Research On Availability Modeling And Optimization For Complex K-out-of-n:G Systems Oriented Mission Requirements

Large modern complex systems,such as inertial confinement fusion laser devices currently in existence or under construction,large nuclear power units and large transport equipment,characterized by complex structure and state,diverse mission requirements,high availability and high cost.As the embodiment of the development of science and the symbol of the strength of the country,the research on the availability and the economy of the large complex systems in the phase of design and operation are popular topics of quality management and reliability engineering theory,and have received extensive attention in academic and engineering fields.This dissertation concentrates on the large complex systems,such as the large laser devices,and the researches of their availability and the economy.The structure and state charactices of the system,including the multi-level redundant structure,the cases of conventional two-state and multi-state are considered respectively.The availability modeling and the optimization of complex k-out-of-n:G system oriented different mission requirements are carried out to solve the problems in availability research of multi-state system.The contents are divided into four parts shown as follows:Firstly,considering the operation time limitation and maintenance time limitation of system,the method of mission availability modeling for the hierarchical k-out-of-n:G system is proposed to evaluate the mission availability of chip amplifier system of large laser facility.The method has broken the assumption that the component operation time obeys exponential distribution and solved the problem that it is difficult to obtain the analytic expression of mission availability when the failure time of components obeys Weibull distribution by using “the principle of equal cumulative failure rate”.The availability of the chip amplifier system considering the multiple failure modes of its components is evaluated by the proposed method.Besides,in order to satisfy the design requirements of the system availability,the redundancy of its module and the boundaries of the reliability and maintenance time of the key components are obtained.The above conclusions can provide useful suggestions for the availability analysis of large laser devices and the reliability management of its components.The research results can refer the references to the designers for the availability analysis of complex systems with time constraints.Secondly,to evaluate and optimize the availability of large complex system with dual performance requirements,the method of availability modeling and the redundancy allocation model of the dual hierarchical multi-state weighted k-out-of-n:G system are proposed.The new availability modeling method combins dual hierarchical operators with universal generating function(UGF),and broadens the current system availability model which is only applicable in the case of a single performance requirement.Then,the redundancy allocation model under availability constraint is constructed to solve the problem of “selection of subsystems and components” in the phase of system design.Unlike existing models that only optimize the underlying components,the proposed model can synthetically optimize the types and quantities of components and subsystems.A power system is taken as an example to verify the correctness and effectiveness of the new model and method,and the effects of single hierarchical performance requirement and dual hierarchical performance requirements on system availability and optimization cost are revealed by the comparative results.Thirdly,to evaluate and optimize the availability of large complex system with hierarchical performance requirements,the method of availability modeling and the redundancy allocation model of the multi-level multi-state weighted k-out-of-n:G system are proposed.The extended new method combining hierarchical operators with UGF can meet either single performance requirement or non-independent performance requirements,which has a wider range of applications.In addition,the recursive algorithm(RA)is suggested to solve the problem of exponential explosion of the state space caused by the increasing number and state of the system units in UGF method.And the upper and lower bounds of the system availability under the hierarchical performance requirements are obtained.Based on this,the system multi-level redundancy allocation model under the availability constraint is constructed to solve the problem of “comprehensive selection of hierarchical components”,and the entire configuration optimization results of the system can be obtained by genetic algorithm(GA)programing.Then,the flexibility of the model and method are verified by the example of an energy system,and the calculation results and efficiency of the UGF and the RA are compared.Finally,to evaluate and optimize the availability of large complex system with random performance requirements,two methods of availability modeling and the reliability allocation model of multi-state k-out-of-n:G system are proposed.By solving the problem that the number requirements and the performance requirements of multi-state k-out-of-n:G system cannot be mapped to each other,the method combining stochastic joint operator with UGF and the improved RA is presented,and the system availability is obtained.The proposed methods relaxe the assumption that the performance requirement is fixed in the existing research,and introduce the requirement of the number of components.In addition,the system reliability allocation models under system availability constraint or cost constraint are constructed respectively.The models solve the optimization problem of reliability and performance of components.Compared with the system optimization models which only considers the performance requirements,the engineering applicability of the new models are stronger.The numerical example of a marine transportation system is analyzed,and the computational efficiency and scope of application of the two availability modeling methods are compared,and the influence of number requirement on system availability and optimization results are expounded.The above works solved the availability evaluation and optimization problems of the complex k-out-of-n:G system under different mission requirements taking into account the multi-level redundant structure and multi-state properties of the complex system.The proposed methods and models are theoretical innovative,which can enrich the availability evaluation methods and optimization models of the multi-state system,and further expand the large complex system availability theory.The achievements have been applied into the reliability design and evaluation of large laser device,which can also offer references for the quality and reliability design and optimization of other complex systems.