Drill flank wear estimation using supervised vector quantization neural networks

摘要: Drill wear detection and prognosis is one of the most important considerations in reducing cost rework scrap to optimize tool utilization hole making industry. This study presents development implementation two supervised vector quantization neural networks for estimating flank-land size a twist drill. The algorithms are; learning (LVQ) fuzzy (FLVQ). input features were extracted from vibration signals using power spectral analysis continuous wavelet transform techniques. Training testing performed under variety speeds feeds dry drilling steel plates. It was found that FLVQ more efficient assessing flank than LVQ. experimental procedure acquiring data extracting time-frequency domain detailed. Experimental results demonstrated proposed network effective drill wear.