Estimation of pressure and viscous drag force for an axisymmetric underwater body by artificial neural network

The present work developed an ANN1 model as a new approach for estimate the pressure and viscous drag force whichhave been most usage in hydrodynamics design for an underwater body. A faster and simpler solution was obtained bysome equations from results of ANN model for calculate data against complex and big differential equations andprogramming or very costly and limited experimental approach. The FFNN2 applied that is the most suitable model andcan predict properly value of pressure and viscous drags force. By CFD approach a dataset for an underwater bodyproduced and used for training the ANN model. In this problem, a dataset involved the different dimensional of thebody that pressure and viscous drag force are two outputs of CFD method. The ANN contains three layers such asinput, hidden, and output layer. For input layer used from the diameter of body, diameter of nose disc, length of body,length of nose, and velocity. In test phase attained the good agreement with unseen data. The present approach showsthis type proposed for the neural network model, could used for extend choice of specific problems to determined moredata and design the reasonable body for their mission.

Keywords:

Feedforward neural networks , backpropagation algorithm , hydrodynamic , pressure drag force , viscous drag force