Research In Tool Wear Monitoring Method Of Drilling Power Head

Drilling is a general kind of manufacturing metal machining process of manufacturing industry. As a general device of drilling, the drilling power head is widely used in industrial automation production line. Twist drill is a common tool of drilling, the wear of which not only results in large deviation of the workplace in size, but also probably causes the halting of the whole line and decline of the efficiency, even the damage of the machine. There is an urgent requirement of the monitoring system for the automation production line that can both monitor wear condition of drilling power head and warn in time before accidents of wear failure. Therefore, it is of great significance in researching and designing the monitoring system for the tool wear of drilling power head.First of all, the paper makes a brief introduction of the structure and the wear condition of Twist Drill, which can be classified into:normal, excessive wear and broken. Then the FEA software, Deform, is applied to simulate and analyze the coefficients of force and energy during the drilling process.Later, with the drilling force and the motor current signal as monitored objects, a survey platform is designed to sample and analyze. Time-frequency method and wavelet transform are used to submit the character of monitored signals.At last, two kinds of artificial neural network, BP model and RBF model, are used to identify the wear condition of the tool. Models are made after the training of several sample data and the identification is further tested by a large sum of data. With the sample data of two kinds of artificial neural network compared, conclusions can be drawn that the convergence and training efficiency of BRP model is superior to that of BP model. The submitted character can better reflect the tool wear condition, In addition, SVM model decline the difference between the experience risk and expected rick.