Improving the Electroplating Coating Uniformity Based on Fuzzy Control of the Setting Value Using the Combined Defuzzification Method with the Presence of Stochastic Influences

Abstract

The article describes a method to improve the uniformity of electroplating using fuzzy logic. This method provides for the replacement of the non-stationary model of the process dynamics in distributed coordinates with the quasi-stationary model of the process dynamics in lumped coordinates. The production knowledge model with "IF-THEN" rules is developed for the stochastic influences taken into account. The dynamic choice of the defuzzification method is justified by solving the problem of minimizing the absolute deviation of the coating thickness average value obtained from the non-stationary model from the predicted value according to the quasi-stationary model at const values of stochastic influences. As an example, fuzzy control of the current for nickel coating is considered, taking into account the stochastic influence of the electroplating time, detail area, temperature and acidity of the electrolyte. The experimental results prove the effectiveness of the combined defuzzification method in comparison with their independent use in the control process.