Reliability Modeling And Analyzing Methods For Lauching Device Under Accelerated Life Testing And Zero-failure Data

As a crucial component of weaponry,launching device plays an important role in defending national sovereignty and safeguarding territorial integrity,its performance,technology and quality have already been the key problem in national defense industry construction.With the rapid development of science and technology and the continuous improvement of manufacturing in our country,the reliability of weaponry tends to be much higher and the lifetime tend to be extremely longer than ever.However,researches for weaponry mainly focus on developing new type weapon equipment and enhancing its working performance,little attention has been paid to the related quality and reliability engineering.This phenomenon has created a great imbalance in weapon development and led to some disconnections between each phase in whole life cycle of weaponry,which seriously restrict the construction level of weapons and equipment.Furthermore,the application of accelerated life testing(ALT)and zero-failure data analyzing is usually for general engineering machinery,which is not suitable for launching device with high reliability in harsh working conditions.Therefore,carrying out the researches on the reliability evaluation and life estimation for launching device is an urgent need at present.The reliability modeling and life estimation for launching device are the specific applications of reliability theory in engineering practice,which is also an indispensable part from design and pre-research to production and service phase in launching device whole life cycle.It provides important bases for the improvement of reliability,the extension of lifetime and the guarantee of quality.Using accelerated life testing and zero-failure data analyzing theory in reliability assessment for launching device,we start from typical life distribution models,combined with the characteristics of the launching device and its key components,establish the corresponding reliability and lifetime mathematical model.Collecting the relevant field testing data then complete the reliability evaluation and life prediction for the launching device.This thesis devotes to the researches on reliability modeling and assessment,accelerated life testing and zero-failure data analyzing for launching device and its key components,the main research contributions are summarized as follows.(1)Development of a small sample ALT reliability and lifetime assessment method for the high speed impeller of air turbine in launching device.In the design and pre-research phase,the samples of high speed impeller are usually very small,general evaluation methods are not suitable for lifetime prediction for this case.Build a life distribution model and collect the historical sample data to obtain the model parameters.By introducing the inverse power law into the model,a reliability assessment model is proposed for the high speed impeller with small sample in ALT.Based on this model,design and implement an ALT for the high speed impeller of a certain launching device and complete reliability model analysis and lifetime evaluation.(2)Development of a nonparametric predictive inference(NPI)method for ALT and extend the research with imprecise probability theory.During the testing phase launching device is under a varied environment,which lead to complicated failure mechanisms in reliability analyzing.In this case,nonparametric method and imprecise probability theory are combined for reliability modeling in ALT.In the presented imprecise NPI method,only a few basic assumptions are used for ALT data analyzing,which make the model works well in different cases.The imprecise probability theory enhances the robustness of the model,compared to general parametric analyzing method,imprecise NPI is more capable of dealing with data errors,which provides a strong theoretical support for the application of imprecise probability in reliability.(3)Development of an imprecise survival analysis method for reliability and discuss the quantification of imprecise probability for ALT.Using accelerated link functions for life distributions and the log-rank test in survival analysis,this thesis presents lower and upper survival functions with a equivalent transforming model of ALT data for the electronic control module of lauching device.The presented survival analyzing method solves the problem of modeling misspecification for different stress levels in ALT and illustrates the feasibility of applying survival analyzing method with imprecise probability for lauching device.(4)Development of an E-Bayesian estimated method using zero-failure data for reliability analysis wrt exponential and success-failure type data.According to the zero-failure situation in the production and service phase of lauching device,different models are used corresponding to exponential or success-failure field data.For lifetime data of electronic control system of lauching device,exponential life distribution model with zero-failure data is included to obtain the failure rate using E-Bayesian estimation.For success-failure type data in the field tests for lauching device,environmental factor is introduced for data fusion in zero-failure case and on that basis E-Bayesian estimation with zero-failure is proposed for success-failure type data at normal stress level.According to different field testing conditions,E-Bayesian method is improved for different data types,reliability assessment model with zero-failure data is built for lauching device.This research offers a solution to the problem for the absence of failure information in reliability modeling and provides some information for determining the value of model hyper parameters in practical application,which improves the reliability analysis and lifetime evaluation method and realizes the reliability assessment for lauching device.