The Research Of Metal Removal Model In Field Controlled Electro-chemical Honing On Artificial Neural Network

Field Controlling Electro-Chemical Honing (FCECH) is a new brand of machining method. Microcomputer is in real tune controlling the electrolytic field at the electro-chemical (EC) dissolution spot of workpiece,which is scanning over the workpiece during the machining process,so that the rate of metal removal is varying during the scanning. With the technology of FCECH,the geometrical errors of the workpiece can be quickly corrected and excellent surface quality can be easily obtained at the same time. The artificial neural network (ANN) technology is used to build the relation between the machining parameters,such as EC current and EC gape,and metal removal rate. The metal removal model of the machining system can be established on ANN.Based on the Faraday's law,research the basic theory and metal removal mechanism in FCECH. According to the basic theory of FCECH,submit the EC current opened-loop control technology and EC current closed-loop controlling technology,and establish the mathematical model of quasi-stable and dynamic machining removal law.According to the experimental requirement,build the FCECH machining system,so that cathode can scan over the surface of the workpiece,electrolyte can be supplied circularly and electrolytic gap can be adjusted in hand. Design the specific machining way of EC current closed-loop control technology on inexact differential PID algorithm,develop the EC current acquirement,amplification and the computer real time collecting passageway,and design relevant controlling program.As the electric medium,electrolyte is the main factor,which affects the machining efficiency,machining precision and surface quality of workpiece. Through analyzing and comparing large of experimental results,compound five different kind of electrolytes. Find out the main factor of affecting the metal removal rate and current efficiency,such as EC gape,EC current and honing pressure. Offer the ideal scope of different electrolytic parameter for experiment.Based on the powerful nonlinear model building capacity,the artificial neural network is used to establish the metal removal model of quasi-stable and dynamic machining removal law hi FCECH,and it is proved on experiments that the metal removal model is correct. Research and establish the dynamic responsive model between EC voltage and EC current,find and reveal the phenomenon of "starting withdelay and immediate stop" taking place between EC current and metal removal. The machining precision can be improved greatly after considering the influence of "starting with and immediate stop" on metal removal.The metal removal model on neural network,which has been built,is to calculate the distribution H(X) of metal removal on the basis of the distribution l(x) of EC current on the workpiece. It is always required in production that calculating the distribution /(x) of EC current according to the error and removal distribution E(X) of workpiece. A practical B - P network model is established to calculate l(x) from E(X) ,and a correlative program is designed to calculate the distribution l(x) of EC current.With the metal removal model built on artificial neural network,serials of experiments are carried out on the machining system. The results show that the machining precision is up to 3jm and the roughness Ra is within 0.03 8/zra. It can be concluded that the principle of FCECH is correct,machining system has high precision and good stability,and FCECH as a special technology will have a bright future.