Reliability Modeling And Assessment Of Dependent Competing Failure For Performance Degradation Self-recovery Products

Degradation-based reliability modeling and analysis technology are based on the products' failure physics mechanism,using product performance degradation data for reliability modeling and analysis,which providing reliability analysis for the product with very few failures or even zero failures.Currently,most degradationbased reliability studies ignore the performance degradation self-recovery process.However,in actual engineering,the performance parameters of many products are easily affected by external sporadic factors.External sporadic factors such as random shock will increase the product performance degradation value,when the external sporadic factors disappear,product performance will be spontaneously recover to the pre-arrival state of accidental factors.The performance degradation self-recovery process is accompanied by the entire service cycle of the products.Ignoring the performance degradation self-recovery process for reliability analysis,cannot truly and reasonably characterize the evolution of product performance,cannot accurately reflect the relationship between product performance and the environment experienced,will have some influence on the accuracy of product reliability analysis,reliability assessment,and life prediction.In addition,besides performance degradation in the course of service,the product may suddenly fail due to random shock and other occasional factors at a certain time,that is,product failure is usually the result of competing between degradation failure and sudden failure.In view of this,this paper conducts in-depth exploratory research on performance degradation self-recovery modeling,dependent competing failure reliability modeling including performance degradation self-recovery process and dependent competing failure reliability assessment including performance degradation self-recovery process.The research includes the following aspects:1.Considering the self-recovery phenomenon of performance degradation caused by occasional factors such as random shock,a performance degradation selfrecovery model is established.Firstly,on the basis of the homogeneous Poisson shock process,the influence of the shock time interval on product performance degradation is analyzed.Secondly,combined with the update theory,the characteristics of the performance degradation self-recovery process are studied,and the conditions for considering performance degradation self-recovery in degradation failure reliability analysis are given.Then,the effects of parameters such as recovery threshold and shock arrival rate on the performance degradation self-recovery amount are analyzed.Finally,the rationality and effectiveness of the performance degradation self-recovery model are verified by simulation,which lays a theoretical foundation for the reliability analysis of performance degradation self-recovery products.2.Aiming at the products with performance degradation self-recovery process and competing failure,a dependent competing failure reliability model including performance degradation self-recovery process is constructed.Firstly,on the basis of the increase in the performance degradation caused by the random shock,it is considered that when the time interval of the shock is greater than a certain threshold,the amount of performance degradation caused by the shock is self-recovery.Thus,a degradation failure reliability model derived from random shock and performance degradation including self-recovery is derived.Secondly,according to the stress intensity interference theory,the sudden failure reliability model under extreme shock is established.Then,combined with the competing failure theory,the competing failure reliability model and analytical expressions including the performance degradation self-recovery process are given.Finally,the necessity of considering performance degradation self-recovery process when reliability analysis is analyze by an example,and the sensitivity of model parameters is further discussed.It provides a theoretical basis for studying the reliability of performance self-healing products.3.Aiming at the products with performance degradation self-recovery process and competing failure,a dependent competing failure reliability assessment method including performance degradation self-recovery process is proposed.Firstly,by analyzing the mechanism and correlation of product degradation failure and sudden failure,combined with the characteristics of different failure modes and many degradation data and fewer failure data,the reliability assessment models are constructed for degradation failure and sudden failure.Then,using Wald's theorem and moment estimation to estimate the degradation failure parameters,and quantify the effect of random shock on the natural degradation process,the support vector regression machine is used to estimate the sudden failure parameters.The reliability assessment of the dependent competing failure under the complete data of sudden failure,degradation failure,and the self-recovery process is realized.Finally,the effectiveness and rationality of the reliability assessment method are verified by simulation,which provides theoretical basis and technical support for product reliability and reasonable maintenance.4.Through the reliability test,the performance degradation self-recovery model,the dependent competing failure reliability model including the performance degradation self-recovery process,and the dependent competing failure reliability assessment method including the performance degradation self-recovery process are verified.Firstly,according to the characteristics of semiconductor laser's competing failure and self-recovery of output optical power degradation,the reliability test is designed,and the semiconductor laser output optical power test device is built,and the test is carried out and analyzed.Then,the semiconductor laser output optical power degradation self-recovery model is obtained.Finally,the competing failure reliability of semiconductor lasers is analyzed and evaluated.The research provides strong guidance and reference for the reliability test design analysis of other competing failure products with performance degradation self-recovery process.