Research On Penetration Detection Control Technology Of EDM Small-hole Machine Tool Based On Classification Algorithm

EDM small hole machining is an important branch of EDM.Due to the constant distance between the tool electrode and the workpiece during the EDM process,it is particularly suitable for machining materials with higher hardness.Therefore,it is widely used in the machining of molds in the manufacturing industry and blades in the aerospace industry.In the actual process of using electric discharge machining for small holes,not only the parts of the workpiece that need to be machined will be corroded by electricity,but the tool electrode itself will also be affected by the energy of electric discharge discharge,resulting in certain losses.This phenomenon makes the commonly used deep machining method only suitable for situations with low machining accuracy,and for situations with high precision requirements,such as the machining of turbine blades,Using traditional methods is prone to the phenomenon of small hole diameters on the back of workpieces not meeting the requirements and damaging the workpiece.Therefore,effective penetration detection of the machining process is the key to improving the efficiency and quality of electric discharge small hole machining.The specific research on the penetration detection method for electric discharge small hole machining in this article is as follows:(1)Based on the actual use of the upper computer numerical control system of the machining machine,it is decided to use FPGA as the main control chip to achieve communication and transmission functions with the upper computer numerical control system,and to achieve compatibility and adaptation with the actual machine tool.(2)By analyzing the processing parameter instructions and discharge functions issued by the actual machine tool upper computer,in order to meet the processing requirements under different processing conditions and workpiece processing,each discharge function is designed with a modular approach,and the accurate implementation of the function is verified through simulation.(3)By analyzing the waveform of inter pole machining voltage during the process of small hole machining,it was found that the number of changes in machining voltage p is influenced by machining parameters,and there is a phenomenon of significant changes before and after penetration.The experimental data was iteratively trained and learned using the BP neural classification algorithm to obtain a penetration detection classification model in MATLAB,and the impact of FPGA on-chip resources and algorithm time consumption was comprehensively considered,The method of hardware description language programming is transplanted into FPGA chip,and the feasibility and accuracy of BP neuron classification algorithm in FPGA are verified through simulation.(4)Based on actual EDM small hole machining machines,conduct corresponding discharge and penetration detection experiments to verify the accuracy of each function during the discharge process.By analyzing the penetration detection signal and processing voltage waveform during the machining process,a penetration detection model is obtained that only needs to collect processing voltage and is suitable for various processing conditions using the BP neural classification algorithm.This simplifies the design of the circuit and the complexity of collecting multiple types of signals.The accuracy of using this classification algorithm model to determine the penetration detection that occurs within 2mm of the tool electrode protruding from the machining workpiece can reach 93.3%.