PSI - Issue 42

Iryna Didych et al. / Procedia Structural Integrity 42 (2022) 1344–1349 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

1348

5

1000. The algorithm of boosted trees is similar to the flowchart that is easy to visualize and interpret. Therefore, it is often used to make consistent decisions. The dependence of the root mean square error on the number of trees in boosted trees is shown in Fig. 5. The error of learning and test sample is stabilized after the construction of 1459 trees.

Fig. 5. Calculated error depending on the number of trees, obtained by the method of boosted trees.

The error of the neural network method is 0.05%, while the error of the method of boosted trees is 0.3%.

4. Conclusions The influence of stress on the deformation in aluminum alloy Al- 6061 at temperatures температур T = 343 and 413  C is investigated by machine learning methods, particularly, neural networks and boosted trees. It is found that the prediction results are in good agreement with the experimental ones. The smallest error of 0.05% is obtained by the method of neural networks in the test sample. The error of the boosted trees method is 0.03%. References Alpayndin, E., 2010. Introduction to Machine Learning, MIT Press, pp. 584. Artymiak, P., Bukowski, L., Feliks, J., Narberhaus, S., Zenner, H., 1999. Determination of S-N Curves with the Application of Artificial Neural Networks, Fatigue and Fracture of Engineering Materials and Structures 22, 723-728. Didych, I., Yasniy, O., Fedak, S., Lapusta, Yu., 2022. Prediction of jump-like creep using preliminary plastic strain. Procedia Structural Integrity 36, 166 – 170. Fletcher, R., 1987. Practical methods of optimization, John Wiley & Sons, New York. Haykin, S., 1999. Neural Networks: A Comprehensive Foundation, Prentice Hall, Hamilton, Ontario, pp.823. Kang, J. Y., Song, J. H., 1998. Neural Network Applications in Determining the Fatigue Crack Opening Load, Int. J. Fatigue 20, 57 − 69. Kim, Y., Kim, H., Ahn, I.-G., 2016. A study on the fatigue damage model for Gaussian wideband process of two peaks by an artificial neural network, Ocean Engineering 111, 310 – 322. Mitchell, T. M., 1997. Machine learning, McGraw-Hill Science/Engineering/Math, London, pp. 432. Mohanty, J. R., Verma, B. B., Parhi, D. R. K., Ray D. R., 2009. Application of artificial neural network for predicting fatigue crack propagation life of aluminum alloys. Archives of Computational Materials Science and Surface Engineering 1, 133 – 138. Pidaparti, R. M. V., Palakal, M., 1995. Neural network approach to fatigue-crack-growth predictions under aircraft spectrum loadings, Journal of Aircraft 32, 825-831.