Integrated System for Intelligent Monitoring of Accuracy and Repeatability on Numerically Controlled Machine Tools Supported by Machine Learning Techniques

The paper presents the results of measurement errors in the position of the kinematic center of the rotary axes of five-axis machine tools with a tilting rotary table, using the modern R-Test calibration system. The use of the system was the basis for obtaining experimental data enabling the assessment of changes in the position of the kinematic center of the rotary axes of the tested machine tools. The estimation was made based on the statistical analysis of the obtained measurements. The measurements made it possible to determine the position errors of the rotary axes A, B, C in the Y-Z, X-Z, X-Y planes and the average values of the deviation of the reference ball in specific axes of the machine tools. The parameters of the reference ball deflection in the machine tool space were simulated using 3 models – the MLP, KAN and MOGP. The simulation results showed that the MOGP model performed best in estimating the deviation parameters. Moreover, the MOGP model achieved the best values in all metrics compared to other tested models. Based on the obtained research results, an expert system will be built to supervise the operation of machine tools, enabling the acquisition of knowledge about the technical condition of CNC machine tools.