摘要
随着加工精度要求不断提高,切削过程中,对刀具在机磨损或崩刃状态进行在机实时监测的需求日益增加。本文以声发射和主轴功率为监测信号,通过提取时域、频域和时频域的有效特征,构建了基于融合信号的平底立铣刀在机崩刃SVM监测模型;采用网格搜索、粒子群和遗传算法优化SVM模型参数,并在实际切削环境中,将平底立铣刀的崩刃监测效果进行对比。结果表明:基于遗传算法优化的SVM模型对铣刀崩刃状态监测效果最佳。
With the continuous improvement of machining accuracy requirements,there is an increasing demand for real-time monitoring of tool wear or blade breakage during cutting.In this paper,acoustic emission and spindle power are used as monitoring signals.By extracting effective features in time domain,frequency domain and time-frequency domain,a SVM monitoring model based on fusion signal is constructed.Mesh search,particle swarm optimization and genetic algorithm were used to optimize SVM model parameters,and the monitoring effect of flat end milling cutter was compared in actual cutting environment.The results show that the SVM model based on genetic algorithm optimization has the best effect on the state monitoring of milling cutter breakage.
作者
张曦
周青峰
张龙佳
郑文妞
ZHANG Xi;ZHOU Qingfeng;ZHANG Longjia;ZHENG Wenniu
出处
《计量与测试技术》
2024年第2期92-95,99,共5页
Metrology & Measurement Technique
关键词
声发射
主轴功率
崩刃检测
遗传算法
SVM模型
acoustic emission
spindle power signals
blade breakage monitoring
genetic algorithm
SVM model
