Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning Algorithm

Tool condition monitoring (TCM) systems are essential in milling operations to guarantee the product’s quality, and when they are paired with indirect measuring techniques, such as vibration or acoustic emission sensors, the monitoring can happen without sacrificing productivity. Some more advanced...

Full description

Saved in:
Bibliographic Details
Main Authors: Giovanni Oliveira de Sousa, Pedro Oliveira Conceição Júnior, Ivan Nunes da Silva, Dennis Brandão, Fábio Romano Lofrano Dotto
Format: Article
Language:English
Published: MDPI AG 2023-11-01
Series:Engineering Proceedings
Subjects:
Online Access:https://www.mdpi.com/2673-4591/58/1/39
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850261372531638272
author Giovanni Oliveira de Sousa
Pedro Oliveira Conceição Júnior
Ivan Nunes da Silva
Dennis Brandão
Fábio Romano Lofrano Dotto
author_facet Giovanni Oliveira de Sousa
Pedro Oliveira Conceição Júnior
Ivan Nunes da Silva
Dennis Brandão
Fábio Romano Lofrano Dotto
author_sort Giovanni Oliveira de Sousa
collection DOAJ
description Tool condition monitoring (TCM) systems are essential in milling operations to guarantee the product’s quality, and when they are paired with indirect measuring techniques, such as vibration or acoustic emission sensors, the monitoring can happen without sacrificing productivity. Some more advanced techniques in tool wear estimation are based on supervised machine learning algorithms, like several other applications in Industry 4.0’s context; however, a satisfactory performance can be obtained with simple techniques and low computational power. This work focuses on an application of tool wear estimation using a simple backpropagation neural network in a milling dataset. Statistical techniques, i.e., the mean, variance, skewness, and kurtosis, were used as features that were extracted from indirect measurements from vibration and acoustic emission sensors’ data in a real milling testbench dataset containing multiple experiments with sensor data and a direct measure of the flank wear (VB) in most instances. The data were preprocessed, specifically to acquire clean and normalized values for the neural network training, assuming that the VB measure would be the target variable used to predict tool wear; all incomplete samples without a VB measure, as well as outliers, were removed beforehand. The train and test subsets were chosen randomly after making sure that the maximum values of every variable were represented in the training subset. A multiple topology approach was implemented to test the configurations of multiple backpropagation neural networks to determine the most suitable one based on two performance criteria, i.e., the mean absolute percent error (MAPE) and variance. Although only a simple backpropagation algorithm was used, the results were adequate to demonstrate a balance between accuracy and computational resource usage.
format Article
id doaj-art-178d9f8856434ec0bfeb11803ecfc078
institution OA Journals
issn 2673-4591
language English
publishDate 2023-11-01
publisher MDPI AG
record_format Article
series Engineering Proceedings
spelling doaj-art-178d9f8856434ec0bfeb11803ecfc0782025-08-20T01:55:27ZengMDPI AGEngineering Proceedings2673-45912023-11-015813910.3390/ecsa-10-15997Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning AlgorithmGiovanni Oliveira de Sousa0Pedro Oliveira Conceição Júnior1Ivan Nunes da Silva2Dennis Brandão3Fábio Romano Lofrano Dotto4Department of Electrical and Computer Engineering, São Carlos School of Engineering, University of São Paulo USP, Av. Trab. São Carlense, 400-Pq. Arnold Schimidt, São Carlos 13566-590, SP, BrazilDepartment of Electrical and Computer Engineering, São Carlos School of Engineering, University of São Paulo USP, Av. Trab. São Carlense, 400-Pq. Arnold Schimidt, São Carlos 13566-590, SP, BrazilDepartment of Electrical and Computer Engineering, São Carlos School of Engineering, University of São Paulo USP, Av. Trab. São Carlense, 400-Pq. Arnold Schimidt, São Carlos 13566-590, SP, BrazilDepartment of Electrical and Computer Engineering, São Carlos School of Engineering, University of São Paulo USP, Av. Trab. São Carlense, 400-Pq. Arnold Schimidt, São Carlos 13566-590, SP, BrazilDepartment of Electrical and Computer Engineering, São Carlos School of Engineering, University of São Paulo USP, Av. Trab. São Carlense, 400-Pq. Arnold Schimidt, São Carlos 13566-590, SP, BrazilTool condition monitoring (TCM) systems are essential in milling operations to guarantee the product’s quality, and when they are paired with indirect measuring techniques, such as vibration or acoustic emission sensors, the monitoring can happen without sacrificing productivity. Some more advanced techniques in tool wear estimation are based on supervised machine learning algorithms, like several other applications in Industry 4.0’s context; however, a satisfactory performance can be obtained with simple techniques and low computational power. This work focuses on an application of tool wear estimation using a simple backpropagation neural network in a milling dataset. Statistical techniques, i.e., the mean, variance, skewness, and kurtosis, were used as features that were extracted from indirect measurements from vibration and acoustic emission sensors’ data in a real milling testbench dataset containing multiple experiments with sensor data and a direct measure of the flank wear (VB) in most instances. The data were preprocessed, specifically to acquire clean and normalized values for the neural network training, assuming that the VB measure would be the target variable used to predict tool wear; all incomplete samples without a VB measure, as well as outliers, were removed beforehand. The train and test subsets were chosen randomly after making sure that the maximum values of every variable were represented in the training subset. A multiple topology approach was implemented to test the configurations of multiple backpropagation neural networks to determine the most suitable one based on two performance criteria, i.e., the mean absolute percent error (MAPE) and variance. Although only a simple backpropagation algorithm was used, the results were adequate to demonstrate a balance between accuracy and computational resource usage.https://www.mdpi.com/2673-4591/58/1/39tool condition monitoringbackpropagation neural networktool wear estimation
spellingShingle Giovanni Oliveira de Sousa
Pedro Oliveira Conceição Júnior
Ivan Nunes da Silva
Dennis Brandão
Fábio Romano Lofrano Dotto
Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning Algorithm
Engineering Proceedings
tool condition monitoring
backpropagation neural network
tool wear estimation
title Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning Algorithm
title_full Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning Algorithm
title_fullStr Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning Algorithm
title_full_unstemmed Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning Algorithm
title_short Tool Wear Estimation in the Milling Process Using Backpropagation-Based Machine Learning Algorithm
title_sort tool wear estimation in the milling process using backpropagation based machine learning algorithm
topic tool condition monitoring
backpropagation neural network
tool wear estimation
url https://www.mdpi.com/2673-4591/58/1/39
work_keys_str_mv AT giovannioliveiradesousa toolwearestimationinthemillingprocessusingbackpropagationbasedmachinelearningalgorithm
AT pedrooliveiraconceicaojunior toolwearestimationinthemillingprocessusingbackpropagationbasedmachinelearningalgorithm
AT ivannunesdasilva toolwearestimationinthemillingprocessusingbackpropagationbasedmachinelearningalgorithm
AT dennisbrandao toolwearestimationinthemillingprocessusingbackpropagationbasedmachinelearningalgorithm
AT fabioromanolofranodotto toolwearestimationinthemillingprocessusingbackpropagationbasedmachinelearningalgorithm