在现代焊接生产中，随着自动化、智能化程度的不断提高，焊缝跟踪技术已成为自动焊接过程中最为关键的技术之一。本文在电弧传感器理论模型及其试验研究的基础上，结合CO2焊自身熔滴过渡的特点，通过对摆动式电弧传感器电流信号的提取、分析，提出了一种可实现焊道坡口边界识别和焊道轨迹识别的焊缝跟踪控制系统，可用于大尺寸和不同形状坡口焊的摆动焊接中。 本文首先通过实验建立了CO2焊电弧传感的静态模型，得到了焊炬高度与焊接电流之间的关系因子，以焊接电流的变化反映的焊炬高度的变化作为边界识别和跟踪控制的参量，为焊炬自动控制奠定了基础。为满足高性能、低成本、经济型焊缝跟踪系统的目的，以Mircochip公司PIC16F877单片机为主控制单元设计了跟踪控制系统中的各外围电路，包括信号采集和处理电路、执行机构的驱动电路等，并确定了基于这些硬件电路的软件结构与思想。 由于短路过渡的CO2气体保护焊不可避免地会受到各种干扰因素的影响，在硬件滤波的基础上，在信号提取及处理时采用了对电弧电流信号进行离散积分差值的判断方法，用以提高整个系统的跟踪精度。步进电机具有实时、易控制、惯量小、精度高等特点，选择步进电机作为跟踪系统的执行机构。 电弧传感器是利用电弧本身作为焊炬高度的传感信号。与其他传感器相比，电弧传感器不需要在焊炬上附加另外传感器元件，而且结构简单，有更好的动态品质及更高的控制精度，相对于其它传感技术来说是一个低成本、经济型的系统。

In modern welding industry, with the continuous improvement of automation and intelligent levels, welding seam tracking technology has become one of the key technologies of the automatic welding process. Based on the study of the arc sensors theory model, considering the characteristics of CO2 welding melt transition and by extracting and analysing the current signals of the swing-type arc sensors, this paper proposes welding seam tracking and control system which can achieve welding path and crevasses borderline identification. This system quite adapts to welding production about large dimension welding seam and diversified shape crevasses. This paper firstly established a static model about CO2 welding arc sensor by experiment, and educed the connection coefficient between welding current and welding torch height. Thus, the parameters of borderline identification and welding seam tracking can achieved by detect welding torch height which welding current’ change reflected. In order to produce a high-performance, low-cost, economic-type welding seam tracking system, we used a SCM with very high performance-price ratio, PIC16F877, produced by Mircochip Corporation. With the PIC chip as the main control modules, designed the external circuits of the tracking and control system, including the signal acquisition, processing circuits and the driving circuits of the executing organs, and determined the software structure and thoughtway based on these hardware curcuits. Considering the short-circuit transitted CO2 gas-protected welding process would inevitably be affected by all kinds of interfering factors, we extracted and disposed the arc current signals by discrete integral determinant method, based on hardware filter, improved the tracking accuracy of the entire system. The stepper motor is used as the executing organs of the tracking system, because of its characteristics of real-time, easy controlled, small inertia and high precision. Arc sensor used the arc itself as the signal of the welding torch height. Compared with other sensors, arc sensors do not need to be attached with other sensor modules. They also have simpler structures, better dynamic qualities and higher control precision. Thus, the welding seam tracking system is a low-cost, economic-type system, which is an important foundation for follow-up research and practice.