Issue 45

S. Harzallah et al, Frattura ed Integrità Strutturale, 45 (2018) 147-155; DOI: 10.3221/IGF-ESIS.45.12 152   n 2 i i i 1 1 RMSE y x n     (10)   n 2 i i i 1 n i i 1 1 y x n rRMSE 100 1 x N        (11) Figure 6 : Inverse problem RN with two inputs and two outputs. The ranges of rRMSE define the model performance as: Excellent if: rRMSE < 10% Good if: 10 % < rRMSE < 20% Fair if: 12 % < rRMSE < 30% Poor if: rRMSE > 30% i i n i 1 1 y x N MAE     (12)     i i i i n i 1 n i 1 y x ² y y ² ² R        (13) where: y i and x i are the estimated and the measured values, i y is the average of estimated values and N is data number. Figure 7 : Cost function (MSE) vs. number of iterations. X position  z Direct Model by FEM Stocking data base ANN Depth of crack Length of cracks Depth of crack estimated Length of cracks estimated