Abstract
As the cutting stone is a wear process, performing this process with diamond pieces' aid can be considered the wear of stone particles bypassing diamond grains on its surface. To better understand this process as well as the conditions governing the cutting diamond grain, it is necessary to familiar with the cutting mechanism along with the affecting parameters. In this matter, predicting the amount of segment consumption plays a prominent role to estimate the production cost as well as to schematize the building stone mines. This paper utilized the data obtained from Carbonate and Granite stones to estimate the amount of consumption of diamond cutting wire segments. To do so, two methods, namely support vector regression (SVR) and genetic algorithm + Multilayer perceptron (GA-MLP) were chosen using the MATLAB software toolboxes in order to estimate the segment erosion. In each of the above algorithms, a lowpass smoothing filter, called Savitzky-Golay was employed on the data. For this purpose, three rock properties including uniaxial compressive strength, Shimazk friction factor, and Young's modulus, were also employed as the model's input. After that, twelve models were constructed and then the segment erosion was estimated as well. Ultimately, the accuracy of the above models was assessed using the coefficient of determination (\({R}^{2}\)), mean square error (MSE), root mean square error (RMSE), mean value absolute error (MAD), mean absolute percentage error (MAPE), and variance of factor analysis (VFA). According to the obtained results, it can be concluded that the SVR approach and the Savitzky-Golay filter with Polynomial Kernel could better estimate the wear rate of the diamond cutting wire segment.