Development of the intelligent algorithm to control on-line bead height for robotic welding process

摘要

Purpose: The demand to increase productivity and quality, the shortage of skilled labour and strict health andsafety requirements finally led to the development of the robotic welding process to deal with many problems ofthe welded fabrication. Many techniques were developed to control process parameters to get the optimal beadgeometry during welding process by minimizes their magnitude in the affected area.Design/methodology/approach: The development of thermo mechanical mechanism in some techniques isnot fully understood. To solve this problem, we have carried out the sequential experiment based on a Taguchimethod and identified the various problems that result from the robotic GMA welding process.Findings: To characterize the GMA welding process and establish guidelines for the most effective joint design.Also using multiple regression analysis with the help of a standard statistical package program, SPSS, on anIBM-compatible PC, a quadratic model has been developed for on-line control which studies the influence ofprocess parameters on bead height and compares their influences on the bead height to see which one of processparameters is most affecting.Originality/value: This model developed has been employed the prediction of optimal process parameters andassisted in the generation of process control algorithms.