Bdd-based Multi-phased Mission System Reliability Modeling Analysis

As the system of aerospace, aviation, ships, chemicals, electronics and other industrial field become increasingly complex and the automation level is continuously improving, the task of complex systems is no longer a simple single phase task, but a task with multiple phases consists of different functions. For a complex system with different phase configurations, in order to achieve the various types of prescribed objectives and complete the corresponding work, the entire system needs to have a high level of reliability. Therefore, phased mission system analysis is becoming more and more important.As a typical phased mission system, satellite includes more than one task phases. With each phase's conversion, the reliability requirements, system configuration, failure criteria, and failure rate of components may be changed. Therefore, how to deal with these phase changes reasonably, analyze the reliability of the satellite accurately, become the key point to the reliability evaluation of satellite.In this paper, Binary Decision Diagram (BDD) based approach was introduced to evaluate the reliability of satellite. First a detailed summary of phased mission system's domestic and international research status and the development trends have been investigated. The concepts and characteristics of PMS were also described, the dependence between phases, and the analysis methods for PMS have been investigated in order to achieve a more systematic summary.Then two types of PMS have been studied, for the phase which is no dynamic behavior, taking the satellite control systems as an example, a BDD-based approach is applied to solve the dependence among phases. It proves that BDD-based approach is more efficient than Fault Tree (FT) method. For the phase with dynamic behaviors between components, taking the whole satellite system as the object of study, modular solution is proposed, which provides a combination of BDD-based solution for static modules, and Markov chain solution for dynamic modules. Compared with traditional Markov chain model, modular solution is more efficient.At last, an efficient method based on the BDD is proposed to evaluate the reliability of PMS with periodical phases.