The number of smart inverters in active distribution networks is growing rapidly, making it challenging to realize a fast, distributed Volt/Var control (VVC). This work proposes a machine learning-assisted distributed algorithm to accelerate the solution of the VVC strategy. We first observe the convergence process of the Alternating Direction Method of Multipliers (ADMM)-based VVC problem and explore the potential relationships between the convergence and time-series regression. Then, the long short-term memory (LSTM) technique is applied to learn the convergence process and regress the converged values of the dual and global variables with previous ADMM observations. After that, the LSTM-assisted ADMM algorithm is proposed, where the regressions are used for ADMM parameter updates. In this algorithm, the inputs of the LSTM-model are carefully designed since the complementary conditions implied in the conventional ADMM should be considered. Unlike existing methods, the proposed method does not use the LSTM to determine the VVC strategy directly, indicating that it is non-intrusive and can satisfy all safety constraints during operations. The proof of its optimality and convergence is also given. The numerical simulations on the 33-bus distribution system demonstrate the effectiveness and efficiency of the proposed method.

Power imbalances between generation and consumption will cause frequency deviations in the grid. In modern power systems, intermittent renewable energy sources (RES) have resulted in more frequent frequency violations, as traditional power plants cannot compensate for power gaps timely. Electric vehicles (EVs) can participate in load frequency control (LFC) through aggregators and are capable of reacting faster to control commands than conventional frequency control reserves (FCR) in generators. Thus EVs hold great promise in assisting with LFC. This paper proposes a composite control scheme that fully utilizes EVs for LFC in both normal scenarios and contingencies. The designed droop control can greatly reduce instantaneous frequency deviations (IFD) in emergencies, while the tube model predictive control (Tube MPC) can ensure smooth frequency trajectories during normal operations. Based on realistic models, simulation results illustrate the effectiveness of the proposed method.

An increasing share of fluctuating and intermittent renewable energy sources can cause over-currents (OCs) in the power system. The heat generated during OCs increases the junction temperature of semiconductor devices and could even lead to thermal runaway if thermal limits are reached. In order to keep the junction temperature within the thermal limit of the semiconductor, the power module structure with heat-absorbing material below the chip is investigated through COMSOL Multiphysics simulations. The upper limits of the junction temperature for Silicon (Si) and Silicon Carbide (SiC) are assumed to be 175 and 250 ∘∘C, respectively. The heat-absorbing materials considered for analysis are a copper block and a copper block with phase change materials (PCMs). Two times, three times, and four times of OCs would be discussed for durations of a few hundred milliseconds and seconds. This article also discusses the thermal performance of a copper block and a copper block with PCMs. PCMs used for Si and SiC are LM108 and Lithium, respectively. It is concluded that the copper block just below the semiconductor chip would enable OC capability in Si and SiC devices and would be more convenient to manufacture as compared to the copper block with PCM.

The wide integration of inverter based renewable energy sources (RESs) in modern grids may cause severe voltage violation issues due to high stochastic fluctuations of RESs. Existing centralized approaches can achieve optimal results for voltage regulation, but they have high communication burdens; existing decentralized methods only require local information, but they cannot achieve optimal results. Deep reinforcement learning (DRL) based methods are effective to deal with uncertainties, but it is difficult to guarantee secure constraints in existing DRL training. To address the above challenges, this paper proposes a projection embedded multi-agent DRL algorithm to achieve decentralized optimal control of distribution grids with guaranteed 100% safety. The safety of the DRL training is guaranteed via an embedded safe policy projection, which could smoothly and effectively restrict the DRL agent action space, and avoid any violation of physical constraints in distribution grid operations. The multi-agent implementation of the proposed algorithm enables the optimal solution achieved in a decentralized manner that does not require real-time communication for practical deployment. The proposed method is tested in modified IEEE 33-bus distribution and compared with existing methods; the results validate the effectiveness of the proposed method in achieving decentralized optimal control with guaranteed 100% safety and without the requirement of real-time communications.

Hydrogen electrolyzers are promising tools for frequency regulation of future power systems with high penetration of renewable energies and low inertia. This is due to both the increasing demand for hydrogen and their flexibility as controllable load. The two main electrolyzer technologies are Alkaline Electrolyzers (AELs) and Proton Exchange Membrane Electrolyzers (PEMELs). However, they have trade-offs: dynamic response speed for AELs, and cost for PEMELs. This paper proposes the combination of both technologies into a Hybrid Hydrogen Electrolyzer System (HHES) to obtain a fast response for frequency regulation with reduced costs. A decentralized dynamic power sharing control strategy is proposed where PEMELs respond to the fast component of the frequency deviation, and AELs respond to the slow component, without the requirement of communication. The proposed decentralized approach facilitates a high reliability and scalability of the system, what is essential for expansion of hydrogen production. The effectiveness of the proposed strategy is validated in simulations and experimental results.

Deep reinforcement learning (DRL) is a promising solution for voltage control of distribution grids with high penetration of inverter-based renewable energy sources (RESs). Yet, when adopting the DRL-based control method, the safe and optimal operation of the system cannot be guaranteed at the same time, as the conventional DRL agent is not designed to solve the hard constraint problem. To address this challenge, this paper proposes a deep neural network (DNN) assisted projection based DRL method for safe control of distribution grids. First, a finite iteration projection algorithm is proposed to guarantee hard constraints by converting a non-convex optimization problem into a finite iteration problem. Next, a DNN assisted projection method is proposed to accelerate the calculation of projection and achieve the practical implementation of hard constraints in DRL problem. Finally, a DNN Projection embedded twin-delayed deep deterministic policy gradient (DPe-TD3) method is proposed to achieve optimal operation of distribution grids with guaranteed 100% safety of the distribution grid. The safety of the DRL training is guaranteed via the embedded Projection DNN in TD3 with participation in gradient return process, which could smoothly and effectively project the DRL agent actions into the feasible area, thus guaranteeing the safety of data driven control and the optimal operation at the same time. The case studies and comparisons are conducted in the IEEE 33 bus system to show the effectiveness of the proposed method.

Data manipulation attacks have become one of the main threats to cyber-physical direct current (DC) microgrids, but how to ensure voltage and current restoration under cyber attacks has not been well explored. In this paper, the event-based attack detection and mitigation problem for DC microgrids is considered. Specifically, an attack detection mechanism is designed to detect whether an attack has occurred. Then the proposed resilient secondary control strategy is only activated when the detection mechanism generates an attack event. For unknown types of attacks that aim at tampering with the information transmitted in the communication network, an adaptive linear quadratic regulator (LQR) based control strategy is designed to mitigate the effects such that the voltage and current restoration is achieved. Finally, the effectiveness of the proposed strategy is verified through simulationthis.

The alternating direction method of multipliers (ADMM) has been recently recognized as well-suited for solving distributed optimization problems among multiple agents. Nonetheless, there remains a scarcity of research exploring ADMM's communication costs. Especially for large-scale multi-agent systems, the impact of communication costs becomes more significant. On the other hand, it is well-known that the convergence property of ADMM is significantly influenced by the different parameters while tuning these parameters arbitrarily would disrupt the convergence of ADMM. To this end, inspired by the preliminary works on incremental ADMM, we propose a fast incremental ADMM algorithm that can solve large-scale multi-agent optimization problems with enhanced communication efficiency and fast convergence speed. The proposed algorithm can improve the convergence speed by introducing an extra adjustable parameter to modify the penalty parameter ? in both primal and dual updates of incremental ADMM. With several mild assumptions, we provide the convergence analysis of our proposed algorithm. Finally, the numerical experiments demonstrate the superiority of the proposed fast incremental ADMM algorithm compared to the other incremental ADMM-type methods.

In this paper, the real-time energy trading problem between the energy provider and the consumers in a smart grid system is studied. The problem is formulated as a hierarchical game, where the energy provider acts as a leader who determines the pricing strategy that maximizes its profits, while the consumers act as followers who react by adjusting their energy demand to save their energy costs and enhance their energy consumption utility. In particular, the energy provider employs a pricing strategy that depends on the aggregated amount of energy requested by the consumers, which suits a commodity-limited market. With this price setting, the consumers' energy demand response strategies are designed under a non-cooperative game framework, where a unique generalized Nash equilibrium point is shown to exist. As an extension, the consumers are assumed to be unaware of their future energy consumption behaviors due to uncertain personal needs. To address this issue, an online distributed energy trading framework is proposed, where the energy provider and the consumers can design their strategies only based on the historical knowledge of consumers' energy consumption behavior at each bidding stage. Besides, the proposed framework can be implemented in a distributed manner such that the consumers can design their demand responses by only exchanging information with their neighboring consumers, which requires much fewer communication resources and would thus be more suitable for the practical operation of the grid. As a theoretical guarantee, the proposed framework is further proved to asymptotically achieve the same performance as the offline solution for both energy provider and consumers' optimization problems. The performance of practical designs of the proposed online distributed energy trading framework is finally illustrated in numerical experiments.

Hybrid energy storage system (HESS) is effective to compensate for fluctuation power in renewables and fast fluctuation loads in DC microgrids. To regulate DC bus voltage, a power management strategy is an essential issue. In the meantime, the increasing integration of constant power loads (CPLs) in DC microgrids brings great challenges to stable operation due to their negative incremental impedance. In this paper, a fast composite backstepping control (FBC) method is proposed for the HESS to achieve faster dynamics, smaller voltage variations, and large-signal stabilization. In the FBC method, a higher order sliding mode observer (HOSMO) is adopted to estimate the coupled disturbances. Furthermore, the FBC method is integrated with the droop control; so that the FBC-based decentralized power allocation (FBC-DPA) strategy for HESS in DC microgrids is developed. The proposed FBC method is designed based on the Lyapunov function to ensure its stability. Moreover, the design guidelines are provided to facilitate the application of the proposed method. Both simulation and experimental studies under different operating scenarios show that the proposed method achieves faster voltage recovery and smaller voltage variations than the conventional backstepping control method.