中文摘要：

熟练焊工通过观察熔池表面三维形态实时调整焊枪姿态从而保证在复杂工况下良好的焊缝成形。这表明焊工经验和技能的实质是焊枪姿态与熔池的交互行为,但目前尚无有效的方法和手段清晰地描述出焊工的这种经验和技能。为了研究焊枪姿态与熔池形貌的交互行为,采用动态倾角传感器实时传感焊枪姿态的方法,结合激光视觉传感技术,搭建焊枪姿态与熔池形貌实时传感系统平台,实现焊枪姿态与熔池形貌的实时传感。通过德国CLOOS弧焊机器人对倾角传感器进行标定,利用搭建的试验平台传感并分析不同水平焊工焊接过程试验,结果表明：倾角传感器的角β与？和的方均根误差C=0.019 6°,角β与α和的方均根误差C=0.031 7°,角α与γ和的方均根误差C=0.105 8°;焊接过程中,熟练焊工焊枪姿态平稳,波动较小,其对最佳熔池形貌维持能力强,焊缝成形好;新手焊工焊枪姿态波动较大,姿态突变较多,缺乏应对熔池形貌变化所必需的焊接经验和技能,对焊枪姿态角度的调整单一,焊缝成形差,余高不均。

英文摘要：

Experienced skilled welders can ensure weld pool shape well by adjusting to the welding torch orientation according to the change of weld pool they observed under complicated working conditions. It indicates that the essence of the welders＇ experience and skills are the interactive behaviours of welding torch orientation and weld pool shape. Unfortunately, there is no effective method and mean to describe them out clearly. To solve this problem, a sensor system of welding torch orientation and weld pool shape is built with the principle of attitude sensing and laser vision sensing technology. The CLOOS arc welding robot is used to verify accuracy of sensor, it indicates that the root mean square value of β and γ is 0.019 6°, β and α is 0.031 7°, α and γ is 0.105 8°, respectively. Through sensing and analyzing welding test operated by different welders, it shows that during the welding process, welding torch of skilled welder is stable and small fluctuate. The skilled welder has a better maintain weld pool shape ability. A well bead weld can be formed with the skilled welder; the welding torch of new welder is fluctuated and mutated frequently. They lack the required welding experience and skills response to the change of weld pool shape so that the welding torch angle is adjusted singly. The weld seam shape is bad and the residual high is uneven.