Harmonic coupling and the stability of modular multilevel converters (MMCs) have been extensively studied under balanced conditions. However, in practical applications, MMCs may operate under unbalanced conditions due to fac-tors such as faults and protection actions. This paper investi-gates the harmonic coupling and stability of MMC under sin-gle-phase disconnection (SPD) conditions, a scenario that, to the best of the authors' knowledge, has been relatively under-explored in the existing literature. To address this gap, the paper develops a comprehensive harmonic state-space model for an MMC-HVDC system under SPD condition. Using this model, a multi-input multi-output harmonic coupling matrix is derived to analyze the interactions between different harmon-ics and the signal-flow graph is depicted to explain the mech-anisms behind these couplings. Furthermore, the multi-input multi-output matrix is transformed into a single-input single-output impedance that encapsulates all harmonic couplings, thereby enabling the visualization of the effects of various couplings on system stability through impedance variations. Nyquist curve based stability analysis, along with the hard-ware-in-the-loop experiments, underscores the importance of incorporating SPD conditions in the parameter design of MMC-HVDC systems.

The integration of renewable energy (RE) systems into data center presents a critical pathway against high electricity costs and carbon emissions. However, this transition faces huge challenges of source-load compatibility and economic feasibility. This study proposes an integrated multi-stage framework addressing three-dimensional synergies among site selection, multi-energy dispatch, and transmission planning for RE-powered data centers. First, a geospatial evaluation methodology is developed to identify optimal data center locations through multi-criteria analysis. Second, a multi-energy complementary dispatch model is formulated to coordinate wind, solar, hydroelectric resources with dynamic computing loads. Third, a techno-economic comparison is conducted between direct-current and alternating-current transmission systems employing life-cycle cost analysis. A case study in Liangshan Prefecture in China demonstrates that the optimized solution achieves 100% renewable energy utilization rate, and identifies direct-current transmission as more economically viable for 110 kV medium-voltage applications. These findings provide actionable insights for policymakers and infrastructure planners seeking to balance sustainability goals with economic feasibility in RE-powered data center deployments.

The operational stability of isolated industrial and mining microgrids is critically challenged by the diverse response characteristics of heterogeneous flexible resources (FRs), such as varying ramp rates and response times. Conventional scheduling methods, which often overlook these multi-granularity attributes, result in significant power imbalances, mismatched execution of the dispatch commands, and degradation of industrial product quality. To address this gap, this paper proposes a novel dispatch command disaggregation strategy for the precise coordination of these resources. The strategy's core innovation begins with establishing a pioneering quantitative flexibility model for lithium mine loads (LMLs), mechanistically exploring the thermo-electric dynamics of the salt-lake lithium extraction process. This specific model is then integrated within a unified framework designed to systematically characterize and quantify the multi-granularity flexibility attributes of diverse resources. Finally, a Discrete Choice Model (DCM) is employed to optimize the dispatch priority of FRs, effectively translating the microgrid's aggregate dispatch command into individualized and feasible setpoints. Validated on a real-world microgrid in Xizang, the proposed strategy reduces system power deviation by 89.05% and decreases the total operating cost by 25.25%. This work provides a practical framework for enhancing the control precision and economic efficiency of isolated microgrids through effective coordination of diverse flexibility assets.

The sending-end system of line-commutated converter based high-voltage direct current (LCC-HVDC) systems is vulnerable to transient voltage disturbances (TVDs), posing a significant threat to voltage stability. This paper proposes a novel strategy to maximize the dynamic voltage support (DVS) capability of LCC-HVDC systems under various TVDs. The physical mechanisms underlying DVS in LCC-HVDC systems are systematically analyzed, forming the basis for an optimization model that maximizes the DVS capability while incorporating security constraints at both the rectifier and inverter ends. To address the challenge of directly solving the model, an optimality analysis with intuitive geometric interpretations is performed. Based on these insights, a two-stage optimal DVS control strategy for LCC-HVDC systems is developed to iteratively approach the optimal solution through coordinated control of the rectifier and inverter stations. The effectiveness and superiority of the proposed strategy in supporting the sending-end system are validated through dynamic simulations, and its applicability under practical operating conditions is discussed.

Deep learning (DL) holds significant potential for distinguishing forced oscillations (FOs) from natural oscillations (NOs) and resonances in power systems. However, most existing studies treat DL models as opaque system classifiers, and these models are often susceptible to challenges such as class imbalance among the three oscillation patterns (i.e., FOs, NOs, and resonances) and the limited size of real-world training datasets. These issues can lead to incorrect diagnoses of oscillation patterns in practical applications. To address these challenges, we propose a FO oscillation recognition framework that integrates balanced data processing, an interpretable recognition network, and a fine-tuning learning strategy. In this framework, DeepSMOTE1D is introduced to address the issue of class imbalance by oversampling features extracted from an encoder-decoder network. The interpretable DWT-CNN-LSTM recognition network incorporates multiple discrete wavelet transforms (DWT) into each convolutional and LSTM layer to enhance both interpretability and performance of the model. Finally, a fine-tuning transfer learning strategy is employed, where the recognition network is first pretrained on simulation data and subsequently retrained on limited real-world data. The effectiveness of DeepSMOTE1D and the fine-tuning strategy in the proposed framework is also evaluated. Experimental results show that the proposed recognition model not only learns effectively from simulated oscillation data but also outperforms existing related networks in real-world scenarios, achieving superior accuracy, precision, recall, and F1-score.

In power systems with clustered renewable power plants (RPPs), global coordinated transient voltage support (TVS) for the RPP clusters under fault conditions is crucial for safeguarding the power quality of critical sensitive loads and ensuring RPP fault ride-through. However, conflicting multi-objective demands and dynamic fault evolution pose significant challenges. Existing strategies often focus on local control or slow-dynamic global regulation, and are unable to simultaneously meet the requirements for system-wide coordination and rapid response. To address these challenges, this study proposes a global coordinated TVS strategy for clustered RPPs considering fault evolution. First, a global current-voltage transfer analytical model considering RPP coupling is established to rapidly map multi-source injections to node voltage response. This approach eliminates timeconsuming iterative calculations, ensuring efficient strategy optimization across massive fault scenarios. Second, a coordinated optimization model based on Pareto hierarchical optimization is developed, resolving the difficult trade-off between critical load TVS and RPP ride-through. Additionally, a fault scenario deduction model is introduced to dynamically correct the strategy, achieving optimal support throughout the fault process. The strategy is computed offline and integrated into RPP controllers to enable rapid response. Validations demonstrated that the proposed strategy significantly enhances the TVS for global nodes and mitigates the risk of cascading RPP disconnections.

With the rapid advancement of power electronics, active power filters (APFs) have emerged as a preferred solution for mitigating harmonic resonance in high-voltage direct current (HVDC) systems. APFs can be broadly categorized into current compensation APFs (I-APFs) and resistive APFs (R-APFs), depending on their control strategies. Although both types have demonstrated effectiveness in practical applications, previous studies have not comprehensively compared their resonance mitigation performance, particularly under capacity-limited conditions. To address this gap, this paper establishes an equivalent harmonic model for LCC-HVDC systems with APFs connected and proposes a novel metric, capacity utilization efficiency (CUE), to compare the resonance suppression effectiveness of APFs. The variation characteristics of CUE are systematically analyzed considering factors such as resonance frequency, system damping, and APF capacity. The findings provide a practical guideline to assist utilities in selecting the optimal APF control, accounting for factors such as varying network impedances and background harmonic distortions. Finally, the developed harmonic model and the APF-selection guideline are validated using a real-life LCC-HVDC system experiencing resonance issues and a modified IEEE 39-bus system.

Wind generation is required to provide dynamic voltage support (DVS) during fault ride-through (FRT) to enhance system voltage stability. However, DVS causes the FRT behavior of wind generation to be segmented, and its impact on voltage stability in the wind farm (WF) integrated linecommutated converter-based high-voltage direct current (LCCHVDC) sending-end system remains unclear. To achieve a quantitative analysis to this gap, a novel quasi-static piecewise model for the sending-end system is developed, in which the segmented FRT behavior of WF is fully incorporated. Beyond the well-known saddle-node bifurcation (SNB), a new FRT-induced nonsmooth bifurcation (FRTNB), is identified. The influence of WF and the LCC-HVDC system on both SNB-related and FRTNBrelated voltage stability is thoroughly analyzed. Based on the acquired insights, a control strategy is proposed for the LCCHVDC system to proactively utilize its DVS capability, which is adaptive to the segmented FRT behavior of WF and can effectively enhance system voltage stability without requiring knowledge of grid parameters. Dynamic simulations validate the theoretical analysis and the effectiveness of the proposed method under both voltage sag and swell conditions

The transformer inrush current has been a potential threat in wind farms connected modular multilevel converter based high-voltage direct current (WF-MMC-HVDC) system due to the low overcurrent capability of power electronic devices. To investigate this issue, this paper develops a complete harmonic state space (HSS) model of the WF-MMC-HVDC system containing saturable transformers. The severity of the inrush current is investigated under different transformer configurations and the result is compared with EMTP simulations. More importantly, key factors that influence inrush current characteristics in a WF-MMC-HVDC system are studied using the single-input single-output impedance model derived from the linearized HSS model. The results indicate that wind farms have a minor impact on the inrush current characteristics, whereas V/ F controlled modular multilevel converter (MMC) reduces its output voltage during transformer energization, thereby mitigating the severity of the inrush current. The severity of the in-rush current largely depends on the resonance point determined by the transmission line. In the case of offshore WFMMC-HVDC system, long submarine cables may cause severe harmonic amplifications and even do not attenuate for a long time.

Tree-related high-impedance faults (THIFs) represent a critical risk source for wildfires and power outages in distribution systems. Existing HIF models do not jointly consider the time-varying resistance of trees and the controllable nonlinear behavior of arcs. Consequently, they cannot support fault characteristic analysis across all time scales, thereby limiting the research and validation of THIF identification algorithms. To address this limitation, this paper presents a dynamic model for THIFs under sustained contact conditions that accurately characterizes fault currents over multiple time scales. First, extensive THIF experiments elucidate the failure mechanism and the evolution of current-waveform distortion characteristics. A piecewise-linear function is employed to represent the time varying electrical resistance of trees. Based on arc thermal equilibrium theory, a THIF arc model with controllable distortion characteristics is proposed, enabling independent regulation of current waveform distortion features through three physically meaningful parameters. This model can reproduce the envelope variations of fault currents over extended time scales and demonstrates superior fitting accuracy for local distortions within short time windows compared to existing models, significantly enhancing the ability to characterize THIF current behavior. Parameter ranges for the model's key variables are established from extensive experimental data, and the model serves as a reference for THIF inversion in simulations without actual measurement data. These findings provide a modeling foundation for early identification and intelligent detection of THIF in distribution systems, as well as for wildfire risk assessment.

In power distribution systems with multiple time-varying harmonic sources, to address the limitation in the accuracy of harmonic power flow (HPF) calculations due to the difficulty in accurately obtaining harmonic source and system models, this paper proposes a high-accuracy time-varying probabilistic harmonic power flow (PHPF) calculation method based on iterative correction of the harmonic coupling matrix model (HCMM). First, the harmonic-coupled HPF equation of the system is established. Second, the HCMM is iteratively corrected using time-series measurements of node voltages and currents, and adaptive time-segmentation is performed according to the variation characteristics of harmonic source voltages and currents. Finally, to overcome the limitations of the independence assumption and symmetric distribution assumption, the point estimation method is improved based on Nataf–Cholesky transform decoupling and asymmetric sampling, achieving high-accuracy PHPF calculation. In the IEEE-33 and IEEE-123 bus test systems, the relative errors of voltage total harmonic distortion obtained by the proposed method are below 1%, significantly better than the traditional PHPF calculation method (about 10%), with an accuracy improvement of about 10 times. In addition, the actual distribution system is used to verify the performance of the proposed method.

This paper presents a new method for asymmetric fault location in distribution network with distributed generations based on the voltage at secondary side of distribution transformers. The ratio of positive and negative voltage is defined as the fault feature to achieve the estimation of fault section and the fault distance. Firstly, the relationship between positive sequence voltage and negative sequence voltage in the fault condition was analyzed with the variation of line distance. Secondly, fault section identification depends on the changes in the defined feature ratio upstream and downstream of the fault point. Thirdly, the fault distance calculation is performed by solving systems of determined equations. Finally, the proposed fault location method is validated on the modified IEEE 34-bus and IEEE 134-bus distribution network with or without distributed generations. Simulation results show the robustness and accuracy of the method under several scenarios with fault resistance, fault types, neutral grounding modes and measurement errors.

Voltage sags can lead to substantial economic losses, making accurate source location the critical first step toward effective mitigation. Given that the absence of detailed network parameters constrains the use of impedance-based approaches, this study explores deep learning-based alternatives for source location. To address the challenge of limited training samples—arising from sparse measurements in distribution networks—this work proposes a knowledge transfer framework employing a collaborative training strategy for voltage sag source location (VSSL). An interpretability-enhanced graph convolutional neural network (IIGCN) is developed, using the voltage magnitude depth difference (VMDD) derived from recorded waveforms as input features. A cross-region knowledge transfer framework is then introduced to capture general mapping relationships between VMDD and corresponding source positions. IIGCN models for different subregions are trained collaboratively, with parameter exchange and compensation mechanisms ensuring robust performance despite limited training data. The proposed approach is evaluated on both synthetic and real-world systems, achieving about 85% accuracy under stringent training conditions. Comparative analyses demonstrate that the method consistently outperforms existing techniques, delivering superior VSSL performance across diverse application scenarios.

Virtual synchronous generator (VSG) control plays a critical role in enhancing grid inertia and providing voltage support. However, during grid faults, its neglect of phase angle jumps (PAJ) can lead to a sudden increase in the phasor difference between the internal electromotive force (EMF) and the point of common coupling (PCC) voltage, resulting in device overcurrent and inadequate reactive power support. To address these issues, a virtual synchronous permanent magnetic synchronous generator (PMSG) model is developed, simulating various fault scenarios accompanied by PAJ. A mapping relationship between PMSG key parameters and fault characteristics is established to identify critical parameters influencing fault current and transient reactive power. To mitigate power steady-state error caused by PAJ, an internal EMF phase angle compensation control is proposed to correct the phase information of the grid voltage. Furthermore, a segmented objective control for the grid side converter is introduced to fully leverage the converter capacity, enabling flexible regulation of both safety margins and reactive power support capabilities. Experimental and simulation results demonstrate the efficacy of the proposed control strategy. Under voltage sags, swells, and continuous faults, it ensures device safety while fully utilizing the converter capacity. Notably, it significantly enhances the transient reactive power support capability of wind turbines, delivering twice the reactive power compared to conventional methods.

Fault direction identification is essential for accurate fault location and effective isolation, thereby enhancing the reliability of the power supply. Traditional power-based directional methods are significantly affected by the bidirectional power flow introduced by distributed photovoltaic (PV) generation. This paper proposes a novel current-only method for fault direction identification in effectively grounded distribution systems, based on the phase-angle difference between post-fault positive-and negative-sequence currents and the change in the negative-sequence current angle from pre-fault to post-fault conditions. For asymmetrical faults - including single-line-to-ground (SLG), line-to-line (LL), and double-line-to-ground (DLG) - a phase difference exceeding 90 degrees indicates an upstream fault, whereas a smaller value indicates a downstream fault. To address weakly asymmetric conditions where the pre-fault negative-sequence current is negligible and its angle becomes unreliable, an auxiliary criterion using the relative polarity of post-fault positive-sequence phase angles is introduced with a grid-side reference. For symmetrical three-phase (LLL) faults, direction is determined from the post-fault positive-sequence phase deviation, where angles larger than 90 degrees reflect reverse current contribution from PV units located near or downstream of the measurement point. Extensive PSCAD/EMTDC simulations over diverse fault locations and resistances, including high-resistance asymmetrical faults up to 1000 Omega, validate the theoretical phase-angle patterns and demonstrate robust, high-accuracy direction identification without voltage or power measurements. These results suggest the proposed method is practical for directional protection in modern PV-rich distribution networks.

DC microgrid is regarded as a promising architecture for integrating renewable energy generators, energy storage devices, and sensitive loads with multiple voltage/current levels in oil/gas field systems. However, with the increasing penetration of renewables and power electronics, voltage sag/swell power quality issues and rapid, high-magnitude over-currents challenges under various transient and fault conditions need urgent attention. This article presents a novel solution of multiport current-limiting interline dynamic voltage restorer (MCL-IDVR), which employs a shared one-battery energy storage unit and an N +1 -port improved active full-bridge converter topology to achieve coordinated compensation among multiple circuits. The multiple output ports of this device can be simultaneously connected in series to dc lines with different voltage and current levels, enabling two coordinated functions of voltage sag/swell compensation and customizable energy-feedback fault current limiting. The working principles, control strategies, and operational modes of the MCL-IDVR are elaborated in detail. Its technical performance is validated through MW-class oilfield dc microgrid simulations and kW-class scaled-down experimental tests. Results demonstrate the device's efficacy in mitigating voltage fluctuations and suppressing fault currents, proving the feasibility and practicality of the MCL-IDVR solution.

P / omega admittance modeling is one of effective techniques to analyze the low-frequency oscillation (LFO). However, existing modeling methods are typically developed for specific control strategies on a case-by-case basis, lacking a general framework. Such case-specific formulations limit the broader applicability of this model in power system studies. To address this limitation, a general P / omega admittance modeling framework is proposed based on the equivalence of controlled source and impedance (ECSI). In the proposed framework, the VSC is modeled as a controlled frequency source, which is regulated by both active power and the frequency at the filter terminal, whereas the reactive power-voltage dynamics are modeled as an equivalent admittance in parallel with the line admittance. Such a modeling strategy enables the construction of P / omega admittance models for VSCs with diversified controls, using a consistent and intuitive framework. Leveraging the model's transparency, the LFO characteristics of various VSCs are identified. The case study of a multi-VSC system demonstrates that the proposed model can accurately reflect the interactions among control loops and reveals the characteristics of LFO caused by different controls of VSCs, providing a powerful tool for stability analysis and control design. Finally, comprehensive simulations under various scenarios are conducted to verify the accuracy of the proposed model.

As a typical electrothermic industrial load (EIL), the adjustable potential of aluminum smelter load has not been fully utilized. The reason is the lack of a precise model to describe the conversion between electricity and thermal energy, and the ignorance of energy consumption related to yield. To address these issues, this paper proposes a price-based demand response (DR) strategy for EIL using an equivalent thermal parameters (ETP) model that considers electro-thermal coupling characteristics and energy consumption changes in the electrolytic process. First, a dynamic thermal balance equation is employed to mathematically depict the energy conversion from electricity to thermal in reduction cells. Second, the ETP model for EIL is derived from the dynamic thermal balance equation based on Taylor expansion, which accurately characterizes the time‑variant relationship between power regulation and electrolysis temperature changes. Finally, an iterative temperature‑power mechanism is embedded into the DR optimization algorithm under TOU tariff to cope with the significant variation of equivalent thermal resistance during regulation. The simulation results demonstrate that the proposed ETP model combined with the iterative algorithm can shift more peak load and reduce more electricity costs for aluminum smelters, especially during an electrolytic aluminum production cycle.

Faced with the dual pressures of increasingly frequent extreme disasters and a high penetration of renewable energy, power systems are encountering significant challenges in maintaining secure and stable operation. Conventional transmission network expansion planning (TNEP) methods, which rely on simulations of typical steady-state scenarios, fall short in ensuring system reliability under severe disaster conditions. To address this issue, this paper introduces an extreme disaster simulation method tailored for the planning stage of hybrid AC/DC power grids, along with an intelligent TNEP strategy based on reinforcement learning. The research follows three main steps: (1) constructing a representative set of extreme scenarios and quantitatively evaluating grid damage in each case; (2) proposing a stepwise Markov decision process (MDP) model suited to the characteristics of TNEP tasks; and (3) developing a reinforcement learning-based intelligent agent to derive optimal wide-area grid expansion schemes. Simulation results on representative systems confirm that the proposed approach significantly improves the reliability of hybrid AC/DC grids in the face of extreme disasters.

Air conditioning loads (ACLs) are important flexible load resources in the power regulation of new power systems. Accurate evaluation of their adjustable capacity is a prerequisite for grid scheduling, and the post-response rebound effect poses a potential threat to the power system. This paper proposes a rebound-suppressing control method for ACLs. Based on the time-varying complementary characteristics of the aggregated power curve under this method, an adjustable capacity assessment model for ACLs is constructed. Firstly, the mechanism of rebound effect caused by traditional control methods is analyzed using a state queue model. Then, a control method is introduced to suppress the rebound effect by restoring the relative temperature position of the air conditioners at the end of the control period. Finally, a time-varying complementary combination model is developed to evaluate and explore the adjustable capacity of air conditioner groups, considering both response duration and power variation. Simulation results show that compared with the other two methods, the proposed rebound-suppressing method can better maintain the temperature diversity of air conditioner groups and effectively suppress the rebound effect. The proposed adjustable capacity evaluation method can still provide considerable adjustable power for response durations of 10 minutes or longer.