Data-Driven Thermal Deviation Prediction in Turning Machine-Tool - A Comparative Analysis of Machine Learning Algorithms

Thermal error significantly impacts the machining precision of machine-tools. deformations in machine-tool structure caused by various machine heat sources is at origin this phenomenon. In order to ensure expected quality parts, manufacturer have run machine-tools for hours before start producing reach thermal stability. This heating phase has a high negative impact on productivity one hand and its ecological footprint other. paper presents data-driven approach model predict correct tool reference position accordingly. The automatic adjustment allows produce parts with regardless state machines, which substantially increase their productivity. For purpose, temperature sensors as well measurement instruments are deployed Tornos SwissNano4 machine-tool. A set experiments conducted collect data related these two measurements. Four major Machine Learning algorithms trained using subset collected tested remaining subset. Quantitative comparative analysis shows that three four prediction mean Absolute Error (MAE) below 1µm Correlation Coefficient higher than 90%. Even classical linear regression models able accuracy. Advanced techniques show potential provide better