Geometric Decentralized Stability Condition for Power Systems Based on Projecting DW Shells
Linbin Huang, Liangxiao Luo, Huanhai Xin, Dan Wang, Ping Ju, and Florian Dörfler · 2025
Read Geometric Decentralized Stability Condition for Power Systems Based on Projecting DW Shells by Linbin Huang, Liangxiao Luo, Huanhai Xin, Dan Wang, Ping Ju, and Florian Dörfler (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-17033.txt.
Introduction
Linbin Huang and co-authors wrote a paper in 2025 on a geometric, decentralized stability condition for power systems. Modern grids mix many kinds of equipment, from wind and solar units to conventional plants. Each unit must be assessed on its own, yet the stability of the whole system must hold.
Earlier decentralized conditions rest on a phase theorem that can fail in some frequency ranges, which makes them overly conservative. The new condition draws on the Davis-Wielandt shell, a geometric shell-shaped representation of how each unit behaves, and on its projections. Those projections give a clear visualization of how units interact in the closed loop. The geometric view also recovers the classical small-gain and small-phase theorems as special cases. The authors argue that the condition scales well and can be applied unit by unit, suiting the grids of the future.
Catalog record
| CREATOR | Linbin Huang, Liangxiao Luo, Huanhai Xin, Dan Wang, Ping Ju, and Florian Dörfler |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 15 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-17033
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Geometric Decentralized Stability Condition for Power Systems Based on Projecting DW Shells Linbin Huang, Liangxiao Luo, Huanhai Xin, Dan Wang, Ping Ju, and Florian Dörfler Abstract The development of decentralized stability conditions has gained considerable attention due to the need to analyze heterogeneous multi-converter power systems. A recent advance is the application of the small-phase theorem, which extends the passivity theory. However, it requires the transfer function matrix to be sectorial, which may not hold in some frequency range and will result in conservatism. This letter tackles this problem by leveraging the Davis-Wielandt (DW) shells for decentralized stability analysis. We develop a geometric decentralized stability condition that visually displays how heterogeneous converters interact with the power grid and enable modular system analysis. I Introduction Modern power systems feature the large-scale integration of heterogeneous power electronics converters. Such heterogeneity comes from the various types of devices (wind, solar, HVDC, energy storage systems, etc.), as well as the various deployed control schemes such as grid-following (GFL) control and grid-forming (GFM) control. The stability analysis of heterogeneous multi-converter systems has always been an intricate task, as it involves high-dimensional dynamics and complex closed-loop interactions. Recent works on this topic include the application of the small-phase theorem, combined with the small-gain theorem, to perform decentralized (small-signal) stability analysis of power electronics-dominated power systems [ 1 ] . The small-phase theorem builds upon the concept of matrix phases, which is defined using the numerical range [ 2 ] . This approach has shown its capability in analyzing highly heterogeneous multi-converter systems in a decentralized manner, that is, if each converter satisfies certain conditions, then the interconnected multi-converter system is guaranteed to be stable. The decentralized stability analysis will be an important cornerstone for future power systems, due to the need to understand how heterogeneous converters interact with the power network and how such interaction results in stability. However, the calculation of matrix phases requires that the matrix is sectorial, which may be conservative under certain scenarios. To tackle this problem, this letter proposes geometric decentralized stability conditions based on the concept of Davis-Wielandt (DW) shell [ 3 , 4 ] , which further extend the conditions in [ 1 ] . We will demonstrate how the DW shell of a matrix is related to its numerical range and further its phases. Moreover, we will show how the small-signal stability of a heterogeneous multi-converter system can be analyzed using DW shells in a decentralized manner. We further prove that with an inductive network, the DW shells can be projected into the x - z plane without resulting in conservatism, allowing better visualization of the geometric stability analysis. II Preliminaries The numerical range of a complex matrix A\in\mathbb{C}^{n\times n} is W(A)=\{x^{*}Ax:x\in\mathbb{C}^{n},\|x\|=1\}, (1) which is a convex subset of \mathbb{C} and contains the eigenvalues of A [ 2 ] . Here, x^{*} denotes the conjugate transpose of x , and \|x\| denotes the two-norm of x . If 0\notin W(A) , then A is said to be sectorial . The smallest phase \underline{\phi}(A) and the largest phase \overline{\phi}(A) of a sectorial matrix A are defined by the two supporting rays of W(A) , and the other phases of A lie in between. The DW shell of a complex matrix A\in\mathbb{C}^{n\times n} is DW(A)=\{(x^{*}Ax,\|Ax\|^{2}):x\in\mathbb{C}^{n},\|x\|=1\}, (2) which extends the numerical range from the 2D complex plane to the 3D space, where the extra z -axis reflects the matrix gain. We refer to [ 3 ] and [ 4 ] for detailed properties of DW shells. Fig. 1 shows the DW shell of an example matrix A , where its projection onto the x - y plane is the numerical range of A . Fig. 1: Illustration of the DW shell DW(A) , numerical range W(A) , and the phases \underline{\phi}(A) , \overline{\phi}(A) of a sectorial matrix A . The (2D) numerical range W(A) is the projection of the (3D) DW shell DW(A) onto the x - y plane. III Stability Conditions Based on DW Shells Fig. 2: Illustration of a multi-converter system and its closed-loop dynamics. Consider a multi-converter system as illustrated in Fig. 2 , where N heterogeneous converters are interconnected via the power network. As detailed in [ 1 ] , such a system can be modeled as a closed-loop feedback system, and its characteristic equation is {\rm det}(I+\widetilde{\bf Y}_{\rm grid}^{-1}(s)\widetilde{\bf Y}_{{\rm C}}^{N}(s))=0 , where, following the notations in [ 1 ] , {\rm det}(A) denotes the determinant of matrix A , \widetilde{\bf Y}_{\rm grid}(s) is the admittance transfer function matrix of the power network, and \widetilde{\bf Y}_{{\rm C}}^{N}(s) is a block-diagonal transfer function matrix capturing the dynamics of the N converters. Note that \widetilde{\bf Y}_{{\rm C}}^{N}(s) contains N blocks (each one is 2\times 2 ), with the i -th one denoted by \widetilde{\bf Y}_{{\rm C},i}(s) , which captures the dynamics of the i -th converter. Both \widetilde{\bf Y}_{{\rm C},i}(s) and \widetilde{\bf Y}_{\rm grid}^{-1}(s) are stable in open loop, and we refer to [ 1 ] for their detailed derivations. According to the generalized Nyquist criterion (GNC), the closed-loop system in Fig. 2 is stable if and only if the Nyquist plot of {\rm det}(I+\widetilde{\bf Y}_{\rm grid}^{-1}(s)\widetilde{\bf Y}_{{\rm C}}^{N}(s)) does not encircle the origin. A sufficient condition to satisfy the GNC is for each \omega\in[0,\infty) , {\rm det}(I+\tau\widetilde{\bf Y}_{\rm grid}^{-1}(j\omega)\widetilde{\bf Y}_{{\rm C}}^{N}(j\omega))\neq 0,\forall\tau\in(0,1]. (3) In what follows, we will demonstrate how the above condition can be guaranteed through the separation of DW shells. Proposition III.1 (Separation of DW Shells) . The multi-converter system in Fig. 2 is stable if for each \omega\in[0,\infty) , DW(\widetilde{\bf Y}_{{\rm C}}^{N}(j\omega))\cap DW(-\tfrac{1}{\tau}\widetilde{\bf Y}_{\rm grid}(j\omega))=\emptyset,\forall\tau\in(0,1]\,, (4) that is, the DW shell of the converters’ collected dynamics \widetilde{\bf Y}_{{\rm C}}^{N}(j\omega) is separated from the DW shell of the power network dynamics \widetilde{\bf Y}_{\rm grid}(j\omega) scaled by -\tfrac{1}{\tau} . Proof. By Theorem 2.1 in [ 5 ] , DW(A)\cap DW(-B)=\emptyset indicates {\rm det}(A+B)\neq 0 . We further have {\rm det}(I+B^{-1}A)\neq 0 since {\rm det}(A+B)={\rm det}(B)\times{\rm det}(I+B^{-1}A) . Hence, we can derive ( 3 ) from ( 4 ) by considering A=\widetilde{\bf Y}_{{\rm C}}^{N}(j\omega) and B=\tfrac{1}{\tau}\widetilde{\bf Y}_{\rm grid}(j\omega) and conclude that the system is stable. ∎ The above proposition shows that the system stability can be guaranteed by the separation of the converters’ DW shell and the power network’s DW shell. Before going to the decentralized version of this stability condition, we first give an illustrative example to show how a converter’s DW shell interacts with a power network’s DW shell. Consider a GFL converter that is connected to an infinite bus, shown in Fig. 3 . The converter’s parameters are the same as those in [ 1 ] , and the short-circuit ratio (SCR) of this system is SCR=1/X_{\rm net}=3.5 where X_{\rm net} is the line reactance. The converter’s admittance matrix \widetilde{\bf Y}_{{\rm C},1}(j\omega) has been derived in [ 1 ] and is thus omitted here. Fig. 4 plots the converter’s DW shells according to \widetilde{\bf Y}_{{\rm C},1}(j\omega) at different frequencies, where each DW shell is an ellipsoid. For the network part, since it is a single line in Fig. 3 , we have \widetilde{\bf Y}_{\rm grid}(j\omega)=X_{\rm net}^{-1}I_{2} , where I_{2} denotes the 2\times 2 identity matrix. Since in this case \widetilde{\bf Y}_{\rm grid}(j\omega) is positive