Abstract:Hybrid energy storage system (HESS) is an effective solution to address power imbalance problems caused by variability of renewable energy resources and load fluctuations in DC microgrids. The goal of HESS is to efficiently utilize different types of energy storage systems, each with its unique characteristics. Normally, the energy management of HESS relies on centralized control methods, which have limitations in flexibility, scalability, and reliability. This paper proposes an innovative artificial neural network (ANN) based model predictive control (MPC) method, integrated with a decentralized power-sharing strategy for HESS. In the proposed technique, MPC is employed as an expert to provide data to train the ANN. Once the ANN is finely tuned, it is directly utilized to control the DC-DC converters, eliminating the need for the extensive computations typically required by conventional MPC. In the proposed control scheme, virtual resistance droop control for fuel cell (FC) and virtual capacitance droop control for battery are designed in a decentralized manner to achieve power-sharing, enhance lifespan, and ensure HESS stability. As a result, the FC is able to support steady state loads, while the battery handles rapid load variations. Simulation results using Matlab/Simulink demonstrate the effective performance of the proposed controller under different loads and input variations, showcasing improved performance compared to conventional MPC.

Abstract:Large-scale development of electric vehicles (EVs) exposes power grids and transportation networks with limited capacity to increased fault risks. In this paper, a method to prevent fault risks in advance by using the flexibility of EV travel to coordinate the operation of distribution and transportation networks is proposed. Since EV travel decisions are influenced by the charging and travel time costs, adjusting charging price and travel time price can help guide behavior and enable coordinated operation of power and transportation networks. First, risk-based distribution locational marginal prices (RDLMPs) are established to restrain the distribution network risks. Second, traffic risks are formulated using origin-destination (OD) risk marginal prices (ODRMPs) considering the degree of traffic congestion fault risks. Under the guidance of the RDLMPs and ODRMPs, EV fleets optimize their travel plans to minimize overall costs, including charging and time costs. Finally, case studies verify that the proposed method can reduce the operational risks of both distribution and transportation networks.

Abstract:Considering the escalating usage of renewable energy and rising frequency of extreme meteorological events, the risk of emergency load shedding (ELS) in power grids due to faults is increasing. The existing ELS strategies fail to provide users with advance warnings and blackout preparation time. To address this issue, an innovative strategy called warning and delayed load shedding is proposed in this study. In this approach, when it becomes necessary to shed load for emergency control, users are immediately notified with a power outage warning. Subsequently, energy storage and other regulatory resources are employed to substitute for load shedding, thus postponing the execution of the load shedding command. This delay equips users with the ability and time to respond and prepare. To implement this strategy, the operational principles supported by energy storage and backup power are further discussed. Five performance indexes are utilized to evaluate the delayed load shedding capability. Moreover, the delayed load shedding switch function and energy storage power balance equation are constructed to determine the relationship between energy storage, backup power sources, and load shedding time. Subsequently, two optimized load shedding models supported by energy storage are established, i.e., maximum and flexible delayed models. For comparison, improvements are made to the conventional load shedding model without delay by incorporating energy storage. An IEEE 30-node network with energy storage is used to test the three load shedding models. Accordingly, the evaluation indexes are calculated and compared. The results of the performance indexes and comparative analysis validate the effectiveness of the proposed methods, indicating that by using energy storage, users can be notified with advance power outage warnings and preparation time.

Abstract:Based on the dual equivalent model of a single-phase two-winding transformer and a single-phase three-winding autotransformer, a method for identifying inrush current in single-phase transformers is proposed. This method distinguishes inrush current from internal fault current using the instantaneous equivalent inductance of the dual model. The setting principle of the method is determined by analyzing the air-core inductance and the equivalent model of the faulty transformer. PSCAD simulations and recorded transformer protection data demonstrate that the proposed method can accurately identify inrush current when the transformer core is deeply saturated, and can quickly discriminate between currents during a critical internal fault. Furthermore, the simulation results show that the proposed method is sensitive to minor turn-to-turn faults but is less sensitive to the equivalent impedance of an external source.

Abstract:Automatic reclosing is widely employed in wind farm transmission lines. However, conventional reclosing strategies cannot identify the nature of the fault before reclosing, thereby posing a risk to the safe operation of the wind farm transmission system and associated equipment, especially during permanent faults. This study focuses on 220 kV wind farm transmission lines without parallel reactors. A single-phase ground fault circuit model is established first, and expressions for faulted phase-end voltages are derived before and after arc extinction. Then, the variations in voltage amplitude before and after fault arc extinction are revealed. The short-time Fourier transform is employed to extract fundamental frequency voltage amplitude from the secondary side of the capacitive voltage transformer (CVT). Subsequently, an adaptive reclosing strategy based on fundamental frequency voltage amplitude detection for wind farm transmission lines is proposed to enhance feature variations through sequential overlapping derivative (SOD) transformation. Finally, through analysis and validation on the RTDS platform, the proposed adaptive reclosing strategy is demonstrated to be simple, feasible, robust against transient resistances, and characterized by high sensitivity.

Abstract:Excessive temperature rise during the operation of the generator system can affect the safety and life cycle of the machine. Therefore, in order to accurately obtain the internal temperature of the switched reluctance generator (SRG), an internal temperature estimation model based on electric heating is established in this paper. First, an improved variable coefficient Bertotti loss separation calculation formula is adopted to solve the iron loss of the generator under various operating conditions. Subsequently, the accurate heat source parameters in the temperature model can be obtained, and the corresponding heat source data can be calculated. Then, based on the obtained heat source data, an equivalent thermal circuit model is established for SRG. Meanwhile, in order to effectively reduce the internal temperature during SRG operation, a new water-cooled structure for direct cooling of SRG stator windings is proposed in this paper, which can effectively reduce the temperature rise during operation, thus improving the reliability of the generator. Finally, by comparing the equivalent thermal circuit model, finite element thermal model, and experimental temperature measurements of SRG, it is found that results from the equivalent thermal circuit model of the SRG are closer to the measured temperatures, while the effectiveness of the water-cooled structure is verified.

Abstract:Inverter-dominated isolated/islanded microgrids (IDIMGs) lack infinite buses and have low inertia, resulting in higher sensitivity to disturbances and reduced stability compared to grid-tied systems. Enhancing the resilience of IDIMGs can be achieved by maximizing the system loadability and/or mitigating the expected disturbances such as line switching operations. This paper proposes a two-stage framework based on the deployment of mobile energy storage (MES) to enhance the resilience of IDIMGs. In the first stage, the network configuration and deployment of MES are optimized to maximize the system loadability. The proposed formulation for this stage is a stochastic multi-period mixed-integer nonlinear program (MINLP) that maximizes a weighted sum of minimax loadabilities. In the second stage, transitional locations of MES, line-exchange execution sequence, and droop control of dispatchable sources are jointly optimized to mitigate line-switching disturbances that occur when transitioning to the new network configuration obtained in the first stage. The second stage model is a multi-objective MINLP. The proposed models are solved within the general algebraic modeling system (GAMS), utilizing a modified IEEE 33-bus system. Simulations are conducted to assess the significance of each proposed model, and the results reveal remarkable improvements in system loadability with the utilization of the first-stage model and significant reductions in the total switched power with the adoption of the second-stage model.

Abstract:Since the effectiveness of the flexible current arc suppression method heavily relies on the accurate measurement of the distribution line-to-ground parameters, the suppression of single line-to-ground (SLG) fault current may deteriorate due to factors such as line switching and other disturbances during SLG fault arc suppression. Additionally, during SLG fault arc suppression, promptly identifying the fault type and rapidly deactivating the flexible arc suppression device (FASD) can reduce the overvoltage risk in non-faulted phase devices. To address these issues, this paper presents a parameter identification method based on recursive least squares (RLS) while a variable forgetting factor strategy is introduces to enhance the RLS algorithm's disturbance rejection capability. Simulations verify that the variable forgetting factor recursive least squares (VFF-RLS) algorithm can accurately identify distribution line-to-ground parameters in real time and effectively suppress SLG fault current. The online identification of grounding transition conductance is simultaneously used to determine the fault type and quickly detect when the SLG fault has been cleared.

Abstract:Integration of inverter-based resources (IBRs) reshapes conventional power swing patterns, challenging the operation of legacy power swing protection schemes. This paper highlights the significant consequences of the distinctive power swing patterns exhibited by grid-forming IBRs (GFM-IBRs). The swing mechanism involving GFM-IBRs is elucidated using analytical modeling of GFM-IBRs with various power synchronization loops (PSLs). With varying GFM-IBR penetration levels and different generator scenarios, the performance of power swing protection functions is examined, including power swing blocking (PSB) and out-of-step tripping (OST). Additionally, the IEEE PSRC D29 test system is employed to present power swing protection results in a large-scale power system integrated with synchronous generators (SGs) and IBRs. The results indicate that, GFM-IBRs with sufficient voltage support capability can provide positive impact on power swing dynamics, and power swing protection performance can be enhanced by emulating the inertia and droop mechanism of SGs and rapidly adjusting the active power output. However, with high penetration levels of GFM-IBRs, OST maloperation may occur in scenarios with inadequate voltage support capability.

Abstract:The rapid and sustained advancement of photovoltaic (PV) power generation technology has introduced significant challenges to the power grid operation, including reduced grid strength and poor damping, thereby causing occurrence of harmonic resonance and potential instability. Conventional stability assessments of PV inverters often overlook critical factors-such as DC-side voltage control and operating mode variations-which arise from the characteristics of the connected PV array. Consequently, these assessments yield inaccurate stability assessments, particularly under weak grid conditions. This study addresses this issue by integrating the characteristics of PV output and DC-side power control into small-signal models while accounting for different control modes, such as constant current (CC), constant voltage (CV), and maximum power point tracking (MPPT). Initially, the study integrates the inverter's output control, LCL filter, and characteristics of the grid to develop a comprehensive framework for the entire PV system. Subsequently, the study uses amplitude and phase stability margins to evaluate how DC-side operating modes, the short-circuit ratio (SCR) of the power grid, and inverter controller parameters influence the system stability. Finally, the accuracy and stability of the model are validated through simulations and a 20-kW three-level prototype PV inverter.

Abstract:This paper employs artificial intelligence and machine learning techniques to predict the dominant oscillation modes in AC microgrids. The dominant modes are highly dependent on the droop gains and only slightly affected by the loading conditions. This paper utilizes the least absolute shrinkage and select operator (LASSO) algorithm to extract the key features contributing directly to dominant modes. The adaptive neuro-fuzzy inference system (ANFIS) is employed as a nonlinear regression technique to train a model that relates the system's key features to the dominant modes of the AC microgrid. The data obtained from a 6-bus AC microgrid test system is used to train the LASSO-based ANFIS model. The results show that the proposed method can substantially reduce the data volume of the training set due to LASSO sparse feature. The precision of the proposed algorithm is determined by comparing its output to the modes determined by the derived small-signal model of the system.