Electrical Parameter Identification of Medium-Frequency DC Resistance Spot Welding System Using Intelligent Algorithm

Abstract

The three-phase medium-frequency dc resistance spot welding (RSW) system has a complex electrical structure that includes various electrical components. As it is a multiparameter coupled, nonlinear, and time-varying system, it has difficulties in achieving the accurate modeling and analyses when combined with the metal melting process. This study provides a parameter identification method for the electrical structure, which only requires the measurement values of voltages and currents in the input and output terminals of the structure. The method is based on four mathematical equations, according to the electrical and magnetic characteristics of the electrical structure. A particle swarming optimization algorithm was employed to solve the equations. This helped achieve the goal of accurately identifying key parameters, which were difficult to measure directly. Using a model established by MATLAB Simulink, a detailed identification process for seven key parameters of electrical components was properly organized. The corresponding comparisons of two types of currents in terms of measurement and calculated values validated the effectiveness of the proposed method. Two specially designed welding experiments using different metal sheets were conducted to validate the reliability, accuracy, and universality of the results. This article can benefit a comprehensive model establishment as well as system and process optimization for RSW operations.