Russian Federation
UDC 378.147
UDC 629.424
The article describes the experience of using generative neural networks (ChatGPT, Claude) in a dissertation research on vibration diagnostics of a diesel locomotive engine. The tasks of signal processing, classification algorithms implementation and code debugging are considered. The limitations of AI assistants are identified: generation of non-existent references and outdated code. Recommendations for integrating neural networks into the training of research engineers are formulated.
generative neural networks, machine learning, vibration diagnostics, classification, engineering education
1. Ivahnenko E. N., Nikol'skiy V. S. ChatGPT v vysshem obrazovanii i nauke: ugroza ili cennyy resurs? // Vysshee obrazovanie v Rossii. 2023. T. 32, № 4. S. 9–22.
2. Dolgiy A. I., Kudyukin V. V., Hatlamadzhiyan A. Yu. Integrirovannyy post avtomatizirovannogo priema i diagnostiki podvizhnogo sostava // Intellektual'nyy transport. 2025. Vyp. 4 (36). S. 76–85.
3. Scikit-learn: Machine Learning in Python / F. Pedregosa [et al.] // Journal of Machine Learning Research. 2011. Vol. 12. Pp. 2825–2830.
4. Garkusha N. S., Gorodova Yu. S. Pedagogicheskie vozmozhnosti ChatGPT dlya razvitiya kognitivnoy aktivnosti studentov // Professional'noe obrazovanie i rynok truda. 2023. T. 11, № 1. S. 6–23.
5. Shevelev A. A. Analiz ispol'zovaniya ChatGPT v processe obucheniya studentov programmirovaniyu // Vestnik KGUSTA. 2024. № 1 (117). S. 99–102.
