面向机器人铣削加工的铣刀姿态动力特性研究  

Research on Dynamic Characteristics of Milling Cutter Posture for Robot Milling Processing

作　　者：周小青[1] 童一飞[2] 万志敏 王婷 ZHOU Xiaoqing;TONG Yifei;WAN Zhimin;WANG Ting(School of Mechanical Engineering,Nantong Vocational University,Nantong 226007,China;School of Artificial Intelligence,Nanjing University of Science and Technology,Nanjing 210094,China)

机构地区：[1]南通职业大学机械工程学院,南通226007 [2]南京理工大学人工智能学院,南京210094

出　　处：《组合机床与自动化加工技术》2024年第11期156-161,共6页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：江苏省高等学校自然科学研究项目(21KJB580005)。

摘　　要：针对铣刀对机器人铣削稳定性的影响,采用回归模型与耦合子结构分析(receptance coupling substructure analysis,RCSA)方法,建立了考虑交叉耦合频响函数的不同刀具姿态动力学模型。通过设计激励试验来测量期望的频响矩阵,将实测频响矩阵与多自由度模型拟合,减少了模态参数的数量。采用多任务高斯过程(multi-task Gaussian process,MTGP)回归模型挖掘不同模态参数之间的物理相关性,与普通单任务高斯过程回归模型(single-task Gaussian process,STGP)相比,所提模型减少了MTGP所需回归模型的数量,提高了预测精度和鲁棒性。通过工业机器人的冲击试验验证了所提方法的有效性。In response to the impact of milling cutters on the stability of robot milling,a regression model and coupled substructure analysis(RCSA)method were used to establish different tool attitude dynamic models considering cross coupling frequency response functions.By designing excitation experiments to measure the expected frequency response matrix,the measured frequency response matrix was fitted to a multi degree of freedom model,reducing the number of modal parameters.The multi task Gaussian process(MTGP)regression model is used to explore the physical correlations between different modal parameters.Compared with the ordinary Gaussian process regression model,the proposed model reduces the number of regression models required for MTGP,improves prediction accuracy and robustness.The effectiveness of the proposed method was verified through impact tests on industrial robots.

关 键 词：机器人铣削 交叉耦合频响 高斯过程回归 刀具姿态 

分 类 号：TH16[机械工程—机械制造及自动化] TG57[金属学及工艺—金属切削加工及机床]