The States Identification Of Oxide Layer On The Wheel In ELID Grinding Process

In recent years, with the rapid developing of microelectronics, optics and computer technique, a lot of material with special function such as silicon, optics glass, project ceramic are widely used in ultraprecision field. ELID ultraprecision grinding is one of the ultraprecision machining methods developed recent years, which resolve the problem that ultrarigid minuteness grain wheel is easily jammed and hard to dress, and it actualize the ultraprecision machining of the hard and brittle materials.The reason why ELID grinding can actualize low roughness is the oxide layer on the wheel which generate in the electrolysing process. The states of the oxide layer directly affect the ultraprecision grinding process. In this dissertation, we design a special wheel with sampling equipment and simulate the process of ELID grinding,in order to analyze the appearance and component of oxide layer in different electrical parameter,we observing the samples with Scanning Electron Microscope. Based on Labview software, we write the program that actualize the sampling, analyzing and saving of current, vibration, grinding force and AE signal. Based on the wavelet analyze theory, using MATLAB software, putting forward the method that using grinding force and current in the ELID grinding process to identify the states of oxide layer.Based on Adaptive Network-based Fuzzy Inference System(ANFIS) theory and Multi-sensor Information Fusion Technique, build a network-based fuzzy inference system, with grinding force and current as input parameter, and use the times of conglutination as the output parameter, to accomplish ANFIS training, and this network pass the checking data's test,and obtain preferable effect in simulation.