A survey of the sensing and control techniques for robotic arc welding

The sensing and control techniques to control the position and orientation of welding torch and to achieve the adjustment capacity of the welding process variables in robotic arc welding are discussed. Various kinds of sensors such as arc sensors, ultrasonic sensors and vision sensors have been used to conduct the weld joint detection work. The visual sensing technology can be can be divided into passive method and active method according to imaging light source. No assistant light source is needed to illuminate the region of interest while assistant light source is needed in an active method. Classical controllers such as PID controllers are used to adjust the welding torch to follow the seam precisely according to the weld joint features. Intelligent control methods are often required to perform the torch adjusting task. Fuzzy control is adopted to improve the stability and control precision of response and decrease system overshoot.