Function finding via genetic expression programming to predict microhardness of Ni/Al2O3 nanocomposite coatings

A new proposing model based on gene expression programming (GEP) to predict the microhardness of Ni/Al2O3 nanocomposite coating was the subject of the present study. Accordingly, a series of the laboratory experiments was designed by the factorial D-optimal array. This was accomplished by considering the most effecting practical electrodeposition parameters including the amount of Al2O3 nanoparticles in the bath, current density, temperature, magnetic stirring rate, time of stirring, and plating time as the input and the microhardness of the coating as the output of model. Various performance criteria including determination (R2) coefficient, the mean absolute error (MAE), and the root relative squared error (RRSE) were utilized to evaluate the developed models. Finally, the model with R2 = 0.9752, MAE = 0.030 and RRSE = 0.158 was developed as the optimum proposed function. Also, the results of the sensitivity analysis confirmed that the current density was the most effective parameter, while the amount of Al2O3 nanoparticles in the bath, plating time, magnetic stirring rate, time of stirring, and temperature had relatively lower effect. In conclusion, the exclusive features of the GEP simulation have been approved to determine Ni/Al2O3 nanocomposite coatings microhardness.