A Virtual-Flux State Observer-Based Inductance Identification Method for Model Predictive Control of Grid-Tied Inverters With a Finite Phase Angle Set-Based PLL

Abstract

To reduce the calculation burden of the conventional finite phase angle set-based phase-locked loop (PLL) and enhance the parameter robustness of the model predictive control for grid-tied inverters (GTIs), an improved finite phase angle set-based PLL and a new inductanceidentification method basedonvirtual-flux state observer (VFSO) are proposed in this paper. First, three improved PLLs are proposed based on the principles of the conventional finite phase angle set-based PLL. Second, the proposed PLLs are further improved by redesigning the cost function, which can search the optimal phase angle within the region of [0, π] instead of [0, 2π], reducing the iteration times. Third, a new VFSO with only one observer parameter is proposed, based on which a new inductance identification method is presented. Fourth, by analyzing the characteristics of the proposed VFSO, it is found that a larger observer gain can improve the dynamic convergence speed of the proposed VFSO, while a smaller one can enhance its robustness against grid frequency deviation. So, an observer gain adaptive method is proposed, aiming to improve both the dynamic and steady state control performance of the proposed inductance identification method. Finally, comparative experimental studies verify the effectiveness of the proposed strategies.