On-line & Real-time Monitoring System Of Micro-hole Drilling Based On Wavelet Technology And Neural Network

With the development of mechanical-electronic products for accurate, integraland miniature applications, there is widely applied for micro-holes. At presentthere are many methods for machining micro-holes, however, machining by usingdrills tends to be the most useful in practice. When drills are used to machinemicro-holes there is a serious drawback in that intensity of micro drills is low andbreak easily when they are wear away. Besides this drawback, the statisticalscatter of micro-drill lives is large and it is very difficulty to predict just how longthe working life of a micro-drill might be. Once drills break, it is difficult toextract trapped part from workpiece, thereby usually causing workpiece wastage.Experiment shows that increment of drilling force when the drill is worn away,including thrust and torque, is main reason. To prevent breakage of micro drills, asuit of the on-line and real-time micro-drilling force monitoring system based onArtificial Neural Network technology was built up in this dissertation.In this paper,On-line & Real-time Monitoring System of Micro-hole DrillingBased on Virtual Instruments is built on LabVIEW 7.0 with the virtualinstruments technology. The system possesses various functions includeddate of drill forces signal collection and controlling step-machine。Using this system ,you can do FFT analysis and Wavelet Transformanalysis to the signal collected and do Neural Network training withthe historical date as well as estimate the state of aiguille and makedecision with the trained Artificial Neural Network.On the based of collecting signal of micro-hole drilling force ,thrust and torque are analyzed , low-frequency component are picked up as thecharacteristic vectors who represent the state of aiguille in this paper.High-frequency component in the torque signal are analyzed in virtue of WaveletTransform, added component of the Wavelet Transform to the dynamic...