Comparison Study the Modeling of Limiting Current in the Magneto Electrodeposition of Vanadium using Neural-Wiener Model and Feed Forward Neural Network

Abstract

Vanadium has long been used as a corrosion-resistant coating, including as a metal alloy for battery cathodes. However, batteries discovered with non-smooth cathode surfaces due to the fabrication process have a short battery life. So, a cathode coating stage is required via the electroplating method under the influence of a magnetic field or Magneto Electro Deposition (MED). Knowing the limiting current in MED is very important because the optimum mass transport achieves at the limiting current (iB). The smoothest and most compact electrodeposit surface will occur at this limiting current. In this study, Feed Forward Neural Network and Neural-Wiener are suggested and compared as a nonlinear modeling approach to determine the ideal limiting current because of their strong capacity to anticipate the link between input and output from experiment data. The Levenberg-Marquadt optimization technique with hidden neurons was used to evaluate and compare the modeling capabilities of two neural networks, the Feed Forward Neural Network, and the Neural Wiener. The results of this study are presented as a comparison of the Mean Square Error (MSE) values obtained from the nonlinear modeling of two artificial neural network algorithms. The algorithm that models the ideal current limiting has the lowest MSE value (iB).