Performance Prediction of Diamond Sawblades Using Artificial Neural Network and Regression Analysis

This paper is concerned with the application of artificial neural networks (ANNs) and regression analysis for the performance prediction of diamond sawblades in rock sawing. A particular hard rock (granitic) is sawn by diamond sawblades, and specific energy (SE) is considered as a performance criterion. Operating variables namely peripheral speed (V (p)), traverse speed (V (c)) and cutting depth (d) are varied at four levels for obtaining different results for the SE. Using the experimental results, the SE is modeled using ANN and regression analysis based on the operating variables. The developed models are then tested and compared using a test data set which is not utilized during construction of models. The regression model is also validated using various statistical approaches. The results reveal that both modeling approaches are capable of giving adequate prediction for the SE with an acceptable accuracy level. Additionally, the compared results show that the corresponding ANN model is more reliable than the regression model for the prediction of the SE.