Equipment remaining useful life prediction method based on multi-scale Mamba

TIAN Congbao; WANG Zhaoqiang; HU Changhua; FAN Hongdong; ZHENG Ziyang; KONG Xiangyu

doi:xx

您当前的位置：

首页 >

文章列表页 >

Equipment remaining useful life prediction method based on multi-scale Mamba

更新时间：2026-01-02

- Equipment remaining useful life prediction method based on multi-scale Mamba
- JOURNAL OF XI’AN JIAOTONG UNIVERSITY (2026)
- 作者机构：
  
  1.火箭军工程大学智控实验室，陕西西安710025
  2.浙江大学流体动力基础件与机电系统全国重点实验室，浙江杭州310058
- 作者简介：
- 基金信息：
- DOI：xx
  CLC： TH17
- Received：20 June 2025，
  
  Revised：2025-08-25，
  
  Accepted：20 October 2025，
- 稿件说明：
移动端阅览
TIAN Congbao, WANG Zhaoqiang, HU Changhua, et al. Equipment remaining useful life prediction method based on multi-scale Mamba[J/OL]. JOURNAL OF XI’AN JIAOTONG UNIVERSITY, 2026.
DOI：

TIAN Congbao, WANG Zhaoqiang, HU Changhua, et al. Equipment remaining useful life prediction method based on multi-scale Mamba[J/OL]. JOURNAL OF XI’AN JIAOTONG UNIVERSITY, 2026. DOI： xx.

摘要

剩余寿命（Remaining Useful Life

RUL）预测是装备故障预测与健康管理（Prognostics and health management

PHM）技术的关键环节，对于保障装备安全可靠运行和降低维修保障成本具有重要的意义。Mamba是一种基于结构化状态空间序列模型的新兴网络架构，在捕捉时序数据长期依赖关系方面具有显著优势。然而，现有基于Mamba的装备RUL预测方法仍存在卷积核单一固定、局部时序特征捕捉能力不足等问题，限制了模型泛化能力和RUL预测精度的提升。鉴于此，本文提出一种基于多尺度Mamba（Multi-scale Mamba

MsMamba）的RUL预测方法。首先构建MsMamba-GRU并行网络，同时捕捉长期退化趋势与短时动态特征，实现了长短期特征互补提取，突破了单一固定卷积核限制；其次，利用互补特征动态融合（Dynamic Fusion of Complementary Features

DFCF）模块，对提取的长期退化趋势与短时动态特征进行动态融合；最后通过全连接层完成退化特征映射及RUL预测。利用NASA提供的发动机公开数据集对所提出的RUL预测方法进行了验证，实验结果表明，所提方法能够显著提升装备RUL预测的准确性，具有良好的可解释性以及优异的工程应用潜力。

Abstract

Remaining Useful Life (RUL) prediction is a crucial part of Prognostics and Health Management (PHM) technology

and it holds significant importance in ensuring the safe and reliable operation of equipment and reducing maintenance and support costs. Mamba is an emerging network architecture based on a structured state space sequence model

which exhibits significant advantages in capturing long-term dependencies in time-series data. However

the existing RUL prediction methods based on Mamba still face issues such as using a single and fixed convolution kernel

and insufficient ability to capture local temporal features

which limit the model's generalization ability and the improvement of RUL prediction accuracy. In view of this

this paper proposes an RUL prediction method based on Multi-scale Mamba (MsMamba). Firstly

the MsMamba-GRU parallel network is constructed to capture the long-period degradation trend and short-time dynamic features at the same time

so as to realize the complementary extraction of short- and long-period features; secondly

the extracted long-period degradation trend and short-time dynamic features are dynamically fused by the Dynamic Fusion of Complementary Features (DFCF) module; finally

the mapping of degradation features and the prediction of RUL are accomplished through the fully connected layer. The proposed RUL prediction method was validated using NASA's publicly available engine data set. The experimental results show that the proposed method can significantly improve the accuracy of equipment RUL prediction

has good interpretability

and has excellent engineering application potential.

关键词

Keywords

references

李响，徐宜销，雷亚国，李熹伟，李乃鹏，杨彬．面向旋转机械装备的智能故障诊断通用基础模型研究 [J/OL ] ．西安交通大学学报 , https://link.cnki.net/urlid/ 61.1069.T.20250102.0812.003 https://link.cnki.net/urlid/61.1069.T.20250102.0812.003 .

Zhao X , Yao J , Deng W , et al . Intelligent fault diagnosis of gearbox under variable working conditions with adaptive intraclass and interclass convolutional neural network [J ] . IEEE Transactions on Neural Networks and Learning Systems , 2022 , 34 ( 9 ): 6339 - 6353 .

Zhao X , Hu Y , Liu J , et al . A novel intelligent multicross domain fault diagnosis of servo motor-bearing system based on domain generalized graph convolution autoencoder [J ] . Structural Health Monitoring , 2025 , 24 ( 3 ): 1485 - 1499 ..

王宇 , 刘秋发 , 彭一真 . 非均匀监测条件下滚动轴承剩余寿命预测方法 [J ] . 机械工程学报 , 2023 , 59 ( 23 ): 96 - 104 .

Yang L , Liao Y . An optimal dual data-driven framework for RUL prediction with uncertainty quantification [J ] . IEEE Sensors Journal , 2025 , 25 ( 3 ): 4943 - 4957 .

Li P , Wu X , Radu G , et al . Applying neural network to health estimation and lifetime prediction of Lithium-ion batteries [J ] . IEEE Transactions on Transportation Electrification , 2025 , 11 ( 1 ): 4224 - 4248 .

Hachem H , Vu H , Fouladirad M . Different methods for RUL prediction considering sensor degradation [J ] . Reliability Engineering & System Safety , 2024 , 243 : 109897 .

Kordestani M , Saif M , Orchard M E , et al . Failure prognosis and applications—A survey of recent literature [J ] . IEEE Transactions on Reliability , 2019 , 70 ( 2 ): 728 - 748 .

Lei Y , Li N , Lin J , et al . A model-based method for remaining useful life prediction of machinery [J ] . IEEE Transactions on Reliability , 2016 , 65 ( 3 ): 1314 - 1326 .

张波 , 胡昌华 , 张浩 , 等 . 机器学习应用于随机退化设备剩余寿命预测的综述 [J ] . 哈尔滨工程大学学报 , 2024 , 45 ( 09 ): 1783 - 1790 .

Sun L , Liang Y , Yan Z . Remaining useful life (RUL) prediction based on the bivariant two-phase nonlinear Wiener Degradation Process [J ] . Entropy , 2025 , 27 ( 4 ): 349 .

Ruan , D , Lin M , Yang Y , et al . Improvement by monte carlo for trajectory similarity-based RUL prediction [J ] . IEEE Transactions on Instrumentation and Measurement , 2024 , 73 : 1 - 11 .

Prathiba R V , Saravanan M , David P W , et al . Machine learning-based direct estimation of remaining useful life (RUL) of IGBT using multi-precursor prognostics [J ] . Electrical Engineering , 2025 , 107 ( 3 ): 3435 - 3449 .

Shi J , Gao J , Xiang S . Adaptively Lightweight spatiotemporal information-extraction-operator-based DL method for Aero-Engine RUL prediction [J ] . Sensors , 2023 , 23 ( 13 ): 6163 .

Wang G C , Wang Y , Li B T , et al . Probabilistic deep learning based on Bayes by backprop for remaining useful life prognostics of consumer electronics [J ] . IEEE Transactions on Consumer Electronics , 2025 , 71 ( 1 ): 839 - 848 .

Liu H , Li Y , Luo L , et al . A lithium-ion battery capacity and RUL prediction fusion method based on decomposition strategy and GRU [J ] . Batteries , 2023 , 9 ( 6 ): 323 .

Yu W , Kim IY , Mechefske C . Analysis of different RNN autoencoder variants for time series classification and machine prognostics [J ] . Mechanical Systems and Signal Processing , 2021 , 149 : 107322 .

Velasco-Gallego C , Lazakis I . Mar-RUL: A remaining useful life prediction approach for fault prognostics of marine machinery . Applied Ocean Research . 2023 , 140 : 103735 .

Sun S , Wang J , Xiao Y , et al . Few-shot RUL prediction for engines based on CNN-GRU model Scientific Reports , 2024 , 14 ( 1 ): 16041 .

Jing T , Zheng P , Xia L , et al . Transformer-based hierarchical latent space VAE for interpretable remaining useful life prediction [J ] . Advanced Engineering Informatics , 2022 , 54 : 101781 .

Ternharz G , Elhalwagy A , Kalganova T . Data-efficient estimation of remaining useful life for machinery with a limited number of run-to-failure training sequences [J ] . IEEE Access , 2022 , 10 : 129443 - 129464 .

段佳俊 , 陆中 , 王捷 . 基于多尺度CNN和Transformer的航空发动机寿命预测方法 [J ] . 航空动力学报 , 2025 , 1 - 12 .

Kim G , Choi J , Lim S . Using Transformer and a reweighting technique to develop a remaining useful life estimation method for turbofan engines [J ] . Engineering Applications of Artificial Intelligence , 2024 , 133 : 108475 .

Wang Z , Kong F , Feng S , et al . Is mamba effective for time series forecasting [J ] . Neurocomputing , 2025 , 619 : 129178 .

A. Gu, T. Dao, Mamba : Linear-time sequence modeling with selective state spaces , 2023 , arXiv preprint arXiv: 2312.00752 .

Chang Y , Lu W , Xue F , et al . Combining market-guided patterns and mamba for stock price prediction [J ] . Alexandria Engineering Journal , 2025 , 113 : 287 - 93 .

Pan Z , Li C , Plaza A , et al . Hyperspectral Image Classification with Mamba [J ] . IEEE Transactions on Geoscience and Remote Sensing , 2024 , 63 .

Ming R , Chen N , Peng J , et al . Semantic tokenization-based Mamba for hyperspectral image Classification [J ] . IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing . 2025 , 18 : 4227 - 4241 .

Shi Z . MambaLithium: Selective state space model for remaining-useful-life, state-of-health, and state-of-charge estimation of lithium-ion batteries . arxiv preprint arxiv : 2403. 05430 . 2024 Mar 8.

Wang H , Dai X , Cui L , et al . A progressive learning residuals based on multivariate Mamba and adaptive singular value decomposition method for remaining useful life prediction of lithium-ion batteries [J ] . Ionics , 2025 , 31 ( 2 ): 1731 - 1746 .

Liu Z , Du X , Shi Y . Application of multi-modal temporal neural network based on enhanced sparrow optimization in lithium battery life prediction [J ] . Scientific Reports , 2024 , 14 ( 1 ): 27476 .

Liang Y , Zhao S . Early Prediction of Remaining Useful Life for Lithium-Ion Batteries with the State Space Model [J ] . Energies , 2024 , 17 ( 24 ): 6326 .

Huang J , Liu L , Zhao H , et al . RUL-Mamba: Mamba-based remaining useful life prediction for lithium-ion batteries . Journal of Energy Storage . 2025 , 120 : 116376 .

Li P , Ao Z , Hou J , et al . Physics-informed mamba neural network with potential knowledge for state-of-charge estimation of lithium-ion batteries [J ] . Journal of Energy Storage , 2025 , 123 : 116546 .

Yue C , Qin Y , Liu X , et al . Data-physics collaborativige fusion prediction method for tool remaining useful life based on Mamba state space and physical description [J ] . Journal of Manufacturing Systems , 2025 , 81 : 1 - 15 .

Li , M , Zhu L , Luo M , et al . Remaining useful life prediction of airplane engine based on bidirectional Mamba and causal discovery . Sensors 2025 , 25 , 3429 . https://doi.org/10.3390/s25113429 https://doi.org/10.3390/s25113429 .

SAXENA A , GOEBEL K , SIMON D , et al . Damage propagation modeling for aircraft engine run-to-failure simulation [C ] . Proceedings of the International Conference on Prognostics and Health Management. Denver , CO, USA , 2008 : 1 ‒ 9 .

许丹阳 , 尚洁 , 蒋琛 , 等 . 基于自适应时空图卷积网络的航空发动机剩余寿命预测 [J/OL ] . 计算机集成制造系统 , 1 - 22 [ 2025-06-15 ] . https://doi.org/10.13196/j.cims.2024.0433 https://doi.org/10.13196/j.cims.2024.0433 .

Al-DULAIMI A , ZABIHI S , ASIF A , et al . NBLSTM: Noisy and hybrid convolutional neural network and BLSTM Based deep architecture for remaining useful life estimation [J ] . Journal of Computing and Information Science in Engineering , 2020 , 20 ( 2 ): 021012 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Equipment Remaining Useful Life Prediction Method Using Multi-Scale Mamba Architecture

Study on the Thermodynamic Properties in the Anomalous Region of Supercritical Water

Numerical Simulation of the Energy Conversion Process in Thermoacoustic Engines with Different Stack Materials

A Structure-Semantics-Residual Collaborative Image Compression Framework：Towards Privacy Protection and High-Fidelity Reconstruction

Diffusion Characteristics and Influencing Factors in Case of Hydrogen Leakage from Fuel Cell Vehicles Running in Tunnels

Related Author

WANG Zhaoqiang

Related Institution

No data

⁰