Grid-Forming and Spatially Distributed Control Design of Dynamic Virtual Power Plants

TL;DR

The paper develops a novel grid-forming and spatially distributed DVPP framework that aggregates heterogeneous grid-forming DERs to deliver fast frequency and voltage support. It introduces an adaptive divide-and-conquer (ADPF) strategy to disaggregate a desired aggregate transfer function $T_\mathrm{des}(s)$ into local targets and employs LPV $\mathcal{H}_\infty$ matching control to realize them at the device level, while respecting DER limitations. It further extends the approach to hybrid DVPPs and to spatially distributed DERs by incorporating rotated powers $p', q'$ and coupled $p'-f$ and $p'-q'$ dynamics, ensuring accurate coordination across non-negligible line impedances. Numerical case studies on the IEEE 9-bus system with an interconnected MV grid validate the method, demonstrating coherent frequency responses, effective ADPF online adaptation, and robust performance under disturbances and outages. The work provides a scalable, controller-architecture framework for next-generation DVPPs to deliver dynamic ancillary services across temporal and spatial scales, with potential impact on grid reliability and renewable integration.

Abstract

We present a novel grid-forming control design approach for dynamic virtual power plants (DVPP). We consider a group of heterogeneous grid-forming distributed energy resources (DER) which collectively provide desired dynamic ancillary services, such as fast frequency and voltage control. To achieve that, we study the nontrivial aggregation of grid-forming DERs to establish the DVPP, and employ an adaptive divide-and-conquer strategy that disaggregates the desired control specifications of the aggregate DVPP via adaptive dynamic participation factors to obtain local desired behaviors of each DER. We then design local controllers at the DER level to realize these local desired behaviors. In the process, physical and engineered limits of each DER are taken into account. We extend the proposed approach to make it also compatible with grid-following DER controls, thereby establishing the concept of so-called hybrid DVPPs. Furthermore, we generalize the DVPP design to spatially dispersed DER locations in power grids with different voltage levels and R/X ratios. Finally, the DVPP control performance is verified via numerical case studies in the IEEE nine-bus transmission grid with an interconnected medium voltage distribution grid.

Figures (19)

• Figure 1: DVPP configurations with heterogeneous DERs connected at one single bus in the transmission grid. DVPPs with spatially dispersed DER locations are considered later in our novel results in \ref{['sec:spatially_distributed']}.
• Figure 2: Comparison of grid-following and grid-forming DVPP controls.
• Figure 3: Exemplary Kron-reduction of a parallel DER interconnection.
• Figure 4: Schematic of the proposed grid-forming DVPP control design from grid-level to device-level. The system operator provides the aggregate DVPP specification to encode grid code requirements in the form of a desired transfer function, while the DVPP operator takes care of all subsequent design steps.
• Figure 5: Grid-forming frequency control architecture for hybrid DVPPs.

Theorems & Definitions (5)

• Remark 1
• Remark 2
• Remark 3
• Remark 4
• Remark 5