On the Robust Control and Optimization Strategies for Islanded Inverter-Based Microgrids

In recent years, the concept of Microgrids (MGs) has become more popular due to a significant integration of renewable energy sources (RESs) into electric power systems. Microgrids are small-scale power grids consisting of localized grouping of heterogeneous Distributed Generators (DGs), storage systems, and loads. MGs may operate either in autonomous islanded mode or connected to the main power system. Despite the significant benefits of increasing RESs, many new challenges raise in controlling MGs. Hence, a three layered hierarchical architecture consisting of three control loops closed on the DGs dynamics has been introduced for MGs. The inner loop is called Primary Control (PC), and it provides the references for the DG’s DC-AC power converters. In general, the PC is implemented in a decentralized way with the aim to establish, by means of a droop control term, the desired sharing of power among DGs while preserving the MG stability. Then, because of inverterbased DGs have no inertia, a Secondary Control (SC) layer is needed to compensate the frequency and voltage deviations introduced by the PC’s droop control terms. Finally, an operation control is designed to optimize the operation of the MGs by providing power setpoints to the lower control layers. This thesis is mainly devoted to the design of robust distributed secondary frequency and voltage restoration control strategies for AC MGs to avoid central controllers and complexity of communication networks. Different distributed strategies are proposed in this work: (i) Robust Adaptive Distributed SC with Communication delays (ii) Robust Optimal Distributed Voltage SC with Communication Delays and (iii) Distributed Finite-Time SC by Coupled Sliding-Mode Technique. In all three proposed approaches, the problem is addressed in a multi-agent fashion where the generator plays the role of cooperative agents communicating over a network and physically coupled through the power system. The first approach provides an exponentially converging voltage and frequency restoration rate in the presence of both, model uncertainties, and multiple time-varying delays in the DGs’s communications. This approach consist of two terms: 1) a decentralized Integral Sliding Mode Control (ISMC) aimed to enforce each agent (DG) to behaves as reference unperturbed dynamic; 2) an ad-hoc designed distributed protocol aimed to globally, exponentially, achieves the frequency and voltage restoration while fulfilling the power-sharing constraints in spite of the communication delays. The second approach extends the first one by including an optimization algorithm to find the optimal control gains and estimate the corresponding maximum delay tolerated by the controlled system. In the third approach, the problem of voltage and frequency restoration as well as active power sharing are solved in finite-time by exploiting delay-free communications among DGs and considering model uncertainties. In this approach, for DGs with no direct access to their reference values, a finite-time distributed sliding mode estimator is implemented for both secondary frequency and voltage schemes. The estimator determines local estimates of the global reference values of the voltage and frequency for DGs in a finite time and provides this information for the distributed SC schemes. This dissertation also proposes a novel certainty Model Predictive Control (MPC) approach for the operation of islanded MG with very high share of renewable energy sources. To this aim, the conversion losses of storage units are formulated by quadratic functions to reduce the error in storage units state of charge prediction.