面向机器人铣削加工的刀尖动态特性分析与稳定性预测

Analysis of Tool Tip Dynamic Characteristics and Stability Prediction for Robotic Milling Tasks

在大型复杂构件机器人铣削加工过程中,由于机器人结构刚度较低、末端负载配置复杂以及工件局部弱刚性等问题,导致机器人铣削加工过程中极易发生颤振现象,进而影响工件的表面加工质量。而机器人铣削加工的稳定性主要取决于其刀尖动态特性,因此提出了一种面向机器人铣削加工的刀尖频响函数(FRF)预测方法,能够实现任意姿态下的刀尖频响预测,进而实现机器人铣削加工稳定性预测,在通过加工参数优化来进一步提升工件表面加工质量。首先,提出了一种基于非线性最小二乘法的加权叠加法,减少了实验量并放宽了对实验位姿的要求。然后,给定任意多个基准实验姿态实测的刀尖频响函数,所提出的模型就能够预测出一定工作空间范围内任意目标姿态下的机器人刀尖频响函数。接着又提出了一种线性化模态参数叠加法,避免了频响函数直接叠加产生的多模态现象,提高了预测准确率。最后,通过开展机器人模态实验与铣削加工试验,验证了该模型的准确性和实用性,进而通过加工参数进一步优化实现了加工质量提升。

The machining vibration is easy to occur in the robotic milling process due to the low stiffness of robot structure,the complex configuration of robotic end load and local weak rigidity of workpiece,which seriously affects the machining surface quality.Robotic milling stability mainly depends on the dynamic characteristic of the tool tip,therefore a tool tip frequency response function(FRF)prediction method for robot milling tasks is proposed,which could predict the tool tip FRF at any robot posture to realize the stability prediction of robotic milling process,then the workpiece machining quality could be largely improved by optimizing the process parameters.Firstly,the experimental amount is reduced and the requirements for experimental postures are lowered by introducing the nonlinear least square weighted superposition method.Then the tool tip FRF of robot at any target posture in a certain workspace can be predicted by giving the tool tip FRF of any amount of benchmark experimental postures.Furthermore,a linearized modal parameter superposition method is proposed to avoid the multi-mode phenomenon caused by the direct superposition of FRF,which could largely improve the prediction accuracy.Finally,the accuracy and practicability of the model are verified by robot milling experiments,and then the workpiece machining quality is largely improved by the further optimization of the robotic milling process parameters.