaknesses, for example, the precise number can't accurately represent the risk factors with fuzzy meaning; the relative importance of risk factors are not considered; and it can't effectively manage FMEA knowledge. With the scale increment of aviation software and the wide use of component-based software system development method, a failure mode of one component may be the cause of the failure mode of another component in the component-based development process, whic h is not considered in traditional FMEA.For these shortages of traditional FMEA, this paper proposes a FMEA method based on component. Firstly, combined with the theory of ontology, the method constructs failure mode description model of the component-based software system. And then, the fuzzy matrix of risk factors is proposed to describe the relative importance of risk factors. Finally, to improve the accuracy of traditional FMEA, the method constructs the hierarchical FMEA graph and the respect influence matrix for each component, and derives the influence vector for the failure modes and leverage s this vector to amend the risk priority numbers of traditional FMEA. The analysis on a practical flight control system shows the usability of the method proposed in this paper. The simulative experimental results show that our compositional approach is more efficient than the non-compositional approach. Finally, an analysis tool for component based software system is designed and implemented.The method proposed in this paper gives an approach to build a FMEA knowledge base of component based software system, and it can effectively and accurately analyze the failure mode with high risk in the component-based software system, and the analysis efficiency is better than other non-component method.