Electrode displacement patterns inferred as the optimal control criteria during the resistance spot welding process

Currently real-time control and online quality estimation of the resistance spot welding (RSW) has benefited a lot from monitoring the electrode displacement of nugget thermal expansion. Based on these emerging monitoring techniques a new approach is proposed to determine the optimal control parameters and help to assess the weld quality. A causal model is built with the offline trained Bayesian Belief Networks (BBN) as a pattern determination net which deals with the optimal pattern of the electrode displacement, i.e. the ideal parameter combination between the maximum electrode displacement and its expansion velocity, to provide more reliable welding process and qualified welds. The experimental results show that the proposed approach is valid and feasible to determine the controlled parameters for welding robots.