Fig 1.
The structure of ANN: (a) Simple ANN, (b) Deep neural network.
Fig 2.
Linear response surface of three IVs: (a) The connection among x1, x2 and y, (b) The connection among x1, x3 and y, (c) The connection among x2, x3 and y.
Fig 3.
Nonlinear response surface of three IVs: (a) The connection among x1, x2 and y, (b) The connection among x1, x3 and y, (c) The connection among x2, x3 and y.
Fig 4.
Test materials: (a) cement (b) sand (c) RBA.
Table 1.
Mix proportion plan of RBAC.
Fig 5.
Influence of volume content of CA and FA on RBACCS.
Fig 6.
The RBACCS with RBA volume fraction in the range of 40% ~ 50%.
Fig 7.
Response surface of Pg, mw/mc and fcu: (a) Three dimensional response surface, (b) Response surface projection.
Fig 8.
Response surface of Ps, Pg and fcu: (a) Three dimensional response surface, (b) Response surface projection.
Fig 9.
Influence of activation function on loss function.
Fig 10.
The effect of the number of hidden layers on the loss function: (a) The calculation steps are 5×104 to 6×106 (b) The calculation steps are 6×105 to 6×106.
Fig 11.
The RSM-DNN model structure.
Fig 12.
MRSF-DNN and DNN training results: (a) The training results of MRSF-DNN and DNN, (b) Probability distribution of MRSF-DNN, (c) Probability distribution of DNN.
Fig 13.
Extended analysis results of DNN and MRSF-DNN: (a) Extended analysis results of DNN, (b) Extended analysis results of MRSF-DNN.
Fig 14.
Extended analysis results of DNN: (a) RBA volume content is in the range of 30%-35%, (b) RBA volume content is in the range of 50%-55%.
Fig 15.
Extended analysis results of MRSF-DNN: (a) RBA volume content is in the range of 30%-35%, (b) RBA volume content is in the range of 50%-55%.