耦合多特征部件的数控装备可靠性建模与评估技术研究

Reliability modeling and assessment are now acting indispensable roles in quality-improvement and innovation-implementation of modern numerical control (NC) equipment. They are also critical supporting techniques for core-coopetition promoting of high-tech manufacturing industry. As the gradual incorporation of automation, IT, and intelligence into classical NC equipment, reliability modeling and assessment are challenged by various characteristics of newly-introduced NC equipment. These challenges include the complexities introduced by multiple functions, coupling components, and dynamic conditions of new NC equipment. Based on a new typical of heavy duty machine tools, this proposal aims to overcome these challenges by improving the method of reliability modeling and assessment of NC equipment. The research is based on the deep investigation of reliability analysis and multiple degradation modeling. Methods of multivariate analysis, Copula based dependence modeling, multi-source information integration, and Bayesian estimation methods are incorporated with reliability modeling and assessment methods. Major attention are laid on the following research contents: (1) Improve a classical reliability analysis method by taking account of the characteristics of multiple functions and dynamic conditions of NC equipment; (2) Develop an integrated reliability modeling method based on multiple degradation processes under dynamic operational and environmental conditions; and (3) Introduce a Bayesian degradation analysis method for reliability assessment. Finally, based on the methods and techniques developed above, a coherent methodology for reliability modeling and assessment for newly-developed NC equipment is constructed. It can further facilitate the development, manufacturing and launching of newly-developed NC equipment, and serve as a fundamental technique for core-competitiveness improvement of high-tech NC equipment as well.

数控装备可靠性评估技术是新型数控装备保证质量、有效创新的关键使能技术，也是我国高端装备制造业核心竞争力提升的关键支撑技术。新型数控装备的自动化、信息化和智能化使得数控装备可靠性评估受到功能特征多样、部件耦合繁复、外部影响突出以及小批量定制等问题的挑战。本项目以某新型重型数控机床为研究背景，以可靠性分析方法和多性能退化建模技术研究为研究切入点，将多变量分析技术、Copula相关建模方法、多源信息融合技术和贝叶斯评估方法整合进数控装备可靠性建模和评估中，重点研究考虑多功能特征和多环境工况应力的可靠性分析方法、考虑多性能退化相关与多环境工况应力的可靠性建模方法、基于贝叶斯多性能退化分析的可靠性评估方法。从而，着力构建系统、全面的数控装备可靠性评估技术，为新型数控装备的设计、研制、使用和市场投放以及高端数控装备质量水平的本质提升提供可靠性基础技术的支撑。

搜集了典型数控装备生产和使用过程的有关可靠性数据，完成了数控装备的系统可靠性分析，识别了其关键功能特性和关键功能部件，确认了关键功能特性对应的性能指标、相关关系以及影响性能指标的环境工况应力因素。建立了与性能指标对应的边缘性能退化模型与基于Copula函数的多性能退化模型。将所建立的性能退化模型对所收集的性能退化数据进行了建模与优化分析。在所收集的适用环境和任务工况的可靠性信息数据基础上，建立了考虑多环境工况应力影响的多性能退化模型。对比、选择并优化了所建立的考虑多环境工况应力影响的可靠性模型，并应用贝叶斯多性能退化分析的两段式求解方法解算了多环境工况应力下多性能退化分析模型的参数估计和可靠性指标。选择典型机床部件进行了计算机数字仿真，修正了多性能退化可靠性模型和优化贝叶斯融合方法。