摘要
为实现少量故障数据样本下五轴联动数控机床精确的可靠性评估,通过对贝叶斯估计方法进行研究,设计实现数控机床可靠性建模。该方法首先设定威布尔模型参数服从伽玛分布,为解决无先验信息问题,引入两层贝叶斯方法。为解决积分求解后验分布计算困难问题,使用MCMC方法计算模型参数估计值,最终建立起威布尔模型。引用某机床一年的历史故障数据,使用所设计方法与最小二乘法实现建模,用拟合优度检验将两个方法结果进行比较判断出所设计方法具有可行性。最后依据建立的可靠性模型对机床进行可靠性评估。
In order to achieve accurate reliability evaluation of five-axis CNC machine tools with a small number of fault data samples, the Bayesian estimation method was studied, and the reliability model of CNC machine tools was designed and realized. Firstly, the parameters of Weibull model are set to follow the gamma distribution. To solve the problem of no prior information, a two-level Bayesian method is introduced. In order to solve the problem of calculating the posterior distribution of integral solution, MCMC method is used to calculate the estimated parameters of the model, and finally Weibull model is established. Based on the one-year historical fault data of a machine tool, the design method and the least square method are used to build the model. The feasibility of the design method is judged by comparing the results of the two methods by goodness-of-fit test. Finally, the machine tool reliability is evaluated based on the established reliability model.
作者
李晨光
王红军
李连玉
王茂
LI Chen-guang;WANG Hong-jun;LI Lian-yu;WANG Mao(School of Mechanical Electrical Engineering,Beijing Information Science and Technology University,Beijing 100192,China;Key Laboratory of Modern Measurement&Control Technology,Beijing Information Science and Technology University,Beijing 100192,China;AVIC Chengdu Aircraft Industrial Co.,Ltd.,Chengdu 610000,China;不详)
出处
《组合机床与自动化加工技术》
北大核心
2020年第4期79-82,87,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家自然科学基金资助项目(51575055)
国家科技重大专项:高档数控机床与基础制造装备(2015ZX04001002)。