基于大数据技术的机床结构动力学在线分析方法研究

With the requirements of high-speed, high-precision and high-efficiency machining, the dynamic characteristics of machine tool structure have become an important performance index. High-precision machining of group machines is an important production mode for large-scale manufacturing enterprises, and it is of great significance to study the dynamics characteristics of the machine tool group. The traditional machine tool dynamic analysis cannot meet the needs of high-precision analysis of the group machine running.. In view of the above problems, this project plans to make use of big data technology, from three aspects of "on-site condition data and on-line dynamic feature extraction", "multi-dimensional feature knowledge mining of machine tool structural dynamics" and "multi-dimensional characterization of machine structural dynamics", to study on-line analysis of methods of machine structural dynamics.. The innovative points of this project are as follows: The deep learning method is used to study the real-time extraction method of dynamic characteristics of group machine tools for the first time; large-scale analysis and mining of structural dynamics knowledge of group machine tools for the first time; the knowledge graph is used to characterize multidimensional dynamics features of group machine tools for the first time.

随着高速、高精、高效加工要求的提高，机床结构动力学特性已经成为一项重要的服役性能指标。群体机床高精加工是上规模制造企业的重要生产方式，研究群体机床结构动力学特性具有重要意义。传统机床动力学分析无法满足群体机床运行状态下的高精度分析需求。. 针对以上问题，本项目拟利用大数据技术，从“现场工况数据清洗与动力学特征提取”、“机床结构动力学多维度特征知识挖掘”以及“机床结构动力学多维度特性表征”三个方面研究机床结构动力学的在线分析与表征方法。. 本项目研究创新点在于：首次利用深度学习方法研究群体机床结构动力学特征实时提取方法；首次利用工况大数据分析与挖掘群体机床结构动力学多维度特征知识；首次利用知识图谱来表征群体机床结构动力学的多维度特征。

随着高速、高精、高效加工要求的提高，机床结构动力学特性已经成为一项重要的服役性能指标。群体机床高精加工是上规模制造企业的重要生产方式，研究群体机床结构动力学特性具有重要意义。传统机床动力学分析无法满足群体机床运行状态下的高精度分析需求。.在本项目中，申请者利用大数据技术，从“现场工况数据清洗与动力学特征提取”、“机床结构动力学多维度特征知识挖掘”以及“机床结构动力学多维度特征表征”三个方面研究了机床结构动力学的在线分析与表征方法。主要研究内容如下：①开展数控机床在变参数切削条件下的振动数据截取和标记研究工作；②开展数控机床不同机床个体之间机床结构动力学特征的变化规律研究；③开展变切削激励和位置相关动力学耦合作用下切削振动表征和抑制的研究工作。.本项目按照研究计划顺利完成。本项目提出了基于端到端深度学习模型的机床变参数切削条件下振动数据截取和标记方法，完成了基于迁移学习的数控机床不同机床个体之间机床结构动力学特征预测工作，实现了基于知识图谱的变切削激励和位置相关动力学耦合作用下切削振动表征和抑制。