PSI - Issue 38

Frédéric Kihm et al. / Procedia Structural Integrity 38 (2022) 12–29 Kihm, Miu, Bonato / Structural Integrity Procedia 00 (2021) 000 – 000

26

15

The results of the Ordinary Least Squares function are presented in Table 8:

Table 8. OLS results.for the case with only Rear_axle_inclination Damage input Dep. Variable: StrainDamage R-squared 0.476 No. Observations 1220 Adj. R-squared 0.476 Df Residuals 1218 Prob (F-statistic): 3.79e-173 Df Model 1

coef

P value

[0.025 0.141 0.479

0.975] 0.161 0.539

const

0.1514

0.000 0.000

Rear_axle_inclinationDamage 0.5089

Also, the p- value of the model’s F - statistic indicates the model parameters’ overall statistical significance. This means that the presence of the model parameters (just Rear_axle_inclinationDamage, in this case) is far more justified than merely predicting the average in the training set. On top of this, the p-value of the slope indicates statistical significance (the values are rounded to three decimals, meaning the p-value is smaller that 1.0E-3). This means that the Rear_axle_inclinationDamage parameter is likely useful in explaining the variance in the output. We now consider a linear model with a different input (the transversal acceleration of the vehicle body). This model registers a mean relative absolute error of 50.3%.

Fig. 9. comparison of damage from Accel_lat_Chassis versus strain-induced damage

Fig. 10 shows the distribution of residuals, which resembles a normal distribution:

UNRESTRICTED