Modeling and Optimization of Combined Methods of Surface Hardening of Machine Parts Based on Neural Networks And genetic Algorithms
DOI:
https://doi.org/10.32515/2664-262X.2026.13(44).273-285Keywords:
surface hardening optimization, neural networks, genetic algorithm, combined hardening methods, wear resistance, heat treatment, protective coatings, predictive maintenance, Pareto optimization, resource-determining partsAbstract
It is shown that the problem of choosing the optimal parameters of surface hardening of resource-determining parts of automotive and agricultural machinery is an urgent task of modern engineering. Traditional approaches to choosing hardening methods are based mainly on empirical experience or expensive experimental studies, which limits the possibility of systematic optimization of combined technologies (heat treatment + protective coating) for specific operating conditions. A neural network system for optimizing the parameters of combined surface hardening is proposed, which integrates a direct predictive model based on a multilayer perceptron (MLP) with an optimization module based on a genetic algorithm (GA). The direct model predicts the wear intensity and residual life of the part according to the parameters of the Universal Description of the Part (UDP) and the hardening characteristics, and the genetic algorithm performs multi-criteria optimization according to the objective function that takes into account wear minimization, processing cost and technological feasibility. The study was conducted on eight types of resource-determining parts of automotive (piston rings, gear boxes, rolling bearings, cylinder liners) and agricultural machinery (plough shares, harrow discs, cultivator tines, chopper knives) using 14 surface hardening options, including 5 types of heat treatment, 4 types of coatings and 5 combined technologies. A dataset of 40,000 samples was formed using two types of synthetic data: 20,000 samples were generated with increased accuracy to simulate experimental conditions, another 20,000 with a wider range of parameters to cover the entire space of possible solutions. The validation of the direct model demonstrated the accuracy of prediction: coefficient of determination R² = 0.33...0.97, RMSE = 0.0...3.1 μm, MAPE = 4.0...21.5%. The optimization module allowed us to determine the optimal combinations of strengthening for each type of part with an increase in the predicted resource by 143...196% compared to the baseline. The constructed Pareto fronts "wear resistance–cost" provide a reasonable choice of technological solution taking into account economic constraints.
References
Список літератури
1. Mobley R. Keith. An Introduction to Predictive Maintenance / R. Keith Mobley. – 2nd ed. – Oxford : Butterworth-Heinemann, 2002. – 532 p. – URL: https://doi.org/10.1016/B978-075067531-4/50002-8 (дата звернення: 21.01.2026).
2. Lee J. et al. Service Innovation and Smart Analytics for Industry 4.0 and Big Data Environment // Procedia CIRP. – 2014. – Vol. 16. – P. 3–8. – URL: https://doi.org/10.1016/j.procir.2014.02.001 (дата звернення: 21.01.2026).
3. Чумак В. М. та ін. Підвищення зносостійкості та надійності ресурсовизначальних деталей транспортної та сільськогосподарської техніки методами інженерної оптимізації // Центральноукраїнський науковий вісник. Технічні науки. – 2025. – Т. 11, № 42, ч. 2. – С. 143–159. – URL: https://doi.org/10.32515/2664-262X.2025.12(43).1.272-288 (дата звернення: 25.01.2026).
4. Чумак В. М. та ін. Універсальний метод формалізації параметрів ресурсовизначальних деталей транспортної та сільськогосподарської техніки для систем предиктивного обслуговування // Центральноукраїнський науковий вісник. Технічні науки. – 2025. – Вип. 12(43), ч. 2. – С. 204–219. – URL: https://doi.org/10.32515/2664-262X.2025.12(43).2.204-219 (дата звернення: 25.01.2026).
5. Hutchings I., Shipway P. Tribology: Friction and Wear of Engineering Materials. – 2nd ed. – Oxford : Butterworth-Heinemann, 2017. – 412 p. – URL: https://doi.org/10.1016/B978-0-08-100910-9.00001-8 (дата звернення: 25.01.2026).
6. Zum Gahr K. H. Microstructure and Wear of Materials. – 1st ed. – Amsterdam : North Holland, 1987. – 560 p. – ISBN 978-0-08-087574-3. – URL: https://www.academia.edu/82805296/Karl_Heinz_Zum_Gahr_Microstructure_and_Wear_of_Materials (дата звернення: 25.01.2026).
7. Archard J. F. Contact and Rubbing of Flat Surfaces // Journal of Applied Physics. – 1953. – Vol. 24, No. 8. – P. 981–988. – URL: https://doi.org/10.1063/1.1721448 (дата звернення: ).
8. ISO 281:2007 Rolling bearings - Dynamic load ratings and rating life. – 2nd ed. – Geneva : International Organization for Standardization, 2007. – URL: https://www.iso.org/standard/38102.html (дата звернення: 25.01.2026).
9. Bhushan B. Modern Tribology Handbook. – 1st ed. – Boca Raton : CRC Press, 2000. – URL: https://doi.org/10.1201/9780849377877 (дата звернення: 25.01.2026).
10. Horvat Z. et al. Reduction of mouldboard plough share wear by a combination technique of hardfacing // Tribology International. – 2008. – Vol. 41, No. 8. – P. 778–782. – URL: https://doi.org/10.1016/j.triboint.2008.01.008 (дата звернення: 25.01.2026).
11. Bayhan Y. Reduction of wear via hardfacing of chisel ploughshare // Tribology International. – 2006. – Vol. 39, No. 6. – P. 570–574. – URL: https://doi.org/10.1016/j.triboint.2005.06.005 (дата звернення: 10.02.2026).
12. Аулін В. В. та ін. Кіберфізичний підхід при створенні транспортно-виробничих систем // Центральноукраїнський науковий вісник. Технічні науки. – 2020. – Т. 3, № 34. – С. 331–343. – URL: https://doi.org/10.32515/2664-262X.2020.3(34).331-343 (дата звернення: 15.02.2026).
13. Головатий А. О. та ін. Вдосконалення математичного моделювання машинобудівних технологій для смарт-підприємств в системі машинного зору // Центральноукраїнський науковий вісник. Технічні науки. – 2025. – Вип. 11(42), ч. 2. – С. 143–159. – URL: https://doi.org/10.32515/2664-262X.2025.11(42).2.143-159 (дата звернення: 15.02.2026).
14. Shah R. et al. Machine Learning in Wear Prediction // ASME Journal of Tribology. – 2025. – Vol. 147, No. 4. – P. 040801. – URL: https://doi.org/10.1115/1.4066865 (дата звернення: 17.02.2026).
15. Carvalho T. P. et al. A systematic literature review of machine learning methods applied to predictive maintenance // Computers & Industrial Engineering. – 2019. – Vol. 137. – P. 106024. – URL: https://doi.org/10.1016/j.cie.2019.106024 (дата звернення: 17.02.2026).
16. Łach Ł. Recent Advances in Laser Surface Hardening: Techniques, Modeling Approaches, and Industrial Applications // Crystals. – 2024. – Vol. 14, No. 8. – P. 726. – URL: https://doi.org/10.3390/cryst14080726 (дата звернення: 17.02.2026).
17. Liu G. et al. Parameters Optimization of Plasma Hardening Process Using Genetic Algorithm and Neural Network // Journal of Iron and Steel Research International. – 2012. – URL: https://doi.org/10.1016/S1006-706X(12)60010-7 (дата звернення: 17.02.2026).
18. Ulas M. et al. A new approach for prediction of the wear loss of PTA surface coatings using artificial neural network and basic, kernel-based, and weighted extreme learning machine // Friction. – 2020. – Vol. 8. – P. 1102–1116. – URL: https://doi.org/10.1007/s40544-017-0340-0 (дата звернення: 17.02.2026).
19. Altay O. et al. Prediction of wear loss quantities of ferro-alloy coating using different machine learning algorithms // Friction. – 2020. – Vol. 8. – P. 107–114. – URL: https://doi.org/10.1007/s40544-018-0249-z (дата звернення: 17.02.2026).
20. Jatavallabhula J. K., Shabana S., Pappula B. Development and Evaluation of Machine Learning Based Predictive Models for Tribological Properties of Blended Coatings at Elevated Temperature // Journal of Bio- and Tribo-Corrosion. – 2025. – Vol. 11. – Art. 25. – URL: https://doi.org/10.1007/s40735-025-00952-7 (дата звернення: 17.02.2026).
21. Natsis A., Papadakis G., Pitsilis J. The Influence of Soil Type, Soil Water and Share Sharpness of a Mouldboard Plough on Energy Consumption, Rate of Work and Tillage Quality // Journal of Agricultural Engineering Research. – 1999. – Vol. 72, No. 2. – P. 171–176. – URL: https://doi.org/10.1006/jaer.1998.0360 (дата звернення: 17.02.2026).
22. Zhang B., Zhang S., Li W. Bearing performance degradation assessment using long short-term memory recurrent network // Computers in Industry. – 2019. – Vol. 106. – P. 14–29. – URL: https://doi.org/10.1016/j.compind.2018.12.016 (дата звернення: 17.02.2026).
23. Zheng J., Li W., Li J. A Comparative Study on the Wear Behavior of Quenched-and-Partitioned Steel (Q&P) and Martensite Steel (Q&T) // Coatings. – 2024. – Vol. 14, No. 6. – P. 727. – URL: https://doi.org/10.3390/coatings14060727 (дата звернення: 19.02.2026).
24. Deshmankar A. P. et al. Review of the Applications of Machine Learning for Prediction and Analysis of Mechanical Properties and Microstructures in Additive Manufacturing // Journal of Computing and Information Science in Engineering. – 2024. – Vol. 24, No. 12. – P. 1–17. – URL: https://doi.org/10.1115/1.4066575 (дата звернення: 19.02.2026).
25. Yan H. et al. Machine Learning-Based Prediction of Tribological Properties of Epoxy Composite Coating // Polymers. – 2025. – Vol. 17, No. 3. – P. 282. – URL: https://doi.org/10.3390/polym17030282 (дата звернення: 19.02.2026).
26. Davis J. R. (ed.). Alloying: Understanding the Basics. – Materials Park : ASM International, 2001. – URL: https://doi.org/10.31399/asm.tb.aub.9781627082976 (дата звернення: 19.02.2026).
References
1. Mobley, R. K. (2002). An introduction to predictive maintenance (2nd ed.). Butterworth-Heinemann. https://doi.org/10.1016/B978-075067531-4/50002-8
2. Lee, J., Kao, H.-A., & Yang, S. (2014). Service innovation and smart analytics for Industry 4.0 and Big Data Environment. Procedia CIRP, 16, 3–8. https://doi.org/10.1016/j.procir.2014.02.001
3. Chumak, V. M., Aulin, V. V., Hrynkiv, A. V., Lysenko, S. V., & Kuzyk, O. V. (2025). Improvement of wear resistance and reliability of resource-determining parts of transport and agricultural machinery by methods of engineering optimization. Central Ukrainian Scientific Bulletin. Technical Sciences, 11(42, part 2), 143–159. https://doi.org/10.32515/2664-262X.2025.12(43).1.272-288 [in Ukrainian].
4. Chumak, V. M., et al. (2025). Universal method for formalizing parameters of resource-determining parts of transport and agricultural machinery for predictive maintenance systems. Central Ukrainian Scientific Bulletin. Technical Sciences, 12(43, part 2), 204–219. https://doi.org/10.32515/2664-262X.2025.12(43).2.204-219 [in Ukrainian].
5. Hutchings, I., & Shipway, P. (2017). Tribology: Friction and wear of engineering materials (2nd ed.). Butterworth-Heinemann. https://doi.org/10.1016/B978-0-08-100910-9.00001-8
6. Zum Gahr, K. H. (1987). Microstructure and wear of materials (1st ed.). North Holland. https://www.academia.edu/82805296/Karl_Heinz_Zum_Gahr_Microstructure_and_Wear_of_Materials
7. Archard, J. F. (1953). Contact and rubbing of flat surfaces. Journal of Applied Physics, 24(8), 981–988. https://doi.org/10.1063/1.1721448
8. ISO 281:2007. (2007). Rolling bearings - Dynamic load ratings and rating life (2nd ed.). International Organization for Standardization. https://www.iso.org/standard/38102.html
9. Bhushan, B. (2000). Modern tribology handbook (1st ed.). CRC Press. https://doi.org/10.1201/9780849377877
10. Horvat, Z., Filipović, D., Kosutic, S., & Emert, R. (2008). Reduction of mouldboard plough share wear by a combination technique of hardfacing. Tribology International, 41(8), 778–782. https://doi.org/10.1016/j.triboint.2008.01.008
11. Bayhan, Y. (2006). Reduction of wear via hardfacing of chisel ploughshare. Tribology International, 39(6), 570–574. https://doi.org/10.1016/j.triboint.2005.06.005
12. Aulín, V. V., Hrynkiv, A. V., & Holovatyi, A. O. (2020). Cyber-physical approach in the creation of transport and production systems. Central Ukrainian Scientific Bulletin. Technical Sciences, 3(34), 331–343. https://doi.org/10.32515/2664-262X.2020.3(34).331-343 [in Ukrainian].
13. Holovatyi, A. O., et al. (2025). Improvement of mathematical modeling of engineering technologies for smart enterprises in the machine vision system. Central Ukrainian Scientific Bulletin. Technical Sciences, 11(42, part 2), 143–159. https://doi.org/10.32515/2664-262X.2025.11(42).2.143-159 [in Ukrainian].
14. Shah, R., et al. (2025). Machine learning in wear prediction. ASME Journal of Tribology, 147(4), 040801. https://doi.org/10.1115/1.4066865
15. Carvalho, T. P., Soares, F. A. A. M., Vita, R., Francisco, R. P., Basto, J. P., & Alcalá, S. G. S. (2019). A systematic literature review of machine learning methods applied to predictive maintenance. Computers & Industrial Engineering, 137, 106024. https://doi.org/10.1016/j.cie.2019.106024
16. Łach, Ł. (2024). Recent advances in laser surface hardening: Techniques, modeling approaches, and industrial applications. Crystals, 14(8), 726. https://doi.org/10.3390/cryst14080726
17. Liu, G., et al. (2012). Parameters optimization of plasma hardening process using genetic algorithm and neural network. Journal of Iron and Steel Research International. https://doi.org/10.1016/S1006-706X(12)60010-7
18. Ulas, M., et al. (2020). A new approach for prediction of the wear loss of PTA surface coatings using artificial neural network and basic, kernel-based, and weighted extreme learning machine. Friction, 8, 1102–1116. https://doi.org/10.1007/s40544-017-0340-0
19. Altay, O., et al. (2020). Prediction of wear loss quantities of ferro-alloy coating using different machine learning algorithms. Friction, 8, 107–114. https://doi.org/10.1007/s40544-018-0249-z
20. Jatavallabhula, J. K., Shabana, S., & Pappula, B. (2025). Development and evaluation of machine learning based predictive models for tribological properties of blended coatings at elevated temperature. Journal of Bio- and Tribo-Corrosion, 11, Article 25. https://doi.org/10.1007/s40735-025-00952-7
21. Natsis, A., Papadakis, G., & Pitsilis, J. (1999). The influence of soil type, soil water and share sharpness of a mouldboard plough on energy consumption, rate of work and tillage quality. Journal of Agricultural Engineering Research, 72(2), 171–176. https://doi.org/10.1006/jaer.1998.0360
22. Zhang, B., Zhang, S., & Li, W. (2019). Bearing performance degradation assessment using long short-term memory recurrent network. Computers in Industry, 106, 14–29. https://doi.org/10.1016/j.compind.2018.12.016
23. Zheng, J., Li, W., & Li, J. (2024). A comparative study on the wear behavior of quenched-and-partitioned steel (Q&P) and martensite steel (Q&T). Coatings, 14(6), 727. https://doi.org/10.3390/coatings14060727
24. Deshmankar, A. P., et al. (2024). Review of the applications of machine learning for prediction and analysis of mechanical properties and microstructures in additive manufacturing. Journal of Computing and Information Science in Engineering, 24(12), 1–17. https://doi.org/10.1115/1.4066575
25. Yan, H., et al. (2025). Machine learning-based prediction of tribological properties of epoxy composite coating. Polymers, 17(3), 282. https://doi.org/10.3390/polym17030282
26. Davis, J. R. (Ed.). (2001). Alloying: Understanding the basics. ASM International. https://doi.org/10.31399/asm.tb.aub.9781627082976
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Vitalii Chumak, Yehor Manko, Serhii Lysenko,
This work is licensed under a Creative Commons Attribution 4.0 International License.
