基于混合威布尔分布的数控折弯机可靠性评估

Reliability assessment of CNC bending machine based on mixture Weibull distribution

在收集国产WEHK 110/3100型数控折弯机故障数据的基础上,选用两参数混合威布尔模型进行数据建模,采用最大期望值(EM)算法进行参数估计,通过拟合优度检验确定,机床故障过程服从混合威布尔分布,最后通过计算获得该型机床的可靠性指标。结果表明,机床可靠性指标与企业实际运行非常契合,可作为该型机床可靠性水平的参考,为国内关于数控折弯机的可靠性评估提供了一种新方法。

A two-parameter mixture Weibull distribution model was established on the basis of failure data gathered from CNC bending machine of WEHK 110/3100, the model was defined by goodness-of-fit test, and the parameter es- timate was solved by EM algorithm. Based on the data, the machine breakdowns following Weibull distribution was proved. In the end, reliability index of the CNC bending machine was achieved. The results show that the index is in conformity with actual operations of enterprise, and it is capable to be used as a reference for bending machine reliabili- ty, hence maybe it is a new approach to reliability assessment of CNC bending machine.