A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearings

Accurately predicting the Remaining Useful Life (RUL) of machinery plays important role for implementing effective predictive maintenance strategies and reducing downtime. However, many existing data-driven approaches rely heavily on supervised learning and treat RUL estimation as a direct regressio...

Full description

Saved in:
Bibliographic Details
Main Authors: Thanh Tung Luu, Duy An Huynh
Format: Article
Language:English
Published: Elsevier 2025-09-01
Series:Results in Engineering
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2590123025028063
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849229235898548224
author Thanh Tung Luu
Duy An Huynh
author_facet Thanh Tung Luu
Duy An Huynh
author_sort Thanh Tung Luu
collection DOAJ
description Accurately predicting the Remaining Useful Life (RUL) of machinery plays important role for implementing effective predictive maintenance strategies and reducing downtime. However, many existing data-driven approaches rely heavily on supervised learning and treat RUL estimation as a direct regression task from sensor data, lacking the ability to model temporal decision-making or adapt to different domains. To address these limitations, this study proposes a deep reinforcement learning framework that integrates a ResNet-based autoencoder for latent feature extraction from raw vibration signals with the Soft Actor-Critic (SAC) algorithm for dynamic RUL prediction. Unlike conventional methods, our approach allows the model to learn RUL dynamically by interacting with the environment, rather than passively mapping inputs to targets. This interaction enables better adaptability to uncertain degradation patterns. Experimental results on the PHM 2012 dataset demonstrate that the proposed SAC-ResNet framework achieves superior accuracy and generalization performance, highlighting its potential as a promising alternative to traditional RUL estimation models.
format Article
id doaj-art-5f047e1b4d6242099acb77abf20bb01b
institution Kabale University
issn 2590-1230
language English
publishDate 2025-09-01
publisher Elsevier
record_format Article
series Results in Engineering
spelling doaj-art-5f047e1b4d6242099acb77abf20bb01b2025-08-22T04:57:24ZengElsevierResults in Engineering2590-12302025-09-012710673910.1016/j.rineng.2025.106739A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearingsThanh Tung Luu0Duy An Huynh1Department of Construction Machinery and Handling Equipment, Faculty of Mechanical Engineering, Ho Chi Minh City University of Technology (HCMUT), 268 Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Vietnam; Vietnam National University Ho Chi Minh City, Ho Chi Minh City, VietnamDepartment of Construction Machinery and Handling Equipment, Faculty of Mechanical Engineering, Ho Chi Minh City University of Technology (HCMUT), 268 Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Vietnam; Vietnam National University Ho Chi Minh City, Ho Chi Minh City, Vietnam; Corresponding author at: Department of Construction Machinery and Handling Equipment, Faculty of Mechanical Engineering, Ho Chi Minh City University of Technology (HCMUT), 268 Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Vietnam.Accurately predicting the Remaining Useful Life (RUL) of machinery plays important role for implementing effective predictive maintenance strategies and reducing downtime. However, many existing data-driven approaches rely heavily on supervised learning and treat RUL estimation as a direct regression task from sensor data, lacking the ability to model temporal decision-making or adapt to different domains. To address these limitations, this study proposes a deep reinforcement learning framework that integrates a ResNet-based autoencoder for latent feature extraction from raw vibration signals with the Soft Actor-Critic (SAC) algorithm for dynamic RUL prediction. Unlike conventional methods, our approach allows the model to learn RUL dynamically by interacting with the environment, rather than passively mapping inputs to targets. This interaction enables better adaptability to uncertain degradation patterns. Experimental results on the PHM 2012 dataset demonstrate that the proposed SAC-ResNet framework achieves superior accuracy and generalization performance, highlighting its potential as a promising alternative to traditional RUL estimation models.http://www.sciencedirect.com/science/article/pii/S2590123025028063AutoencoderResidual networksReinforcement learningSoft actor-criticRemaining useful life (RUL)
spellingShingle Thanh Tung Luu
Duy An Huynh
A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearings
Results in Engineering
Autoencoder
Residual networks
Reinforcement learning
Soft actor-critic
Remaining useful life (RUL)
title A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearings
title_full A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearings
title_fullStr A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearings
title_full_unstemmed A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearings
title_short A ResNet-based deep reinforcement learning framework using soft actor-critic for remaining useful life prediction of rolling bearings
title_sort resnet based deep reinforcement learning framework using soft actor critic for remaining useful life prediction of rolling bearings
topic Autoencoder
Residual networks
Reinforcement learning
Soft actor-critic
Remaining useful life (RUL)
url http://www.sciencedirect.com/science/article/pii/S2590123025028063
work_keys_str_mv AT thanhtungluu aresnetbaseddeepreinforcementlearningframeworkusingsoftactorcriticforremainingusefullifepredictionofrollingbearings
AT duyanhuynh aresnetbaseddeepreinforcementlearningframeworkusingsoftactorcriticforremainingusefullifepredictionofrollingbearings
AT thanhtungluu resnetbaseddeepreinforcementlearningframeworkusingsoftactorcriticforremainingusefullifepredictionofrollingbearings
AT duyanhuynh resnetbaseddeepreinforcementlearningframeworkusingsoftactorcriticforremainingusefullifepredictionofrollingbearings