Importance Based Reliability Allocation Modelling And Algorithm For Phased Mission System

In some engineering applications,some systems can be divided into a member of connective independent phases with different success criteria and structure function.This kind of systems is normally termed as the phased-mission systems(PMS).Mission reliability allocation of PMS may have to consider the following problems:(1)components in different phases have differents reliability level;(2)different phases have different component combinations and structure functions;(3)components reliability parameters are untertain,for example,the component reliability parameters may vary with different environment stress;(4)PMS may have to complete several missions.In addition,PMS may have parts of components needing to be upgraded or be eliminated.This paper presents two kinds of reliability allocation model for PMS reliability allocation problem with uncertain component parameters and PMS reliability allocation problem with both new and exists components.For evaluating the mission reliability of PMS,this paper also provides an extended object-oriented Petri net(EOOPN)model.The major contributions and innovations of this dissertation can be concluded as follows:(1)An EOOPN model for reliability modeling and anlysis of PMSAnalysis methods for PMS reliability evaluation have problem to be directly applied to actural engineerings with many model assumptions.Simulation methods typically offer greater generality in system representation.Petri net(PN)combines the merit of Markov method and simulation method,moreover,it allows simple graphical representation with significant modeling power.However,PN model will become extremely complex and inconvenient in describing the complex system due to its limited elements.This paper provide an EOOPN(extended objected-oriented Petri net)model based on the colord PN to deal with mission reliability modeling problems of PMS.EOOPN can have a hierarchically nested network structure.As the basic element of EOOPN,subnet uses an extended colored Petri net(CPN)to represent internal behavior of a component or a phase.Objects of subnet communicate and collaborate with each other using information carried by arcs.EOOPN introduces the concept of logic transitions to describe the complicated logic relations between the various system components.In EOOPN,broadcast place is introduced to transmit information to the large number of objects in the model simultaneously.Four kinds of sub-models are defined to ensure the reusability of EOOPN model.(2)Reliability allocation model and algorithm for phased mission systems with uncertain component parameters based on importance measureReal value of the reliability of a component may become lower than its designed value due to such uncertainty,thus making the whole system fail to meet the required reliability level.Therefore,this paper presents a model that incorporates the component uncertainty in the system reliability allocation process.The model uses components' MTBF instead of component's reliabilities in different phases as variables to avoid the problem caused by using components reliability as variables.Additionally,the model sets a confidence level constraint to make the designed mission reliability requirement be satisfied at a specified probability level.Thie paper also proposes a variance-based global importance hybrid heuristic algorithm for its solution.The variance-based global importance measure is used to evaluate the importance of a component to the mission reliability of PMS while the reliability of all components vary randomly.To illustrate the effectiveness of the proposed model and algorithm,two examples of PMS are presented and the allocation solutions are validated through Monte-Carlo simulation method.Illustrative examples show that reliability allocation solutions produced by the proposed model and algorithm can meet the design requirement.(3)Reliability allocation model and algorithm for phased mission systems with both upgraded and existed componentsGenerally,PMS consists several subsystems with different life expectancy.When a new task needs to be performed,it is necessary to evaluate whether the reliability of new task can reach the setted level.If not,the system need to be improved.For the subsystem consists of components that have been eliminated by market for some reasons,decision makers need to choose to upgrad it or abandon it.For other subsystems,decision makers can purchase them from the market.In this case,the system will be made up with both existed and new upgraded components.In this paper,a reliability allocation model considering PMS consists of both existed and new upgraded components is proposed.The model divides the components into two categories,namely,the components that meed to be upgraded and the components that are normal.The components need to be upgraded can futher be divided into two sub-categories,namely redesigned parts and eliminated parts.In order to improve the reliability of the system during the reliability allocation,the reliability parameters can be improved for the redesigned components,and the redundant design can be applied to the components that are normal.A cost-importance is presented to solve the model with taking both component importance and cost into consideration.An example shows that the model and algorithm are suitable for solving the reliability allocation problem of this kind of PMS.In addition,the second reliability allocation model and algorithm are applied to the reliability allocation of space tracking,telemetry and command system for manned space mission whose effectiveness are futher to be verified.