Graduation Year

2021

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Electrical Engineering

Major Professor

Zhixin Miao, Ph.D.

Committee Member

Lingling Fan, Ph.D.

Committee Member

Chung Seop Jeong, Ph.D.

Committee Member

Achilleas Kourtellis, Ph.D.

Committee Member

Kaiqi Xiong, Ph.D.

Keywords

Dynamic modeling, Frequency domain analysis, System identification, Inverters, Weak grids

Abstract

More and more stability issues have been observed in real-world inverter-based resources (IBRs) power systems. Stability analysis is highly requested in the industrial area. To have an insight into the root causes of stability issues within IBR integrated systems, e.g, PV, wind, and battery storage systems, the dynamic characteristics should be fully understood.

The approaches are different when considering the presupposition of investigated IBR systems. A reduced-order analytical model of the IBR integrated system can be built if the IBR is treated as a white box. Since the parameters and control structures of converter control can be accessed. This analytical model can be used to reveal stability issues and propose improvement accordingly. However, completely replicating the real-world PV farms is not feasible. Some details should be ignored in a simplified model to conduct mathematical analysis. Meanwhile, the PV model must be accurate enough to capture major dynamic characteristics from the original PV system. Utilizing small-signal analysis of the linearized mathematical model, eigenvalue analysis can reveal the oscillation modes and identify the most impact states via participation factor analysis. Additionally, the stability impact of IBR controller parameters can be examined by open-loop analysis.

On the other hand, the internal details of original equipment manufacturers (OEMs) will not be provided for most scenarios due to confidential reasons. Under this condition, the IBRs are like black boxes. The dynamic characteristics can only be identified via measurement data. Consequentially, a $s$-domain admittance model can be found to lead the eigenvalue analysis. With measured admittance model, two things can be achieved. The first one is that extended mode shape analysis can be carried out to explain the dynamic phenomenons on various measurements, i.e., voltage magnitudes, current magnitudes, frequencies, real and reactive powers at different buses. In this way, the interactions among IBRs in a power system can be figured out. Another one is carrying out dynamic simulation using the admittance model only. Numerical Laplace transform (NLT) is adopted to convert the frequency-domain data to fast time-domain simulation.

Download

Share

COinS