Degradation analysis is crucial in product reliability assessment and lifetime prediction, especially when the time to failure is difficult to observe directly. In this paper, a Bayesian approach to degradation analysis of inverse Gaussian (IG) process models is used to provide a more informative and reliable prediction of product lifetime distributions based on the flexibility of the Bayesian framework, which incorporates prior knowledge and uncertainty into the analysis. Furthermore, a comprehensive sensitivity analysis of the prior distribution and sample size was performed through simulations, and the calculations were performed using OpenBUGS, an open-source software for Bayesian analysis. Finally, a classical example is used to illustrate the applicability of the Bayesian approach in the degradation analysis of inverse Gaussian process models.
OpenBUGS软件是一种使用马尔可夫链蒙特卡罗(Markov chain Monte Carlo,MCMC)方法进行贝叶斯分析的强大工具[17]。贝叶斯分析可以将先验知识与当前数据结合起来,当无法直接计算模型参数的后验分布时,可以使用MCMC方法来近似这些分布。利用MCMC方法,OpenBUGS软件将未知参数设定为随机变量,随后对概率模型进行求解,得到参数的轨迹图、均值、方差、置信区间以及收敛性诊断结果。这款软件操作灵活简洁,具体的操作步骤如图1所示。
WANGX, XUD H. An inverse gaussian process model for degradation data[J]. Technometrics, 2010, 52(2): 188-197.
[2]
PENGC Y. Inverse Gaussian processes with random effects and explanatory variables for degradation data[J]. Technometrics, 2015, 57(1): 100-111.
[3]
PENGW W, LIY F, YANGY J, et al. Bayesian degradation analysis with inverse Gaussian process models under time-varying degradation rates[J]. IEEE Transactions on Reliability, 2017, 66(1): 84-96.
[4]
GUOJ, WANGC, CABRERAJ, et al. Improved inverse Gaussian process and bootstrap: degradation and reliability metrics[J]. Reliability Engineering & System Safety, 2018, 178: 269-277.
[5]
HAOS H, YANGJ, BERENGUERC. Degradation analysis based on an extended inverse Gaussian process model with skew-normal random effects and measurement errors[J]. Reliability Engineering & System Safety, 2019, 189: 261-270.
[6]
PENGW W, ZHUS P, SHENL J. The transformed inverse Gaussian process as an age- and state-dependent degradation model[J]. Applied Mathematical Modelling, 2019, 75: 837-852.
[7]
SUNB, LIY, WANGZ L, et al. An improved inverse Gaussian process with random effects and measurement errors for RUL prediction of hydraulic piston pump[J]. Measurement, 2021, 173: 108604.
[8]
FANGG Q, PANR, WANGY K. Inverse Gaussian processes with correlated random effects for multivariate degradation modeling[J]. European Journal of Operational Research, 2022, 300(3): 1177-1193.
[9]
YEZ S, CHENN. The inverse Gaussian process as a degradation model[J]. Technometrics, 2014, 56(3): 302-311.
[10]
LIY F, XIANGY Y, PANB S, et al. A hybrid remaining useful life prediction method for cutting tool considering the wear state[J]. The International Journal of Advanced Manufacturing Technology, 2022, 121(5): 3583-3596.
[11]
GUOJ Y, LIY F, ZHENGB, et al. Bayesian degradation assessment of CNC machine tools considering unit non-homogeneity[J]. Journal of Mechanical Science and Technology, 2018, 32(6): 2479-2485.
[12]
ZHAOQ, JIAX, GUOB. Parameter estimation for the two-parameter exponentiated Weibull distribution based on multiply type-Ⅰ censored data[J]. IEEE Access, 2019, 7: 45485-45493.