Content area
Over the past decade, renewable energy sources (RESs) have been widely researched and applied because of their environmental friendliness and increasingly competitive prices compared to traditional fossil fuel-based energy sources. To effectively coordinate a large number of RESs, the microgrid (MG) concept has been developed, which adopts a hierarchical control structure to manage the operation of the whole system at different time scales. This structure includes three control hierarchies: primary control, secondary control, and tertiary control. While primary control uses a decentralized approach to stabilize the system, a centralized communication network is needed in conventional secondary and tertiary control levels to synchronize voltage/frequency and exchange power with the utility grid, respectively. Due to the reliance on the MG central controller, the centralized control method exposes limitations in controlling and scaling MG systems. Conversely, distributed control has been intensively developed to coordinate the RESs without the need for the MG central controller, making the MG systems more reliable and expandable. This dissertation aims to improve the distributed secondary control to tackle the remaining problems in autonomous MGs regarding frequency and voltage synchronization, power-sharing, trade-offs between voltage and reactive power, and MG operation costs. The major contribution of the dissertation includes:
• Develop a novel distributed secondary frequency controller to address the convergence response speed of the islanded MG system. The proposed controller restores the system’s frequency to its rated value while ensuring the proper power-sharing ratio regulated by the primary control level. Unlike previous studies, where the response speed of frequency and active power relied on complex control schemes, our approach allows users to easily predetermine the convergence speed using a single control parameter. Furthermore, the convergence rate is independent of both the communication network types and the initial states of the MG system.
• Based on the concept of free-will arbitrary time (FWAT) control, a distributed secondary control is further developed for both frequency and voltage to achieve timely synchronization. Simultaneously, a cost-based distributed active power control is introduced to reduce the total operating cost of the entire MG system, considering the operation cost of individual RES. The active power response can reach a steady state within the pre-setting convergence time.
• A distributed FWAT containment-based voltage control is proposed to limit the voltage magnitudes within a controllable range while providing headroom for reactive power regulation from the RESs.
In addition, a state-space model of the entire MG system and Lyapunov functions are built to demonstrate the stability of the proposed controller and prove its convergence time characteristic. Different simulation scenarios have been performed in the Matlab/SIMULINK software environment to demonstrate the efficiency of the proposed distributed control strategies. Experimental MGs are executed to test the feasibility of the proposed control schemes.