PSI - Issue 48

Oleh Yasniy et al. / Procedia Structural Integrity 48 (2023) 183–189 Yasniy et al / Structural Integrity Procedia 00 (2019) 000 – 000

186 4

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.00 0.02 0.04 0.06 0.08 0.10 Strain (pred), mm/mm Strain (exp), mm/mm

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.00 0.02 0.04 0.06 0.08 0.10 Strain (pred), mm/mm Strain (exp), mm/mm

Fig. 3. The predicted and experimental data of linear region (0.01-0.08) of the strain for different temperatures obtained by the method of k-nearest neighbors (left) and random forests (rigth) The predicted and experimental dependences of the linear region of the strain on the stress for different temperatures obtained by the methods of machine learning (Fig. 4). By using the methods of machine learning, we plotted the dependences of the experimental data on the predicted values of the non-linear region (0-0.01) of the strain for different temperatures (Fig. 5). The predicted and experimental dependences of the non-linear region (0-0.01) of the strain for different temperatures obtained by the methods of machine learning (Fig. 6). In addition, the dependences of the root mean square error on the number of trees for the linear (a) and non-linear (b) region obtained by the random forests method are shown in Fig. 7. The dependences of importance of input parameters for the linear (a) and non-linear (b) region obtained by the random forests method are shown in Fig. 8.

300

300

Exp Pred _20 Pred _100 Pred _150 Pred _200 Pred _250 Pred _300

Exp Pred_ 20 Pred_ 100 Pred_ 150 Pred_ 200 Pred_ 250 Pred_ 300

250

250

200

200

150

150

100 Stress, MPa

100 Stress, MPa

50

50

0

0

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

Strain, mm/mm

Strain, mm/mm

Fig. 4. The predicted and experimental dependences of the linear region of the strain on the stress for different temperatures obtained by the method of k-nearest neighbors (left) and random forests (right)