Multiport Dissipativity Finds SST Instability Root Cause

Solid-state transformers in grid-connected power systems sit at an awkward control boundary: they must regulate AC-side synchronization and DC-side power transfer while the surrounding grid impedance can move the operating point into poorly damped regimes. The paper argues that low-frequency oscillations in single-phase SSTs are not always explainable as a bad local loop or a single negative-resistance port. The contribution is a multiport stability diagnosis that treats the AC $d$ axis, AC $q$ axis, and primary DC port as one coupled admittance object, then uses dissipativity to ask which part of that object is responsible for instability.

Core Contribution

The main result is diagnostic rather than just stabilizing. Prior impedance and passivity methods can flag that a converter-grid system is unsafe, but the authors want to identify whether the failure comes from a self-admittance term at one port or from cross-coupling among ports. Their decomposition of the passivity condition into self-dissipativity and coupling-dissipativity indices gives the design workflow a target: in the studied SST, the dominant problem is a coupling-dissipativity failure created by synchronization-loop dynamics, not a localized self-dissipativity defect.

That distinction matters because it changes the controller one would build. If the instability were local, retuning a voltage or current loop might be the natural response. Here, the paper points instead to the dynamics introduced by second-order generalized integrator filtering in the synchronization path. The proposed mitigation, dynamics-free orthogonal signal reconstruction, is meant to remove that coupling mechanism rather than add another compensator around it.

Technical Approach

Figure 1 lays out the target system: a cascaded H-bridge dual-active-bridge SST power stage and its control block diagram. The paper models this system as a small-signal multiport admittance matrix spanning the AC dq variables and the primary DC port, so the stability object includes both same-port terms and cross-port transfer paths.

The useful technical move is shown by the equivalent circuit in Figure 2. The self-admittances define how each port behaves in isolation, while the cross-coupling admittances separate $d$-$q$, $d$-DC, and $q$-DC interactions. The dissipativity criterion is then decomposed so that a failed condition can be traced to a specific coupling class rather than treated as a scalar pass/fail result.

The design flow in Figure 3 turns that diagnosis into a controller choice. After identifying the unstable channel, the authors replace the problematic synchronization signal path with DE-OSR, a dynamics-free orthogonal signal reconstruction scheme. In the paper's interpretation, this reshapes the SST admittance where the multiport test found the deficit, avoiding the extra parameter tuning that an added stabilization loop would introduce.

Results and Analysis

The reported evidence has two parts: small-signal stability prediction and hardware validation on a down-scaled SST prototype. The criterion predicts coupling-induced oscillations under weak-grid conditions, and the enhanced controller is reported to maintain stable operation in those same challenging cases. The paper's abstract and conclusion frame the result as a root-cause match: the analysis identifies coupling dissipativity as the failing mechanism, and the controller targets that mechanism directly.

The article is strongest as a control-design workflow. It does not merely say that weak-grid operation is unstable; it gives the engineer a way to distinguish a self-port problem from a cross-port problem before changing the controller. That is a practical improvement over a stability margin that cannot explain where to intervene. The trade-off is that the publicly visible summary of the experiments is more qualitative than metric-heavy: the paper validates predicted oscillation and mitigation on hardware, but the provided material does not expose a table of oscillation frequencies, damping ratios, grid-strength thresholds, or efficiency penalties.

Limitations

The authors state one important assumption directly. The proposed stabilization relies on reference tracking and is suited to weak-grid conditions and small-signal disturbances. During severe grid-side faults, where the current can deviate substantially from the reference, the method may not be sufficient; a dedicated fault-ride-through strategy would be required. The evaluation is also scoped to the studied single-phase SST architecture, so transfer to other SST topologies or protection regimes would need separate modeling and tests.