Abstract
A method of tool condition monitoring based on the second generation wavelet transformation and hyper-sphere support vector machine was proposed to improve classifying precision of tool condition in the process of machining. Cutting force signal and vibration signal were filtered by the second generation wavelet transformation. Lots of features were extracted by wavelet packet transformation in the time domain and frequency domain analysis. The sensitive features to tool condition as the average value, the mean square root, wavelet coefficient were selected by principal component analysis. The feature vector made up by the features was inputted to hyper-sphere support vector machine, which built the relation between tool condition and features to predicting tool condition automatically. The experimental results show hyper-sphere SVM are of excellent study ability, generalization ability, and of high recognized precision with small training samples.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call