Answers are grounded in this paper and cite their sources.
Physical equations provide an interpretable temporal structure for drying, while variations in operating conditions are represented through parameter relationships. This study examines whether a small multi-layer perceptron (MLP) can improve Newton–Lewis predictions for unseen operating conditions. The dataset comprises 14665 time points from 79 experiments covering 27 combinations of temperature, dry-air mass flow rate and humidity ratio . A capacity parameter is calculated for each experiment from its initial and reference masses. A log-linear map and a feedforward neural network share the same exponential drying equation and are calibrated using normalized curve errors. In 27-fold leave-one-condition-out validation, the neural network reduces the median curve root mean square error (RMSE) from 279.63 to 254.48 g and the median absolute error in the time required to remove 90% of the reference removable water from 59.51 to 51.84 minutes. Three random initializations yield similar results. The nonlinear representation of operating conditions improves rate prediction, while stage-dependent changes throughout drying appear in the residuals. Embedding parameter learning within analytical kinetics provides a compact computational link between operating conditions, mass trajectories and target times.
No takes yet. Share an insight, caveat, or question.
Qiyong Yang (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: