Traditional phase selection method is influenced by control strategy of double-side converters in Low-Frequency Transmission System of Offshore Wind Power, resulting in reduced reliability and difficulty in adaptation. Aiming at to the above issues, proposes a transient fault phase selection method suitable for LFTSOWP. Then, a reasonable frequency range for traveling wave analysis is selected, and a phase-selection method based on the Generalized S-Transform using the energy ratio of line voltage traveling waves is proposed. In addition, the positive and negative energy ratios of phase voltages are adopted as auxiliary criteria to solve the problem of identifying the faulty phase in phase-to-phase ground faults under a certain fault startup Angle. Finally, a typical LFTSOWP model is built in PSCAD/EMTDC. The verification results indicate that the proposed method can rapidly identify the fault phase, has good robustness.

Apparent impedances of conventional distance relays cannot reflect the fault distance correctly in low frequency transmission system (LFTS) since the uncertain phase relationship between local and remote currents makes the voltage drop across fault resistance unpredictable. This paper reveals the mechanism by which a widely used phase selectors based on sequence voltage angles mis-operates in LFTS and indicates that negative-sequence current is the sufficient condition to ensure selectors reliable. Then a distance protection is proposed to eliminate the effect of fault resistance. By means of zero-sequence network, distance protection for single-line-to-ground and line-to-line-to-ground faults can be designed as a source-agnostic scheme. As for line-to-line faults, a uniform calculation expression for fault distance is derived to cope with various control operations. A control-based distance protection for a symmetrical fault is proposed to make the angle of local current be consistent with that of remote current. Simulation tests prove the effectiveness and precision of the proposed distance protection. The above solutions are also applicable to systems consisting of 100% converter-based sources.

Identifying fault properties before reclosing mitigates secondary impacts caused by traditional three-phase automatic reclosing onto permanent faults in distribution networks. While existing literature proposes detection-based reclosing for identification, low sensitivity and the influence of neutral grounding modes remain problematic. This paper proposes a sequential reclosing strategy with permanent fault identification capability for both grounding and phase-to-phase faults. First, an analytical model for detection-based reclosing is developed to demonstrate that the fault current surge from single-phase reclosing is substantially lower than that of three-phase reclosing. Second, zero-sequence impedance expressions are derived for grounding faults; a criterion is then established based on the characteristic phase angle differences between transient and permanent faults. For phase-to-phase faults, a criterion utilizing the average Hausdorff distance algorithm is proposed to quantify line voltage waveform similarities. Finally, RTDS-based closed-loop tests under low-resistance grounding, ungrounded, and arc-suppression-coil grounding systems verify that the proposed scheme can reliably identify fault properties, tolerate high fault resistance, and remain independent of neutral grounding modes.

To mitigate secondary damage caused by reclosing a distribution line with a permanent fault, a highly sensitive fault property identification scheme using detection-based reclosing is proposed in this paper. Firstly, the analytical model for single-phase detection-based reclosing in distribution networks is established to reveal the differences between transient and permanent faults. Given that the equivalent fault resistance of permanent ground faults differs significantly from the insulation resistance of transient faults, an equivalent fault resistance calculation method is proposed to improve the sensitivity of fault property identification. Furthermore, to reduce the impact of fault resistance and load capacity on phase-to-phase fault property identification sensitivity, a permanent fault identification method utilizing the line voltage ratio is proposed. Finally, a comprehensive detection-based reclosing scheme is developed. Validation results based on PSCAD simulations and RTDS closed-loop tests demonstrate that the proposed scheme exhibits both wide applicability across various scenarios and high sensitivity.

Accurate fault location is crucial for power system emergency repair. However, the existing fault analysis-based fault location methods for the outgoing line of inverter-based resources use the traditional synchronous generator-based source model as the remote-end source model. The reliance on traditional source characteristics limits the adaptability of these methods in low-frequency transmission system (LFTS) used for offshore wind farm integration, where both ends of the system are inverter-based resources. Therefore, a novel fault location method is proposed for LFTS, which incorporates the source equations and fault network equations. Firstly, the source equations of LFTS are derived from control characteristics, while the network equations are derived from sequence networks and fault boundary conditions. Then, the fault location is converted into an optimization problem. The constraints of source equations, network equations and actual measured values are integrated in the objective functions, thereby allowing the fault distance and fault resistance to be identified by solving optimization problem. Finally, the performance of proposed fault location method is verified in PSCAD/EMTDC. The proposed fault location method exhibits an error of less than 1% in most cases, with a maximum error of less than 4%, and a fault resistance tolerability up to 300 Omega.

This work develops a cascaded intelligent diagnosis strategy to meet the dual demands of high precision and fast response for fault identification in DC microgrids. First, a temporal convolutional network is adopted to effectively extract local features and expand the feature dimension. Then, a bidirectional long short-term memory network integrated with an attention mechanism is used to capture long-term temporal dependencies and dynamically focus on key time-series features to construct global representations. Finally, a multilayer perceptron is employed to achieve rapid classification via its nonlinear decision-making capability. The proposed cascaded structure fully exploits the advantages of each model, effectively reduces the internal complexity of individual models, and features a lightweight and clear architecture. Simulation results demonstrate that the method can quickly and accurately identify various fault types and impact load fluctuations, with significantly higher diagnostic accuracy than traditional methods and extremely fast processing speed. It exhibits strong stability under small-sample conditions, noise interference, and different operating conditions of DC microgrids. Furthermore, the method only requires single-point measurement data, avoiding communication-related issues, and offers the merits of fast response, low cost, and easy deployment.

The rapid expansion of wind energy necessitates highly accurate ultra-short-term power prediction techniques to address integration and dispatch challenges caused by wind volatility. While point forecasting provides a single estimate, probabilistic forecasting offers a more comprehensive view by capturing the distribution of future outputs, enabling risk-aware decision-making. However, direct modeling of such distributions remains challenging due to complex spatio-temporal dependencies and high network optimization burdens. To address these issues, this study proposes the spatio-temporal mixture-of-experts diffusion (MOE-STD) framework for wind power probabilistic forecasting. The framework introduces a Channel Attention Enhancement Module (CAEM) to identify key dependencies between turbine output and influencing factors. A Mixture of Spatial Experts (MOSE) dynamically learns both simple and complex spatial correlations among wind farms, while a Mixture of Temporal Experts (MOTE) captures continuous trends and sudden fluctuations in time-series data. These modules are embedded within a diffusion framework, which learns to reconstruct future output distributions via multi-step prediction, rather than direct forecasting. Validated on real-world wind farm datasets, the proposed model consistently outperforms both point and probabilistic forecasting methods in terms of accuracy and reliability, demonstrating a superior ability to capture uncertainty in ultra-short-term wind power forecasting.

The increasing integration of power electronic equipment into modern power systems has led to the gradual weakening or elimination of boundary components, posing significant challenges for single-ended whole-line fault identification. To address this issue, this article employs coordinated active boundary control and signal injection control during the fault steady state to enhance fault characteristics and establish seriesconnected boundary properties. A fault analysis is established using modal network modeling for both pole-to-ground (P2G) and pole-to-pole (P2P) faults. The results demonstrate that the injected characteristic frequency current magnitude at the protection point increases or remains unchanged in the case of internal faults, while it decreases under external faults. Based on this, a novel single-ended setting-free protection criterion is proposed, which is independent of system parameters, simulations, and complex calculations. Its performance is validated on a multiterminal hybrid HVDC transmission system. Simulation and experimental results confirm that the method achieves wholeline fault identification under fault resistances up to 500 ohm, noise levels up to 20 dB, and parameter variations, demonstrating high robustness and reliability even in weak or nonboundary scenarios.

As the penetration of power electronic converters and renewable generation increases, voltage and current waveforms in power systems are increasingly distorted and nonstationary. Accurate estimation and fast tracking of electric power quantities are essential for energy metering and billing, online monitoring, and safe operation of power systems. However, most existing estimation methods assume that the waveform within an observation window is stationary and periodic, which often leads to a slow dynamic response under nonstationary conditions and introduces estimation errors when the fundamental frequency deviates from its nominal value. To address these challenges, this paper introduces a framework termed operating point fluctuations (OPF). In the OPF framework, a nonstationary waveform is reinterpreted by treating the time-varying fundamental component as a dynamic operating point and describing all distortions as fluctuations around it. Based on this framework, a set of single-phase and three-phase power quantity expressions is derived. Furthermore, an estimation method is proposed. Comprehensive simulation studies demonstrate that the proposed method achieves high accuracy under stationary conditions, fast dynamic tracking under nonstationary conditions, low computational burden, and robustness to noise and fundamental frequency deviation. Validation using field measurement data further confirms these results and demonstrates its suitability for practical applications.

Since the PV-side voltage source converter (PVVSC) and the sending-end modular multilevel converter (SEMMC) are connected by AC lines, they constitute a double-ended weak-feed system (DEWFS). Under the traditional negative sequence cooperative control strategy, it will induce incorrect action of traditional distance protection. Therefore, an improved distance protection considering the negative-sequence control coordination strategy is proposed. Firstly, the adaptability of distance protection of DEWFS under the traditional fault cooperative control strategy is analyzed. Secondly, considering the influence of fault severity on the control effect, a switch strategy between negative sequence voltage suppression (NSVS) and negative sequence impedance angle reconstruction (NSIAR) on the MMC side is proposed. Finally, combined with the composite sequence network analysis to obtain the additional impedance angle provoked by the fault resistance, and the correct action of the traditional distance relay is realized. The simulation results suggest that the proposed method has high resistance to fault resistance and can realize accurate identification of internal and external faults within 60 ms, and it has certain robustness. Moreover, the proposed negative sequence control strategy did not affect other fault ride-through (FRT) control of the DEWFS while ensuring the safe and stable operation of the DEWFS.

Active distribution networks (ADNs) exhibit characteristics such as topological flexibility and bidirectional power flow, which distinguish them from traditional distribution networks (DNs). These intrinsic characteristics present significant challenges to traditional single-ended protection schemes deployed in conventional DNs. Boundary protection, which is based on the coordination between the relay on one side of the line and the boundary element at the opposite terminal, can achieve rapid fault detection across the entire line utilizing exclusively local information, thereby offering a promising solution for addressing the protection challenges in ADNs. In this paper, a method for constructing a resonant-type boundary element (RBE) is proposed, based on the analysis of the performance of various boundary element types. Furthermore, the fault characteristics for internal and external faults are analyzed. Within the specified blocking frequency band (BFB) of the RBE, the voltage during an internal fault exhibits significantly higher magnitudes compared to those during external faults. Based on this distinction, a voltage energy criterion incorporating fault phase angle correction is subsequently proposed. Simulation and experimental results demonstrate that the proposed protection method, in conjunction with the designed RBE, can accurately and rapidly distinguish internal and external faults.

Reliable fault location in converter-dominated DC distribution networks is challenging because fast converter control actions may suppress or reshape the natural fault characteristics. This paper proposes a single-ended fault-location method for radial DC distribution networks interfaced by a dual active bridge (DAB) converter. The method exploits the intrinsic switching harmonics generated during active current-limiting operation. Based on these characteristic harmonics, a frequency-domain equivalent fault-loop model is established using distributed-parameter line sections, and the fault distance is estimated through nonlinear impedance matching using only local voltage and current measurements. To account for fault-resistance and load-parameter uncertainties, a multi-parameter estimation procedure is developed to jointly identify the fault distance and associated equivalent parameters. Finally, PSCAD/EMTDC simulation results verify the effectiveness of the proposed method under different fault distances and fault resistances.

The sensitivity of traditional current differential protection may decrease or even lead to maloperation due to the frequency deviation and limited amplitude characteristics of fault currents from power electronic-based power supplies. To solve aforementioned challenges, this study presents a fast frequency-domain pilot protection method based on line model consistency identification. This approach distinguishes between internal and external faults through transient voltage and current measurements while comprehensively accounting for the frequency response characteristics of instrument transformers. First, the mathematical relationships governing the secondary voltage and current under internal and external faults are rigorously established. Subsequently, the complex frequency-domain components of the transient voltage and current are extracted by applying of the numerical Laplace transform. Then, a consistency criterion for the line model is established, accompanied by a comprehensive protection scheme. Finally, extensive numerical simulations are conducted to validate the effectiveness of the method. Compared with existing approaches, the proposed method functions independently of the output power level of the power supplies at both ends and demonstrates superior robustness against high fault resistance and strong noise interference. In addition, the proposed method exhibits a light computational burden and requires low sampling rates.

With the increasing integration of power electronic converters and renewable energy resources, power system voltage and current waveforms have become increasingly distorted and nonstationary. Rapid tracking of power quality (PQ) indices is crucial for the safe operation of equipment. Conventional Fourier transform (FT)-based methods suffer from limited accuracy under nonstationary conditions because of spectral leakage, while existing time-frequency methods still suffer from limited estimation accuracy and high computational burden. To address these challenges, this paper introduces a conceptual framework termed operating point fluctuations (OPF), in which the time-varying fundamental component is regarded as a dynamic operating point and all distortions are modeled as fluctuations around it. Based on this framework, a set of instantaneous PQ indices is reformulated, and a real-time estimation method is proposed using recursive least squares (RLS) filters and the Hilbert transform (HT). Simulation studies demonstrate that the proposed method achieves high steady-state accuracy, fast dynamic tracking, robustness against a certain degree of frequency estimation error, interharmonics, and noise, and low computational burden. Hardware-in-the-loop (HIL) experiments and field measurement data verification further demonstrate its practical potential.

With the improvement of power electronic equipment, boundary components are gradually weakened or even eliminated, making it difficult to realize fault identification across the whole line solely based on single-ended electrical characteristics. Therefore, this paper introduces the concept of active boundaries, leveraging the high controllability of power electronic equipment to develop virtual diversion and suppression controls. These controls can help to amplify the differences between internal and external faults, effectively mitigating protection dead zones at the terminal of the line. Additionally, using the refraction coefficient as an indicator, the influence of equivalent resistance, inductance, and capacitance parameters of the converter is analyzed. Building on this, the energy ratio of high to low frequency band criterion considering frequency information is proposed to achieve single-ended whole-line protection in power electronic systems. The proposed scheme is validated using a multi-terminal hybrid HVDC transmission system model. Simulation results demonstrate that even in weak or non-boundary scenarios, the scheme can accurately identify faults across the whole line without relying on communication or coordination with other protection schemes, while maintaining high sensitivity and reliability.

The inherent flexible and controllable characteristics of the multi-terminal flexible DC grid based on modular multilevel converters (MMCs) have opened up new avenues for the research on fault current limitation and line fault identification. Existing active current limitation control strategies and fault identification methods operate independently, without considering their interplay and mutual influence, leading to an inability to fully integrate them. To enhance the synergy between active current limitation control and protection, this paper introduces a two-stage active current limiting control coordinated protection scheme for half-bridge MMCs. Firstly, the interdependence between current limitation control and fault identification is established; building on this, to prevent rapid blocking of MMCs post short-circuit faults and the impact of fault characteristic changes on the reliability of fault identification, the first stage of current limitation control is implemented using AC virtual impedance control. This approach limits the arm current without modifying the DC-side fault characteristics, thus creating favorable conditions for existing protection. Secondly, line boundary properties are leveraged to delineate the scope for the second stage of DC-side current limitation control, enabling selective deployment of current limitation measures and curbing the propagation of fault impacts. Finally, by integrating the current limitation areas with directional elements, reliable fault identification is accomplished. Simulation outcomes indicate that the proposed scheme effectively limits both AC and DC currents, providing a novel perspective for the convergence of current limitation control and fault identification. (c) 2017 Elsevier Inc. All rights reserved.

With the widespread integration of distributed generators (DGs), traditional relay protection strategies, which are based on the assumption of a single-sided power supply, fail to meet the operational requirements of the new power systems. In this context, whole-line rapid protection based on single-ended measurements has emerged as a promising solution to address protection challenges in the new distribution networks, due to its advantages such as independence from remote-end information and fast response speed. This method constructs a boundary element at the remote end of the protected line and achieves rapid fault identification across the entire line by utilizing abrupt changes in impedance characteristics on both sides of the boundary element. However, existing distribution networks lack naturally available boundary elements, which limit the practical application of this principle. To address this issue, this paper draws inspiration from the design of a single-frequency line trap used in high-voltage transmission systems and proposes the construction of a resonance-based boundary element capable of effectively isolating signals of specific frequencies. Furthermore, to accommodate space constraints commonly encountered in field installations, an optimization scheme involving the filling of the boundary element's magnetic core with soft magnetic materials is introduced. This significantly reduces the element's size and enhances its engineering applicability, offering a novel technical approach for the compact design of power equipment. In addition, considering the frequency band characteristics of the designed boundary element, this paper proposes a voltage high- and lowfrequency energy ratio criterion integrated with a phase correction mechanism. This criterion serves as the primary basis for distinguishing internal faults from external ones, thereby establishing a complete set of whole-line rapid protection criteria based on single-ended measurements. Finally, an electromagnetic transient simulation model based on MATLAB/Simulink, combined with magnetic field analysis using ANSYS Maxwell, is developed to validate both the accuracy of the proposed modeling method for the boundary element and the effectiveness of the protection criteria.

The LCL-type three-level grid-connected inverter is extensively employed in photovoltaic (PV) power generation systems, which has multiple individually controlled objectives. To this end, an integrated division-summation (I-D- $\Sigma $ ) control strategy is proposed in this article, which can consider grid-connected current tracking, neutral-point potential (NPP) balance control, resonance suppression, and low-frequency common mode voltage (CMV), simultaneously. To compensate for the fact that the dual-division-summation (D-D- $\Sigma $ ) method only considers grid-connected current and resonance suppression, a virtual zero potential point is added for NPP balance control. Additionally, a low-frequency CMV estimation method based on fully connect-convolutional neural network (FC-CNN) is proposed, which perfectly avoids the unobservability of sensorless CMV. The I-D- $\Sigma $ method can effectively reduce the low-frequency CMV, which is compensated by the low-frequency CMV. The effectiveness of the proposed strategy was verified through comparative experimental results.

The performance of conventional distance protection is degraded by the source characteristics of power electronic-based power sources (PEPS) in the low-frequency transmission system (LFTS) of offshore wind farm. Furthermore, as a dual-terminal PEPS system, offshore LFTS facing the inadaptability of improved distance protection scheme originally proposed for general single-terminal PEPS system. To address this, a novel distance protection scheme based on control characteristics of PEPS is proposed for offshore LFTS. The source characteristics of both PEPSs in LFTS are derived from their control strategies and expressed as constraints between output voltages and currents. Building on this, a calculation method for fault electrical quantities is proposed to quantify the influence of PEPS on distance protection. Subsequently, the apparent impedances representing the source characteristics are precomputed and stored in a data matrix using the proposed calculation method. Finally, the proposed protection scheme estimates the fault distance and distinguishes between internal faults and external faults by searching for the best matching values in the data matrix against the actual measured values of apparent impedances. The performance of proposed protection scheme is tested via a simulation model in PSCAD/EMTDC. Simulation results demonstrate that the proposed protection scheme exhibits a high tolerance to fault resistance under different fault types, control strategies and operating states of wind farm.

Low-voltage direct current (LVDC) system incorporates power electronic converters and new energy sources, which is different from traditional AC distribution systems. Identification of disturbance is crucial for system state monitoring and fault diagnosis. Firstly, this paper theoretically analyzed the transient information generated by capacitor switched on and fault process, clarifying the differences in their frequency characteristics. Secondly, wavelet decomposition and reconstruction techniques were employed to extract transient features in different frequency bands, and the classification and identification of capacitor switched on and fault disturbances were realized based on the differences in frequency characteristics. Finally, the effectiveness of the proposed method was verified through simulation.