cpomaemКоррозия: защита материалов и методы исследованийTitle in englishИФХЭ РАН10.61852/2949-3412-2024-2-1-41-59cpomaem-39Research ArticleСтатьиМодель прогноза коррозионных потерь углеродистой стали за первый год экспозиции на основе алгоритма «случайный лес»A model for predicting corrosion losses of carbon steel for the first year of exposure based on the random forest algorithmГаврюшинаМ. А.GavryushinaM. A.

119071, Москва, Ленинский проспект, д. 31, корп. 4

31 Leninsky Prospekt, 4, Moscow, 119071

maleeva.corlab@yandex.ruПанченкоЮ. М.PanchenkoYu. M.

119071, Москва, Ленинский проспект, д. 31, корп. 4

31 Leninsky Prospekt, 4, Moscow, 119071

МаршаковА. И.MarshakovA. I.

119071, Москва, Ленинский проспект, д. 31, корп. 4

31 Leninsky Prospekt, 4, Moscow, 119071

Институт физической химии и электрохимии им. А.Н. Фрумкина РАНРоссияFrumkin Institute of Physical Chemistry and Electrochemistry of the Russian Academy of SciencesRussian Federation202411042024014159Copyright © Гаврюшина М.А., Панченко Ю.М., Маршаков А.И., 20242024Гаврюшина М.А., Панченко Ю.М., Маршаков А.И.Gavryushina M.A., Panchenko Y.M., Marshakov A.I.This work is licensed under a Creative Commons Attribution 4.0 License.https://www.cpmrm.ru/jour/article/view/39

На основе алгоритма «случайный лес» (RF) получены две модели для прогноза первогодовых коррозионных потерь (К1) углеродистой стали в открытой атмосфере в различных регионах мира. Первая модель RF_общая получена с использованием объединенных баз данных международных программ ISO CORRAG, MICAT, ECE/UN и испытаний на территории России и предназначена для оценки К1 в различных типах атмосферы в различных регионах мира. Вторая модель RF_конт позволяет предсказать К1 в континентальных районах мира. Проведено сравнение точности предсказаний К1 по моделям RF и двум функциям «доза-ответ»: представленной в стандарте ISO 9223 и новой версии, разработанной ИФХЭ РАН для континентальных регионов. Показано, что достоверность обеих моделей RF существенно лучше, чем функций «доза-ответ», за исключением предсказаний коррозионных потерь стали на территории России с холодным климатом.

Based on the random forest (RF) algorithm, two models have been obtained for predicting first-year corrosion losses (K1) of carbon steel in an open atmosphere in various regions of the world. The first RF_general model was obtained using the combined databases of the international ISO CORRAG, MICAT, ECE/UN programs and tests in Russia and is designed to assess K1 in various types of atmosphere in different regions of the world. The second RF_cont model allows you to predict K1 in the continental regions of the world. The accuracy of K1 predictions based on RF models and two dose-response functions was compared: the one presented in ISO 9223 standard and the new version developed by IPCE RAS for continental regions. It is shown that the reliability of both RF models is significantly better than the dose-response functions, with the exception of predictions of corrosion losses of steel in Russia with a cold climate. 

малоуглеродистая стальмоделивероятная скорость коррозииметоды машинного обученияlow-carbon steelmodelsprobable corrosion ratemachine learning methodsReferencesISO 9223:2012(E). Corrosion of metals and alloys — Corrosivity of atmospheres — Classification, determination and estimation, International Standards Organization, Geneve, 2012.ISO 9223:2012(E). Corrosion of metals and alloys — Corrosivity of atmospheres — Classification, determination and estimation, International Standards Organization, Geneve, 2012.ISO 9224:2012(E) Corrosion of metals and alloys —Corrosivity of atmospheres — Guiding values for the corrosivity categories, 2012.ISO 9224:2012(E) Corrosion of metals and alloys —Corrosivity of atmospheres — Guiding values for the corrosivity categories, 2012.Yu.M. Panchenko, A.I. Marshakov, Prediction of First-Year Corrosion Losses of Carbon Steel and Zinc in Continental Regions, Materials, 2017, 10, no. 4. 422. doi: 10.3390/ma10040422Yu.M. Panchenko, A.I. Marshakov, Prediction of First-Year Corrosion Losses of Carbon Steel and Zinc in Continental Regions, Materials, 2017, 10, no. 4. 422. doi: 10.3390/ma10040422Yu.M. Panchenko, A.I. Marshakov, L.A. Nikolaeva, VV. Kovtanyuk, Prediction of first-year corrosion losses of copper and aluminum in continental regions, AIMS Materials Science, 2018, 5, no. 4, 624−649. doi: 10.3934/matersci.2018.4.624Yu.M. Panchenko, A.I. Marshakov, L.A. Nikolaeva, VV. Kovtanyuk, Prediction of first-year corrosion losses of copper and aluminum in continental regions, AIMS Materials Science, 2018, 5, no. 4, 624−649. doi: 10.3934/matersci.2018.4.624Yu.M. Panchenko, A.I. Marshakov, L.A. Nikolaeva, T.N. Igonin, Evaluating the Reliability of Predictions of First-Year Corrosion Losses of Structural Metals Calculated Using Dose-Response Functions for Territories with Different Categories of Atmospheric Corrosion Aggressiveness, Protection of Metals and Physical Chemistry of Surfaces, 2020, 56, no. 7. 1249–1263. doi: 10.1134/S207020512007014XYu.M. Panchenko, A.I. Marshakov, L.A. Nikolaeva, T.N. Igonin, Evaluating the Reliability of Predictions of First-Year Corrosion Losses of Structural Metals Calculated Using Dose-Response Functions for Territories with Different Categories of Atmospheric Corrosion Aggressiveness, Protection of Metals and Physical Chemistry of Surfaces, 2020, 56, no. 7. 1249–1263. doi: 10.1134/S207020512007014XYu.M. Panchenko, A.I. Marshakov, L.A. Nikolaeva, T.N. Igonin, Development of models for the prediction of first-year corrosion losses of standard metals for territories with coastal atmosphere in various climatic regions of the world, Corrosion Engineering Science and Technology, 2020, 55, no. 8. 655−669. doi: 10.1080/1478422X.2020.1772535Yu.M. Panchenko, A.I. Marshakov, L.A. Nikolaeva, T.N. Igonin, Development of models for the prediction of first-year corrosion losses of standard metals for territories with coastal atmosphere in various climatic regions of the world, Corrosion Engineering Science and Technology, 2020, 55, no. 8. 655−669. doi: 10.1080/1478422X.2020.1772535D. Knotkova, P. Boschek, K. Kreislova, Results of ISO CORRAG Program: Processing of One Year Data in Respect to Corrosivity Classification, ASTM Special Technical Publication, West Conshohocken, PA, 1995, 38.D. Knotkova, P. Boschek, K. Kreislova, Results of ISO CORRAG Program: Processing of One Year Data in Respect to Corrosivity Classification, ASTM Special Technical Publication, West Conshohocken, PA, 1995, 38.M. Morcillo, Atmospheric corrosion in Ibero-America. The MICAT project. In Atmospheric Corrosion, ASTM Special Technical Publication, Philadelphia, PA, USA, 1995, 257–275.M. Morcillo, Atmospheric corrosion in Ibero-America. The MICAT project. In Atmospheric Corrosion, ASTM Special Technical Publication, Philadelphia, PA, USA, 1995, 257–275.J. Tidblad, A.A. Mikhailov, V. Kucera, Acid Deposition Effects on Materials in Subtropical and Tropical Climates. Data Compilation and Temperate Climate Comparison, SCI Report 2000:8E, Swedish Corrosion Institute, Stockholm, Sweden, 2000, 1–34.J. Tidblad, A.A. Mikhailov, V. Kucera, Acid Deposition Effects on Materials in Subtropical and Tropical Climates. Data Compilation and Temperate Climate Comparison, SCI Report 2000:8E, Swedish Corrosion Institute, Stockholm, Sweden, 2000, 1–34.Ю.М. Панченко, Л.Н. Шувахина, Ю.Н. Михайловский, Атмосферная коррозия металлов в регионах Дальнего Востока, Защита металлов, 1982. 18, № 4, 575–582.Ю.М. Панченко, Л.Н. Шувахина, Ю.Н. Михайловский, Атмосферная коррозия металлов в регионах Дальнего Востока, Защита металлов, 1982. 18, № 4, 575–582.L. Breiman, Random forests, Machine Learning, 2001, 45, 5−32. doi: 10.1023/A:1010933404324L. Breiman, Random forests, Machine Learning, 2001, 45, 5−32. doi: 10.1023/A:1010933404324Y. Zhi, D. Fu, D. Zhang, T. Yang, X. Li, Prediction and Knowledge Mining of Outdoor Atmospheric Corrosion Rates of Low Alloy Steels Based on the Random Forests Approach, Metals, 2019, 9. no. 3, 383. doi: 10.3390/met9030383Y. Zhi, D. Fu, D. Zhang, T. Yang, X. Li, Prediction and Knowledge Mining of Outdoor Atmospheric Corrosion Rates of Low Alloy Steels Based on the Random Forests Approach, Metals, 2019, 9. no. 3, 383. doi: 10.3390/met9030383L. Yan, Y. Diao, K. Gao, Analysis of environmental factors affecting the atmospheric corrosion rate of low-alloy steel using random forest-based models, Materials, 2020, 13, no. 15, 3266. doi: 10.3390/ma13153266L. Yan, Y. Diao, K. Gao, Analysis of environmental factors affecting the atmospheric corrosion rate of low-alloy steel using random forest-based models, Materials, 2020, 13, no. 15, 3266. doi: 10.3390/ma13153266Y. Zhi, Z. Jin, L. Lu, T. Yang, D. Zhou, Z. Pei, D. Wu, D. Fu, D. Zhang, X. Li, Improving atmospheric corrosion prediction through key environmental factor identification by random forest-based model, Corrosion Science, 2021, 178, no. 109084. doi: 10.1016/j.corsci.2020.109084Y. Zhi, Z. Jin, L. Lu, T. Yang, D. Zhou, Z. Pei, D. Wu, D. Fu, D. Zhang, X. Li, Improving atmospheric corrosion prediction through key environmental factor identification by random forest-based model, Corrosion Science, 2021, 178, no. 109084. doi: 10.1016/j.corsci.2020.109084A.A. Mikhailov, J. Tidblad, V. Kucera, The classification system of ISO 9223 standard and the dose-response functions assessing the corrosivity of outdoor atmospheres, Protection of Metals, 2004, 40, no. 6, 541−550. doi: 10.1023/B:PROM.0000049517.14101.68A.A. Mikhailov, J. Tidblad, V. Kucera, The classification system of ISO 9223 standard and the dose-response functions assessing the corrosivity of outdoor atmospheres, Protection of Metals, 2004, 40, no. 6, 541−550. doi: 10.1023/B:PROM.0000049517.14101.68J. Tidblad, V. Kucera, A.A. Mikhailov, D. Knotkova, Outdoor and Indoor Atmospheric Corrosion, ASTM Special Technical Publication, West Conshohocken, PA, USA, 2002, 73.J. Tidblad, V. Kucera, A.A. Mikhailov, D. Knotkova, Outdoor and Indoor Atmospheric Corrosion, ASTM Special Technical Publication, West Conshohocken, PA, USA, 2002, 73.Scikit-learn, Machine Learning in Python, https://scikit-learn.org/stable/index.htmlScikit-learn, Machine Learning in Python, https://scikit-learn.org/stable/index.htmlScikit-learn, Sklearn.model_selection.GridSearchCV, https://scikitlearn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.htmlScikit-learn, Sklearn.model_selection.GridSearchCV, https://scikitlearn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.htmlYu.M. Panchenko, A.I. Marshakov, I.V. Bardin, A.V. Shklyaev, Use of Statistical Analysis Methods for Estimating the Reliability of First-Year Carbon Steel and Zinc Corrosion Loss Predictions Calculated Using Dose-Response Functions, Protection of Metals and Physical Chemistry of Surfaces, 2019, 55, no. 4, 753–760. doi: 10.1134/S2070205119040142Yu.M. Panchenko, A.I. Marshakov, I.V. Bardin, A.V. Shklyaev, Use of Statistical Analysis Methods for Estimating the Reliability of First-Year Carbon Steel and Zinc Corrosion Loss Predictions Calculated Using Dose-Response Functions, Protection of Metals and Physical Chemistry of Surfaces, 2019, 55, no. 4, 753–760. doi: 10.1134/S2070205119040142C. Leygraf, I.O. Wallinder, J. Tidblad, T. Graedel, Atmosheric Corrosion. Published by John Wiley & Sons, Inc., Hoboken, New Jersey, 2016, 397 р.C. Leygraf, I.O. Wallinder, J. Tidblad, T. Graedel, Atmosheric Corrosion. Published by John Wiley & Sons, Inc., Hoboken, New Jersey, 2016, 397 р.Yu. Panchenko, A. Marshakov, T. Igonin, L. Nikolaeva, V. Kovtanyuk, Corrosivity of atmosphere toward structural metals and mapping the continental Russian territory, Corrosion Engineering, Science and Technology, 2019, 54, no. 5, 369–378. doi: 10.1080/1478422X.2019.1594526Yu. Panchenko, A. Marshakov, T. Igonin, L. Nikolaeva, V. Kovtanyuk, Corrosivity of atmosphere toward structural metals and mapping the continental Russian territory, Corrosion Engineering, Science and Technology, 2019, 54, no. 5, 369–378. doi: 10.1080/1478422X.2019.1594526

The authors declare that there are no conflicts of interest present.