目的 针对医疗设备存在的退化加速现象,提出对退化过程进行监控的方法,旨在有效监测到真实情况中退化加速的现象并进行预警,降低设备运维成本.方法 无论是退化率恒定还是退化率加速的维纳过程,其增量概率分布都是独立于其他时间区间增量的,本文运用上侧累积和(Cumulative Sum,CUSUM)控制图监控策略,并分析两类测量误差(乘性测量误差和加性测量误差)对控制图监控效率的影响.结果 运用蒙特卡洛仿真得到不同加速系数和测量误差下的平均运行链长(Average Run Length,ARL)和运行链长标准差,结果表明对于退化率恒定的固定效应模型,CUSUM对退化加速的监控具有稳健性,ARL的降低率均维持在43%左右;而在退化加速的随机效应模型中,监控性能会随着乘性测量误差(当乘性误差>0.1后)的增加而下降.结论 CUSUM控制图对医疗设备退化加速现象具有良好的监控能力,随着退化率的加速系数的增加,控制图能够越早发现异常退化加速状态.
Monitoring Design of Wiener Process with CUSUM Control Chart Considering Measurement Errors
Objective To propose a method to monitor the degradation process to detect the acceleration of equipment degradation process and reduce the cost of equipment operation and maintenance,in response to the degradation acceleration of medical equipment.Methods Considering that the incremental probability distribution of the Wiener process with either constant degradation rate or accelerated degradation rate was independent of time interval increments,this paper proposed a control chart monitoring strategy utilizing the upper-side cumulative sum(CUSUM).The effects of two types of measurement errors(multiplicative and additive measurement errors)on the efficiency of control chart were analyzed.Results Monto Carlo simulation was applied to obtain the average run length(ARL)and standard deviation of run length under different acceleration rate and measurement errors.The results showed that for the fixed-effects Wiener process with constant degradation rate,CUSUM was robust to the monitoring of degradation acceleration and the reduction rates of ARL were maintained at around 43%;whereas,in the random-effects model of degradation acceleration,the monitoring performance decreased with the multiplicative measurement error increased(when the multiplicative error was greater than 0.1).Conclusion The CUSUM control chart performs well for the degradation acceleration phenomena of medical devices.As the acceleration coefficient of the degradation rate increases,the earlier the control chart is able to detect the abnormal degradation acceleration state.
medical equipment maintenancedegradation accelerationWiener processCUSUM control chartmeasurement error
国家科技期刊平台
NSTL
万方数据
维普
