PSI - Issue 59

Victor Aulin et al. / Procedia Structural Integrity 59 (2024) 436–443

440

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

sufficient to verify the adequacy of the built model based on the ANN method and the quality of training. The size of the test sample was usually taken equal to 10...25% of the total number of samples. 3. Experimental results The obtained results of observations based on the algorithm of recognition of combinations of defects for the resource-determining elements of the YMZ-238 engines were carried out by the method of iteratively increasing the number of neurons in the hidden layer. The results are shown in Table 1. Table 1. Value of errors in recognition of combinations of defects of engine elements (the cylinder-piston group (CPG), the slider-crank linkage(SCL), the valvetrain(V)) depending on the number of neurons in the hidden layer Number of neurons in the Recognized educational samples, % Value of the maximum error during training Recognized test samples, % Value of the maximum error during testing

hidden layer, l

SCL

CPG

V

SCL

CPG

V

SCL

CPG

V

SCL

CPG

V

2 3 4 5 6 7 8 9

85

82 95

83

0,014 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000

0,016 0,020 0,006 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000 0,000

34 76

0

0

0,046 0,030 0,016 0,004 0,006 0,008 0,012

0,178 0,069 0,054 0,033 0,022 0,006 0,015 0,018 0,020

0,170 0,080 0,046 0,032 0,016 0,008 0,012 0,015 0,022

100 100 100 100 100 100 100 100

100 100 100 100 100 100 100 100

24 50 82

30 54 92

100 100 100 100 100 100 100

100 100 100 100 100 100 100

100 100 100 100 100

100 100 100 100 100

0,16 0,18

10

The graphic representation of the convergence of the recognized defects of the resource-determining elements (the cylinder-piston group, the slider-crank linkage, the valvetrain) of the engine during training and testing by the ANN method is shown in fig. 2.

100%

90%

80%

70%

60%

50%

40%

30%

2

3

4

5

6

7

8

9

10

a - number of neurons in hidden layer Кількість нейронів у прихованому шарі

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

2

3

4

5

6

7

8

9

10

b - Number of neurons in hidden layer Кількість нейронів у прихованому шарі