Method For Reliability Assessment Of Warm Stand-by System

In reliability design, the redundant system is an important means to improve system reliability.The systems are designed based on the known reliability of components, but in many cases these reliabilities are in dark. In previous studies, scholars estimated the reliability of components with small samples, or gived system reliability assessment with a large sample, but few estimate the system reliability with the small sample cases.The reason may be the system reliability evaluation is necessary to involve small samples of complex components reliability equivalent to the system reliability.In the actual bad conditions, because the components or in reserve in case the work will be more failures, so the reliability of the system design more consideration to the warm standby. With the development of reliability engineering and significant improvement of component reliability, reliability testing is generally a small samples.It is necessary to study warm standby system reliability assessmentIn this paper, with the reliability data obtained on small samples of components the following three aspects of system reliability is derived by the existing research results:1)In the second chapter, Reliability for warm standby system with two same exponential components and failure exponential form switch is assessed.First of all, the failure rate of component, failure rate of reserve component and the reserve switch failure rate is only one unknown, were obtained accurate system Fiducial confidence level lower confidence limits; then the failure rate and reserve component failure rate, switch failure rate are unknown, converting the information of component and switch with Beta distribution,system reliability approximate confidence limit is derived.And, the approximate lower confidence limit of system reliability is derived by bootstrap. Finally, the above demand system reliability lower confidence limit of the method similar to the random simulation, and simulation results are compared.2)In the third chapter,based on non-replacement of regular censoring obtained form the work of Warm Standby System failure rate and failure rate of the Bayesian reserve estimates obtained using life testing samples of the success of switch failure rate of Bayesion estimates. Conclusions are given warm standby system with Bayes estimates. The first moment, the variance of the hyper parameters were estimated with history sample. Use of gain or loss when the times demand a sample of ultra-efficient maximum likelihood parameter estimation. Bayesian method of use and multi-temperature analysis of standby system reliability. Finally, using the method of stochastic simulation results obtained are analyzed and compared.3)In the fourth chapter, the exponential distribution, Weibull distribution mixed warm standby system reliability has been evaluated. Using WCF method, warm standby system random polynomial is obtained, the switch entirely reliable,0-1 failure, exponential failure case of system reliability confidence limits is derived; using the ootstrap method various types of system reliability is assessed; Finally,using stochastic simulation to compare two methods of reliability assessment.