Resilience becomes vital for power grids facing the increasingly frequent extreme weather events. Microgrid formation is a promising way to achieve resilient distribution networks (RDN) when the utility power is unavailable. This paper proposes a RDN-oriented microgrid formation (RoMF) method based on the deep reinforcement learning (DRL) technique, which integrates the OpenDSS as an interaction object and searches for optimal control policies in a model-free fashion. Specifically, we formulate the microgrid formation problem as a Markov decision process, taking into account complex factors such as unbalanced three-phase power flow and microgrid operation constraints. Next, a simulator-based RoMF environment is constructed and integrated into the OpenAI Gym, providing a standard agent-environment interface for applying DRL algorithms. Then, the deep Q-network is used to search for optimal microgrid formation strategies, and an offline-training and online-application framework of the DRL-based RoMF is given. Finally, extensive numerical results validate the effectiveness of our proposed method.

Due to the energy transition process, distribution systems will feature a high penetration of distributed renewable energy sources (RESs). The multiple distributed generation can provide emergency power supply to critical loads against blackouts caused by natural disasters and malicious attacks. However, the uncertainty of RESs, the control mode variation of RESs together with energy storage systems (ESSs), and the interaction among distribution system operator (DSO) and RESs add increasing difficulties to load restoration decisions. This paper applies the coordination control strategy to enable RESs to regulate the frequency and voltage during restoration process. Then, distributed energy resource management systems (DERMSs) and its aggregation characteristic are studied in load restoration. Finally, a two-step scenario-based stochastic optimization with the DSO-DERMS interaction framework is formulated to utilize RESs for resilient distribution systems. At last, the proposed critical load restoration optimization method is validated on a modified IEEE 37 and 123 node test feeder to verify the effectiveness.

Radiality constraints are widely embedded in optimization problems associated with distribution systems. In this letter, a novel mathematical formulation of radiality constraints is proposed based on the maximum density, and proved to be the necessary and sufficient condition for radial topology. Compared with the existing formulation, the number of involved binary variables and constraints can be reduced, and thus the computational performance of related optimization problems can be improved. The proposed constraints are specifically applied to the resilient reconfiguration problem, and numerical results on 33-, 123-, 201-, 873-node test systems verify the effectiveness of the proposed constraints.

Increasingly frequent natural disasters and man-made malicious attacks threaten the power systems. Improving the resilience has become an inevitable requirement for the development of power systems. The importance assessment of components is of significance for resilience improvement, since it plays a crucial role in strengthening grid structure, designing restoration strategy, and improving resource allocation efficiency for disaster prevention and mitigation. This paper proposes a component importance assessment approach of power systems for improving resilience under wind storms. Firstly, the component failure rate model under wind storms is established. According to the model, system states under wind storms can be sampled by the non-sequential Monte Carlo simulation method. For each system state, an optimal restoration model is then figured out by solving a component repair sequence optimization model considering crew dispatching. The distribution functions of component repair moment can be obtained after a sufficient system state sampling. And Copeland ranking method is adopted to rank the component importance. Finally, the feasibility of the proposed approach is validated by extensive case studies.

Load rebound characteristics (LRC), which means that restoration of distribution systems with large thermostatically controlled load (TCL) after an outage may create significantly higher load demand than normal (namely the demand if this outage does not occur), will make the reliability index-expected energy not supplied (EENS) of distribution systems obtained by traditional methods pessimistic and inaccurate. Since EENS is one of the base points of investment planning for distribution system operator, it is necessary to consider the LRC in the reliability evaluation and provide a more accurate evaluation on EENS. This paper proposed a new reliability evaluation method of smart distribution systems based on a nonnegative k-singular value decomposition (NN-K-SVD) algorithm and a sequential Monte Carlo (SMC) simulation. The NN-K-SVD algorithm is used to identify the TCL in total load. A reliability model considering the LRC is established based on an electric water heater model. The SMC simulation is used to mimic the operation of distribution systems based on the established reliability evaluation model, and calculate reliability indices. Furthermore, the feasibility of the proposed method is validated by extensive cases studies.

The frequent occurrence of natural disasters and malicious attacks has exerted unprecedented disturbances on power systems, accounting for the extensive attention paid to power system resilience. Combined with the evolving nature of general disasters, this paper proposes resilience assessment approaches for power systems under a tri-stage framework. The pre-disaster toughness is proposed to quantify the robustness of power systems against potential disasters, where the thinking of area division and partitioned multi-objective risk method (PMRM) is introduced. In the case of information deficiency caused by disasters, the during-disaster resistance to disturbance is calculated to reflect the real-time system running state by state estimation (SE). The post-disaster restoration ability consists of response ability, restoration efficiency and restoration economy, which is evaluated by Sequential Monte-Carlo Simulation to simulate the system restoration process. Further, a synthetic metric system is presented to quantify the resilience performance of power systems from the above three aspects. The proposed approaches and framework are validated on the IEEE RTS 79 system, and helpful conclusions are drawn from extensive case studies.