Remaining Useful Life Prediction Method for Bearings Based on Multi-Scale Feature Fusion and Bi-ResLSTM

Di Xuan; Xiao Wang; Shan Song; Xu Yuan; Wang Qingfeng

doi:10.12473/CPM.202504090

您当前的位置：

首页 >

文章列表页 >

Remaining Useful Life Prediction Method for Bearings Based on Multi-Scale Feature Fusion and Bi-ResLSTM

OIL AND GAS ARTIFICIAL INTELLIGENCE | 更新时间：2026-10-15

- Remaining Useful Life Prediction Method for Bearings Based on Multi-Scale Feature Fusion and Bi-ResLSTM
- China Petroleum Machinery Vol. 54, Issue 9, Pages: 1-9(2026)
- 作者机构：
  
  1.北京化工大学高端压缩机及系统技术全国重点实验室
  2.中国石油天然气股份有限公司长庆油田分公司工程技术监督中心
  3.国家管网西部管道有限公司
  4.中石化（天津）石油化工有限公司
- 作者简介：
- 基金信息：
- DOI：10.12473/CPM.202504090
  CLC： TE969/TH133/TP18
- Received：27 April 2025，
  
  Revised：2026-03-28，
  
  Accepted：12 September 2025，
  
  Published：25 September 2026
- 稿件说明：
移动端阅览
第轩，肖旺，山崧，等.多尺度特征融合和Bi-ResLSTM网络的轴承剩余寿命预测方法［J］.石油机械，2026，54（9）：1-9.

Di Xuan，Xiao Wang，Shan Song，et al.Remaining useful life prediction method for bearings based on multi-scale feature fusion and Bi-ResLSTM［J］.China Petroleum Machinery，2026，54（9）：1-9.
第轩，肖旺，山崧，等.多尺度特征融合和Bi-ResLSTM网络的轴承剩余寿命预测方法［J］.石油机械，2026，54（9）：1-9. DOI： 10.12473/CPM.202504090.

Di Xuan，Xiao Wang，Shan Song，et al.Remaining useful life prediction method for bearings based on multi-scale feature fusion and Bi-ResLSTM［J］.China Petroleum Machinery，2026，54（9）：1-9. DOI： 10.12473/CPM.202504090.

摘要

滚动轴承是旋转机械核心承载部件，对其退化趋势精准评估与剩余寿命预测是设备视情运维的关键支撑。现有数据驱动的故障预测方法普遍存在健康指标构建难度大、长时序退化特征挖掘不充分、剩余使用寿命预测精度不足等缺陷，难以适配工业现场复杂噪声、样本稀缺的监测工况。针对以上技术痛点，提出一种融合多尺度特征筛选与双向残差长短期记忆网络（Bi-ResLSTM）的滚动轴承剩余寿命预测方法。对设备采集的振动信号开展多尺度混合特征提取，借助动态关联窗口分析筛选高敏感特征，再通过深度去噪自编码器（DDA）完成多源特征降噪融合，以此构建能够直观反映轴承退化状态的健康指标曲线；依托自编码器的重构误差识别轴承早期故障发生时刻，精准确定寿命预测的起始时间点；将构建完成的健康指标时序曲线输入Bi-ResLSTM网络，深度挖掘指标随时间推移的劣化演变规律，结合设备预设失效阈值完成轴承剩余使用寿命定量计算。采用公开标准轴承数据集开展对比验证试验，试验结果表明，相比传统粒子滤波寿命预测算法和基础单层长短期记忆网络（LSTM）预测算法，所提方法整体预测误差百分比分别下降了15.57和5.75个百分点，表明其具有更高的预测精度和抗干扰鲁棒性能；充分证实了该预测方案即使在现场可获取训练故障样本有限的场景下，依旧能够维持稳定且准确的预测效果。研究结果可为工业旋转机械故障预测与预防性运维工作开展提供技术参考。

Abstract

Rolling bearing is the core load-bearing component of rotating machinery， and accurate assessment of its degradation trend and prediction of its remaining useful life （RUL） provide key supports for condition-based operation and maintenance of equipment. Existing deep learning-driven fault prediction frameworks generally suffer from difficulties in defining health indicators， insufficient mining of long-term degradation features， and inadequate RUL prediction accuracy， making them poorly suited to industrial monitoring conditions with complex noise and scarce samples. To address these technical weaknesses， a RUL prediction method for rolling bearings was proposed by integrating multi-scale feature screening and a bidirectional residual long short-term memory （Bi-ResLSTM） network. The method consists of three main procedures. First， multi-scale hybrid features are extracted from the vibration signals collected from the equipment， highly sensitive features are screened by dynamic correlation window analysis， and multi-source feature denoising and fusion are completed by using a deep denoising autoencoder （DDA）， so as to construct a health indicator curve that intuitively reflects the degradation state of the bearing. Second， the reconstruction error of DDA is used to identify the occurrence time of early bearing faults and accurately determine the starting point of RUL prediction. Finally， the constructed health indicator time-series curve is fed into the Bi-ResLSTM network to deeply capture the degradation evolution pattern of the indicator over time， and the RUL of the bearing is quantitatively calculated according to the pre-specified failure threshold of the equipment. The experimental results show that， compared with the traditional particle filter life prediction algorithm and the basic single-layer long short-term memory （LSTM） prediction algorithm， the overall prediction percentage error of the proposed method decreases by 15.57 and 5.75 percentage points， respectively， indicating that it has higher prediction accuracy and anti-interference robustness. It is fully demonstrated that the proposed scheme can deliver stable and accurate prediction even when only limited training samples of faults are available on site， and has high engineering application potential for fault prediction and preventive maintenance of industrial rotating machinery.

关键词

Keywords

references

Randall R B ， Antoni J . Rolling element bearing diagnostics： a tutorial ［J］. Mechanical Systems and Signal Processing ， 2011 ， 25 （ 2 ）： 485 - 520 .

莫琦，陈立定，冯太合．小波包在旋转机械故障振动信号处理中的应用［J］．计算技术与自动化， 2005 ， 24 （ 2 ）： 60 - 62 ．

Mo Qi ， Chen Liding ， Feng Taihe . Application of wavelet package in vibration signals processing for rotating mechanical failure ［J］. Computing Technology and Automation ， 2005 ， 24 （ 2 ）： 60 - 62 .

褚东亮．旋转机械故障信号分析及诊断技术研究［D］．北京：华北电力大学， 2017 ．

Chu Dongliang . Research on fault signal analysis and diagnosis technology of rotating machinery ［D］. Beijing ： North China Electric Power University ， 2017 .

Zhang Guobin ， Du Ronghua ， Xin Xiaogang ， et al . Early fault detection for power plant fans based on dynamic neural network ［J］. Journal of Physics Conference Series ， 2021 ， 2005 （ 1 ）： 012203 .

Zhai L ， Han N . Remaining life forecast model based on kalman filter ［J］. Electronic Science and Technology ， 2013 ， 26 （ 9 ）： 28 - 30 .

Li Yuxiong ， Huang Xianzhen ， Ding Pengfei ， et al . Wiener-based remaining useful life prediction of rolling bearings using improved kalman filtering and adaptive modification ［J］. Measurement ， 2021 ， 182 ： 109706 .

Wang Biao ， Lei Yaguo ， Li Naipeng ， et al . A hybrid prognostics approach for estimating remaining useful life of rolling element bearings ［J］. IEEE Transactions on Reliability ， 2020 ， 69 （ 1 ）： 401 - 412 .

Ren Lei ， Sun Yaqiang ， Wang Hao ， et al . Prediction of bearing remaining useful life with deep convolution neural network ［J］. IEEE Access ， 2018 （ 6 ）： 13041 - 13049 .

韩光洁，史国华，缑林峰，等 . 一种用于航空发动机RUL预测的推理网络模型［J］. 小型微型计算机系统， 2022 ， 43 （ 6 ）： 1217 - 1220 ．

Han Guangjie ， Shi Guohua ， Gou Linfeng ， et al . Inference network model for RUL prediction of aero-engine ［J］. Journal of Chinese Computer Systems ， 2022 ， 43 （ 6 ）： 1217 - 1220 .

王欣，孟天宇，周俊曦．基于注意力与LSTM的航空发动机剩余寿命预测［J］．科学技术与工程， 2022 ， 22 （ 7 ）： 2784 - 2792 ．

Wang Xin ， Meng Tianyu ， Zhou Junxi . Remaining useful life prediction for aircraft engine using attention based LSTM ［J］. Science Technology and Engineering ， 2022 ， 22 （ 7 ）： 2784 - 2792 .

He Ke ， Su Zuqiang ， Tian Xiaoqing ， et al . RUL prediction of wind turbine gearbox bearings based on self-calibration temporal convolutional network ［J］. IEEE Transactions on Instrumentation and Measurement ， 2022 ， 71 ： 1 - 12 .

王玉静，李少鹏，康守强，等．结合CNN和LSTM的滚动轴承剩余使用寿命预测方法［J］．振动、测试与诊断， 2021 ， 41 （ 3 ）： 439 - 446 ．

Wang Yujing ， Li Shaopeng ， Kang Shouqiang ， et al . A method for predicting remaining useful life of rolling bearings combining CNN and LSTM ［J］. Journal of Vibration ， Measurement & Diagnosis， 2021 ， 41 （ 3 ）： 439 - 446 .

雷亚国，韩天宇，王彪，等． XJTU-SY滚动轴承加速寿命试验数据集解读［J］．机械工程学报， 2019 ， 55 （ 16 ）： 1 - 6 ．

Lei Yaguo ， Han Tianyu ， Wang Biao ， et al . XJTU-SY rolling element bearing accelerated life test datasets： a tutorial ［J］. Journal of Mechanical Engineering ， 2019 ， 55 （ 16 ）： 1 - 6 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

PDC Bit Selection and Parameter Optimization Based on Big Data Analysis and Intelligent Technology

Intelligent Early Kick Detection Method Based on Cost-Sensitive Learning

Research and Application of Intelligent Lost Circulation Diagnosis Model Integrating Logic Rules

Drillstring Vibration Identification Based on Data Augmentation and Automated Machine Learning

Analysis and Optimization of Shear Capacity of Ram Blowout Preventer and Matching of Drill Pipe Parameters

Related Author

Ye Shanlin

Song Xianzhi

Zhu Zhaopeng

Wang Chaochen

Zhu Lin

Liu Zihao

Li Qingfeng

Zou Guanggui

Related Institution

College of Mechanical and Transportation Engineering， China University of Petroleum

College of Petroleum Engineering， China University of Petroleum

College of Artificial Intelligence， China University of Petroleum

PetroChina Tarim Oilfield Company

Xinjiang Key Laboratory of Ultra-deep Oil and Gas

AI问答

⁰