摘要
通过电流传感器分别测量驱动正常铣刀以及发生破损铣刀旋转的主轴电机电流,将其作为原始采集信号,并对该信号进行消噪处理。其次,利用消噪后的电流信号数据,以MATLAB分析时域均方根和信号功率谱密度随频率的分布情况,从而提取刀具破损特征。最后,针对提取到的样本特征,应用最小二乘支持向量分类机模型进行刀具破损模式识别,实验结果表明,通过小波消噪、burg功率谱特征提取和应用最小二乘支持向量分类机判断,刀具破损状态的识别准确率可达95%。
The current of spindle motor respectively with normal tool and broken tool is measured by an electric current transducer, which is taken as original signal, and de-noised. Then the current signal data after de-noising is used in MATLAB to analyze the distribution of root mean square in time domain and signal power spectral in frequency domain, so as to obtain tool breakage features. Through the extracted sample characteristics, the model of the tool breakage pattern is identified by the least squares support vector classifier model. The experimental results show that, after wavelet noise reduction, burg power spectral extraction and the least squares support vector classifier, the accuracy rate reaches 95%.
作者
王欣
强云玥
Wang Xin;Qiang Yunyue(School of Mechanical Engineering,Shanghai University for Science and Technology,Shanghai 200093,China)
出处
《农业装备与车辆工程》
2018年第10期65-70,共6页
Agricultural Equipment & Vehicle Engineering
关键词
刀具破损
电流测量
特征提取
小波消噪
最小二乘支持向量机
tool breakage
electric current measurement
feature extraction
wavelet noise reduction
the least squares support vector classifier
