Condition monitoring and health management of deteriorating systems have attracted significant attentions in reducing failure rate and ensuring normal operation of systems.However,related existing methodologies usuall...Condition monitoring and health management of deteriorating systems have attracted significant attentions in reducing failure rate and ensuring normal operation of systems.However,related existing methodologies usually confront difficulty when applied to systems with multivariate performance characteristics(PCs)in the presence of model uncertainty.In this work,a periodic inspection and maintenance policy based on bivariate degradation process is proposed,where each PC is modeled by a gamma process with time-scale transformation.Further,copula functions are used to model the dependency between degradation increments.The uncertainty of model parameters is estimated by the Bayesian Markov chain Monte Carlo(MCMC)algorithm.Subsequently,the maintenance policy is optimized by minimizing the expected cost under parameter uncertainty.The expectation and variance of maintenance cost are computed over a finite-time horizon.Finally,a real example of heavy machine tools is applied to illustrate the proposed maintenance model.Numerical results show that the proposed maintenance policy provides a more robust and accurate solution in maintenance decision-making for deteriorating systems,leading to lower maintenance cost compared with other policies.展开更多
基金supported by the National Natural Science Foundation of China[Grant NO.72032005,71872123,72002149,71802145].
文摘Condition monitoring and health management of deteriorating systems have attracted significant attentions in reducing failure rate and ensuring normal operation of systems.However,related existing methodologies usually confront difficulty when applied to systems with multivariate performance characteristics(PCs)in the presence of model uncertainty.In this work,a periodic inspection and maintenance policy based on bivariate degradation process is proposed,where each PC is modeled by a gamma process with time-scale transformation.Further,copula functions are used to model the dependency between degradation increments.The uncertainty of model parameters is estimated by the Bayesian Markov chain Monte Carlo(MCMC)algorithm.Subsequently,the maintenance policy is optimized by minimizing the expected cost under parameter uncertainty.The expectation and variance of maintenance cost are computed over a finite-time horizon.Finally,a real example of heavy machine tools is applied to illustrate the proposed maintenance model.Numerical results show that the proposed maintenance policy provides a more robust and accurate solution in maintenance decision-making for deteriorating systems,leading to lower maintenance cost compared with other policies.
