💻 computer science

On Theoretical Stability Proof and Stability Margin Analysis of Enhanced Droop-Free Control Schemes for Islanded Microgrids

This paper provides the first comprehensive theoretical proof of asymptotic stability for two normalized active power consensus (NAPC)-based droop-free control schemes in islanded microgrids and analyzes how system parameters like DER capacity and network topology influence stability margins to guide optimal control selection.

Original authors: Weipeng Liu, Upendra Prasad, Yutian Liu, Yong Dong, Haoran Zhao, Lei Wu

Published 2026-03-05

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Weipeng Liu, Upendra Prasad, Yutian Liu, Yong Dong, Haoran Zhao, Lei Wu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a small, isolated island power grid (a microgrid) that doesn't connect to the main national power line. This island relies on a mix of energy sources: some are reliable batteries or generators (called DERs or Distributed Energy Resources), and others are unpredictable, like solar panels that stop working when clouds pass or wind turbines that stop when the wind dies.

The big challenge for this island is: How do we make sure everyone shares the workload fairly and keeps the lights on without the whole system crashing?

This paper tackles that problem by introducing a new, smarter way to manage these energy sources. Here is the breakdown in simple terms:

1. The Problem: The "Old Way" vs. The "New Way"

The Old Way (Droop Control): Think of this like a group of rowers in a boat. If one rower gets tired, the others have to work harder, but the boat slows down (frequency drops). The old method tries to fix this, but it's sensitive to bad measurements and requires a "captain" (a central computer) to tell everyone what to do. If the captain's radio breaks, the whole team panics.
The New Way (Droop-Free Control): This is like a team of rowers who can talk to their immediate neighbors. They don't need a captain. They just whisper to the person next to them, "Hey, I'm doing 80% of my max effort; you doing 80%?" If everyone is at 80%, they are sharing the load fairly. This paper focuses on a specific version of this called NAPC (Normalized Active Power Consensus).

2. The Big Question: Is it Safe?

For years, engineers have used this "neighbor-whispering" method (NAPC) and it seemed to work great. But, they were essentially guessing that it would always be stable. They didn't have a mathematical "proof" that it wouldn't suddenly go crazy and shut down the island during a storm.

This paper is the first to provide that "proof."
The authors did the heavy math (using something called "eigenvalues," which are like the heartbeat of the system) to prove that:

The system will always settle down after a disturbance (like a cloud covering a solar panel).
It won't oscillate or crash.
They proved this for two specific versions:
- O-NAPC (Ordinary): The standard version.
- A-NAPC (Amplifier-equipped): A version with a "volume booster" (amplifier) to help smaller energy sources speak up louder.

3. The "Goldilocks" Discovery: Which Team Member is Best?

The researchers found that which control method works best depends on the "size" of the energy sources on the island.

Scenario A: Big, Strong Energy Sources.
If your island has huge batteries and generators (like a giant truck), the Ordinary (O-NAPC) method is best. It's like a team of strong athletes; they don't need a volume booster to coordinate.
Scenario B: Small, Weak Energy Sources.
If your island relies on many small batteries or small solar panels (like a fleet of bicycles), the Amplifier (A-NAPC) method is better. The "volume booster" helps these small sources communicate effectively so they don't get drowned out by the noise.

4. Finding the Weak Links (Vulnerability Analysis)

The paper also acts like a "stress test" for the island. They asked: "What happens if we break a wire? What if someone hacks the communication network? What if a battery dies?"

They found three main things that determine how stable the island is:

The Wires (Electrical Network): If the wires are long and thin (high resistance), the system is weaker. Short, thick wires make it stronger.
The Talk (Communication Network): If the energy sources can only talk to one neighbor, it's slow. If they can talk to everyone (a dense network), the system is very stable.
The Capacity (The Batteries): This is the most important factor. If the batteries are small and weak, the system is very fragile. If they are large, the system is robust.

The "Achilles' Heel":
The study found that the capacity of the batteries is the most critical weak point. If a battery fails or its capacity drops, it hurts the system more than a broken wire or a bad communication link. They also identified specific "hub" nodes in the communication network; if a hacker attacks those specific hubs, the whole system is in danger.

Summary Analogy

Imagine a choir singing in a room without a conductor.

The Goal: Everyone must sing at the exact same volume relative to their lung capacity.
The Old Method: Everyone tries to guess the volume based on how loud they feel. It's messy and often off-key.
The New Method (NAPC): Everyone whispers to their neighbor, "I'm singing at 80% of my max."
The Paper's Contribution:
1. Proof: It mathematically proves that this whispering method will never cause the choir to scream or go silent unexpectedly.
2. Strategy: It tells you that if you have a choir of giants, use the standard whisper. If you have a choir of children, add a microphone (amplifier) so they can be heard.
3. Safety Check: It identifies that if the children get sick (lose capacity), the whole choir collapses faster than if someone in the back row stops whispering.

In short: This paper gives engineers the confidence to build smarter, self-sustaining power grids that can survive storms and blackouts, provided they choose the right control strategy for the size of their energy sources.

1. Problem Statement

Islanded microgrids rely on Distributed Energy Resources (DERs) for power balance and frequency stability. Traditional hierarchical control structures (primary/secondary/tertiary) often use droop control, which suffers from inherent limitations:

Sensitivity to measurement errors requiring high-quality sensors.
Reliance on a Microgrid Central Controller (MGCC), creating a single point of failure and high communication costs.
Complexity in parameter tuning for distributed secondary control.

Droop-free control has emerged as an alternative, utilizing sparse neighboring communication to achieve zero frequency deviation and active power sharing without an MGCC. While Active Power Consensus (APC) (equal power sharing) is well-understood, Normalized Active Power Consensus (NAPC) (sharing proportional to available capacity) is more practical for heterogeneous DERs. However, existing literature lacks a comprehensive theoretical proof of asymptotic stability for NAPC-based schemes and offers limited analysis on their stability margins under varying system conditions.

2. Methodology

The paper proposes and analyzes two enhanced droop-free control schemes:

Ordinary NAPC (O-NAPC): Standard normalized power sharing.
Amplifier-equipped NAPC (A-NAPC): An enhanced version adding an amplifier in the control loop to improve performance for specific system characteristics.

Key Methodological Steps:

State-Space Modeling: The authors derive state-space models for both O-NAPC and A-NAPC, incorporating the electrical network (susceptance matrix $B$ ), communication network (Laplacian matrix $L_W$ ), and DER capacities (diagonal matrix $D_p$ ).
Theoretical Stability Proof: The core contribution is a rigorous mathematical proof of asymptotic stability. The authors utilize:
- Eigenvalue Analysis: Proving that all effective eigenvalues of the system matrices have negative real parts.
- Matrix Theory: Using continuity arguments, Perron-Frobenius theorems, and properties of Positive Semi-Definite (PSD) and M-matrices to demonstrate that the system matrices possess one simple zero eigenvalue (due to power balance constraints) and $(N-1)$ eigenvalues with strictly negative real parts.
Stability Margin & Sensitivity Analysis: The study quantifies stability margins based on the dominant pole (the eigenvalue closest to the imaginary axis). It analyzes the sensitivity of these poles to three key metrics:
- $B_{avg}$ : Average electrical admittance (inverse of impedance).
- $L_{sparse}$ : Sparseness of the communication network.
- $P_{avg}$ : Average available capacity of controllable DERs.
Vulnerability Analysis: The paper calculates eigenvalue sensitivity to identify specific components (lines, communication links, or DER capacities) that, if perturbed, most severely destabilize the system.
Hardware-in-the-Loop (HIL) Validation: A 4-node and 8-node microgrid testbed (using OP5707 simulator and physical hardware) is used to validate theoretical findings under load and generation disturbances.

3. Key Contributions

First Theoretical Proof of NAPC Stability: The paper provides the first comprehensive proof that both O-NAPC and A-NAPC schemes are asymptotically stable for symmetric communication networks, resolving a gap in existing literature where stability was often assumed.
Introduction of A-NAPC: Proposes an amplifier-equipped scheme specifically designed to enhance stability for microgrids with smaller DER capacities.
Stability Margin Guidelines: Establishes clear guidelines for selecting control schemes based on system parameters:
- O-NAPC is more suitable for microgrids with larger available DER capacities.
- A-NAPC is more suitable for microgrids with smaller available DER capacities, offering faster convergence in these scenarios.
Vulnerability Identification: Identifies that the available capacity of controllable DERs is the most decisive factor influencing stability margins. The analysis pinpoints specific "weak links" (e.g., specific communication hubs or electrical lines) that require enhanced monitoring.

4. Results

Steady-State Performance: Both O-NAPC and A-NAPC successfully achieve normalized active power sharing (load shared proportional to capacity) and maintain nominal frequency, unlike APC which shares load equally regardless of capacity.
Convergence Speed:
- In Case 1 (Small DER capacities): A-NAPC converged fastest ( $10.1s$ ), followed by O-NAPC ( $10.9s$ ), then APC ( $12.2s$ ).
- In Case 2 (Large DER capacities): APC converged fastest ( $7.9s$ ), followed by O-NAPC ( $10.3s$ ), while A-NAPC was slowest ( $14.2s$ ).
Sensitivity Findings:
- Electrical Network: Lower impedance (higher admittance) improves stability.
- Communication: Denser communication networks improve stability margins.
- DER Capacity: The stability margin decreases as the average available capacity increases for NAPC schemes because the normalized power signal ( $p_p(t)$ ) becomes weaker.
Vulnerability Analysis: The analysis revealed that perturbations in DER capacities ( $D_p$ ) have a higher sensitivity impact on stability than perturbations in electrical lines ( $B$ ) or communication weights ( $L_W$ ). Specific nodes were identified as critical hubs for communication and generation stability.

5. Significance

This paper is significant for the practical deployment of islanded microgrids because:

Reliability: It moves NAPC-based control from a "heuristic" approach to a theoretically guaranteed stable solution, essential for critical infrastructure.
Design Optimization: It provides engineers with a decision-making framework to choose between O-NAPC and A-NAPC based on the specific capacity profile of their DER fleet.
Resilience: The vulnerability analysis offers a proactive tool for operators to identify and reinforce weak points in the microgrid (e.g., securing specific communication nodes or maintaining specific DERs) against cyber-physical attacks or component aging.
Scalability: The proof holds for general symmetric networks, and the Appendix extends the validity to scenarios with non-zero R/X ratios, filter dynamics, and heterogeneous node types, ensuring the theory applies to real-world complex microgrids.