数控机床主轴系统可靠性评价

Reliability evaluation of spindle system of CNC

以40台某国产数控机床主轴系统现场试验数据为基础,针对数控机床主轴系统故障数据少,导致截尾时间所占比重比较大的特点,为提高建模的准确性,采用故障总时间法对数据进行预处理,通过参数估计、d检验和游程检验确定故障过程服从二参数威布尔分布;针对某些样本无故障的特点,采用平均秩次法评估影响顾客满意度的指标即平均首次故障时间。针对主轴系统的故障特点,采用以上方法能够得到更精确的可靠性评价指标。

On the basis of failure data gathered from forty spindle systems of CNC for evaluating the reliability of spindle system of CNC, because of few data of spindle system, resulting in a relatively large proportion censored time, it is difficult to evaluate, for enhancing the accuracy of the modeling, the total failure time method was used for data pre-processing, then a two-parameter Weibull distribution model which passed d test and Heteroscedasticity testing was established. Because some samples have no fault, mean rank order method was used to assess the impact on average customer satisfaction index-mean time to first failure. The results show that with the failure characteristics of the spindle system, using this method can get more accurate reliability evaluation.