Discrete-Model Identification for Nonlinear Laser Welding Process

This paper introduces the identification results of a discrete-time model of a laser welding process whose input is the reciprocal of the welding speed and output is the width of the weld pool. To model this nonlinear process, a Hammerstein model is used. A series of step response experiments have been conducted to obtain the experimental data. A method has been proposed to identify the discrete Hammerstein model in two steps. MATLAB programs are utilized with Least Square method for parameters identification. Experimental results confirmed that the proposed identification method for the nonlinear dynamic discrete-time as described by a Hammerstein model is effective to approximate the laser welding process.