Robust Ultra-Local Model Control for Single-Axis Servomechanism Actuated Through PMSM Drive Via Optimal Extended State Observer

This paper proposes an robust ultra-local model control (RULMC) scheme for high-accuracy positioning of a single-axis ball-screw servomechanism driven by permanent-magnet synchronous motor (PMSM) drive system. Owing to the uncertainties in real-time applications, friction forces, backlash nonlinearities, parameter disparities and exterior disturbances, the design of a model-based controller cannot accomplish this control objective. Ultra-local model control (ULMC) schemes are model-free controllers that provide good rejection for different disturbances and model uncertainties. The parameters of the ULMC schemes are determined by trial-and-error methods. Therefore, it is difficult to determine the optimal parameters of the ULMC that provide the best dynamic performance of the servomechanism. Hence, to optimize the behavior of system, the proposed composite control structure combines ULMC, an extended state observer (ESO) and an optimal H∞ controller. The ULMC is designed to stabilize the servomechanism, the ESO is developed to estimate the uncertain dynamics and states of the single-axis servomechanism whereas the optimal H∞ controller is adopted as the auxiliary controller to compensate the estimation error of the ESO and improve system stability. The experimental results confirmed the robustness of the proposed RULMC with optimized ESO in case of parameter uncertainties and system disorders existence compared with the conventional ULMC.