Automatic well test interpretation based on convolutional neural network for a radial composite reservoir

LI Daolun,LIU Xuliang,ZHA Wenshu,YANG Jinghai,LU Detang

Table 3 Parameter interpretation values of the 6 samples selected and their absolute errors.

Sample lg(M) lg(F) lg(RfD) lg(CDe2S) Real value Interpreted value Absolute error Real value Interpreted value Absolute error Real value Interpreted value Absolute error Real value Interpreted value Absolute error Val1 0.404 0.418 -0.014 -0.261 -0.362 0.101 3.142 2.994 0.148 3.773 3.676 0.097 Val2 -0.124 -0.127 0.003 -0.314 -0.018 -0.296 5.877 5.839 0.038 8.296 8.390 -0.094 Val3 0.318 0.607 -0.289 0.797 0.773 0.024 2.631 2.874 -0.243 2.782 3.283 -0.501 Test1 0.882 0.872 0.01 0.592 0.641 -0.049 1.059 1.057 0.002 0.143 0.148 -0.005 Test2 0.994 0.052 0.942 -0.618 -0.004 -0.614 3.336 3.207 0.129 2.000 2.335 -0.335 Test3 0.589 0.568 0.021 -0.472 -0.417 -0.055 2.826 2.783 0.043 2.687 2.655 0.032