Sammendrag
Droop control is the most common approach for controlling microgrids (MGs) and interfacing distributed generators
(DGs) due to its inherent power-sharing characteristics and low
dependence on communication systems. However, the stability
of MGs operated with this control paradigm is sensitive to the
droop parameter values and system topology. This situation will
be even more challenging in future grid scenarios, where system
reconfiguration and parameter uncertainty will become more
relevant. In this paper, a centralized controller for improving
the stability margins of a MG is proposed. Firstly, the apparent
impedance of the MG is periodically identified by injecting
a multi-tone current perturbation. Then, the matrix fitting
technique is applied to calculate the system eigenvalues based on
the estimated apparent impedance. Depending on the location
of the dominant eigenvalues, the droop gains are modified
by using a control law to ensure sufficient stability margins.
No previous information of the system topology or parameter
values is required. It is shown that this method can greatly
improve stability margins without compromising the inherent
power sharing, transient performance and low dependence on
communication systems of the droop control. Simulation and
experimental results for a MG based on DGs rated at 60 kVA
are included to validate the proposed control scheme.
Index Terms—Microgrid Control, Impedance Estimation,
Adaptive Control, Apparent Impedance Analysis.
(DGs) due to its inherent power-sharing characteristics and low
dependence on communication systems. However, the stability
of MGs operated with this control paradigm is sensitive to the
droop parameter values and system topology. This situation will
be even more challenging in future grid scenarios, where system
reconfiguration and parameter uncertainty will become more
relevant. In this paper, a centralized controller for improving
the stability margins of a MG is proposed. Firstly, the apparent
impedance of the MG is periodically identified by injecting
a multi-tone current perturbation. Then, the matrix fitting
technique is applied to calculate the system eigenvalues based on
the estimated apparent impedance. Depending on the location
of the dominant eigenvalues, the droop gains are modified
by using a control law to ensure sufficient stability margins.
No previous information of the system topology or parameter
values is required. It is shown that this method can greatly
improve stability margins without compromising the inherent
power sharing, transient performance and low dependence on
communication systems of the droop control. Simulation and
experimental results for a MG based on DGs rated at 60 kVA
are included to validate the proposed control scheme.
Index Terms—Microgrid Control, Impedance Estimation,
Adaptive Control, Apparent Impedance Analysis.