Impedance modeling, analysis, and adaptation of grid-connected inverters with PLL

Abstract

This thesis presents impedance modeling, analysis, and adaptation of three-phase grid-connected VSC to address their high-frequency control instability and resonance problems. The focus is grid-connected inverters that use a phase-locked loop (PLL) to synchronize with the grid voltages. The primary application considered is the integration of wind energy into the power grid, but the technique can also be extended to other systems such as grid-scale solar energy and energy storage. The harmonic linearization method is used to model the positive-sequence and negative-sequence impedance of the three-phase inverter. The models are further used to analyze resonance and control instability of grid-connected inverters, and to optimize PLL design to minimize the potential for such resonance problems. Online measurement of the grid impedance by using the grid-connected inverter as the measurement device is also presented, and the measured grid impedance is used to demonstrate adaptive control that could maintain inverter control stability in face of large grid impedance variations. The developed impedance models and the proposed control methods are verified and demonstrated by experiments on the Distributed Generation Test-Bed.

Three-phase voltage-source converters (VSC) are basic building blocks for many applications in power systems, including grid integration of renewable energy and energy storage, high-voltage dc transmission, and flexible ac transmission systems. High-frequency pulse-width modulation control is essential for VSCs to perform these functions, and makes VSCs more advantageous over conventional line-commutated converters in terms of weight, cost, speed of response and operational flexibility. On the other hand, high-frequency control also creates potentials for new stability and resonance problems above the grid fundamental frequency that are not common in conventional power systems. The impedance-based stability theory in Ref. [1]-[2] has proven to be an effective tool for characterizing and solving such high-frequency problems.