Cyber-physical systems (CPSs) are vulnerable to cyber-attacks. Nowadays, the detection of cyber-attacks in microgrids as examples of CPS has become an important topic due to their wide use in various practical applications, from renewable energy plants to power distribution and electric transportation. In this paper, we propose a new artificial intelligence (AI)-based method for detection of cyber-attacks in direct current (DC) microgrids and also identification of the attacked distributed energy resource (DER) unit. The proposed method works based on time series analysis and a nonlinear auto-regressive exogenous model (NARX) neural network, which is a special type of recurrent neural network, for estimating DC voltages and currents. In the proposed method, we consider the effect of cyber-attacks named false data injection attacks (FDIAs), which try to affect the accurate voltage regulation and current sharing by affecting voltage and current sensors. In the presented strategy, first, a DC microgrid is operated and controlled without any FDIAs to gather enough data during the normal operation required for the training of NARX neural networks. It is worth mentioning that, in the data generation process, load changing is also considered to have distinguishing datasets for load changing and cyber-attack scenarios. Trained and fine-tuned NARX neural networks are exploited in an online manner to estimate the output DC voltages and currents of DER units in DC microgrid. Then, based on the error of estimation, the cyber-attack is detected. To show the effectiveness of the proposed method, offline digital time-domain simulation studies are performed on a test DC microgrid system in MATLAB/Simulink environment, and the results are verified using real-time simulations using OPAL-RT real-time digital simulator (RTDS).

Many techniques used and still in usage for solving the problem of islanding detection are intrinsically passive, active, or hybrid of both. Each one of them has its own benefits and drawbacks. In this paper, we propose a method which takes the advantage of a ML-based algorithm, namely SVM, in order to produce the results more efficiently. The results of the simulations based on the model and experimentally-measured parameters of a real-life practical PV plant, gives much better output than the traditional reported methods. During the tests and simulations, an additional problem, namely grid fault emerged, posing new challenges for the proposed method. Occurrences of islanding and grid fault are grouped together with same kernel dimension and no custom hyper plane bordering. Discrimination between islanding and grid fault events is an essential dilemma which is handled by the proposed SVM-based algorithm to achieve more precision in islanding detection and simultaneously, detect the grid faults authentically. NDZs and DT are tested using two dimensions, namely the generated active energy from PV plant (0-110% of Pn), and distribution network voltage levels (± 10% of Un). Simulations based on the model and parameters of a real-life practical PV power plant are performed in MATLAB/Simulink environment and several tests are executed for several scenarios, Finally, comparisons with previously-reported techniques prove the effectiveness, authenticity, selectivity, accuracy and precision of the proposed islanding and grid fault detection strategy with allowable impact on power quality according to UL1741 and its superiority over other methods.

This paper proposes a novel distributed noise-resilient secondary control for voltage and frequency restoration of islanded microgrid inverter-based distributed generations (DGs) with an additive type of noise. The existing distributed methods commonly are designed as secondary control system systems that operate on the assumption of ideal communication networks among DGs. However, the channels are prone to stochastic noise, whereas each DG obtains noisy measurements of the states of its neighbors via environmental noises. The existing distributed noise-resilient methods, ignore a complete model of the system. In contrast, this paper proposes consensus protocols that take into account both the noisy measurements and a complete nonlinear model of the system, examines the mean-square average consensus for voltage and frequency restoration of islanded AC microgrids in an uncertain environment, and provides accurate proportional real power sharing. Our proposed consensus protocol contains two parts: the state feedback of the agent and the relative states of the DG and its neighboring DGs. Finally, simulation studies are carried out in MATLAB/SimPowerSystems to evaluate the performance of the control laws. Simulation results and comparison with previous work reveals the effectiveness of the proposed method in regulating microgrid voltage and frequency and providing accurate proportional real power sharing.

This paper presents a new decentralized robust strategy to improve small and large-signal stability and power-sharing of hybrid AC/DC microgrids and improve its performance for nonlinear and unbalanced loads. In addition to the sliding mode controller for DC/DC converters, for the sake of improving power sharing and regulating active and reactive powers injected by distributed energy resources, and moreover, controlling harmonic and negative-sequence current in the presence of nonlinear and unbalanced loads, two separate controllers for positive sequence power control and negative sequence current control are designed based on the sliding mode control and Lyapunov function theory, respectively. The theoretical concept of the proposed robust control strategy, including mathematical modeling of microgrid components, basic theorems, controller design procedure, and robustness and closed loop stability analysis are outlined. Also, this direct power/current/voltage control scheme is governed by a new hybrid AC/DC hierarchical control scheme that exploits a harmonic virtual impedance loop and voltage compensation scheme. To show the effectiveness of the proposed robust control scheme, offline time-domain simulations are done on a hybrid AC/DC wind/photovoltaic/fuel-cell microgrid with nonlinear and unbalanced loads in MATLAB/Simulink environment, and the results are experimentally verified by OPAL-RT real-time digital simulator.

This paper presents a new islanding detection strategy for low-voltage (LV) inverter-interfaced microgrids based on adaptive neuro-fuzzy inference system (ANFIS). The proposed islanding detection method exploits the pattern recognition capability of ANFIS and its nonlinear mapping of relation between inputs. The ANFIS monitors seven inputs measured at point of common coupling (PCC), namely root-mean square (RMS) of voltage and current (RMSU and RMSI), total harmonic distortion (THD) of voltage and current (THDU and THDI), frequency (f), and also active and reactive powers (P, Q) that are experimentally obtained based on practical measurement in a real-life microgrid. The proposed method is composed of passive monitoring of the mentioned inputs and therefore, does not influence power quality (PQ); but considerably decreases non detection zones (NDZs). In order to cover as much situations as possible, minimizing false tripping and still remaining selective, type and number of samples are introduced. Here, one of the main goals is reducing NDZ by still keeping PQ in order. Based on the sampled frequency and number of samples, we find that the proposed method has less detection time and better accuracy, compared to the reported methods. Simulations performed in MATLAB/Simulink software environment and several tests performed based on different active load conditions and multiple distributed generation (DG)s, prove the effectiveness, authenticity, selectivity, accuracy and precision of the proposed method with allowable impact on PQ according to UL1741 standard.

All control methods for the decentralized control of distributed energy resources (DERs) in the microgrid need to calculate power to decide whether the power produced will be able to stabilize the system. Unlike the previous research that is limited to the primary and secondary control levels, the presented decentralized droop-based control scheme includes detailed modeling for three hierarchical control levels for either grid-connected or autonomous modes. A new complementary control loop that is added to the hierarchical droop-based control scheme determines and controls the reactive power reference by a novel application of radial basis function neural networks (RBFNNs) for a fast, authentic, and accurate calculation of power to improve power sharing and enhance microgrid stability margins in facing with small and large signal disturbances. This method suppresses the low-pass filter that is normally used to determine high-frequency components of power and replaces it by the power flow nonlinear equation set that is solved by a novel application of RBFNNs, and consequently, power sharing to loads and network is done sufficiently. The simulation studies that have been performed on a microgrid consisting of four DERs and local loads using MATLAB/SIMULINK software demonstrate the effectiveness of the proposed control scheme.