Neural based hybrid metaheuristic technique for computing rotating transport of Falkner-Skan flow

• Published in: Alexandria engineering journal Vol. 57; no. 3; pp. 2123 - 2132
• Main Authors: Iqbal, Z., Azhar, Ehtsham, Mehmood, Zaffar, Sabir, Mirza Muhammad
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.09.2018; Elsevier
• Subjects: Artificial neural network; Falkner-Skan flow; Genetic algorithm; Interior-point algorithm; Rotating transport; Interior-point algorithm; Artificial neural network; Falkner-Skan flow; Genetic algorithm; Rotating transport
• ISSN: 1110-0168
• DOI: 10.1016/j.aej.2017.06.011

In present analysis we are focused on neural based hybrid metaheuristic technique for computing rotating transport of Falkner-Skan flow. We incorporate modeling of present problem using artificial neural network (ANN). An approximate model of the equation in an unsupervised manner is built to subjugate effectiveness of ANN. Stochastic solvers based on hybrid approach are used to verify the accuracy of the designed scheme. The hybrid metaheuristic scheme involves genetic algorithm (GA) along with interior point algorithm (IPA) for the training and weights optimization of ANN. The proposed scheme is evaluated for Falkner-Scan boundary value problem by taking different values of material and rotation parameters. Moreover, comparative studies of the proposed solution are made with standard numerical algorithm in order to establish reliability of present neural based hybrid metaheuristic technique.