Optimizing the RSW Process Quality of Auto-Body via the Taguchi Method and a Neural-Genetic Approach
This paper applies an integrated approach using the Taguchi method, neural network (NN) and genetic algorithm (GA) to optimize the tensile-shear strength of resistance spot welding (RSW) specimens in automotive industry. The proposed approach consists of two stages. First stage executes initial optimization via Taguchi method to construct a database for the NN. In second stage, a NN with Levenberg-Marquardt back-propagation (LMBP) algorithm is used to provide the nonlinear relationship between factors and the response. Then, a GA is applied to obtain the optimal factor settings. The experimental results showed that the tensile-shear strength of the optimal welding parameter via the proposed approach is better than apply Taguchi method only.