Abstract
Modern power grids are increasingly reliant on inverter-based renewable energy sources, which lack the physical inertia of traditional synchronous generators. This paper proposes a unified mathematical framework and a two-layer hierarchical control strategy to stabilize Active Distribution Networks (ADNs). By modeling both grid-following and grid-forming generators under a single state-space representation, the proposed system enables rapid local recovery from voltage and frequency disturbances while ensuring equitable power sharing across the network.
Introduction
The transition from massive, spinning coal or gas turbines to inverter-based Distributed Generators (DGs) like solar and wind has introduced significant stability challenges. Traditional generators possess physical inertia that naturally resists sudden changes in frequency and voltage. In contrast, inverters do not inherently provide this damping, leading to several issues in ADNs:
- Instability: Weather-driven power surges or drops cause frequency and voltage overshoots.
- Heterogeneity: Different control requirements for grid-following (PQ control) and grid-forming (droop control) units make fleet management difficult.
- Power Circulation: Mismatches in local settings can cause energy-wasting loops between nearby generators.
Methodology
The authors address these challenges through a two-pronged approach involving unified modeling and a hierarchical control architecture.
Unified Modeling
To bridge the gap between different DG types, the authors introduced an inertia compensation loop for grid-following units. This allows both grid-forming and grid-following generators to be represented by the same state-space equations. This unified model simplifies the design of controllers that must operate across a heterogeneous network.
Hierarchical Control Strategy
The control system is divided into two layers:
- Local Layer (Robust Control): Each DG utilizes a local robust controller based on $H_\infty/H_2$ theory. Operating without the need for communication signals, it provides instantaneous suppression of frequency and voltage fluctuations, even under uncertain line parameters.
- System Layer (Collaborative Power Control): A centralized PI controller monitors the network on a slower timescale to ensure proportional load sharing among DGs, preventing power circulation.
Experimental Validation
The strategy was validated using MATLAB/Simulink simulations on modified IEEE 33-bus and 69-bus test feeders. Real-world EPRI PV data was used to simulate realistic power fluctuations.
Key results include:
- Voltage Regulation: During a solar output jump from 100kW to 700kW, the strategy maintained voltage within a tight 0.98–1.02 p.u. range, whereas uncontrolled systems spiked past 1.05 p.u.
- Frequency Stability: Improvements were noted in both the Frequency Nadir and the Rate of Change of Frequency (RoCoF).
- Efficiency: The proposed method outperformed Model Predictive Control (MPC) in suppression speed by avoiding complex real-time optimization.
Conclusion
By treating inverter inertia as a programmable "virtual" property and decoupling fast stability control from slow coordination, this framework provides a resilient solution for modern grids. While highly effective for local and system-level stability, future work will explore tertiary control for global economic dispatch and multi-region coordination.
Summary updated