Research Of Grinding Surface Roughness Based On Artificial Neural Networks

The technology of NC profile grinding is a new kind of process technologies developed rapidly both at home and abroad in recent years, its high-efficiency and accuracy has extremely extensive application prospects on the mechanical engineering, but the research on surface quality of NC profile grinding is fewly reported. With the development of computer simulation technology, the artificial neural network is applied in grinding field. Based on that, this paper discussed the relation between the surface roughness of ground workpiece and grinding parameteres. It is convenient for the research of surface roughness and appliance with the combine of the computer technology and artifical neural networks, comparing to the traditional regress analysis.Firsly, based on the literature reviews, this paper introduces the development tendency of the grinding process technology. The paper summarizes the actual status of profile grinding, and briefly introduces research on the surface roughness recently. Also, the experiment on the surface roughness is introduced in this paper. The orthogonal experiment is selected to design the experiment in order to reduce the experimental number of times, save the experimental expense, also can obtain the quite precise experimental result. The regress equation which reflects the relation of surface roughness and grinding parameteres is modeled based on the experiment data. It is proved to be viable of the regress equation by compare to the experiment data. According to the research, the influence of velocity of the grinding wheel to the surface roughnss behaves the most evidence, followed by the grinding depth, and the last by the velocity of the work piece.By summarizing the application of artificial neural networks in grinding field in recent years, BP network is selected to model the relation between the surface roughness and grinding parameteres. Compared to the regress model, it is thought that the BP network model has certain superiority in establishing the model and predicting the surface roughness. Meanwhile, for convenient to establish BP network model, reduce the time of programming, improve working efficiency, it is briefly introduced that how to use the GUI to realize establishing the model and predicting the surface roughness in this paper.