⚡ electrical engineering

Autonomous and Distributed Synchronization and Restoration of an Islanded Network of Microgrids

This paper proposes and validates an autonomous, distributed synchronization and restoration scheme for islanded networks of microgrids using Distributed-Averaging Proportional-Integral (DAPI) control, which effectively addresses synchronization instability and regulation challenges through high-fidelity simulations on a modified IEEE 123-bus system.

Original authors: Ahmed Saad Al-Karsani, Maryam Khanbaghi

Published 2026-02-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ahmed Saad Al-Karsani, Maryam Khanbaghi

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 the power grid not as a single, massive highway, but as a neighborhood of small, independent communities called Microgrids. These communities can run on their own (like a house with a generator) or connect to the big city grid.

Now, imagine a massive blackout hits the city. The big grid goes dark. These Microgrids have to wake up, start their own generators, and then—crucially—learn how to hold hands and reconnect with each other to form a larger, stronger "Network of Microgrids" (NMG) without causing a chaotic crash.

This paper is about teaching these digital power generators how to do that dance autonomously (without a human conductor) and safely.

Here is the breakdown of their solution, using some everyday analogies:

1. The Problem: The "Orphaned" Generators

When the main grid dies, these Microgrids become islands. They have to start up from scratch (a "black start").

The Challenge: Think of these generators as dancers. If they start dancing alone, they might be slightly out of step with each other. If they try to join hands (connect their circuits) while one is spinning fast and the other is slow, or if they are facing different directions, they will trip, fall, and break the equipment.
The Old Way: Usually, you need a central "boss" (a supervisory control) to tell everyone when to start and when to connect. But if the boss's phone line is cut, the whole system fails.
The Risk: If they connect too early or too roughly, it causes a "power surge" (like a sudden shock) that can destabilize the whole network.

2. The Solution: The "DAPI" Teamwork Algorithm

The authors propose a new way for these generators to talk to each other using a method called DAPI (Distributed-Averaging Proportional-Integral).

The Analogy: A Group of Hikers
Imagine a group of hikers lost in a foggy forest (the islanded network). They need to meet up at a specific spot and start walking together in perfect sync.

No Leader: There is no single leader giving orders.
Whispering Neighbors: Each hiker only talks to the people immediately next to them.
The Goal: They want to agree on the same speed (Frequency) and the same direction (Phase).
How it works: If Hiker A is walking too fast, they listen to Hiker B and Hiker C. They average their speeds and gently slow down. If Hiker A is facing North but Hiker B is facing East, they gently turn until they are all facing the same way.

This "whispering" happens electronically between the inverters (the devices that turn battery power into grid power). It ensures that before they ever touch wires, they are already moving in perfect harmony.

3. The "Soft Start" and The "Green Light"

The paper describes a three-step safety protocol:

The Soft Start (Waking Up):
When the power is off, the generators don't just blast on. They slowly ramp up their voltage, like turning up the volume on a radio very gently so it doesn't blast your ears. They stabilize their own little island first.
The "Sync Check" (The Traffic Light):
Before two Microgrids connect, they run a digital check.
- Is our speed the same? (Frequency)
- Is our voltage the same? (Voltage)
- Are we facing the same direction? (Phase)
The authors set these rules to be extremely strict. While standard rules might allow a 10% difference, their system demands less than 1%. It's like a race car driver waiting until they are exactly parallel with the car next to them before merging lanes, rather than just "close enough."
The Distributed "All Clear" (The Handshake):
This is the clever part. A breaker (the switch that connects the grids) won't close unless both sides agree.
- Microgrid A checks its own status.
- Microgrid A asks Microgrid B, "Are you ready?"
- If Microgrid B says "Yes," and Microgrid A says "Yes," then the switch clicks.
- It's a double-check system. If one side is confused or the communication line is broken, the switch stays open. No one gets hurt.

4. The Simulation: A Digital Crash Test

To prove this works, the authors built a massive digital model of a real-world power system (based on the famous IEEE 123-bus system, which is like a complex map of a city's power lines).

They simulated 7 different Microgrids waking up from a blackout.
They watched them talk to each other, adjust their speeds, and finally click together.
The Result: It worked perfectly. The system stabilized, shared the power load evenly (like a group of friends sharing a heavy box), and kept the lights on without any crashes.

Why This Matters

In the future, as we rely more on solar panels and batteries (which don't have the heavy spinning parts of old power plants), the grid becomes more fragile.

Resilience: If a storm knocks out the main grid, these Microgrids can wake up, find each other, and form a new, temporary power grid automatically.
No Single Point of Failure: Because they don't rely on one central boss, if one computer fails, the others can still talk and keep the power on.

In a nutshell: This paper teaches digital power generators how to be a well-coordinated team, whispering to their neighbors to get in perfect sync before they ever shake hands, ensuring that when the lights go out, they can turn themselves back on without causing a disaster.

1. Problem Statement

The transition to renewable energy has led to the proliferation of Inverter-Based Resources (IBRs) and the formation of Networks of Microgrids (NMGs). While NMGs offer enhanced resilience, their operation in islanded mode presents significant challenges:

Lack of Inertia: Unlike traditional Synchronous Machines (SMs), IBRs lack rotational inertia, making them highly sensitive to disturbances and prone to instability.
Synchronization Instability: Restoring power (black start) and reconnecting islanded Microgrids (MGs) requires precise synchronization. Small deviations in voltage, frequency, or phase at the Point of Common Coupling (PCC) can cause inrush currents and small-signal instability.
Inadequate Regulation: Existing secondary-level control schemes often fail to address the trade-offs between voltage regulation and reactive power sharing in weak, islanded grids.
Modeling Gaps: There is a lack of high-fidelity, large-scale simulation models for islanded NMGs that incorporate realistic unbalanced distribution systems and detailed IBR dynamics.

2. Methodology

The authors propose a comprehensive autonomous and distributed synchronization and restoration scheme utilizing Distributed-Averaging Proportional-Integral (DAPI) control. The methodology is structured as follows:

A. System Modeling

Test System: A high-fidelity model of a modified, unbalanced IEEE 123-bus test feeder was created in MATLAB®/Simulink®.
Topology: The system is isolated from the main grid and consists of 7 Microgrids (MGs), each equipped with a single Grid-Forming (GFM) IBR interfaced with a Battery Energy Storage System (BESS).
Control Hierarchy:
- Primary Level: Droop control ( $P-f$ and $Q-V$ ) to establish a linear relationship between power and frequency/voltage.
- Secondary Level: DAPI control is employed to eliminate steady-state errors in frequency and voltage while ensuring power sharing.
- Phase Consensus: A novel phase synchronization dynamic is introduced (Eq. 8) where IBRs exchange phase angle information ( $\theta$ ) with neighbors to align phases before connection.

B. Control Strategy

Black Start & Soft-Start: Each MG performs an autonomous soft-start, ramping voltage from 0 to setpoint ( $480\sqrt{2}/\sqrt{3}$ V) over 0.5 seconds.
Decentralized Regulation: Initially, communication links are disabled ( $a_{ij}=0$ ), allowing each MG to regulate frequency and voltage independently.
Distributed Synchronization Logic:
- Local Check: Each IBR monitors local deviations in frequency ( $\Delta \omega$ ), voltage ( $\Delta V$ ), and phase ( $\Delta \theta$ ).
- Neighbor Check: IBRs exchange synchronization status with physically connected neighbors.
- Breaker Logic: A two-stage AND gate logic (Fig. 2-4) ensures that a breaker only closes if both local and neighboring synchronization checks pass. This acts as a redundancy to prevent simultaneous breaker closures across the NMG unless all conditions are met.
Post-Connection Regulation: Once connected, the DAPI control activates fully to share active power (based on rated capacity) and regulate frequency/voltage across the entire NMG.

3. Key Contributions

Leaderless Phase Synchronization: Implementation of a distributed, leaderless phase consensus scheme that allows IBRs to synchronize autonomously without a central supervisory controller.
Integrated Restoration Scheme: A unified approach that handles soft-start, synchronization, connection, and regulation simultaneously, addressing the specific challenges of weak islanded grids.
High-Fidelity Validation: The proposal is validated on a large-scale, unbalanced NMG model (7 MGs, IEEE 123-bus) using detailed electromagnetic transient (EMT) simulations, moving beyond simplified static power flow studies.
Strict Synchronization Criteria: The authors propose tighter synchronization tolerances than IEEE 1547-2018 standards (e.g., $|\Delta f| < 0.01$ Hz vs. 0.1 Hz) to ensure stability in weak islanded networks.

4. Simulation Results

The simulation was conducted over a 10-second window in MATLAB®/Simulink®:

Black Start (0–3s): Each MG successfully performed a soft-start. Decentralized secondary control regulated frequency and voltage to setpoints ( $f^*=60$ Hz, $V^*=$ setpoint).
Synchronization (t=3s): Local and neighbor checks passed. The distributed logic triggered breaker closures.
Steady State (t>4s):
- Frequency: All MGs synchronized to 60 Hz.
- Power Sharing: Active power was shared equally among IBRs based on their ratings.
- Voltage/Reactive Power: With a voltage deviation gain ( $\xi$ ) of 0.05, the system achieved a trade-off where voltage remained within $\pm 5\%$ of the setpoint, while reactive power was shared. Without DAPI, reactive power circulation was observed.
Stability Analysis: The system remained stable with a 11.1 ms control time step. However, increasing the time step to 22.2 ms caused instability post-closure, highlighting the need for robust control or faster communication (e.g., IEC 61850 GOOSE).

5. Significance

Resilience: The proposed scheme enables NMGs to recover autonomously from blackouts without relying on external grid support or centralized control, which is critical for disaster recovery and remote areas.
Scalability: The distributed nature of the DAPI control and consensus logic ensures the system scales well without single points of failure.
Practicality: By using a high-fidelity unbalanced model and addressing communication delays, the study bridges the gap between theoretical control design and real-world distribution system implementation.
Future Directions: The work identifies the need for robust control strategies to handle longer time steps and communication failures, as well as improved handling of unbalanced loads during the connection process.

In conclusion, this paper demonstrates that DAPI-based distributed control combined with autonomous synchronization logic is a viable and effective solution for the reliable restoration and operation of islanded Networks of Microgrids.