Fitting Methods based on Custom Neural Network for Relaxation Modulus of Viscoelastic Materials
He Yun,Li Haibin,Du Juan
Table 3 Fitting parameters and error after custom neural network training
Parameters n
2 3 4 5 6
Ge(kPa) 2.2038E+3 2.1583E+3 2.1652E+3 2.1617E+3 2.1712E+3
G1(kPa) 3.9513E+6 4.2752E+3 3.6837E+3 2.4103E+3 2.4435E+3
T1(s-1) 800.00 14.226 14.003 14.469 14.515
G2(kPa) 7.4018E+3 2.8018E+4 1.6278E+4 2.8425E+4 2.1148E+4
T2(s-1) 16.909 316.13 330.00 357.62 369.01
G3(kPa) 3.0525E+3 3.6643E+3 2.1675E+4 2.4622E+3
T3(s-1) 13.828 14.389 357.62 14.072
G4(kPa) 2.1810E+4 2.4752E+3 2.4639E+3
T4(s-1) 330.00 14.792 14.959
G5(kPa) 2.4743E+3 2.1172E+4
T5(s-1) 14.105 369.02
G6(kPa) 1.6275E+4
T6(s-1) 369.01
Max relative error 0. 87% 0. 46% 0. 48% 0. 51% 0. 52%
Parameters n
7 8 9 10 11
Ge(kPa) 2.1754E+3 2.1776E+3 2.1742E+3 2.1703E+3 2.18E+3
G1(kPa) 1.8623E+3 1.8458E+3 1.8582E+3 1.8471E+3 1.24E+3
T1(s-1) 14.407 14.748 14.932 14.938 15.1827
G2(kPa) 2.4658E+4 2.6102E+4 2.3974E+4 1.8582E+3 1.24E+3
T2(s-1) 409.50 420.23 433.48 15.088 15.1826
G3(kPa) 1.8675E+3 2.6145E+4 2.3952E+4 1.8418E+3 1.24E+3
T3(s-1) 15.310 420.23 433.48 14.933 15.1826
G4(kPa) 1.8302E+3 2.6174E+4 1.8574E+3 2.2273E+4 1.25E+3
T4(s-1) 14.768 420.23 14.928 443.49 15.1816
G5(kPa) 3.1345E+4 1.8525E+3 1.8541E+3 1.7888E+4 2.7462E+4
T5(s-1) 409.50 14.620 15.068 443.49 469.21
G6(kPa) 3.1356E+4 1.8574E+3 1.8542E+3 2.2275E+4 3.3874E+4
T6(s-1) 409.50 15.453 14.928 443.49 469.21
G7(kPa) 1.8328E+3 1.8445E+3 2.3741E+4 2.1575E+4 3.3855E+4
T7(s-1) 14.767 14.713 433.48 443.49 469.21
G8(kPa) 2.0708E+4 1.9008E+4 2.1577E+4 3.3842E+4
T8(s-1) 420.23 433.48 443.49 469.21
G9(kPa) 2.3754E+4 1.8546E+3 3.3864E+4
T9(s-1) 433.48 15.094 469.21
G10(kPa) 2.1563E+4 1.2180E+3
T10(s-1) 443.49 14.656
G11(kPa) 1.2133E+3
T11(s-1) 15.692
Max relative error 0. 57% 0. 58% 0. 60% 0. 61% 0. 64%