Artificial neural network based production forecasting for a hydrocarbon reservoir under water injection
NEGASH Berihun Mamo,YAW Atta Dennis
Table 7
Information related to the optimum architecture of ANN model for gas flow rate prediction.
Items
Value/content
Number of inputs
13
Number of hidden Neurons
15
Output
Gas flow rate
Training algorithm
Bayesian regularization
R
2
(Training)
0.949
R
2
(Testing)
0.971
Number of epochs
844
Training time
96 s