PSI - Issue 59

Victor Aulin et al. / Procedia Structural Integrity 59 (2024) 444–451

451

8

Victor Aulin et al. / Structural Integrity Procedia 00 (2019) 000 – 000

vehicles are listed. 3. Conclusions

1. The analysis of possible production scenarios for the organized provision of operation complexes based on maintenance and repair of the technical conditions of the parts of units, systems, and assemblies of automobiles at an automotive transportation enterprise allows for reducing internal production losses by decreasing the number of errors in defect recognition in components, systems, and assemblies and their allocation to the technological routes of maintenance and repair operations. 2. The choice of the ANN method as a mathematical tool aimed at solving the problem of reducing the number of errors dealing with recognition of defects in components, systems, and assemblies and their allocation to maintenance and repair technological routes is justified by the ability of this mathematical apparatus for learning, analysis, memorization of results, and high adaptability to solve the specified task. 3. A prerequisite for using the error backpropagation algorithm in training the ANN method is the fact, that the chosen activation function must be differentiable. Therefore, the active sigmoidal function is used, which is differentiable over the entire range of values. 4. The refined algorithm for scaling output data into the range of activation function values during the training of the ANN method is universal and allows for increased accuracy in recognition, prediction, and reduction of training time of this mathematical apparatus. 5. The developed algorithm for defect recognition and prediction in the components of assemblies, systems, and units using the mathematical apparatus of the ANN method can be applied to a wide range of systems with sufficient structural complexity during maintenance and repair operations based on the actual technical condition. References Aulin, V.V., Hrynkiv, А.V., 2019. Solution of the Problem of Operation Reliability Increase Using Information Technologies El ements. Kramarovski Chytannya: Materials of the V th International Scientific-Technical Conference. Kyiv: NULESU. 91-94. Aulin, V.V., Hrynkiv, А.V.,Holovatyi, А.О., 2020. Methodological Bases of Intelligent Transport and Production Systems Design and Functioning: Monograph. (Kropyvnytskyi: LysenkoV.F.). Аulin,V.V., Zamota, Т.М., Hrynkiv, А.V., Zamotа, О.M., Chernai, А.Е., 2018.Advantages of Intelligent Strategy of Technical Ma intenancein Terms of Economic Efficiency. Visnyk of Kharkiv national technical university of agriculture named after P. Vasylenko 192, 29-40. Bodyanskyi, Y.V., Rudenko, О.G., 2004. Artificial Neural Networks: Architectures, Training, Application: Monograph. (Kharkov) . Dia, H., 2001.Anobject-oriented neuraln etwork approach to short- term traffic fo recasti ng. Eur. J. Oper. Res.131, 253-261. Hevko, B.M., Hevko, R.B., Klendii, O.M., Buriak, M.V., Dzyadykevych, Y.V., Rozum, R.I., 2018. Improvement of machine safety devices. Acta Polytechnica. 58(1), 17-25. Hinton, G. E.,Osindero, S.,Teh, Y., 2006. A Fast Learning Algorithm For Deep Belief Nets. Neural Computation 18(7), 1527-1554. Kim, P., 2017.Convolutional Neural Network; MATLAB Deep Learning Apress: Berkeley, CA, USA. Kolmogorov-Arnold-Hecht- Nielsen Teorem, 2015. Е -resource. Access mode: https://lektsii.org/12-28326.html. Date of updating: January 18, 2015. Date of Application:January 28, 2021. Król, A., 2016,The Application of the Artificial Intelligence Methods for Planning of the Development of theTransportation Net work. Transp. Res. Procedia14, 4532-4541. Li, X.,Shi, X.,He, J.,Liu, X., 2011. Coupling Simulation And Optimization To Solve Planning Problems In Afast-Developing Area. Ann. Assoc. Am. Geogr. 101, 1032-1048. Maruschak, P., Panin, S., Zakiev, I., Poltaranin, M., Sotnikov, A., 2016. Scale levels of damage to the raceway of a spherical roller bearing. Engineering Failure Analysis 59, 69-78. Mikhailenko, V.M., Тereikovska, L.О., Теreikovskyi, І.А., Аkhmetov, B.B., 2017. Heural Network Modelsand Methods of Phoneme Recognition in Voice Signalin Dista nce Learning System: Моnograph. Кyiv: Коmprynt. Nguyen, М.Т., 2019.Car Engine DiagnosticsBasedontheNeuralNetwork, Molodyi Vchenyi 26(264), 76 -81. Nuzzolo, A.,Comi, A., 2016. Advanced Public Transport And Intelligent Transport Systems: New Modelling challenges. Transp. A Transp. Sci., 674-699. Setlak,G., 2004.Intelligent Systems of Decision- Making Support К. Logos. Shuklin, D.Y., 2003. Models of Semantic Neural Networks and their Application in the Artificial Intelligence Systems: 05.13.23. PhD (Engineering) Thesis. Kharkiv, 196. Теreikovskyi, І., 2007. Neural Networks in Computer Information Security: Monograph. К. Poligraf Consulting. Vishnukumar,H.J., 2017. Machine Learning and Deep Neural Network – Artificial Intelligence Corefor Lab and Real-World Testand Validation for ADAS and Autonomous Vehicles. Intell. Syst. Conf, 714-721.