Milling process diagnosis using computational intelligence methods

Abstract

The paper presents the results of research aimed at developing a method for hard-to-machine metal alloy milling process diagnosis using computational intelligence methods. To diagnose the process, a signal from an accelerometer mounted on the spindle of a CNC machine was used. The data were recorded during milling of Inconel 625 alloy workpieces, performed by sharp and blunt cutters. The acceleration signal metrics, both in the time and frequency domains were used to develop the classifiers. Based on the experiments, it has been demonstrated that it is possible to effectively diagnose Inconel alloy workpieces milling process using shallow computational intelligence methods (decision trees, k-NN and linear support vector machines). Python was used for data processing and visualisation as well as classifiers development and testing.