A stable policy network for multi-agent reinforcement learning in voltage control of power systems

Digital-analog dual drive for smart voltage control of highly volatile renewable energy systems

Resilient and privacy-preserving distributed dynamic event-triggered secondary voltage control of islanded AC microgrids

Emergent dynamics, which arise from local interactions between many elementary components, are central to complex physical and biological systems. However, realizing such dynamics in artificial materials, particularly under low-energy stimuli, remains a challenge. While dipole-dipole interactions are typically suppressed in magnetic storage, here we amplify and use them as a core mechanism to construct a strongly dipolar-coupled network of SmCo5 macrospins at the wafer scale, which can exhibit intrinsic interaction-driven collective dynamics in response to voltage pulses. This network integrates three key ingredients: the strong dipole-dipole interaction, giant voltage control of coercivity over nearly 1000-fold, and a network topology with a frustrated Ising-like energy landscape. As a result, the network, when stimulated by ∼1 V pulses, transitions from a high-coercivity memory regime into a low-coercivity regime in which internal dipolar fields alone trigger collective magnetic reconfiguration. In this regime, the network exhibits emergent behaviors absent at the level of isolated macrospins, including spontaneous demagnetization, greatly enhanced magnetization modulation, reversible "freeze and resume" evolution, and stochastic convergence toward low-energy magnetic configurations. Furthermore, as a conceptual illustration, micromagnetic simulation of such a strongly dipolar-coupled network shows that the resulting high-dimensional collective dynamics can support temporal information processing, such as accurate chaotic Mackey-Glass prediction and multiclass drone-signal classification. Our work suggests a conceptually distinct route toward scalable, energy-efficient neuromorphic computing, one rooted in local physical interaction-driven emergent dynamics at the network level rather than merely mimicking individual neurons and synapses.

This paper investigates control strategies for large-scale offshore wind power high-voltage direct current (OWP-HVDC) transmission systems employing a series-connected diode rectifier unit and modular multilevel converter (DRU-MMC) topology. First, a comprehensive onshore-offshore coordinated reduced-voltage startup strategy is proposed. By leveraging the bidirectional conduction capability of the MMC, this method enables smooth system energization and safe grid connection of WTGs. Furthermore, to enhance steady-state performance under wind power variations, an enhanced DC voltage control strategy is developed for the offshore MMC, supported by a detailed small-signal stability analysis and rigorous validations under DC and AC grid disturbances. A comparative analysis is then conducted among three distinct grid-forming control schemes: fixed AC voltage control, active power control, and the proposed DC voltage control. Simulation studies on the large-scale OWP-HVDC system based on the hybrid converter reveal that fixed AC voltage control fail to properly regulate DC voltages of the hybrid converter, resulting in severe MMC DC voltage ripples of up to 14.4 %. In contrast, the proposed DC voltage control strategy precisely stabilizes the MMC DC voltage, limiting the ripple to a mere 2.9 % while maintaining bus voltage fluctuations within acceptable 1.3 %. These results indicate that DC voltage control is the most robust grid-forming strategy for series-connected DRU-MMC systems, offering superior power allocation and enhanced protection against over-modulation.

Cu-based Prussian blue analogues (PBAs) are promising material platforms for selective ammonium intercalation due to their vacancy-rich framework. However, their limited cycling stability has hindered practical applications. Here, we design core-shell heterostructured PBAs using copper hexacyanoferrate (CuHCF) as an ammonium-selective core, with a nickel hexacyanoferrate (NiHCF) as an outer shell that enhances electrochemical stability. The nanostructural design of these mixed core-shell particles preserved ammonium selectivity while extending cycling longevity. The CuHCF@NiHCF nanoparticles were synthesized by a two-step coprecipitation, with the NiHCF shell thickness precisely controlled to 10 and 20 nm. Both core-shell particles retained over 98% capacity after 1000 charge-discharge cycles at 1 A g-1. Electrosorption with the 10 nm shell particles achieved a high separation factor of 9.2 for NH4+ over Na+ with 0.42 mmol g-1 NH4+ uptake, whereas the 20 nm shell particles exhibited a lower selectivity of 3.9. Furthermore, the selectivity of the 10 nm shell particles was enhanced to exceed 20 through voltage control. Electrosorption in a flow cell using the 10 nm shell particles demonstrated efficient NH4+ extraction from municipal wastewater, enriching the molar % of NH4+ among total cations from 29% to 61%. This work establishes shell thickness control as a strategy to couple high NH4+ selectivity with stability for sustainable resource recovery.

Navigating the complexities of modern power systems with a high share of renewable energy sources (RES) requires effective control strategies. This research presents a new Coordinated Voltage Control (CVC) framework for RES-integrated grids by combining Ant Colony Optimization (ACO) with a Deep Q-Network (DQN) controller. In this hybrid approach, ACO optimizes voltage profiles while DQN adjusts reactive power support in real-time, allowing for quick responses to grid conditions. The proposed method significantly improves voltage stability and reduces system losses, showing strong performance under different renewable generation and load scenarios. The ACO and DQN framework also shows good computational efficiency, achieving convergence in about 45 to 50 iterations. By using historical and simulated data for DQN training, the controller predicts the best reactive power actions, ensuring scalable and reliable voltage regulation. This work provides a practical and flexible solution for modern power systems rich in renewables, supporting better grid stability and operational efficiency.

ABSTRACT Closed-loop control of dielectric barrier discharge (DBD) loads remains challenging because the load voltage is usually difficult to measure directly in practical high-voltage systems, while the discharge behavior is strongly nonlinear and operating-condition dependent. To address this problem, this paper proposes a control-oriented linearized observation model for estimating the peak load voltage from the peak load current within a specified operating range. First, an equivalent electrical model of the DBD load is established based on the microscopic discharge processes, including capacitive behavior before breakdown, rapid ionization after gas-gap breakdown, and charged-particle recombination during current decay. Based on this model, the relationship between the peak load voltage and the peak load current is investigated through numerical analysis and experimental validation. The results show that, under fixed load structure and limited operating conditions, the peak voltage and peak current exhibit an approximately linear relationship. The proposed model is then applied as a voltage observer in the closed-loop voltage control of a DBD ozone generator. Experimental results demonstrate that the model can estimate the peak load voltage with acceptable accuracy in the tested operating range and can support stable closed-loop control. It should be noted that the model coefficients need to be re-identified when the DBD load structure, gas condition, or operating range changes significantly.

Interconnected power systems are frequently exposed to load disturbances, parameter uncertainties, nonlinear effects, and dynamic coupling between frequency and voltage regulation loops, which make it difficult to maintain stable and well-coordinated operation. Conventional control strategies may suffer from slow damping, large oscillations, and reduced robustness when applied to such highly coupled environments. To address this problem, this paper proposes a novel cascaded controller for the coordinated regulation of Load Frequency Control (LFC) and Automatic Voltage Regulation (AVR) in a two-area interconnected power system. The proposed controller combines a two-degree-of-freedom proportional-derivative controller with filtered derivative action (2DOF-PDf) and an inner leaky tilt-integral (LTI) stage to improve transient shaping, damping, and robustness while reducing sensitivity to disturbances and parameter variations. The controller parameters are optimally tuned using the Crayfish Optimization Algorithm (CrOA) based on the Integral of Time-weighted Squared Error (ITSE) criterion. The effectiveness of CrOA is first verified through comparison with several well-known optimization methods, including the Chimp Optimization Algorithm (ChOA), Dingo Optimization Algorithm (DOA), Sine Cosine Algorithm (SCA), Gorilla Troops Optimizer (GTO), and Gradient-Based Optimizer (GBO). The proposed control scheme is then assessed under various operating scenarios, including step load disturbances, stochastic load variations, practical nonlinearities such as Generation Rate Constraint (GRC) and Governor Dead Band (GDB), time-varying voltage reference tracking, and parametric uncertainty of up to ± 40%. MATLAB/Simulink results demonstrate that the proposed CrOA-tuned cascaded (2DOF-PDf)-(LTI) controller provides faster damping, smaller frequency and tie-line power deviations, and more accurate terminal-voltage regulation than several benchmark controllers, including Proportional-Integral-Derivative (PID), Tilted Integral Derivative (TID), Fractional-Order Proportional-Integral-Derivative (FOPID), and Fractional-Order Proportional-Integral with Proportional-Integral-Double Derivative Squared (FOPI-PIDD2) controllers. Overall, the proposed controller offers a robust and effective solution for coordinated frequency and voltage regulation in interconnected power systems.

Multimode GaN transistors achieve multimodality by exploiting a broadly tunable and ultrawide threshold voltage (VTH) that covers both depletion-mode (D-mode) and enhancement-mode (E-mode) operation. This VTH tunability is enabled by epitaxial growth of ultrathin AlN barriers with high interface quality and strong electrostatic control, combined with tensile stress introduced by PECVD-SiN passivation. Through stress-strain-polarization coupling (SSPC), strain along [0001] approaching -4.2% was generated, which increased the 2DEG density by more than 1 order of magnitude and thus enabled an ultrawide VTH window. The evolution among background-carrier, MOS-like, and D/E-mode high-electron mobility transistor (HEMT) conduction modes under different passivation thickness (PT) was further clarified by simulations, accounting for the observed multimode behavior. SSPC was thus supported as a compact and low-damage method for etch-free tuning of polarization, carrier density, and VTH. The continuously tunable conduction modes provide multimode GaN transistors multiple functionalities, enabling applications in RF amplification, multimode neuromorphic computing, efficient power conversion, and complex circuit optimization.

The accelerating transition toward sustainable transportation has led to a rapid deployment of Plug-in Hybrid Electric Vehicles (PHEVs), introducing significant operational challenges for active distribution networks, particularly in terms of voltage regulation and network flexibility. High and spatially concentrated charging demand, combined with stochastic vehicle behavior, can substantially reduce voltage headroom and compromise grid integrity if not properly managed. To address these challenges, this paper proposes a novel Active Distribution Network Management (ADNM) framework based on the Voltage Network Flexibility Index (VNFI) for coordinated PHEV charging and discharging. The VNFI is employed as an actionable steering signal to identify voltage-critical buses and time periods, enabling flexibility-aware scheduling decisions under strict network-security constraints. Stochastic PHEV arrival, departure, and energy demand are modeled using probabilistic distributions, and the proposed framework is implemented and validated through a high-fidelity MATLAB–OpenDSS co-simulation on a modified IEEE 33-bus distribution system. Numerical results demonstrate that the VNFI-driven coordination improves the voltage flexibility margin by up to 44.2% and reduces total power losses by 29.4% compared with a conventional TOU-based charging strategy. Moreover, even under 100% PHEV penetration, the maximum voltage deviation remains within 0.055 p.u., confirming the robustness and scalability of the proposed approach. The results highlight the effectiveness of VNFI-based management in transforming PHEVs into flexibility resources for future smart grid operations. • The study examines the operational challenges posed by electric vehicles (EVs) on distribution systems, focusing on voltage control and power losses. • A probabilistic model for aggregating plug-in hybrid electric vehicles (PHEVs) is proposed, based on parameters derived from the National Household Travel Survey (NHTS). • The concept of voltage flexibility is introduced, with the evaluation of a newly proposed index to assess it. • A smart charging/discharging approach is developed for optimal PHEV scheduling within active distribution networks (ADNs), addressing power demand throughout different hours. • A nonlinear optimization method for mixed integers is used to formulate and solve the scheduling problem. • The proposed method is tested on an IEEE 33-bus distribution system, demonstrating its effectiveness and validating the proposed index against previously reported strategies.

Active magnetic bearing (AMB) systems serve as a key enabling technology for high-speed rotating machinery. As power ratings continue to rise, the DC-link voltage, phase current, and physical dimensions of the power converter and cabling also increase, which results in larger common-mode voltage (CMV) swings and parasitic capacitive currents. Consequently, power–signal coupling issues, especially common-mode electromagnetic interference (CM EMI), become more severe and have attracted growing research attention in practical applications. In this paper, we propose a zero common-mode voltage (ZCMV) modulation and control strategy combined with an active neutral-point (NP) voltage balancing scheme for the power electronic converter of a high-power AMB system. The proposed solution is implemented on a four-leg three-level half-bridge neutral-point-clamped (NPC) topology, which is used to control a single-plane, two-degree-of-freedom AMB and can be readily extended to multi-DOF magnetic bearing systems. The ZCMV modulation strategy significantly suppresses common-mode EMI and reduces interference with position sensors in the AMB control loop, thereby enhancing position control accuracy. Experimental results from a high-load-capacity AMB motor testbed demonstrate the effectiveness of the proposed approach.

Hybrid PSO-MPC-based dynamic tuning of battery management parameters for enhanced lithium-ion battery performance in electric vehicles

Grid-forming control in Type-IV wind turbine generators can exacerbate drivetrain torsional oscillations due to rapid torque variations from DC-link voltage regulation. This paper proposes an active power control-based damping strategy that introduces damping torque indirectly via the grid-side converter, thereby avoiding direct interference with the machine-side DC-link voltage control action. A frequency-domain modeling framework is developed to guide a structured controller design procedure. In this process, the torsional mode is extracted from the rotor speed using a band-pass filter, the phase lag is compensated with a lead element, and the damping gain is determined iteratively using frequency-domain guidelines to balance damping performance and stability margins. Comparative analysis shows that the proposed method reduces the torsional oscillation envelope decay time to about 3 s, whereas existing methods exhibit slower decay and achieve over a 50% reduction in DC-link peak deviations under both phase-jump and voltage-dip scenarios. Real-time experiments on a 2<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\sim ~$</tex-math></inline-formula>kW hardware setup complement the simulation studies, and an additional wind farm-level case study confirms the scalability of the approach beyond the single-machine infinite bus assumption, providing strong practical validation of the proposed damper.

Series-connected SiC MOSFETs is considered a promising approach for medium-voltage (MV) DC applications, providing an effective balance between cost and performance. In particular, based on quasi-two-level (Q2L) modulation strategy, the hybrid-clamping structure with indirect series connection is attractive due to its inherent voltage self-balancing capability, eliminating the need for additional voltage control loop. However, the issue that the switched-capacitor loop introduced for self-balancing can cause inrush current and additional turn-on switching loss in certain device has not been analyzed, which limits the potential of this topology. In this article, the switching process of the device within the switched-capacitor loop is first analyzed, clarifying the mechanisms behind the inrush current and additional losses. Then, an improved hybrid-clamping topology is proposed, which maintains the self-balancing capability while reducing inrush current and turn-on switching loss. In the end, 2000 V Double Pulse Tests (DPTs) demonstrate effective mitigation of inrush current to the level of the load current, along with an average 23% reduction in turn-on switching loss. Furthermore, continuous experiment validated the effectiveness of the proposed method.

Enhancing UPFC compensation performance in simultaneous power and voltage control of the system with self-tuning PI controllers based on ANFIS design

Free-standing magnetic field energy harvesters (FSMFEH) have proven potential for powering sensors on power transmission lines. To realize the application FSMFEH, a key challenge is to satisfy the power supply requirements of sensors, i.e., the output power and the constant output voltage requirements. However, the output power and voltage of FSMFEH systems are highly dependent on the power line current, which typically varies over a wide-range. Therefore, the stabilized power supply to the sensor will be broken, and the optimal output power will not be maintained. To address these issues, an energy management circuit with dual-coil FSMFEH and bidirectional DC-DC is proposed in this work. Firstly, the influence of power line currents and the constant voltage output to the system are ana-lyzed. Then, the optimal operating modes of the system under different load conditions are clarified. Finally, the energy management strategies are presented based on the mode switching boundaries of the dual-coil system and the voltage control of the bidirectional DC-DC. Experimental results show that when the power line current varies from 30A to 500A, the output power can be improved from 0.6mW to 3.2mW at the minimum power line current, and from 195mW to 365mW at the maximum power line current.

Nonlinear Passivity-Based Dual-Loop Voltage Control with Reduced Sensor Requirement for Three-Phase Voltage Source Inverters

Partial shading conditions (PSC) greatly reduce the performance of photovoltaic (PV) systems by adding many local maxima to the PV characteristics. This makes it very hard for traditional maximum power point tracking (MPPT) methods to work. To solve this problem, this research presents a comparative analysis of five hybrid MPPT algorithms: PO-Fuzzy, PSO-PO, GWO-Fuzzy, PSO-Fuzzy, and GWO-PSO, executed with a SEPIC converter functioning at 10kHz with a 250W PV module. We use MATLAB/Simulink to test how well these methods work in different partial shading situations. It examines the important elements, including PV voltage, PV power, output voltage, and output power. The results indicate that whereas PSO-PO and GWO-PSO enable faster initial convergence, the GWO-Fuzzy method achieves superior tracking accuracy, reduced steady-state oscillations, and reliable voltage control. The proposed GWO-Fuzzy approach ensures effective and reliable tracking of the global maximum power point (GMPP) with variations in sunlight intensity.

ABSTRACT We present three X‐band transmit/receive (T/R) switches utilizing single‐pole double‐throw (SPDT) structures in 250 nm GaN HEMT technology. The first T/R switch was designed to achieve low insertion loss by constructing the SPDT structure with a series λ/4 transmission line (T‐line) and meticulously selected two HEMT switches in parallel for each transmit and receive path. Based on the proposed low‐loss architecture, the second T/R switch integrated 2 nd and 3 rd harmonic rejection notches to suppress harmonic distortions effectively. The third T/R switch was implemented to operate with a positive control voltage, achieved by employing a positive bias voltage at the source of the HEMT with AC coupling capacitors at the drain and source of the normally‐on HEMT. Measurements demonstrated that the fabricated three T/R switches achieved insertion losses of 0.44 , 0.54 , and 0.66 dB, corresponding IP 0.2dB of 44 , 44 , and 43 dBm, respectively. With a compact die size of 5.4 mm 2 , the three T/R switches exhibited excellent RF performance, surpassing the recently reported novel X‐band SPDT switches in 250 nm GaN HEMT.