Neural network design for cubic autocatalysis chemical processes, the flow of a Darcy‐Forchheimer viscous fluid is optimized for entropy

Abstract

AbstractIn this article, neural network backpropagation (NNBP) is used to present two‐dimensional Darcy‐Forchheimer viscous fluid (DFVF) of viscous liquid against a stretching sheet under the influence of heat sink/source and viscous dissipation. Mathematical modelling and numerical simulation are also discussed. Thermal polymer processing, engineering processes, and industrial processes all benefit from the dynamic entropy generation phenomenon. Scientists are primarily interested in finding ways to decrease entropy formation to polish up the thermal performance of industrial systems. In this continuation, the optimal frame for the Darcy‐Forchheimer flow with a curved surface has been developed. Non‐linear partial systems are converted to dimensionless differential systems through the employment of the appropriate variables. Utilizing tools from the bvp4c method, the problem's highly nonlinear equations are numerically solved. These ODEs are solved by applying the bvp4c solution approach to produce the reference dataset for NNBP. Using this reference dataset in MATLAB, the graphs for the solution and error analysis for changing various parameters are analyzed. The performance of NNBP is validated using mean squared error data, regression analysis, and error histogram. The DFVF solution is examined using the testing, validation, and training procedures. Examined through influential variables the effects of fluid flow, thermal field, Bejan number, and concentration.