Figures & data
Table 1. Comparable studies in the literature. The figures of merit are reported in the cases in which the end point of sulfur ([S]end) is used as the output variable or the predictions have been reported.
Figure 1. The overall grey–box model structure.
Figure 2. The training of the black-box model and the use of the pre–trained black–box model as a part of the overall grey–box model.
Table 2. Operators for the genetic algorithm.
Table 3. Referenced studies to estimate the sulfide capacity model.
Table 4. The average values before the studied treatments.
Table 5. The characteristic PSD’s of the reagents with corresponding RRS distribution parameters. The values are given in micrometers (μm).
Figure 3. The parity plots and the histograms of the estimated model parameters.
Table 6. The CI95 of the estimated parameters.
Table 7. The estimated model parameters with the RCGA and corresponding figures of merit computed for the test set.
Figure 4. The simulated sulfur trajectories for 6 randomly selected treatments with corresponding confidence intervals (min, max and 95% limits).
Figure 5. The reagent rates for 6 randomly selected treatments with corresponding confidence intervals (min, max and 95% limits).
Figure 6. The 1-fP with corresponding 95% CI.
Figure 7. The simulated sulfur trajectories and their corresponding 95% CI with two different PSD’s with the histograms of the end sulfur contents.
