CALIBRATION OF MACHINE COMPONENTS EQUIPPED WITH STRAIN GAUGES USING MACHINE LEARNING ALGORITHMS TO CONTROL THE ACCURACY OF MACHINING

Morev Artem Sergeevich

When strain gauges are used, measurement errors occur due to thermal and physical deformations at different force application points in the workspace. This paper examines whether machine learning approaches are suitable for compensating such measurement errors. Two approaches, neural network and multiple linear regression, are investigated. The developed programmes receive as input signals from a force measurement unit equipped with strain gauges and produce as output calculated values for the acting force vector. Finally, the results of the two algorithms used are compared.

Keywords:strain gauges; calibration of machine tool systems; machine learning; multi-axis sensors; measurements