Monitoring of tool wear using feature vector selection and linear regression
ICNC'05 Proceedings of the First international conference on Advances in Natural Computation - Volume Part II
Effect of SVM kernel functions on classification of vibration signals of a single point cutting tool
Expert Systems with Applications: An International Journal
| Hi-index | 0.00 |
When neural networks are utilized to identify toolstates in machining process, the main interest is often onthe recognition ability. It is usually believed that a higherclassification rate from pattern recognition can improvethe accuracy and reliability of tool condition monitoring(TCM), thereby reducing the manufacturing loss.Nevertheless, the two objectives are not identical in mostpractical manufacturing systems. The aim of this paper isto address this issue and propose a new evaluationfunction so that the recognition ability of TCM can beevaluated more reasonably. On this basis, two kinds ofmanufacturing loss due to misclassification are analyzed,and both of them are utilized to calculate correspondingweights in the evaluation function. Then, the potentialmanufacturing loss is introduced in this work to evaluatethe recognition performance of TCM. On the basis of thisevaluation function, a modified support vector machine(SVM) approach with two regularization parameters isutilized to learn the information of every tool state. Theexperimental results show that the proposed method canreliably carry out the identification of tool flank wear,reduce the overdue prediction of worn tool conditions andits relative loss.