Design Of The Fuzzy Control Seam Tracking System For Arc Welding Based On DSP

In order to improve the level of automatic welding in CO₂ shielded welding, the tracking system of welding seam based on arc sensor for CO₂ shielded welding is developed in this paper. Through researching the theory of arc sensor combining with characteristics of CO2 shielded welding, a method of calculating the difference of discrete integral for arc current is adopted to judge the direction of deviation and offset between welding line and welding torch. The static and dynamic physical models of arc sensor are also deduced and analyzed. To meet the precision and real-time performance together with achieving good man-machine interface, a master-slave control structure is adopted in this paper. The personal computer (PC) is chosen as host computer and the digital signal processor (DSP) is chosen as the core chip of slave controller. The slave controller also includes the necessary peripheral circuit, extension circuit and the actuating mechanism—welding oscillator. The process of short circuit transition in CO₂ shielded welding is non-linear and non-determinacy, so a method of fuzzy control that needn't accurate mathematic model is used in this paper. Meanwhile, to cover the shortage of general fuzzy controller, a kind of parameters self-adjusting fuzzy controller is put forward and relative simulation is done by MATLAB. Based on the simulation data, a method of modular programming is adopted to finish the software of slave controller. The software of host computer is programmed under Borland Delphi 6.0. Utilizing the serial communication, the data acquired are transmitting to host computer and then be displayed or analyzed. Finally, the feasibility and practicability of parameters self-adjusting fuzzy controller are validated by a series of welding experiments. Test results indicate that the effect of seam tracking is fine when warp angle is within 8 degree.