The objective of this study is to develop an automated tool condition monitoring scheme for PCB drilling.
Vibration signals are used to distinguish micro drill wear stages with proper features extraction and classifier design. Then a tool condition monitoring system is built up through a back propagation neural network (BPNN).
Experimental results show that BPNN is a practical method of modeling tool wear, and with this method a tool condition monitoring system is built up using energy ratio, root mean square (RMS) and kurtosis coefficient that transformed by vibration signals.
In the further investigation, more signal samples should be computed as monitoring features for BPNN modeling. In addition, in order to build the best monitoring model, it is necessary to evaluate the performance of the BPNN model in advance, and optimize the process.
The paper provides a method and a system for PCB drill wear monitoring. The method and system can achieve on‐line monitoring of PCB drill condition.
