Research On Reliability Assessment Method Suiting For The Characteristic Of Machining Center

The reliability assessment of CNC machine tools is very important to various stages in the entire life cycle of feasibility studies, project development, production and delivery, and maintenance. The reliability assessment are very important to measure whether the availability of the machine tools to achieve the desired design goals, verify the reliability of machine tools design is reasonable, pointed out that the weak links of the machine tools and direction to improve the machine tools design, manufacturing, process and maintenance optimization. Therefore, the research of CNC machine tools reliability assessment techniques is very important for the development of China’s machine tools industry. Although the research of CNC machine tools reliability assessment techniques have achieved remarkable results through the exploration and research in recent years, the problem persists. The CNC machine tool is a complex mechatronic system, which includes a large number of subsystems and components. Each system failure product could occur under a variety of failure mechanisms of joint action, so the failure datas of CNC machine tool have multiple failure modes characteristics, different failure modes, and its distribution type is generally different in different failure modes degree of influence on machining centers are different. In addition, because of the development cycle and funding limitations, the amount of investment in the test sample is very limited, resulting failure data processing center also has a kid-like characteristics. In the research literature of reliability assessment, it is generally considered separately for the characteristics of CNC. It has difficulties considering for both of these characteristics when solve the model parameters, but reliability assessment will be more practical engineering significance considering for the characteristics of the two fault data.Under this background, this paper combined with the National Specialized Science and Technology Fund and the National Natural Science Fund. In a series of domestically horizontal machining centers for the study, to obtain a series of fault data processing center by simulating the actual test conditions, in order to conduct a more in-depth research based on the machining center reliability assessment methods. The research work of the dissertation mainly includes:① Reliability evaluation of machining center based on small sample failure data and multiple failure modes. A mixed double weibull distribution model is builded considering for the characteristics of multiple failure modes of complex mechanical system; Using bayes estimation to solve the mixed double weibull distribution’s parameters; At last the result of reliability evaluation of machining center will be obtained through the mixed double weibull distribution model by bayes estimation.② Reliability evaluation of machining center based on the Radial Basis Function(RBF) neural networks. Considering for the small sample failure data of reliability evaluation, the small sample reliability data which are collected through simulate actual working conditions are simulated and extended by the Radial Basis Function(RBF) neural networks, the extended reliability data have the same failure statistical law with the small sample reliability data. A mixed double weibull distribution model is builded considering for the characteristics of multiple failure modes of complex mechanical system; MLE is used to solve the mixed double weibull distribution’s parameters by the large sample data. Then the result of reliability evaluation of machining center will be obtained.③ Reliability evaluation of machining center based on the failure Severity. Different failure modes have different impact on machining centers, starting from the failure mechanism of failure to establish the appropriate distribution model for each type of failure mode distribution test, parameter estimation, to determine the reliability indicators for each fault. Then the impact that different failure modes have on machining center will be sorted. Then consider the failure modes role in the integration of different Bayes data fusion method to calculate the reliability of the overall distribution model machining centers, and ultimately to obtain the reliability index of machining centers.