摘要
机器人由于低刚度特性导致加工中极易产生颤振,颤振抑制中最具挑战性的任务之一是预测整个空间的动态特性,传统的基于有限元分析或者试验方法在获取全域模态参数时要么耗时,要么不准确。针对这一问题,提出了一种基于随机森林的机器人加工系统模态预测方法。采用LMS-Test-Lab对机器人加工系统开展模态试验,设计试验测试组得到加工平面内有限位姿点刀具末端的频响曲线;利用有理分式多项式法辨识了多阶位姿相关的模态参数;基于随机森林法建立预测模型,最终实现工业机器人工作空间内位姿相关的模态参数的预测。试验结构表明,所提出的随机森林模态预测方法预测精度达到80%以上,该方法仅需几次试验就能覆盖整个加工区域的激励试验数据。
Robots are prone to chatter during processing due to their low stiffness characteristics.One of the most challenging tasks in chatter suppression is to predict the dynamic characteristics of the entire space.Traditional finite element analysis or experimental methods are either time-consuming or inaccurate in obtaining global modal parameters.A modal prediction method for robot machining systems based on random forest is proposed to address this issue.Using LMS Test Lab,modal experiments were conducted on the robot machining system,and a testing group was designed to obtain the frequency response curve of the finite pose point tool end in the machining plane;The modal parameters related to multiple poses were identified using the rational fraction polynomial method;Based on the random forest method,a prediction model is established to ultimately achieve the prediction of modal parameters related to the pose of industrial robots in the workspace.The experimental structure shows that the proposed random forest modal prediction method has a prediction accuracy of over 80%,and this method only requires a few experiments to cover the excitation test data of the entire processing area.
作者
余倩倩
张浩
李宝红
YU Qianqian;ZHANG Hao;LI Baohong(School of Mechanical and Power Engineering,Nanjing Tech University,Nanjing 211816,China;Jiangsu Province Key Laboratory of Industrial Equipment Manufacturing and Digital Control Technology,Nanjing 211899,China)
出处
《组合机床与自动化加工技术》
北大核心
2025年第5期44-47,54,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
江苏省科技成果转化专项资金资助项目(BA2022021)。
关键词
机器人磨削
加工位姿
模态预测
随机森林预测模型
robot grinding
processing pose
modal prediction
random forest prediction model
