Optimisation of spot-welding process using Taguchi based Cuckoo search algorithm

Abstract

The present work evaluated the efficiency of a Taguchi-based Cuckoo Search (CS) algorithm for optimizing the spot-welding process. The L9 orthogonal array of the Taguchi method is used for the conduction of required experiments. During the study input parameters are welding current (kA), hold time (cycles), welding time (cycles), and electrode pressure (kPa) while Peak Load, kN has been considered as an output parameter. The required objective function is developed through regression model formulation. Initially, CS operating parameters such as maximum number of iterations, number of nests, and probability to discover a cuckoo egg by the host bird is optimized through the Taguchi method and are found as 40, 20, and 0.5 respectively. That optimized CS further optimizes the spot-welding process. The maximum peak load of 34.5 kN is obtained if the welding current is 30.7 kA, welding time is 32 cycles, hold time is 20 cycles, and electrode pressure is 480 kPa respectively. Experimental validation yields a very low % error of 1.93% during optimization with an optimized CS method and the error is substantially high if optimization is conducted using a non-optimized CS method.