Application of a wavelet fuzzy neural network in microdrilling online monitoring

A wavelet fuzzy neural network has been developed in this study for monitoring microdrilling, based on the combinative application of fuzzy logic and error back-propagation network, whose inputs come from some characteristic signals by means of time-domain analyzing and wavelet decomposing. The monitoring model has the advantages of fast convergences and exact description of the nonlinear relationship between signal features and microdrill breakage. Data sampled in real time from the thrust and the torque during microdrilling were used for training and testing the model, the online monitoring system of microdrilling on virtual instrument was processed to monitor the real-time microdrill wear. The experiment results show that if the threshold is properly selected, the microdrill breakage will be effectively avoided by monitoring.