In this letter, a novel autonomous control framework “Grid Mind” is proposed for the secure operation of power grids based on cutting-edge artificial intelligence (AI) technologies. The proposed platform provides a data-driven, model-free and closed-loop control agent trained using deep reinforcement learning (DRL) algorithms by interacting with massive simulations and/or real environment of a power grid. The proposed agent learns from scratch to master the power grid voltage control problem purely from data. It can make autonomous voltage control (AVC) strategies to support grid operators in making effective and timely control actions, according to the current system conditions detected by real-time measurements from supervisory control and data acquisition (SCADA) or phasor measurement units (PMUs). Two state-of-the-art DRL algorithms, namely deep Q-network (DQN) and deep deterministic policy gradient (DDPG), are proposed to formulate the AVC problem with performance compared. Case studies on a realistic 200-bus test system demonstrate the effectiveness and promising performance of the proposed framework.

Short‐term voltage stability (SVS) is a serious issue in modern power systems. In China, the East China Power Network is especially vulnerable to short‐term voltage instability due to its increasing dependence on electrical power from external through high‐voltage direct current (HVDC) transmission lines. To study the SVS, a criterion/index is first required to evaluate the SVS of power systems. However, the currently used practical criteria cannot effectively evaluate the influence of controlling strategies (such as regulating dynamic VAR reservation) of power systems. Therefore, a practical and continuous SVS index (SVSI) based on voltage curves is proposed in this study. The proposed SVSI can quantify the extent of SVS in power systems, thus can be used in the optimisation target of preventive control for improving the SVS of power systems. Moreover, the SVSI consists of three components, and each component reflects a feature of the voltage curve in terms of SVS. The SVSI is essentially a model‐free method for assessing the SVS of power systems, so it is robust. Three verifying cases and one application case are presented to show the validity of the SVSI and the feasibility of the SVSI in practical applications.

Parameter identification in load models is a critical factor for power system computation, simulation, and prediction, as well as stability and reliability analysis. Conventional point estimation based composite load modeling approaches suffer from disturbances and noises, and provide limited information of the system dynamics. In this work, a statistics (Bayesian Estimation) based distribution estimation approach is proposed for both static and dynamic load models. When dealing with multiple parameters, Gibbs sampling method is employed. The proposed method samples all parameters in each iteration and updates one parameter while others remain fixed. The proposed method provides a distribution estimation for load model coefficients and is robust for measuring errors. The proposed parameter identification approach is generic and can be applied to both transmission and distribution networks. Simulations using a 33-feeder system illustrated the efficiency and robustness of the proposal.

The ever-increasing penetration of centralized and distributed renewable energy, power electronics-based transmission equipment and loads, advanced protection and control systems, storage devices and new power market rules all contribute to the growing dynamics and stochastic behaviors being observed in today’s grid operation. Understanding operational risks and providing prompt control actions are of great importance to ensure secure and economic operation of a bulk power system. In this paper, a novel integrated online dynamic security assessment system (DSAS) is developed that intakes real-time EMS snapshots combining both bus/branch and node/breaker network models, performs dynamic contingency analysis under various conditions, calculates real-time transfer limits, and provides online control suggestions to mitigate operational risks. Several unique and innovative features are developed to address practical challenges, including: (1) real-time stitching of power flow information from both node/breaker and bus/branch models covering different geographical regions of interest; (2) online corrections and enhancements to power flow and dynamic models; (3) equipment-name-based modelling approaches for complex contingencies and control systems; and (4) distributed computing capabilities for significant computational speed enhancement. The developed DSAS has been deployed in the control center of State Grid Jiangsu Electric Power Company with hundreds of HVAC and 8 (U)HVDC transmission lines, which has been running reliably since Sept. 2018, achieving satisfactory performance in improving situational awareness and economic operation of the power grid.

To effectively tackle the operational challenges facing today’s bulk power systems with growing uncertainties, this paper presents a novel data-driven method for transient stability assessment method of the AC/DC hybrid power systems using auto-encoder-based algorithms for feature extraction and convolutional deep belief networks and Boltzmann machines for training accurate and robust models. First, a feature set is established from power grid operational states, and a stack-based noise reduction automatic encoder (SDAE) is used to extract key features that significantly affect the transient stability of the system. Next, transient stability assessment models are trained based on attention mechanisms, improved convolutional deep belief networks (CDBN), and multiple convolution-constrained Boltzmann machines to learn effective features from the input data adaptively. With such a design, the attention mechanism improves the model’s attention to instability samples, and the high-precision assessment of the systems’ transient stability is achieved. The experimental results conducted on the New England 39-bus system and the WECC 179-bus system show that the proposed method can accurately assess the transient stability of the system with multi-infeed DC by outperforming traditional supervised learning models considering various noise levels.

With the increased penetration of wind energy into the power system, the fault ride-through ability of doubly fed generator (DFIG) has got more and more attention. Recent grid codes request DFIG stay connected to grid when system fault occurs and play an active role in grid voltage recovery. In this paper, a new voltage control strategy without crowbar is to improve fault ride-through characteristic. The transient characteristics of DFIG are analyzed. The fault current contributions DFIG with crowbar protection and proposed voltage control strategy are analyzed and compared based on mathematical expressions and equivalent circuits. The conclusion that fault current contributions of DFIG under two control strategies are different in amplitude and phase is obtained. At last, a 10KV distribution network with DFIG model under two control strategies set up in PowerFactory DIgSILENT. Simulation results show the different behavior of DFIG under different control strategies which will impact distribution network protection in different way.