Robust control for a direct-driven permanent magnetic synchronous generator without mechanical sensors based on model reference adaptive backstepping control method

Abstract

This article is concerned with the robust control issues of direct-driven permanent magnetic synchronous generator without mechanical sensors. In this study, a novel nonlinear model reference adaptive backstepping control method is proposed. Extended electromotive force estimation-based rotor position and rotational velocity estimation methods are adopted, in which the stator parameters are identified online. Since the identified parameters are merely used in extended electromotive force estimator, the mathematically controlled plant cannot reflect the actual model, which decreases the robustness of the control. To provide a high-performance robust controller against uncertainties in stator parameters and mechanical torque fluctuation, a torque observer is adopted, and an adaptation law is derived with the identified parameters and estimated mechanical torque, in the sense of the Lyapunov theory. Meanwhile, the feasibility of the substitution of the actual stator parameters and state vector with estimated values is mathematically demonstrated based on the Lyapunov theory. Simulation experiments have been carried out both in surface permanent magnetic synchronous generator and interior permanent magnetic synchronous generator, and the results show that compared to the conventional process-interaction decoupling controller, the proposed controller has better disturbance rejection performance against wind speed perpetuation and parameter uncertainties.