Stability-Constrained Decentralized Control and Energy Management for Grid-Forming Inverter-Based Microgrids

Technical Summary: Stability-Constrained Decentralized Control and Energy Management for Grid-Forming Inverter-Based Microgrids

Problem Statement
The global transition toward renewable energy has led to microgrids dominated by inverter-based resources, resulting in low system inertia, high power imbalances, and unpredictable operational dynamics. While Grid-Forming Inverters (GFIs) offer a promising solution for voltage and frequency regulation in islanded or weak-grid conditions, existing literature typically treats control design and energy optimization as separate domains. Conventional approaches often rely on grid-following inverters that require external references, or they utilize centralized energy management systems that suffer from scalability issues, communication overhead, and single points of failure. Furthermore, current optimization models frequently simplify inverter dynamics, failing to capture the interaction between control mechanisms and aggregate dispatch decisions. There is a distinct lack of a cohesive framework that integrates decentralized GFI control, optimal energy management, and sustainability analysis within a single system.

Methodology
The paper proposes a unified framework that combines decentralized, consensus-based GFI control with Mixed-Integer Linear Programming (MILP) for energy management in PV-wind-battery microgrids.

• System Modeling: The study models Photovoltaic (PV) systems using a single-diode model and wind turbines using a power coefficient model dependent on wind speed. Energy storage is modeled via State of Charge (SOC) dynamics.
• Decentralized Control: The control layer employs a consensus-based algorithm where inverters communicate with neighbors to achieve proportional power sharing without a central controller. GFIs utilize active and reactive power droop control to regulate frequency and voltage, respectively.
• Stability Analysis: To ensure robustness, the framework incorporates Lyapunov-based stability analysis to prove global asymptotic stability and small-signal stability analysis using Laplacian matrix eigenvalues to guarantee convergence and damping.
• Optimization (MILP): A MILP-based energy management system operates on a slower time scale to co-optimize power dispatch, energy storage scheduling, and inverter operation. The objective function minimizes a weighted sum of power imbalances, operational costs, energy losses, and carbon emissions.
• Coupling: The architecture features a two-layer structure where the decentralized control layer ensures real-time stability, while the optimization layer provides setpoints subject to stability constraints derived from the control dynamics.

Key Contributions

1. Integrated Framework: The study presents a novel architecture that simultaneously addresses decentralized GFI coordination, time-coupled energy storage dynamics, and system-level optimization, bridging the gap between control theory and energy management.
2. Stability-Constrained Optimization: Unlike previous works, this framework directly incorporates stability constraints (Lyapunov and eigenvalue-based) into the energy management process, ensuring that dispatch decisions do not compromise system stability.
3. Techno-Economic-Environmental Evaluation: The paper introduces a unified evaluation framework that assesses performance not only technically (efficiency, stability) but also economically (operational costs) and environmentally (carbon footprint).
4. Scalable Decentralized Operation: The proposed consensus-based approach eliminates the need for a central supervisor, enhancing scalability and resilience against communication failures.

Results
Simulation results over a 24-hour horizon, considering renewable variability, load fluctuations, and fault conditions, demonstrate significant performance improvements over conventional grid-following and centralized control mechanisms:

• Efficiency: System efficiency increased from 88% (conventional) to 94%. Inverter efficiency improved from 85% to 95% through advanced topology and control strategies.
• Loss Reduction: Energy losses were reduced by approximately 50% (from 120,000 kWh to 60,000 kWh annually).
• Carbon Emissions: Carbon emissions were reduced by up to 92%, with a calculated reduction of 460,000 kg CO₂ compared to a traditional grid scenario.
• Stability and Dynamics: The system demonstrated rapid recovery from fault conditions (voltage and frequency restoration) and achieved a settling time of 8 seconds compared to 15 seconds for conventional methods. Eigenvalue analysis confirmed asymptotic stability with well-damped dynamic behavior.
• Economic Viability: Despite a 50% increase in Capital Expenditure (CAPEX), the system achieved a net annual benefit of $4.5 million due to a 25% reduction in Operational Expenditure (OPEX), energy savings, and carbon credit value.

Significance and Claims
The authors claim that this work represents one of the earliest studies to successfully integrate decentralized grid-forming inverter control with MILP-based energy management and sustainability analysis. The significance of the proposed framework lies in its ability to provide a scalable, energy-efficient, and carbon-neutral solution for inverter-dominated microgrids. By addressing the fragmentation in current research—where control, optimization, and sustainability are often studied in isolation—the paper offers a holistic approach that enables stable, resilient, and economically viable operation under islanded and disturbance conditions. The study concludes that such integrated frameworks are essential for the transition to next-generation, carbon-neutral power systems, though it acknowledges limitations regarding the reliance on simulation models and the need for future hardware-in-the-loop validation.