基于故障数据的医疗设备可靠性预测模型研究OA

Objective To propose a medical equipment reliability prediction model based on failure data to enhance the reliability and maintenance strategies of medical equipment.Methods Firstly,173 maintenance reports for 93 ventilators of the same brand were collected from a certain hospital and underwent data cleaning and preprocessing.Secondly,failure intervals were analyzed using seven probability distribution models of Exponential,Weibull,q-Weibull,Exponential-Weibull(Exp-Weibull),Log-Normal,three-component additive Weibull(3CAW)and modified power generalized Weibull(MPGW),and the maximum likelihood estimation(MLE)method was used to determine model parameters.Finally,the model fit was comprehensively evaluated using the log-likelihood(lnℒ),mean square error(MSE),Bayesian information criterion(BIC),Akaike information criterion corrected(AICc),Kolmogorov-Smirnov(K-S)test and p-value,and the optimal model was selected accordingly.Results When compared with other models,the q-Weibull model performed the best in terms of AICc(794.904 7),BIC(801.607 3),the K-S test(0.075 6)and the p-value(0.741 8),who conformed to the aging failure pattern of the equipment with the cumulative distribution function deviating the least from the empirical distribution function,the peak of the probability density function closely matching the data histogram and the failure rate function curve exhibiting a monotonically increasing trend.The probability-probability plots for each model showed that the theoretical cumulative probability of the q-Weibull model was closely consistent with the observed cumulative probability.Conclusion By introducing the shape parameter q,the q-Weibull model effectively overcomes the limitations of traditional models and accurately describes aging-dominated failure patterns,thereby being highly applicable to reliability prediction for medical equipment.[Chinese Medical Equipment Journal,2026,47(3):80-85]