3D Convolutional Neural Network to Predict the Energy Consumption of Milling Processes

Christoph Wald, Thomas Jung, Frank Schirmeier

2025

Abstract

With increasingly fluctuating energy prices, the energy-flexible operation of electrical consumers, including machine tools, has recently gained attention. This study aims to predict the energy consumption from the shape of the volume removed during a time step in milling, generated using a time-discrete simulation environment. A 3D residual network is used to analyze the voxel representation of these ”removed volumes”. In total, 48 unique combinations of cutting depth and feed rate are recorded on a three-axis mill to evaluate the proposed model. The results indicate that energy consumption prediction using these shapes is possible.

Paper Citation

in Harvard Style

Wald C., Jung T. and Schirmeier F. (2025). 3D Convolutional Neural Network to Predict the Energy Consumption of Milling Processes. In Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA; ISBN 978-989-758-758-0, SciTePress, pages 21-30. DOI: 10.5220/0013457000003967

in Bibtex Style

@conference{data25,
author={Christoph Wald and Thomas Jung and Frank Schirmeier},
title={3D Convolutional Neural Network to Predict the Energy Consumption of Milling Processes},
booktitle={Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA},
year={2025},
pages={21-30},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013457000003967},
isbn={978-989-758-758-0},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Conference on Data Science, Technology and Applications - Volume 1: DATA
TI - 3D Convolutional Neural Network to Predict the Energy Consumption of Milling Processes
SN - 978-989-758-758-0
AU - Wald C.
AU - Jung T.
AU - Schirmeier F.
PY - 2025
SP - 21
EP - 30
DO - 10.5220/0013457000003967
PB - SciTePress