Emerging grid-forming standards, such as the VDE-FNN Grid-Forming Guideline 2025, impose stringent dynamic performance requirements on grid-connected converters. These standards mandate fast and consistent responses across wide variations in grid strength (SCR ≈ 1–20) and voltage sags (up to 80% depth), including, for example, reactive-current settling times below 80 ms and apparent-current rise times below 15 ms during low-voltage ride-through (LVRT) events. In other words, compliance with these grid-code requirements highlights the need for GFM control strategies whose dynamic performance is effectively decoupled from grid-side parametric variations.

Unfortunately, conventional grid-forming (GFM) control with fixed-gain PI regulators have closed-loop dynamics that are strongly coupled to grid strength and PCC voltage. As short-circuit ratio (SCR) or voltage decreases, the effective control bandwidth is reduced, leading to slower transients and inconsistent performance. To remain stable under weak-grid conditions, controllers are typically tuned for worst-case scenarios, which degrade dynamic performance under nominal conditions making it difficult to comply with grid code requirements at all operating points. Gain-scheduling can alleviate this, but it relies on fast and accurate online grid-impedance estimation, which is impractical in real-world applications.

To overcome these limitations, this paper proposes a model reference adaptive control (MRAC) framework that treats grid impedance and voltage variations as plant uncertainties. The adaptive controller is embedded in a GFM architecture with inner current and voltage loops. The controller parameters are continuously updated to accurately track predefined reference models with specified dynamic characteristics regardless of grid-side conditions, while closed-loop stability is established via Lyapunov analysis. At fast adaptation rates, classical MRAC implementations are sensitive to non-idealities such as delays and saturation, which can excite high-frequency oscillations and degrade power quality. To improve robustness, the MRAC scheme is augmented with an L1-adaptive architecture that decouples fast adaptation from robustness through a filtered control channel. This constitutes a novel application of L1-adaptive control to grid-forming converters explicitly targeting grid-code-driven dynamic performance.

Electromagnetic transient (EMT) simulations demonstrate uniform performance of the proposed control strategy across a wide range of SCR (≈ 1–20) and voltage faults (up to 80% voltage dips) meeting grid code requirements. The proposed L1-adaptive controller is also benchmarked against a conventional fixed-gain PI-based GFM controller and consistently exhibits superior transient performance.

Remark 1: Experimental validation of the simulation results in laboratory test bench is underway over the spring/summer period and expected to be included in the final paper.

Remark 2: The final abstract of the paper will be naturally more concise. At this stage, the abstract is comprehensive in order to accurately portray the scope of the contribution for evaluation by the reviewing committee.