Combined Artificial Neural Network and Taguchi Technique for Selection of Optimal Process Parameters in Steel Roll Grinding

In industrial manufacturing, grinding processes are used for finishing of workpieces. The work rolls used in Sendzimir cold roll mills were ground in the roll grinding shop to remove the marks formed in the surface of the work rolls during rolling process. Since Sendzimir mills were driven by contact friction, the work roll should have suitable roughness for thickness reduction. To obtain the average roughness in work rolls consistently, the process has to be optimized and the optimal level of the factors has to be determined. This paper deals with optimization of grinding parameters to obtain desired surface roughness in the work rolls using Neural Network-Taguchi approach. In this study, there are six factors wheel speed, workspeed, traverse speed, infeed, dress depth and dressing lead were considered to obtain a optimum grinding condition using Taguchi Techniques with L27 orthogonal array. An attempt was made to minimize the number of experimental runs and increase the reliability of experimental results. Artificial neural network (ANN) model was developed to enhance the prediction accuracy. The confirmation test was conducted to verify the results obtained from Taguchi Technique and Neural network model. Analysis of Variance (ANOVA) was conducted to identify the significant parameters during grinding process.