Data-driven prediction of tool wear using Bayesian regularized artificial neural networks

Tam T. Truong; Jay Airao; Faramarz Hojati; Charlotte F. Ilvig; Bahman Azarhoushang; Panagiotis Karras; Ramin Aghababaei

doi:10.1016/j.measurement.2024.115303

Data-driven prediction of tool wear using Bayesian regularized artificial neural networks

Tam T. Truong, Jay Airao, Faramarz Hojati, Charlotte F. Ilvig, Bahman Azarhoushang, Panagiotis Karras, Ramin Aghababaei^*

^*Corresponding author for this work

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaper › Journal article › Research › peer-review

Abstract

The prediction of wear in cutting tools is pivotal for boosting productivity and reducing manufacturing costs. Although current data-driven models in machine learning and deep learning have advanced predictive capabilities, they often lack generality and demand substantial data training. This paper presents a novel approach using Bayesian Regularized Artificial Neural Networks (BRANNs) to precisely forecast wear in milling tools. Unlike conventional machine learning models, BRANNs merge the strengths of artificial neural networks (ANNs) and Bayesian regularization, yielding a more robust and generalized predictive model. We utilized three open-access datasets from the literature alongside an in-house dataset generated by our milling setup. Initially, we assessed the model's predictive ability by training and testing it against individual open-access datasets. We investigated the impact of input features, training data size, hidden units, training algorithms, and transfer functions on the model's predictive capability. Subsequently, we trained the model using three open-access datasets and tested it against our in-house data.

Original language	English
Article number	115303
Journal	Measurement: Journal of the International Measurement Confederation
Volume	238
ISSN	0263-2241
DOIs	https://doi.org/10.1016/j.measurement.2024.115303
Publication status	Published - Oct 2024

Keywords

Artificial neural network
Bayesian regularization
Cutting force
Machine learning
Tool wear

Access to Document

10.1016/j.measurement.2024.115303Licence: CC BY

Cite this

@article{aecf44852c93452a94fb57f86e1494c7,

title = "Data-driven prediction of tool wear using Bayesian regularized artificial neural networks",

abstract = "The prediction of wear in cutting tools is pivotal for boosting productivity and reducing manufacturing costs. Although current data-driven models in machine learning and deep learning have advanced predictive capabilities, they often lack generality and demand substantial data training. This paper presents a novel approach using Bayesian Regularized Artificial Neural Networks (BRANNs) to precisely forecast wear in milling tools. Unlike conventional machine learning models, BRANNs merge the strengths of artificial neural networks (ANNs) and Bayesian regularization, yielding a more robust and generalized predictive model. We utilized three open-access datasets from the literature alongside an in-house dataset generated by our milling setup. Initially, we assessed the model's predictive ability by training and testing it against individual open-access datasets. We investigated the impact of input features, training data size, hidden units, training algorithms, and transfer functions on the model's predictive capability. Subsequently, we trained the model using three open-access datasets and tested it against our in-house data.",

keywords = "Artificial neural network, Bayesian regularization, Cutting force, Machine learning, Tool wear",

author = "Truong, {Tam T.} and Jay Airao and Faramarz Hojati and Ilvig, {Charlotte F.} and Bahman Azarhoushang and Panagiotis Karras and Ramin Aghababaei",

note = "Publisher Copyright: {\textcopyright} 2024 The Author(s)",

year = "2024",

month = oct,

doi = "10.1016/j.measurement.2024.115303",

language = "English",

volume = "238",

journal = "Measurement: Journal of the International Measurement Confederation",

issn = "0263-2241",

publisher = "Elsevier BV",

}

Data-driven prediction of tool wear using Bayesian regularized artificial neural networks. / Truong, Tam T.; Airao, Jay; Hojati, Faramarz et al.
In: Measurement: Journal of the International Measurement Confederation, Vol. 238, 115303, 10.2024.

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaper › Journal article › Research › peer-review

TY - JOUR

T1 - Data-driven prediction of tool wear using Bayesian regularized artificial neural networks

AU - Truong, Tam T.

AU - Airao, Jay

AU - Hojati, Faramarz

AU - Ilvig, Charlotte F.

AU - Azarhoushang, Bahman

AU - Karras, Panagiotis

AU - Aghababaei, Ramin

PY - 2024/10

Y1 - 2024/10

N2 - The prediction of wear in cutting tools is pivotal for boosting productivity and reducing manufacturing costs. Although current data-driven models in machine learning and deep learning have advanced predictive capabilities, they often lack generality and demand substantial data training. This paper presents a novel approach using Bayesian Regularized Artificial Neural Networks (BRANNs) to precisely forecast wear in milling tools. Unlike conventional machine learning models, BRANNs merge the strengths of artificial neural networks (ANNs) and Bayesian regularization, yielding a more robust and generalized predictive model. We utilized three open-access datasets from the literature alongside an in-house dataset generated by our milling setup. Initially, we assessed the model's predictive ability by training and testing it against individual open-access datasets. We investigated the impact of input features, training data size, hidden units, training algorithms, and transfer functions on the model's predictive capability. Subsequently, we trained the model using three open-access datasets and tested it against our in-house data.

AB - The prediction of wear in cutting tools is pivotal for boosting productivity and reducing manufacturing costs. Although current data-driven models in machine learning and deep learning have advanced predictive capabilities, they often lack generality and demand substantial data training. This paper presents a novel approach using Bayesian Regularized Artificial Neural Networks (BRANNs) to precisely forecast wear in milling tools. Unlike conventional machine learning models, BRANNs merge the strengths of artificial neural networks (ANNs) and Bayesian regularization, yielding a more robust and generalized predictive model. We utilized three open-access datasets from the literature alongside an in-house dataset generated by our milling setup. Initially, we assessed the model's predictive ability by training and testing it against individual open-access datasets. We investigated the impact of input features, training data size, hidden units, training algorithms, and transfer functions on the model's predictive capability. Subsequently, we trained the model using three open-access datasets and tested it against our in-house data.

KW - Artificial neural network

KW - Bayesian regularization

KW - Cutting force

KW - Machine learning

KW - Tool wear

UR - http://www.scopus.com/inward/record.url?scp=85199876993&partnerID=8YFLogxK

U2 - 10.1016/j.measurement.2024.115303

DO - 10.1016/j.measurement.2024.115303

M3 - Journal article

AN - SCOPUS:85199876993

SN - 0263-2241

VL - 238

JO - Measurement: Journal of the International Measurement Confederation

JF - Measurement: Journal of the International Measurement Confederation

M1 - 115303

ER -

Data-driven prediction of tool wear using Bayesian regularized artificial neural networks

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this