ConvLSTM deep learning signal prediction for forecasting bending moment for tool condition monitoring

Stephanie Hall; Stephen Newman; Evros Loukaides; Alborz Shokrani

doi:10.1016/j.procir.2022.05.110

ConvLSTM deep learning signal prediction for forecasting bending moment for tool condition monitoring

Stephanie Hall
, Stephen Newman
, Evros Loukaides
, Alborz Shokrani

Research output: Contribution to journal › Conference article › peer-review

18 Citations (SciVal)

32 Downloads (Pure)

Abstract

This paper presents a new method using a Convolutional Long Short-Term Memory (ConvLSTM) network to predict the future sensor signals for cutting tool bending moment in milling. Tool bending moment is directly correlated to cutting forces and can be used for tool condition monitoring. Bending moment from a sensory tool holder was obtained and transformed into uniform sized time-frequency domain data to be used as frame inputs to the network. Future frames of sensor signals are predicted and validated by comparing the predicted frames with the ground truth. The investigations show the potential of the proposed method for predicting future signals which can be used for tool condition monitoring.

Original language	English
Pages (from-to)	1071-1076
Number of pages	6
Journal	Procedia CIRP
Volume	107
Early online date	26 May 2022
DOIs	https://doi.org/10.1016/j.procir.2022.05.110
Publication status	Published - 26 May 2022

Funding

The authors acknowledge the support of Scorpion Tooling UK Ltd. and the UK Engineering and Physical Sciences Research Council under grant number EP/R513155/1 project reference 2103874.

Keywords

Deep learning
End milling
Sensors
machining

ASJC Scopus subject areas

Control and Systems Engineering
Industrial and Manufacturing Engineering

Access to Document

10.1016/j.procir.2022.05.110Licence: CC BY-NC-ND

Cite this

@article{7fed5f3784be4d3da2a386d3e4693ca7,

title = "ConvLSTM deep learning signal prediction for forecasting bending moment for tool condition monitoring",

abstract = "This paper presents a new method using a Convolutional Long Short-Term Memory (ConvLSTM) network to predict the future sensor signals for cutting tool bending moment in milling. Tool bending moment is directly correlated to cutting forces and can be used for tool condition monitoring. Bending moment from a sensory tool holder was obtained and transformed into uniform sized time-frequency domain data to be used as frame inputs to the network. Future frames of sensor signals are predicted and validated by comparing the predicted frames with the ground truth. The investigations show the potential of the proposed method for predicting future signals which can be used for tool condition monitoring.",

keywords = "Deep learning, End milling, Sensors, machining",

author = "Stephanie Hall and Stephen Newman and Evros Loukaides and Alborz Shokrani",

year = "2022",

month = may,

day = "26",

doi = "10.1016/j.procir.2022.05.110",

language = "English",

volume = "107",

pages = "1071--1076",

journal = "Procedia CIRP",

issn = "2212-8271",

publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - ConvLSTM deep learning signal prediction for forecasting bending moment for tool condition monitoring

AU - Hall, Stephanie

AU - Newman, Stephen

AU - Loukaides, Evros

AU - Shokrani, Alborz

PY - 2022/5/26

Y1 - 2022/5/26

N2 - This paper presents a new method using a Convolutional Long Short-Term Memory (ConvLSTM) network to predict the future sensor signals for cutting tool bending moment in milling. Tool bending moment is directly correlated to cutting forces and can be used for tool condition monitoring. Bending moment from a sensory tool holder was obtained and transformed into uniform sized time-frequency domain data to be used as frame inputs to the network. Future frames of sensor signals are predicted and validated by comparing the predicted frames with the ground truth. The investigations show the potential of the proposed method for predicting future signals which can be used for tool condition monitoring.

AB - This paper presents a new method using a Convolutional Long Short-Term Memory (ConvLSTM) network to predict the future sensor signals for cutting tool bending moment in milling. Tool bending moment is directly correlated to cutting forces and can be used for tool condition monitoring. Bending moment from a sensory tool holder was obtained and transformed into uniform sized time-frequency domain data to be used as frame inputs to the network. Future frames of sensor signals are predicted and validated by comparing the predicted frames with the ground truth. The investigations show the potential of the proposed method for predicting future signals which can be used for tool condition monitoring.

KW - Deep learning

KW - End milling

KW - Sensors

KW - machining

UR - https://www.scopus.com/pages/publications/85132291001

U2 - 10.1016/j.procir.2022.05.110

DO - 10.1016/j.procir.2022.05.110

M3 - Conference article

SN - 2212-8271

VL - 107

SP - 1071

EP - 1076

JO - Procedia CIRP

JF - Procedia CIRP

ER -

ConvLSTM deep learning signal prediction for forecasting bending moment for tool condition monitoring

Abstract

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this