This paper proposes a model for hierarchical combination of deep reinforcement learning (DRL) with quadratic programming for distribution system restoration after major outages. In the proposed model, optimal power dispatch of a collection of distributed energy resources, called integrated hybrid resources (IHRs), is determined by a DRL-trained controller, while a grid-level quadratic programming problem checks grid constraints and performs critical restoration operation. DRL is implemented using Soft Actor-Critic (SAC) algorithm, which is shown to outperform the common Deep Deterministic Policy Gradient in continuous action spaces. The numerical studies, performed on the 123-bus test distribution system, demonstrates that the hierarchical combination of DRL and quadratic programming not only speeds up the local operation of multiple IHRs, but also ensures that the network constraints are satisfied during the restoration operation.

Offshore wind farms connected to the grid via MMC-HVDC should have fault ride-through (FRT) ability. However, without high-speed communications, it is difficult to achieve coordinated control between wind power converters and the MMC-HVDC converters. To address this problem, a coordinated control method that is designed based on the injection of harmonics is proposed. During the fault, the offshore MMC converter detects that the DC voltage has exceeded the threshold value, and then injects sequence harmonics into the offshore AC lines, which enables the wind power converters to cooperatively limit the power injected into the grid. Therefore, the coordinated FRT of multiple converters can be realized without additional communication between wind power converters and the offshore MMC converters. A detailed model of MMC-HVDC connected offshore wind farms is constructed in a real time digital simulator (RTDS), and the proposed method is verified by hardware-in-loop dynamic simulation experiments. In addition to the traditional method of only limiting the output voltage of the offshore MMC, the proposed method also coordinately limits the output current of the wind farm to reduce the output power of the wind farm. Therefore, the proposed method allows the DC voltage to be quickly limited within the allowable range, and the system can achieve safe and reliable FRT without communication.

Clouds passing over solar photovoltaic (PV) power system causes power fluctuations, which contributes to power quality issues. Power fluctuations are usually compensated by an energy storage system (ESS) integrated with a filtering or smoothing controller. However, there is a great burden to tune and determine the filtering time constant (TF) that leads to better alleviation of PV power fluctuations because the irradiance variability varies from day to day. Lack of cloud information leads to incorrect selection of TF values which leads to poor smoothing performance when setting low values on cloudy days, or to unnecessary ESS operation when setting high values on clear days. This paper introduces a novel power smoothing framework to avoid the arbitrary selection of TF values and reduce the stress on ESS based on predictive and adaptive smoothing mechanisms. The proposed controller contains two layers, the first layer predicts the next day's irradiance profile and uses a clear sky model to identify the cloud class. Next, the controller sets the initial values of the filter time constants for cloudy, moderate, and mild days, and disables smoothing on clear and overcast days. The second layer adaptively adjusts the initial filter time constant based on the current power ramp rate and shares power for a hybrid energy storage system (HESS). The performance of the proposed controller is tested against the forecasting errors and compared to fixed time constant-based smoothing techniques and ramp rate control in a range of scenarios, including smoothing, ramp rate reduction, and battery degradation assessment.

Climate change may increase the risk of an area being hit by multiple extreme weather events, which brings significant challenges for distribution system planners in an increasing renewable penetration era. There is an urgent need for planning approaches to be more flexible and allow for adaptive adjustments in the future to hedge against high uncertainties in extreme weather event scenarios. In this work, we propose a resilience-oriented distribution system planning approach that considers multiple extreme weather events. A multi-stage hybrid-stochastic-and-robust formulation is developed to model decisions not only for initial investments, but also for adaptive investments and emergent operations in response to particular extreme events, meanwhile considering both long-term and short-term uncertainties. Our model is solved by a novel progressive hedging algorithm that is embedded with a nested column-and-constraint generation method. Case studies demonstrate the benefits of the proposed approach in making flexible and affordable planning decisions to protect distribution systems against multiple extreme weather events.

Hydrogen production has become an emerging technology and offered a pathway to sustainable energy utilization in transportation systems. In this paper, we propose a coordinated planning of the electricity network and hydrogen refueling and production station planning strategy for hydrogen fuel cell electric vehicles (FCEVs). In the proposed system structure, a hydrogen supply chain for FCEVs mainly composes hydrogen refueling stations (HRSs), hydrogen production stations (HPSs), hydrogen storage, and hydrogen delivery network. To enhance the system's flexibility, hydrogen delivery combines both gas pipeline delivery and truck logistics delivery. In the mathematical models, the synergistic effect and coupling relationship of the electricity network, gas network, and transportation network are presented. To ensure the efficiency and effectiveness of the hydrogen refueling system, a quality of service (QoS) and reliability assessment method is presented. The proposed methodology is verified in case studies. It is found out that compared with the single delivery methods, the proposed multi-network coordinated planning has the lowest total cost with less unserved energy. Besides, through QoS and reliability assessment, enough hydrogen reserves can be maintained to cope with contingencies.

With the increasing penetration of renewable energy generation, grid-forming (GFM) converters that simulate the dynamics of synchronous generator (SG) and actively participate in grid voltage and frequency regulation are increasingly employed. In GFM converter dominated power system, similar transient synchronization stability that is common in conventional power system also exists, which is closely related with the control dynamics of GFM converters. In this paper, the transient synchronization stability of GFM converters with virtual synchronous generator (VSG) control is focused on, with the influence of transient switched operation mode (TSOM) introduced by current saturation limitation considered especially. First, a novel transient equivalent motion model of VSG is developed. The model is a switched system and TSOM significantly changes the power angle characteristics. Based on power angle curve analysis and equal area method, the transient stability analysis is carried out from non-existence of equilibrium point and non-matching of accelerating-decelerating area. Analysis indicates that TSOM imposes another constraint to equilibrium point and reduces deceleration area, which worsens the transient stability. Further, employing the stability region theory of nonlinear dynamic systems, the stability region of VSG is visually displayed in phase plane. It is identified that the TSOM reduces the stability region and worsens the stability. Finally, the influence factors including power reference, grid voltage sag and grid impedance are discussed.

Integrating sustainable energy resources transforms the distribution grid into an active system with higher variations observed in load and generation. Estimating distributed generation, gross load, and cold load pick-up (CLPU) become more challenging with behind-the-meter (BTM) distributed energy resources (DERs), especially in case of outages caused by extreme events. This work proposes a resilience-driven restoration scheme using the most updated information from an integrated and enhanced situational awareness tool (ESAT) using kernelized Bayesian state-space inference (KBSI) with Markov Chain Monte Carlo (MCMC) and multiple optimization algorithms. ESAT consists of the BTM load/ DER estimation and disaggregation, CLPU estimation, and network topology estimation with de-energized islands. The proposed work provides solutions to establish informed restoration schemes considering resilience criteria for a quick recovery of high-priority loads. A resilience metric is utilized after outages to measure the effectiveness of ESAT-driven restoration for the supposed threats. The performance of developed ESAT is demonstrated using actual field datasets and validated using the emulated real scenarios on a benchmark model.

The twenty-two papers in this special section focus on innovative research advances tackling fundamental challenges of achieving more resilient grids to real-world demonstrations of leveraging sustainable energy resources to enhance grid resiliency.

Present-day inverter control mainly employs phase locked loop (PLL) based current-controlled strategies. With growing penetration of inverter-based resources in power systems, stability issues associated with inverter control have been reported in weak grid scenarios. A natural and fundamental question to answer is whether this instability is inherent to PLL based current-controlled strategies or it can be resolved with proper frequency and voltage control loop design. This paper addresses this question by analyzing the small signal stability of the current-controlled inverter in island and weak grid operation scenarios. Torque analysis, which is well known for synchronous machine stability analysis, is adapted to uncover the stability enhancement effect of properly designed and parameterized outer-loop controls. Different types of loads are considered and investigated in the study. Electromagnetic transient (EMT) simulations are conducted on a test network to validate the analytical results. The results from this study indicate that with appropriate outer-loop voltage and frequency control design, the PLL based current-controlled inverter can be small signal stable with high inverter penetration, even in a 100% inverter-based system.

Particle swarm optimization (PSO) is envisioned as potential solution to overcome maximum power point tracking (MPPT) problems. Nevertheless, conventional PSO suffers from large transient oscillation, slow convergence and tedious parameter tuning when tracking global MPP (GMPP) under partial shading conditions (PSC), leading to poor efficiency and significant power loss. Therefore, a modified PSO hybridized with adaptive local search (MPSO-HALS) is designed as a robust, real-time MPPT algorithm. A modified initialization scheme that leverages grid partitioning and oppositional-based learning is incorporated to produce an evenly distributed initial population across P-V curve. Additionally, a rank-based selection scheme is adopted to choose best half of population for subsequent global and local search modes. A modified global search method with fewer parameters is devised to rapidly identify approximated location of GMPP. Finally, a modified local search method using Perturb and Observe with adaptive step size method (P&O-ASM) is proposed to refine the near-optimal duty cycle and track GMPP with negligible oscillations. MPSO-HALS is implemented into low-cost microcontroller for real-time application. Extensive studies prove the proposed algorithm outperforms bat algorithm (BA), improved grey wolf optimizer (IGWO), conventional PSO and P&O, with convergence time shorter than 0.3 s and tracking accuracy above 99% under different complex PSCs.