Modelling and simulation of coal gases in a nano-porous medium: a biologically inspired stochastic simulation

Abstract

This paper aims to study the dynamics of the unsteady pressure flow of coal gases caused by the temperature conditions and compressibility in the presence of a nano-porous medium using soft computing technique. To immaculately understand the mechanism, a novelty in the partial differential equation is augmented by considering the fractional-order Caputo derivative, which produces theoretically significant and accurate approximation. Subsequently, the constructed model is experimentally simulated by means of artificial neural network (ANN) and a stochastic process based on a firefly algorithm (FFA). ANN has the ability to approximate and transform the differential equation into an error minimization problem, while FFA efficiently minimizes the error function and optimizes the unknown weights of the constructed network. Furthermore, two error measuring tools; mean absolute error and root mean square error, is also formulated to evaluate the performance index of the designed scheme. Accordingly, the designed scheme is systematically elaborated to assess the pressure sorption of coal gases such as nitrogen (N2) and carbon dioxide (CO2). The accuracy of the obtained approximation shows the competitiveness of the considered scheme. Notably, the deliberation provides substantial indications about the dynamical behaviour of coal gases, which can be implemented significantly on various dynamical problems.