Adaptive Fuzzy Wavelet Backstepping Control for Two-axis Motion Trajectory Application

Abstract The product quality of a machining tool often corresponds to the accuracy of contour tracking. To achieve the precision tracking control of a biaxial Permanent Magnet Synchronous Motor (PMSM) motion system, attempts made in this paper are to introduce the Adaptive Fuzzy Wavelet Backstepping Controller (AFWBC). Considering the uncertainties of this model, the dynamics used in the control law is estimated by the fuzzy wavelet neural network. The method under consideration used the temporal positioning property of wavelets in the time domain, coupled with the reasoning abilities of a fuzzy system, to effectively estimate a reliable plant model. Then, the estimated model is integrated in the backstepping controller with the adaptive law to tackle system’s nonlinearity and uncertainties. The stability of overall system is guaranteed via a rigorous Lyapunov stability analysis. To better validate the system, both simulation and experimental setups are conducted. During the experimental evaluation, non-uniform rational B-spline (NURBS) is employed for contour planning due to its high-accuracy interpolation capabilities. The results showed that the AFWBC method effectively could reduce errors and so significantly improve trajectory tracking accuracy and localization.