文章详情

题名 · 作者 · 摘要 · 关键词 · DOI · 参考文献

炼化装置大气腐蚀预测研究进展

林壮敦1,袁武飞1,吴海泓2,黄琛淏3,刘桂雄3

(1. 广东省特种设备检测研究院揭阳检测院,广东 揭阳 522031;2. 广东省特种设备检测研究院惠州检测院,广东 惠州 516003;3. 华南理工大学 机械与汽车工程学院,广东 广州 510640)

栏目:信息化应用 · 第 2026 年第 19 期 · 页码:120-128 · 2096-4706(2026)19-0120-09 · 中图分类号:TP39;TG172;TE986 · 文献标识码:A · DOI:10.19850/j.cnki.2096-4706.2026.19.023

摘要

炼化装置长期服役于高湿、高盐雾、污染物沉降和干湿交替环境中,金属设备易发生大气腐蚀并引发泄漏、失效与安全风险。随着在线监测、传感器网络和机器学习技术发展,腐蚀研究正由经验评估转向数据驱动预测。文章围绕炼化装置大气腐蚀环境与监测数据基础、腐蚀数据增强、腐蚀速率预测建模和少样本迁移预测四个方面进行综述,分析多源环境因子对腐蚀演化的影响,梳理传统增强、生成式增强、机器学习、深度时序网络及迁移学习方法的研究进展。最后指出,面向目标域环境适配的物理感知增强、机理融合预测、跨域迁移、可解释不确定性评估和监测—预测—预警—维护闭环系统将是未来重要发展方向。

关键词

炼化装置;大气腐蚀;数据增强;腐蚀预测;迁移学习

参考文献

  1. [1] Olawole O O,Gbadeyan O J,Deigh C,et al. Advanced Sensor Systems and Machine Learning for Pipeline Integrity Management: a Review of Corrosion Monitoring and Prediction Strategies[J/OL].npj Materials Degradation,2026,10:48.https://doi.org/10.1038/s41529-026-00761-4.
  2. [2] Coelho L B,Zhang D W,Van I Y,et al. Reviewing Machine Learning of Corrosion Prediction in a Data-oriented Perspective[J/OL].npj Materials Degradation,2022,6:8.https://www.nature.com/articles/s41529-022-00218-4.
  3. [3] Xu L,Wang Y F,Mo L,et al. The Research Progress and Prospect of Data Mining Methods on Corrosion Prediction of Oil and Gas Pipelines[J/OL].Engineering Failure Analysis,2023,144:106951.https://doi.org/10.1016/j.engfailanal.2022.106951.
  4. [4] 周梦鑫,吴军,樊志彬,等. 大气腐蚀在线监测技术研究现状与展望[J]. 中国腐蚀与防护学报,2023,43(1):38-46.
  5. [5] 王莎莎,马帅杰,车琨,等. 机器学习在自然环境腐蚀评估与预测领域的应用现状[J]. 中国腐蚀与防护学报,2023,43(3):441-451.
  6. [6] 姚勇,刘国军,黎石竹,等. 金属材料腐蚀预测模型研究进展[J]. 中国腐蚀与防护学报,2023,43(5):983-991.
  7. [7] 沈坚,吴柯娴,何晓宇,等. 我国不同地区钢材大气腐蚀预测算法评估与筛选[J]. 中国腐蚀与防护学报,2024,44(4):939-948.
  8. [8] Vangrunderbeek V,Bertolucci C L,Meeusen M,et al. A Field-based Computer Vision and Corrosion Sensor Approach to Study Electrolyte Dynamics in Atmospheric Corrosion[J/OL].npj Materials Degradation,2026,10:3.https://www.nature.com/articles/s41529-025-00714-3.
  9. [9] Wu W T,Xu D,Liu L A,et al. Fusion of Multi Source Data for Atmospheric Corrosion Evaluation Using Sensors and Image Recognition[J/OL].npj Materials Degradation,2025,9:13.https://www.nature.com/articles/s41529-025-00555-0.
  10. [10] Li Q,Xia X J,Pei Z B,et al. Long-term Corrosion Monitoring of Carbon Steels and Environmental Correlation Analysis via the Random Forest Method[J/OL].npj Materials Degradation,2022,6:1.https://doi.org/10.1038/s41529-021-00211-3.
  11. [11] Li R Y,Wang H C,Zhu Y H,et al. Machine Learning-assisted in Situ Corrosion Monitoring: a Review[J/OL]. Corrosion Reviews,2026,44(1):20250014.https://www.degruyterbrill.com/document/doi/10.1515/corrrev-2025-0014/html.
  12. [12] 张昊,陈俊航,胡为峰,等. 大气腐蚀监测Cu-Zn探针环境因素敏感性和腐蚀影响[J]. 中国腐蚀与防护学报,2024,44(5):1339-1344.
  13. [13] 王远明,宫克,毛飞雄,等. 基于ACM及机器学习的腐蚀监测分析系统在重大基础设施上的应用[J]. 中国表面工程,2024,37(6):205-215.
  14. [14] 夏晓健,杨国威,林德源,等. 基于腐蚀大数据的碳钢海洋大气腐蚀行为[J]. 中国表面工程,2024,37(2):58-71.
  15. [15] 刘雄伟,蒋婉娟,张志宏,等. 耐蚀涂层失效监测方法及失效机制研究进展[J]. 腐蚀与防护,2024,45(7):50-60.
  16. [16] 夏晨瀚,杨小佳,李清,等. 碳钢大气腐蚀的大数据分析[J]. 腐蚀与防护,2024,45(2):75-84.
  17. [17] Wen Q S,Sun L,Yang F,et al. Time Series Data Augmentation for Deep Learning: A Survey[C]//Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence,2021:4653-4660.
  18. [18] Iwana B K,Uchida S. An Empirical Survey of Data Augmentation for Time Series Classification with Neural Networks[J/OL].PLOS ONE,2021,16(7):e0254841.https://doi.org/10.1371/journal.pone.0254841
  19. [19] Iglesias G,Talavera E,González-Prieto Á,et al. Data Augmentation Techniques in Time Series Domain: a Survey And Taxonomy[J].Neural Computing and Applications,2023,35:10123-10145.
  20. [20] 徐越遥,朱建新,乔松,等. 基于标准抽样与工况大数据的承压设备损伤智能识别技术[J]. 腐蚀与防护,2025,46(10):35-42.
  21. [21] Woldesellasse H,Tesfamariam S. Data Augmentation Using Conditional Generative Adversarial Network (Cgan): Application for Prediction of Corrosion Pit Depth and Testing Using Neural Network[J/OL].Journal of Pipeline Science and Engineering,2023,3(1):100091.https://doi.org/10.1016/j.jpse.2022.100091.
  22. [22] Brophy E,Wang Z W,She Q,et al. Generative Adversarial Networks in Time Series: A Systematic Literature Review[J].ACM Computing Surveys,2023,55(10):199:1-31.
  23. [23] Yoon J,Jarrett D,Van D S M. Time-series Generative Adversarial Networks[C]//Advances in Neural Information Processing Systems,2019,32:5508-5518.
  24. [24] Liu H B,Zhu Z H,Zhang J Y,et al. Submarine pipeline corrosion rate prediction model based on high-dimensional mapping augmentation and residual update gradient forest[J/OL].Applied Ocean Research,2025,155:104432.https://doi.org/10.1016/j.apor.2025.104432.
  25. [25] Wang Q K,Lu H F,Li F,et al. Advancing LightGBM with Data Augmentation for Predicting the Residual Strength of Corroded Pipelines[J/OL].npj Materials Degradation,2025,9:128.https://doi.org/10.1038/s41529-025-00673-9.
  26. [26] 秦术杰,王欣. 热带海岛大气环境下Q235钢的腐蚀规律及GM(1,1)模型预测[J]. 材料导报,2025,39(16):183-191.
  27. [27] 李丰范,匡健隆,季佳浩,等. 机器学习在金属材料服役性能预测中的应用[J]. 工程科学学报,2024,46(1):120-136.
  28. [28] 王明慧,党鹏飞,杨铮鑫,等. 基于粒子群优化和最小二乘支持向量机的储罐腐蚀速率预测[J]. 腐蚀与防护,2024,45(8):71-76.
  29. [29] 杨彪,肖佳,欧阳晨,等. APSO-BPNN模型在滨海环境中铁质材料腐蚀速率预测中的应用[J]. 腐蚀与防护,2024,45(12):72-79.
  30. [30] Yang X J,Yang J K,Yang Y,et al. Data-mining and Atmospheric Corrosion Resistance Evaluation of Sn- and Sb additional Low Alloy Steel Based on Big Data Technology[J]. International Journal of Minerals,Metallurgy and Materials,2022,29(4):825-835.
  31. [31] Dong Z,Ding L,Meng Z,et al. Machine Learning based Corrosion Rate Prediction of Steel Embedded in Soil[J/OL]. Scientific Reports,2024,14:18194.https://doi.org/10.1038/s41598-024-68562-w.
  32. [32] Zhang H,Su H,Wang J,et al. Comparative Study on Optimization Algorithms for Atmospheric Corrosion Map of Q235 Steel in Hunan, China[J].Process Safety and Environmental Protection,2024,184:105-115.
  33. [33] Kuang J G,Long Z L. Prediction Model for Corrosion Rate of Low-alloy Steels Under Atmospheric Conditions Using Machine Learning Algorithms[J].International Journal of Minerals,Metallurgy and Materials,2024,31(2):337-350.
  34. [34] Chen Q,Wang H,Ji H D,et al. Data-driven Atmospheric Corrosion Prediction Model for Alloys Based on a Two-stage Machine Learning Approach[J].Process Safety and Environmental Protection,2024,188:1093-1105.
  35. [35] Rajendran M,Subbian D. Deep Learning in Corrosion Assessment and Control: a Critical Review of Techniques and Challenges[J/OL].Corrosion Reviews,2025,44(1):20240060.https://doi.org/10.1515/corrrev-2024-0060.
  36. [36] Jiang J X,Wan X L,Zhu F,et al. A deep learning framework integrating Transformer and LSTM architectures for pipeline corrosion rate forecasting[J/OL].Computers & Chemical Engineering,2025,204:109365.https://doi.org/10.1016/j.compchemeng.2025.109365.
  37. [37] Seghier B M E A,Truong T T,Feiler C,et al. A Hybrid Deep Learning Model for Predicting Atmospheric Corrosion in Steel Energy Structures Under Maritime Conditions Based on Time-series Data[J/OL].Results in Engineering,2025,25:104417.https://doi.org/10.1016/j.rineng.2025.104417.
  38. [38] Tan L,Yang Y,Zhang K M,et al. Prediction of Internal Corrosion Rate for Gas Pipeline: a New Method Based on Transformer Architecture[J/OL].Computers & Chemical Engineering,2025,198:109084.https://doi.org/10.1016/j.compchemeng.2025.109084.
  39. [39] Wu T T,Miao X Y,Song F L. Residual Strength Prediction of Corroded Pipelines Based on Physics-informed Machine Learning and Domain Generalization[J/OL].npj Materials Degradation,2025,9:12.https://doi.org/10.1038/s41529-025-00561-2.
  40. [40] Xu L,Wen S M,Huang H F,et al. Corrosion Failure Prediction in Natural Gas Pipelines Using an Interpretable Xgboost Model: Insights and Applications[J/OL].Energy,2025,325:136157.https://doi.org/10.1016/j.energy.2025.136157.
  41. [41] 马腾,董江洁,雷江辉,等. 基于物理机制引导的埋地管道外腐蚀速率预测研究[J]. 石油科学通报,2025,10(4):809-818.
  42. [42] 周涛涛,刘迎正,郑文培,等. 基于物理信息神经网络的油气管道内腐蚀预测方法[J]. 中国腐蚀与防护学报,2025,45(5):1320-1330.
  43. [43] 陈潜,黄伟,张昌会,等. 基于物理引导的集输管道内腐蚀速率预测及可解释性分析[J]. 中国腐蚀与防护学报,2025,45(3):720-730.
  44. [44] Canonaco G,Roveri M,Alippi C,et al. A Transfer-learning Approach for Corrosion Prediction in Pipeline Infrastructures[J].Applied Intelligence,2022,52(7):7622-7637.
  45. [45] Ji H D,Tian Y,Fu C Q,et al. Transfer Learning Enables Prediction of Steel Corrosion in Concrete Under Natural Environments[J/OL].Cement and Concrete Composites,2024,148:105488.https://doi.org/10.1016/j.cemconcomp.2024.105488.
  46. [46] Long P J,Yang G P,Liu H Y,et al. Transfer Learning-driven Prediction of Oil and Gas Pipeline Corrosion Rates in Small Sample Scenarios[J/OL].PLOS ONE,2025,20(10):e0333914.https://doi.org/10.1371/journal.pone.0333914.
  47. [47] 李中玲,王海瑞. 基于机器学习的腐蚀图像特征参数表征方法[J]. 腐蚀与防护,2025,46(8):39-44.
  48. [48] Ma H N,Geng M Y,Wang F,et al. Data Augmentation of a Corrosion Dataset for Defect Growth Prediction of Pipelines Using Conditional Tabular Generative Adversarial Networks[J/OL].Materials,2024,17(5):1142.https://doi.org/10.3390/ma17051142.
  49. [49] 叶正荣,张少刚,周祥,等. 基于大数据的油田腐蚀数据智能管理平台研发[J]. 腐蚀与防护,2025,46(9):80-87.
  50. [50] Seghier B M E A,Höche D,Zheludkevich M. Prediction of the Internal Corrosion Rate for Oil and Gas Pipeline: Implementation of Ensemble Learning Techniques[J/OL].Journal of Natural Gas Science and Engineering,2022,99:104425.https://doi.org/10.1016/j.jngse.2022.104425.

基金项目

2024年揭阳市社会发展领域科技计划项目(SHFZ2024020)

作者简介

林壮敦(1972.12—),男,汉族,广东揭阳人,高级工程师,本科,研究方向:特种设备检验及安全管理与评估;袁武飞(1978.02—),男,汉族,广东揭阳人,高级工程师,本科,研究方向:特种设备检验及安全管理与评估;吴海泓(1974.08—),男,汉族,广东河源人,高级工程师,硕士研究生,研究方向:特种设备检验及安全评估;黄琛淏(2001.11—),男,汉族,广西南宁人,硕士研究生在读,研究方向:智能检测与仪器仪表;通信作者:刘桂雄(1968.10—),男,汉族,广东揭阳人,教授,博士生导师,博士研究生,研究方向:先进传感与仪器。