پديدآورندگان :
Azarbarmas Morteza azarbarmas@sut.ac.ir Assisstant Prof., Faculty of Materials Engineering, Sahand University of Technology, Tabriz, Iran; , Mirjavadi Seyed Sajad s.mirjavadi@ut.ac.ir M.S., School of Mechanical Engineering, College of Engineering, University of Tehran, Tehran, Iran; , Ghasemi Ali s.mirjavadi@ut.ac.ir Ph.D., Department of Mechanical Engineering, Faculty of Engineering, North Tehran Branch, Islamic Azad University, Tehran, Iran;
كليدواژه :
Corrosion Potential , Coating , Artificial Neural Network (ANN) , Modeling
چكيده فارسي :
In this investigation, the corrosion behavior of 1020 steel plates in term of corrosion potential was modeled using the ANN approach. Input variables were the surface pre-treatments, anticorrosive coatings and the time of immersing in the corrosive environment. A feedforward multi-Layer perceptron neural network was used. The reliability and speed of “Levenberg–Marquardt”, “Scaled conjugate gradient”, and “Resilient backpropagation” algorithms were also compared, and it was concluded that the Levenberg–Marquardt is the most accurate and the fastest algorithm for modeling. The results showed that the estimated corrosion potentials of samples are in good agreement with the actual data.
