数控车床故障分布的两重威布尔分段模型

为了定量地分析数控车床故障分布规律,同时也为提高其可靠性水平提供理论依据,在历时2年多收集14台数控车床故障数据的基础上,对数控车床故障率曲线的特征进行了分析,应用概率统计数学方法,建立了数控车床故障分布的两重威布尔分段模型,运用威布尔概率纸图分析法进行参数估计,据此对观测值进行了曲线拟合,拟和结果证明了分段模型的正确性.采用Laplace趋势检验法对模型进行了检验验证.从建立的两重威布尔分段模型和相应的故障率曲线上都可以区分出数控车床的早期故障期和偶然故障期,经计算早期故障期大约是5个多月,这与机床用户反映的使用初期的半年内故障频繁、故障率高相一致.该系列数控车床在使用中存在早期故障期,说明其早期故障在出厂前未被有效排除,降低了其可靠性水平.

Sectional model involving two Weibull distributions for CNC lathe failure probability

To quantificationally analyze failure distribution law of CNC(computerized numerical control) lathes, and to provide the theoretical basis for improving their reliability level, the failure data from 14 CNC lathes over two years were collected, and the failure probability character of CNC lathes was analyzed. Probability and statistics method was introduced, and the sectional model involving two Weibull distributions was established. The analytical method of Weibull plotting paper was applied to estimate the parameters, hereby curve fitting was implemented for the observed values, and the results validated the sectional model. Laplace tendency test method was used to examine the model. The earlier failure period and the random failure period of CNC lathes can be distinguished based on the sectional model and the relevant failure rate curve. It can be calculated that earlier failure period is about 5 months, which is coincident with the consumer feedback that CNC lathe failure occurs frequently at about the first semiannual period with high failure probability. The occurrence of earlier failure period in this series of lathes shows that the early period faults are not totally excluded before leaving factory, so that the reliability level is degraded.