Navigation of Welding Torch for Arc Welding Process

Abstract

Torch position, traveling speed, and attitude play an important role in welding quality control for both manual and robotic arc welding process. Besides, their detection may facilitate welding process monitoring, welder training, as well as novel ways for many other interesting and useful applications. Yet, few research has been done in the torch navigation in manual arc welding process. In this paper, we describe and implement a Kalman-based framework to estimate the position, traveling speed, and attitude of a torch in manual welding process. The proposed framework makes use of an inertial navigation system (INS) mechanization algorithm, a Zero Velocity Update (ZUPT) methodology, and a unscented Kalman filter (UKF). The proposed measurement employs a low-price miniature wireless inertial measuring unit (WIMU) consisting of a tri-axial accelerometer, one tri-axial gyroscope and a magnetometer. The performance of the proposed scheme has been evaluated by welding simulations with different types of fit-ups. Statistics shows that the position errors are about 1% of the total traveled distance, which are considered acceptable for the intended manual welding applications.