摘要
非齐次泊松过程适用于具有故障趋势可修系统的可靠性分析,其中幂律过程较之对数线性过程更为常见。但当研究的数控机床初始故障率不为零时,幂律过程模型则不适用。以对数线性过程模型为基础,综合考虑故障率的变化,应用虚龄概念,研究任意维修质量数控机床的可靠性及维修性,提出了数控机床可靠性评估及不完全预防维修的对数线性虚龄过程方法。该方法可通过故障数据的综合趋势检验确定故障的维修类型,基于赤池信息准则、贝叶斯信息准则,选择非齐次泊松过程模型,利用最大似然估计的渐近对数正态分布特性,用Fisher信息矩阵法给出对数线性过程模型参数及数控机床可靠性指标的点估计及区间估计,在平均维修费用最小、可用度最大及一定可靠度的情况下,确定数控机床不完全预防维修的维修周期及最佳次数。
Non-homogeneous Poisson process is suitable for reliability analysis of repairable systems with fault tendency.Among them,power-law process is more common than log-linear process.However,when the initial failure rate of the studied CNC machine tools is not zero,the power-law process model is not applicable.Based on the log-linear process model,with the change of the failure rate comprehensively considered,concept of virtual age adopted to study the reliability and maintainability of CNC machine tools with arbitrary maintenance quality,a log-linear virtual age process method for CNC machine tool reliability evaluation and incomplete preventive maintenance was put forward.This method can determine the maintenance types of faults through comprehensive trend test of fault data.Based on Akaike information criterion and Bayesian information criterion,a non-homogeneous Poisson process model was selected,and for the asymptotic log-normal distribution characteristics estimated using maximum likelihood,the Fisher information matrix method was used to give point estimates and interval estimates of logarithmic linear process model parameters and reliability indicators of CNC machine tools.Under the conditions of average maintenance cost,largest availability and certain reliability,the maintenance cycle and the optimal number of incomplete preventive maintenance of CNC machine tools were determined.
作者
赵中敏
ZHAO Zhong-min(Applied Technology School of Jiangsu Ocean University,Lianyungang Jiangsu 222069,China)
出处
《林业机械与木工设备》
2020年第2期48-52,共5页
Forestry Machinery & Woodworking Equipment
关键词
可靠性
预防维修
对数线性过程
数控机床
reliability
preventive maintenance
log-linear process
CNC machine tool