基于模糊支持向量机的刀具磨损检测

Tool Wear Detection Based on Fuzzy Support Vector Machine

提出一种基于模糊支持向量机(FSVM)的切削过程中刀具磨损检测方法,对切削加工过程中的刀具磨损状态进行诊断与预测。提取切削加工过程中刀杆的振动信号和切削刀具的切削力信号,对其进行分帧处理,提取FFT特征量,对该特征向量进行模糊支持向量机的学习和训练。实验结果表明,该方法能够充分发挥模糊支持向量机的权系数作用,有效检测切削过程刀具的磨损程度,与同类识别方法的识别结果相比较,具有一定的优越性。

A tool wear detection method based on Fuzzy Support Vector Machine （FSVM） is presented for diagnosing and predicting the tool wear in the cutting process. The signal of the toll holder vibration and the cutting force in cutting process is extracted and performed sub - frame processing. Then the FFT feature vectors are achieved to realize the training and learning of FS- VM. The experiment results show that in the proposed method the weight coefficients of FSVM play an important role. The method can effectively detect the tool wear degreen and outperforms the existing recognition method.