An SRF-PLL-Based Sensorless Vector Control Using the Predictive Deadbeat Algorithm for the Direct-Driven Permanent Magnet Synchronous Generator

This paper proposes an enhanced sensorless vector control strategy using the predictive deadbeat algorithm for a direct-driven permanent magnet synchronous generator (PMSG). To derive favorable sensorless control performances, an enhanced predictive deadbeat algorithm is proposed. First, the estimated back electromotive force (EMF), corrected by a cascade compensator, was put into a deadbeat controller in order to improve the system stability, while realize the null-error tracking of the stator current at the same time. Subsequently, an advance prediction of the stator current based on the Luenberger algorithm was used to compensate the one-step-delay caused by digital control. Maintaining the system stability, parameters of the controller were optimized based on discrete models in order to improve the dynamic responses and robustness against changes in generator parameters. In such cases, the proposed methodology of synchronous rotating frame phase lock loop (SRF-PLL), which applies the estimated back EMF, can observe the rotor position angle and speed without encoders, realizing the flux orientation and speed feedback regulation. Finally, the simulation and experimental results, based on a 10-kW PMSG-based direct-driven power generation system, are both shown to verify the effectiveness and feasibility of the proposed sensorless vector control strategy.