Monitoring Drilling Wear States by a Fuzzy Pattern Recognition Technique

Abstract

This paper introduces a new approach for on-line monitoring of drill wear states by using a fuzzy C-means algorithm. Experimental and simulation results have shown that drill wear conditions can be represented by four fuzzy grades. They are: “initial,” “small,” “normal,” and “severe.” The grade “severe” is proposed to be used as the prediction of tool replacement. This fuzzy technique is more adequate than conventional pattern recognition technqiues.