Abstract:In gas metal arc welding,it is difficult to establish an accurate model of arc length nonlinear systems.Moreover,the system has various uncertainties and is vulnerable to the problem of internal and external disturbance.In this paper,an improved model-free adaptive control method was proposed to address the above issues.First of all,a dynamic linear generalized model with measurable and unmeasurable disturbances and uncertain factors was designed.A parameter estimation algorithm was used to estimate pseudo partial derivative online.Then,the disturbance observer was introduced to compensate the internal and external disturbances and uncertainties in real time. Furthermore,the control law of arc length system was designed by constructing a criterion function.On this basis,a model-free adaptive control scheme was developed.Finally,the robust stability of the method was validated.The simulation results showed that the method had better tracking performance and faster response speed than the existing model-free adaptive control schemes.It also had a stronger robustness when the unmeasured disturbances and uncertain factors are encountered in the welding process,which ensuring the stability of the arc length in welding process and the better control of quality.