XJTU-SY 轴承全寿命（西安交大）

XJTU-SY 滚动轴承全寿命加速退化数据集由西安交通大学（XJTU）联合昌兴苏荣科技有限公司（SY）于 2015–2018 年采集并发布（IEEE Transactions on Reliability, 2020, DOI: 10.1109/TR.2018.2882682）。数据集包含 15 只 LDK UER204 轴承在 3 种工况（转速 2100/2250/2400 RPM，径向力 12/11/10 kN）下的全寿命振动信号，每分钟采集 1.28 秒数据，采样频率 25.6 kHz，失效判定阈值为振动超过 20 g。三种工况各 5 只轴承，涵盖内圈、外圈、保持架等多种失效模式。数据集自发布以来成为轴承 RUL 预测领域被引用最广泛的基准之一，在 IEEE TII、TIE、MSSP、Reliability Engineering 等顶刊上有大量以其为验证集的研究方法论文。

RUL 引用最多的真实轴承全寿命公开集。

• XJTU-SY
  J. Mech. Eng., 2019

• Remaining useful life prediction of rolling bearing based on multi-region hypergraph self-attention network
  Mechanical Systems and Signal Processing, 225:112293, 2025 — Various authors
  多区域超图自注意力网络捕获图结构特征，XJTU-SY 验证
• Multi-scale time series analysis using TT-ConvLSTM technique for bearing remaining useful life prediction
  Mechanical Systems and Signal Processing, 206:110888, 2024 — Various authors
  多尺度时频特征 + 时空 ConvLSTM，XJTU-SY 和 FEMTO 双数据集验证，MSSP 2024
• Remaining useful life prognostics of bearings based on convolution attention networks and enhanced transformer
  Heliyon, 10(19), 2024 — Various authors
  MDSCT 模型：多尺度深度可分离卷积+增强 Transformer，XJTU-SY 上比对比方法降低 MAE 达 10.1pp
• RULSurv: A probabilistic survival-based method for early censoring-aware prediction of remaining useful life in ball bearings
  arXiv 2024, 2024 — Various authors
  存活分析框架处理截尾数据，在 XJTU-SY 上评估早期预测能力
• Bearing remaining useful life prediction using self-adaptive graph convolutional networks with self-attention mechanism
  Mechanical Systems and Signal Processing, 188:110010, 2023 — Yupeng Wei, Dazhong Wu, Janis Terpenny
  自适应图卷积网络+自注意力，在 XJTU-SY 上 RMSE 最低达 0.119，MSSP 代表作
• Remaining Useful Life Prediction of Rolling Bearings Using GRU-DeepAR with Adaptive Failure Threshold
  Sensors (MDPI), 23(3):1144, 2023 — Various authors
  GRU-DeepAR 概率预测框架 + 自适应失效阈值，在 XJTU-SY 上进行全寿命预测验证
• Optimized Random Forest Model for Remaining Useful Life Prediction of Experimental Bearings
  IEEE Transactions on Industrial Informatics, 19(6):7771–7779, 2022 — Alfarizi M.G., Tajiani B., Vatn J., Yin S.
  EMD+随机森林+贝叶斯优化框架，在 XJTU-SY 和 PHM 2012 数据集上验证，IEEE TII 代表作
• Remaining Useful Life Prediction Based on Adaptive SHRINKAGE Processing and Temporal Convolutional Network
  Sensors (MDPI), 22(23):9088, 2022 — Various authors
  自适应软阈值收缩 + TCN，在 XJTU-SY 上 MAE 最高降低 52%，RMSE 降低 64%

• https://github.com/WangBiaoXJTU/xjtu-sy-bearing-datasets

• GitHub
• 官网