模式识别在刀具状态监控中的应用

Application of Pattern Recognition in Power Monitoring of Tool Condition

通过大量试验，研究了刀具磨损、破损对功率信号时域和频域特性的影响．抽取与刀具状态相关的特征量──功率的时域幅值和主轴转频处的幅值，并将模式识别技术应用于刀具状态多特征监控，能较好地识别刀具的磨损和破损，显著提高了监控的准确性．

The effect of tool wear and breakage on the feature of power signal in time and frequency domain is studied based on a large amount of experiments, and it is indicated that there is a good relationship between tool condition and power consumption or magnitude at the spindle rotation frequency. The application of pattern recognition in tool condition monitoring using the above two relationships is explored, and it is proved that both tool wear and breakage can be identified more precisely by this method.