Prediction of transient chemistry effect during fuel pyrolysis on the pressure drop through porous material using artificial neural networks