具备高适应性的高速铣削刀具磨损状态监测系统

Intelligent Tool Wear Condition Monitoring System with High Adaptability in High Speed Milling Process

为提高加工监测系统的适应性和智能化程度,提出基于刀具磨损曲线的实时刀具状态监测系统。自学习能力的引入使该系统可自动进行不同刀具状态的识别和磨损程度的估计,较大程度上摆脱对系统事先"教学"的依赖。同时为有效抑制切削参数变化带来的干扰,提出一种特征提取方法来自动提取敏感特征,减少监测系统开发时间和成本。针对高速铣削过程的刀具磨损监测,采用切削力和声发射传感器来采集信号,并运用时域、频域和小波分析技术来对信号进行处理,试验结果证明了所提出的自动特征提取方法的有效性和智能刀具状态监测系统的高适应性。

To enhance the adaptability of tool condition monitoring（TCM） system, an novel and intelli- gent method is proposed for automatic identifying the different tool wear states and estimating the wear value with no need of the pre-designed ＂teaching＂ or ＂training＂ phase. Automatic sensory feature selection method is used to aid the systematic design of TCM, and to suppress interference introduced by changes of cutting parameter. Force and acoustic emission sensors are used in high speed milling opera- tions. The time domain, frequency domain and wavelet analysis techniques are applied to processing the signals. The real-time intelligent monitoring system is built on the cycle process of linear fitting and Ma- halanobis distance （MD） calculating. A series of experiment application on a CNC vertical milling ma- chine tool show that the proposed method is accurate for feature extraction and efficient for condition monitoring of cutting tools.