| Modern complex systems have many characteristics, such as: multi function, multi phase, multi mission, multi status, high reliability, long life cycle, small sample, complexity correlation and high uncertainty. It became more difficult for us to depict complex system and undergo reliability quantitative analysis and qualitative analysis. Traditional reliability method focus on the simplification of complex system structure and function; analyze and solve reliability problem by using acquired approximate simple model. However the result tends to be quite different from the actual situation. So, according to structure and function of complex requirement, discover and propose proper method for complex system reliability modeling and analysis become one of the current research hotspots and issues in reliability domain by using systematic engineering theory.So far, the development of the system reliability analysis method has become more and more mature based on the combinational of classical probability theory considering static characteristics of system. However, due to the characteristics of complex system characteristics of small sample, component failure correlation and dynamic failure in modern complex system, it is often unable for us to obtain more accurate results by using the traditional system reliability modeling and analysis methods. At the same time, in the actual project, because of the cost, time, management and human factors and so on, there is a fuzzy uncertainty in the process of obtaining the failure data of the complex system. Therefore, it becomes urgent for us to carry out the research of reliability modeling and analysis methods in the aspects of dynamic failure correlation and fuzzy uncertainty.In the view of reliability engineering, the complexity of system reliability modeling and analysis is reflected in two aspects: the complexity of the system structure and the failure mechanism, the complexity of system reliability problems to be solved. The complexity of the system structure and the failure mechanism mainly refers to the description of system reliability behavior. For example, this complexity is reflected in system reliability measurement, reliability modeling, and vague index in quantitative reliability analysis. In the aspect of reliability modeling, it becomes difficult to process the reliability analysis considering failure correlation, common cause failure, redundancy, unmonotonicity and fault-tolerance. The complexity of system reliability problems to be solved mainly refer to four parts: how to ensure the system long life cycle and high reliability requirement; how to adapt to the development in technology and application of new materials; how to meet the requirement of rapid manufacturing of products; how to carry out reliability modeling, analysis, evaluation in case of small sample and inadequate failure data.Based on the above issues, this paper conducts the research and application as follows: reliability modeling of phased mission system; reliability analysis of phased mission system; risk analysis and evaluation of phased mission system. Specifically, reliability modeling of phased mission system is conducted by using extended object-oriented Petri net(EOOPN) which emerge object-oriented programming and Petri net principle. This modeling method can not only retain the ability of Petri net which can describe the dynamic behavior of a system, but also enhance the reusability, scalability and feasibility of system model. Therefore this modeling method can reduce the complexity of the reliability modeling significantly; in the view of the mechanical parts which mostly obeys the Weibull distribution, failure distribution is affected by the state transition process of phased mission system. Therefore, reliability quantitative analysis of phased mission system is conducted by semi-markov process; reliability analysis of phased mission system is conducted by using recursive algorithm when considering propagated failure between components in redundant design, this method does not require composition of decision diagram and the analysis process can achieve fully automation; Because of fuzzy uncertainty, multi failure modes and component failure should be systematic qualitative analyzed. To find out the risk source of complex system comprehensively, risk analysis and evaluation of phased mission system is conducted based on fuzzy set theory and similarity value measure method.The following four parts works are carried out in this dissertation:(1) Reliability modeling and analysis of satellite propulsion system based on extended object-oriented Petri net and semi markov process. Modern satellite propulsion systems are generally designed to fulfill multiphase-missions. Traditional reliability modeling methods have problems of inadequate depict capacity when considering complex system such as satellite propulsion system. An extended object-oriented Petri net(EOOPN) method and semi markov process is proposed to facilitate the reliability modeling of satellite propulsion system in the paper. The proposed method is specified for modeling of phased mission system, and it can be implemented by generating combination of Petri net principles and object-oriented programming. The effectiveness of the proposed method is demonstrated through the reliability modeling of a satellite propulsion system with EOOPN and semi markov process. The major advantage of the proposed method is that the dimension of net model can be reduced significantly, and phase mission system at system, phase, or component levels can be respectively depicted. Furthermore, the state-space explosion problem is solved by the proposed efficiently.(2) Recursive Algorithm Based Reliability Analysis of Multiphase Satellite Systems with Propagated Failures. Modern satellite systems are generally designed to fulfill multiphase-missions. Component/subsystem redundancies are commonly used to achieve high reliability and long life of modern satellite systems. These characteristics have leaded to a critical issue of reliability analysis of satellites that is how to deal with the reliability analysis with multiphase-missions and propagated failures of redundant components. Traditional methods based on the binary decision diagram can hardly cope with these issues efficiently. Accordingly, a recursive algorithm method was introduced to facilitate the reliability analysis of satellites. This method was specified for the analysis of static fault tree and it was implemented by generating combination of component failures and carrying out a backward recursive algorithm. The effectiveness of the proposed method was demonstrated through the reliability analysis of a multiphase satellite system with propagated failures. Compared to PMS-BDD method, the major advantage of the proposed method is that it does not need composition of binary decision diagrams and its computational process is automated.(3) Risk analysis of propulsion system based on similarity measure value and weighted fuzzy risk priority number evaluation in FMEA. The propulsion system is important component of the satellite system; it is the execution systems of a satellite to keep track of the orbit or to complete the orbit-conversion mission. Therefore modern satellite propulsion system is designed to fulfill multiphase-missions. Although the risk analysis of single failure mode using failure mode and effect analysis(FMEA) has been done by using risk priority number(RPN), multiple failure modes and component correlations in propulsion system have not been effectively evaluated. Because of epistemic uncertainty, it is difficult for the experts to give precise numerical parameters in RPN evaluation. To overcome this drawback a hybrid method integrating the concepts of fuzzy set theory, weight analysis and similarity value measure of fuzzy numbers. To identify the hazard source comprehensively, the analysis process is divided into two phases: first phase focus on the fuzzy fault tree analysis(FTA) accordance with the FMEA, then the main potential failure cause can be acquired and importance analysis of basic event can be finished by using fuzzy set theory and weight analysis; second phase focus on the multiple failure modes and component correlations by using fuzzy RPN(FRPN) evaluation and similarity measure value method(SMVM). The effectiveness of the proposed method is demonstrated through the risk analysis of a satellite propulsion system with hazard importance of basic events and cut set importance. The major advantage of the proposed method is the de-fuzzification to prioritize the failure modes does not be necessary. |
