TL;DR
This paper introduces a novel secondary control layer for DC microgrids that ensures stable current sharing and voltage balancing. It leverages a consensus-based approach and Plug-and-Play (PnP) principles, allowing for seamless integration and removal of distributed generation units without complex reconfigurations. This work is crucial for building more robust and scalable DC microgrid systems.
What Problem Does This Paper Address?
Modern power systems are increasingly incorporating distributed energy resources (DERs) like solar panels and batteries, leading to the rise of microgrids. While AC microgrids have been extensively studied, DC microgrids are gaining traction due to their efficiency in handling DC loads and storage devices. However, controlling these DC microgrids presents challenges, primarily in achieving stable voltage and ensuring that all DERs contribute proportionally to the load (current sharing). Existing solutions often require centralized control or complex re-parameterization when DERs are added or removed, hindering scalability and flexibility. This paper addresses the need for a decentralized, scalable, and robust control strategy that can maintain voltage stability and current sharing in DC microgrids, even with dynamic changes in the grid topology.
Methods
The authors propose a hierarchical control structure. The primary control layer, assumed to be a decentralized Plug-and-Play (PnP) voltage controller, ensures local voltage stability. The core contribution is the secondary consensus-based control layer. This layer adjusts the reference voltages of the primary controllers to achieve global current sharing and voltage balancing.
Key methodological aspects include:
- Consensus Dynamics: A consensus algorithm is employed to allow DERs to exchange information and converge to a common state. This is formulated as a set of differential equations that govern the adjustment of voltage references based on local measurements and neighbor information.
- Approximation of Primary Controllers: To simplify analysis, the complex dynamics of the primary controllers are approximated by either ideal unit gains or first-order transfer functions. This allows for rigorous mathematical analysis of the closed-loop system's stability and performance.
- Plug-and-Play (PnP) Design: The secondary control layer is designed to be PnP. This means that when a new DER is added or an existing one is removed, only local updates to the secondary controllers are required, preserving the overall stability and performance of the microgrid. This is a significant advantage over methods requiring global re-configuration.
- Mathematical Analysis: The paper uses linear system theory, including matrix analysis (Laplacian matrices, eigenvalues, invariant subspaces), to prove exponential stability, current sharing, and voltage balancing.
Datasets
Not discussed in the paper. The paper relies on simulations and experimental validation using a custom-built DC microgrid setup.
Key Insights and Ideas
- Decentralized Control for Scalability: The PnP nature of the secondary control layer is a major insight, enabling scalable microgrid architectures where DERs can be added or removed without disrupting the entire system.
- Consensus for Global Objectives: Applying consensus algorithms, typically used for agreement in distributed systems, to achieve global objectives like current sharing and voltage balancing in a physical power grid is a powerful concept.
- Hierarchical Control Synergy: The paper demonstrates how a well-designed secondary consensus layer can effectively coordinate decentralized primary controllers to achieve system-wide goals.
- Robustness to Topology Changes: The PnP design ensures that the control system remains stable and functional even when the communication or electrical topology of the microgrid changes.
- Mathematical Rigor: The use of advanced linear algebra and control theory provides strong theoretical guarantees for the proposed control scheme.
Key Results
- Exponential Stability: The proposed control scheme guarantees exponential stability of the DC microgrid.
- Current Sharing: The system achieves asymptotic current sharing, meaning that DERs contribute to the load proportionally to their capacity (or equally, depending on the configuration). This was demonstrated in simulations with a 7-DGUs microgrid and experiments with a 3-DGUs microgrid.
- Voltage Balancing: The scheme ensures that the voltages at the Point of Common Coupling (PCC) are balanced around a common reference voltage.
- PnP Operation: The PnP design for secondary controllers was successfully demonstrated, showing that DGUs can be added or removed with only local controller updates.
- Simulation and Experimental Validation: Theoretical results were validated through simulations using Simulink/PLECS and experiments on a laboratory-scale DC microgrid.
Limitations and Future Work
- Approximations: The analysis relies on approximations of primary control loops (unit gain or first-order transfer functions). While these are stated to be mild, the impact of more complex primary dynamics on the secondary control performance could be further investigated.
- Non-idealities: The paper mentions that future work will study the impact of non-idealities such as transmission delays, data quantization, and packet drops on the performance of closed-loop microgrids.
- Cybersecurity: The authors highlight the importance of addressing security against cyberattacks in networked microgrid control architectures as a future challenge.
- Communication Network: While the communication graph is assumed to be connected, the impact of communication failures or bandwidth limitations could be explored.
Summary updated