Font Size: a A A

Research On Component State Degradation And Fault Correlation Based On Monitoring Data

Posted on:2018-04-06Degree:MasterType:Thesis
Country:ChinaCandidate:Y LiuFull Text:PDF
GTID:2322330515468628Subject:Statistics
Abstract/Summary:PDF Full Text Request
Now the modern high-speed trains are equipped with the vehicle information recorder,can record real-time status and fault information and component,but is now often the fault information mining enough,the grasp of the component failure rule is not accurate enough,so how to use these components fault data correlation rules and component degradation the failure to make research is the focus of this paper.Firstly,the fault data of single parts of the system were processed by analyzing failure data of the homogeneous test,the trend of the judgment of a preliminary processing of fault data,then according to the test results,choose the Weibull process to fit,and then using the maximum likelihood method to estimate the parameters of Weibull process,then are given the first failure and other failures the best preventive maintenance strategy based on age replacement policy,the best preventive maintenance cycle,so the train after long-term operation,the expected loss per unit time and a preventive replacement costs than the minimum train fault and combined with actual data of listed.The attenuation law of vehicle reliability is studied.Then,considering the complex system of multiple components,component failure is generally not independent,often has certain correlation.It discusses the parallel system is composed of two parts,considering the failure of each component in the system will no longer be independent,The failure rate of the component under consideration is affected by the average number of faults occurring before the time of the component and the average number of failures of the other component before the time.An self interactive failure model is proposed,and the parameters of the model using the method of least squares estimate,based on the model,put forward the preventive maintenance strategy model based on minimal repair,The best preventive maintenance cycle is obtained,and make the expected loss of unit time is minimized.The last train based on the actual fault data,compares under the components of independent failure and self interaction failure,the difference of failure rate and preventive maintenance period,The influence of self interactive failure on maintenance strategy is explained.
Keywords/Search Tags:maintenance strategy, non-homogeneous Poisson process, interactive failure, expected loss, maintenance period
PDF Full Text Request
Related items