Online Monitoring of Micro-Hole Drilling Based on Data-Driven Force Analysis

Due to the high quality, high efficiency, and low-processing cost, micro-hole drilling is widely used, but it is difficult to solve the problem that the micro-drill is low in strength, poor in rigidity, and is easy to break. Based on a large number of drilling experimental data, it was found that the main indictor for micro-drill fracture was increased drilling force as drill wear progressed. Since the drilling force signals can accurately represent the degree of wear of the micro-drill. It is proposed to establish a wavelet fuzzy neural network monitoring model based on data-driven drilling force analysis. The model integrates wavelet analysis technology, neural network decision technology, and fuzzy control technology, so that the monitoring system can simultaneously have low-level learning, computing power, fuzzy system’s high-level reasoning, and decision-making ability of the neural network. Through the offline network training and testing of a large number of experimental data, the robust monitoring model is obtained, and the online monitoring of micro-hole drilling is realized. The results show that it is feasible to use the wavelet fuzzy neural network for the online monitoring and interpretation of micro-hole drilling force, and by the appropriate selection of the monitoring threshold, micro-drill fracture can more effectively be avoided.