基于SABO优化VMD与K-means++的机器人磨削颤振识别

Chatter Recognition of Robotic Grinding Process Based on SABO Optimized VMD and K-means++

机器人由于低刚度特性导致加工中极易产生颤振,针对颤振特征频率提取与颤振识别问题,提出基于减法平均优化算法(SABO)对变分模态分解(VMD)中关键参数进行优化,筛选颤振敏感IMF分量并重组;根据颤振信号的频谱特性构建基于功率谱熵差(ΔPSE)的颤振识别指标,采用K-means++算法对不同颤振类型进行辨识。实验结构表明,所提出的SABO-VMD-K-means++方法可以准确识别机器人磨削加工颤振类型,为机器人磨削颤振监测提供一定的指导。

Due to low stiffness characteristics,the robot is susceptible to chatter vibration during machining.To address the issues of feature frequency extraction and recognition of chatter vibration,a subtraction-average-based optimizer(SABO)is proposed to optimize key parameters in VMD,allowing for the selection and recombination of chatter-sensitive IMF components.Furthermore,a vibration recognition index based on the power spectral entropy difference(ΔPSE)is constructed,taking into account the spectral characteristics of the vibration signal.The K-means++algorithm is employed to distinguish different types of chatter vibrations.Experimental results demonstrate SABO-VMD-K-means++method can accurately identify the types of chatter vibration in robot grinding processes,providing valuable guidance for chatter vibration monitoring in robot grinding operations.