Search

When investigating voltage and current stresses in critical main circuit components during faults of the converter, a detailed equivalent model capable of representing the balancing control strategies of the capacitor voltages on the submodule level, along with blocking and delocking, is always necessary. Among previously proposed equivalent models of the modular multilevel converter (MMC), only submodule averaged models (SAMs) can capture interested inner dynamics inside each arm. However, the simulation accuracy of SAMs is not always satisfactory, especially when the time step is larger than 10 $\mu$s. In order to further improve the simulation accuracy with guaranteed simulation efficiency, the shifted frequency modeling of the half- and full-bridge hybrid MMC is proposed in this paper. Therein, each submodule is represented by Thévenin equivalents derived by submodule dynamic phasors. The arm of the MMC is represented by Norton equivalents to guarantee the efficiency, considering both normal and dc-blocking conditions. The effectiveness of the proposed model in terms of accuracy and efficiency is validated by simulating an MMC-based HVdc transmission. [ABSTRACT FROM AUTHOR]

This paper proposes a novel multirate co-simulation method to improve simulation efficiency of large-scale AC plus multiterminal DC (MTDC) grids. In this method, the whole system is partitioned into different AC and MTDC subsystems. They each are simulated with different time steps according to their requirements of accuracy. Unlike the existing methods where simplified models are adopted, the proposed approach fully reserves both the detailed behavior of converters and the nonlinear dynamics of large-scale AC systems. To realize the coordination between different subsystems, updating and discretization of the interface models are significant to guarantee the overall numerical accuracy and stability of the co-simulation method. Accordingly, the interface models of the partitioned AC and DC networks are represented by time-varying Thevenin and Norton equivalents. To eliminate the aliasing or time-delay errors, their parameters are derived using moving-window prediction, stepwise correction, and averaging techniques. A numerical oscillation suppression algorithm is developed for the discretization of the interface model. Thus, numerical stability is well guaranteed. The variables inside each subsystem as well as the interfaces, which are derived by augmented network equations (ANEs), are calculated simultaneously. The rate ratio of large AC and MTDC systems is determined based on our proposed criterion of electromagnetic transient simulation accuracy. The effectiveness of the cosimulation method is validated on a practical system integrating large AC and modular-multilevel-converter-based MTDC grids. [ABSTRACT FROM AUTHOR]

The dynamics of load, especially induction motors, are the driving force for short-term voltage stability (STVS) problems. In this paper, the equivalent rotation speed of motors is identified online and its recovery time is estimated next to realize an emergency-demand-response (EDR) based under speed load shedding (USLS) scheme to improve STVS. The proposed scheme consists of an EDR program and two regular stages (RSs). In the EDR program, contracted load is used as a fast-response resource rather than the last defense. The estimated recovery time (ERT) is used as the triggering signal for the EDR program. In the RSs, the amount of load to be shed at each bus is determined according to the assigned weights based on ERTs. Case studies on a practical power system in China Southern Power Grid have validated the performance of the proposed USLS scheme under various contingency scenarios. The utilization of EDR resources and the adaptive distribution of shedding amount in RSs guarantee faster voltage recovery. Therefore, USLS offers a new and more effective approach compared with existing under voltage load shedding to improve STVS. [ABSTRACT FROM PUBLISHER]

To make better use of local generators' dynamic var reserve to improve short-term voltage stability (STVS), this paper proposes an integrated high side var-voltage control (IHSV^2C) for power plants in receiving-end power systems. The IHSV^2C consists of a plant-level multi-machine var coordinator (MMVC) and several unit-level high side voltage controllers (HSVCs). MMVC coordinates the reactive power output among generators and provides control parameters for HSVCs, while HSVCs can maintain the voltage of a pre-defined voltage control point (VCP) by regulating the voltage reference of the excitation control of each generator. Therefore, when the system suffers a serious fault, IHSV^2C can drive the generators to provide stronger var support. Consequently, the voltage stability can be improved. A conceptual model of the receiving-end power system in China Southern Power Grid (CSG) is established, and the IHSV^2C has been validated through a simulation analysis on the system. Both time-domain simulation results and the voltage sag severity index (VSSI) have fully demonstrated its performance. Therefore IHSV^2C offers a new and effective approach to improve STVS of receiving-end power systems. [ABSTRACT FROM AUTHOR]

Real-time decision-making in power system scheduling is imperative in response to the increasing integration of renewable energy. This paper proposes a novel framework leveraging Reinforcement Learning from Demonstration (RLfD) to address complex unit commitment (UC) and optimal power flow (OPF) challenges, called GridZero-Imitation (GZ-I). Unlike traditional RL approaches that require complex reward function designs and have limited performance insurance, our method employs intuitive rewards and expert demonstrations to regularize the RL training. The demonstrations can be collected from asynchronous reanalysis of an expert solver, enabling RL to synergize with expert knowledge. Specifically, we conduct a decoupled training approach, employing two separate policy networks, RL and expert. During the Monte Carlo Tree Search (MCTS) process, action candidates from the expert policy foster a guided search mechanism, which is especially helpful in the early training stage. This framework alleviates the speed bottleneck typical of physics-based solvers in online decision-making, and also significantly enhances control performance and convergence speed of RL scheduling agents, as validated by substantial improvements in a 126-node real provincial test case. • Developed GridZero-Imitation, enhancing RL with expert demonstrations for power scheduling. • Demonstrates substantial improvements in control performance and speed in a real-case study. • Employs Monte Carlo Tree Search for guided action selection, improving early-stage training efficiency. • Significantly accelerates decision-making over traditional methods by over 100x, with high reliability. • Adapts to topology changes, ensuring robustness against operational uncertainties in power grids. [ABSTRACT FROM AUTHOR]

• A two-stage available transfer capacity evaluation framework is proposed considering flexible demand response. • An integrated dispatch model of generation and demand side resources is formulated. • A demand side bidding scheme is developed to enable flexible demand response programs. • A real-world power system in Sichuan, China is tested to validate the effectiveness of the proposed method. Available transfer capability (ATC) has been widely adopted as a crucial measure to guarantee free and reliable power trading. The existing literature generally evaluates the ATC of a power grid considering inelastic load. However, the impacts of responsive load have not yet been systematically explored, which underestimates the ability of demand side resources in ATC enhancement. Therefore, we propose a two-stage ATC evaluation framework in this paper integrating flexible demand response (FDR). In the first stage, i.e., a day-ahead market, a demand side bidding scheme is developed in which electricity users can make flexible choices to participate in demand response (DR) programs according to their preferences. In this paper, FDR is categorized into three typical patterns: (i) deferrable DR, (ii) switchable DR and (iii) adjustable DR. The objective is to minimize the total generation and dispatch costs, while respecting the constraints for FDR, the minimal accommodation of renewable energy, the inter-area power trading, etc. Then based on generation and demand side bidding, a day-ahead unit commitment model is formulated to schedule thermal generators, interchange power and FDR. In the second stage, a real-time ATC evaluation model based on the aforementioned unit commitment results is proposed to quantify the contributions of FDR on ATC enhancement. Case studies based on IEEE 14-bus system and a provincial power grid in China demonstrate that FDR can effectively improve real-time ATC by load shifting and peak shaving, and facilitate the accommodation of renewable energy. [ABSTRACT FROM AUTHOR]