The rapid proliferation of inverter-based resources—including solar, wind, and battery storage systems—has fundamentally altered power system dynamics. Unlike synchronous generators, IBRs rely on electronic controllers that must work seamlessly across vendors and topologies. However, these controllers are typically designed in isolation, creating coordination gaps that threaten grid stability.
This research addresses a critical gap by introducing a decentralized control synthesis framework grounded in block-diagonal dominance theory. Rather than requiring a centralized coordinator, the approach enables each IBR to design its own controller while guaranteeing overall system stability. The method leverages grid frequency response signals—natural information available at each resource location—to inform local control decisions.
The framework advances beyond simple decentralized schemes by incorporating rigorous small-signal stability guarantees through MIMO control design constrained by block-diagonal dominance conditions. A key innovation is the introduction of a minimum decay rate requirement, ensuring that system disturbances dissipate within acceptable timeframes rather than merely remaining bounded. The authors also define a novel metric to quantify conservatism in the stability certificate, helping engineers understand the gap between theoretical guarantees and practical system margins.
Validation on the IEEE 9-bus test system demonstrates that the approach successfully stabilizes multiple IBRs operating under realistic conditions. This standardized test case provides a benchmark for comparing this method against other decentralized approaches and validates the theoretical underpinnings.
The implications are substantial for grid operators managing high renewable penetration. Decentralized synthesis eliminates vendor coordination bottlenecks while maintaining provable stability margins. As IBR deployment accelerates, such vendor-agnostic control frameworks become essential infrastructure. Future work should extend the method to larger, more complex networks and validate against hardware-in-the-loop testing to confirm real-world performance.
