Volume 41, Issue 2 (2026)
Young Scholars Column
Review of research on entry path and safety assessment of equipotential live working on extra-high voltage/ultra-high voltage transmission lines
Peng LI, Li HE, Lingxuan GAN, Tao XIE, Houming SHEN, and Tian WU
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.001
With the transformation of China 's energy structure and the accelerated construction of new power systems, extra-high voltage/ultra-high voltage transmission projects are gradually being developed into a backbone network to achieve cross-regional energy deployment and support new energy consumption. In this context, as equipotential live working is the core technology to ensure the uninterrupted operation of power grids, the optimization of entry paths and dynamic safety assessment for it become the key focus of current research. Firstly, the key technology systems of equipotential live working on extra-high voltage/ultra-high voltage transmission lines are systematically reviewed, and a complete research framework is constructed. Secondly, based on the development history and current status of equipotential live working technology in China and abroad, the research progress of live working in recent years in the directions of safety distance, path planning, safety assessment, and potential transfer is reviewed. Finally, combined with the challenges faced in the construction of new power systems and the application practices of emerging technologies such as artificial intelligence and digital twins in power systems, the key directions and technical paths for the future development of equipotential live working technology are prospected.
A review of vehicle-grid interaction with introduction of virtual power plants
Yanxia WANG, Yuchen WANG, and Shaojun GAN
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.002
The advent of virtual power plants (VPPs) promotes the development of the interaction between electric vehicles (EVs) and the power grid, and plays an important role in promoting the load balance of urban power grids. By systematically reviewing the relevant literature on EVs, integrated photovoltaic storage and charging stations, and VPPs and their aggregators, this paper analyzes the vehicle-grid interactions with the introduction of VPPs, summarizes the commonly used EV load forecasting methods, and describes the siting and capacity planning model of integrated photovoltaic storage and charging stations. Further, it summarizes the game relationship among the grid layer, the VPP and its aggregator, and the user layer in the process of vehicle-grid interaction. Finally, it summarizes the existing studies, and provides an outlook on the future development of vehicle-grid interaction by combining with the case studies of the VPPs of different countries.
Arc resonant trajectory soft starting strategy of LLC resonant converter and its experimental verification
Di WU, Faqiang WANG, and Baohai ZHANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.003
LLC converters offer outstanding advantages when used i n data center power supply applications. To solve the problems of slow start-up of LLC resonant converters during traditional down-frequency soft starting and high MCU calculation and storage pressure in the process of optimal trajectory control soft starting, an arc resonant trajectory soft starting strategy based on trajectory control is proposed. The two charging stages of an LLC converter under the strategy are analyzed, and an adjustment stage, an entry stage, and an exit stage are designed based on the charging stages, which effectively suppresses the transient current overshoot phenomenon in the soft starting process. A PLECS simulation model and a prototype with 30 V output are developed. The simulation and experimental results show that the converter can complete the charging of a 400 μF output filter capacitor within 350 cycles, that is, 3.5 ms, which verifies the correctness and effectiveness of the proposed soft starting strategy with low computing pressure.
Smart grid
Risk assessment of power system frequency based on dynamic inertia factor
Jie FU and Qian ZHANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.004
Against the background of the continuous integration of a high proportion of renewable energy into the power system, the system inertia provided by traditional synchronous units shows a continuous downward trend, which significantly increases the risk of power system frequency instability. A frequency risk assessment method based on the dynamic inertia contribution factor (DICF) is proposed to solve the problem of insufficient spatiotemporal resolution in traditional static inertia assessment. By defining the dynamic equivalent inertia, the real-time contribution of each unit to the system inertia is quantified, and the weighted inertia-risk index (WIRI) is constructed. Combined with the frequency deviation rate, the dynamic classification and early warning of risks are achieved. Simulation experiments show that the DICF and the WIRI can track the inertia contribution status of each generator in real time and the risks during faults.
Method for enhancing resilience of smart grids based on adaptive graph attention multi-agent reinforcement learning
Peng CHANG, Yun WANG, Fei MENG, Qing WANG, and Yang SUN
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.005
Enhancing the resilience of smart grids is crucial for maintaining the security and reliability of power systems. An adaptive graph attention multi-agent reinforcement learning (AGA-MARL) method is proposed, through which the learning efficiency and collaborative ability of the system in complex grid environments are improved by an adaptive learning rate and a dynamic task allocation mechanism, thereby enhancing the resilience and interpretability of smart grids. First, adaptive multi-agent deep reinforcement learning (AMA-DRL) and dynamic spatial-temporal graph convolutional networks (DST-GCN) are combined to enhance the information interaction among multiple agents and utilize the dynamic graph structure to capture the complex dependencies of the grid system. Second, an interpretability module is integrated to provide more intuitive decision-making explanations by combining attention weights and Shapley value (SHAP). Finally, the effectiveness of the proposed method is verified through experiments. The research results show that compared with the traditional AMA-DRL method, AGA-MARL performs better in aspects such as grid fault recovery time, system stability, and interpretability.
Coordinated optimal configuration of grid-forming inverters and STATCOMs for enhancing transient voltage security of regional power grid
Xiaofeng YU, Ronghao TAN, Binbin TANG, Shunjiang LIN, Ye CHENG, and Weikun LIANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.006
At present, renewable energy stations in a regional power grid mainly adopt grid-following control. Since it has the current source control characteristic and does not have the voltage support capability similar to that of traditional synchronous generators, it will reduce the transient voltage security of the regional power grid. When short circuit faults occur or the impact loads start up frequently in the regional power grid, it is more likely to result in the transient voltage instability and thus affect the secure and reliable power supply to users. In response, a coordinated optimal configuration model of grid-forming inverters and static synchronous compensators (STATCOMs) is established to enhance the transient voltage security of regional power grids. The objective function is to minimize the investment costs, and transient safety constraints are required to be satisfied when the system is subjected to a short circuit fault or impact load startup. It is also required to ensure that the output current of the grid-forming inverters in renewable energy stations does not exceed the secure limit during the transient process, thereby achieving low voltage ride through. To solve the optimization model with differential equation constraints quickly and accurately, a fourth-order implicit Adams method based on fast startup is proposed, which reduces the number of differential periods in the startup of the traditional fourth-order implicit Adams method, and transforms the differential equation constraints into a smaller number of algebraic equation constraints, so that the optimization model can be solved quickly and accurately. Finally, a case study on an actual regional power grid verifies the effectiveness of the proposed optimal configuration model and solution algorithm.
Hybrid attack detection method for LFC based on enhanced Kalman filter and residual analysis
Yunning ZHANG, Yuxuan FU, Zhenxing LI, and Fayun FU
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.007
Against the backdrop of extensively interconnected cyber-physical power systems (CPPS), this paper proposes an attack detection method based on an enhanced Kalman filter to address hybrid attacks involving denial-of-service (DoS) attacks and false data injection (FDI) attacks. First, a load-frequency control model is established for a two-region interconnected power system, and a hybrid attack model incorporating both DoS and FDI attacks is constructed. Second, an enhanced Kalman filter is introduced, incorporating an adaptive process noise adjustment mechanism and a residual history sliding window to enhance the robustness of state estimation. Subsequently, an improved attack detector is designed, integrating chi-square detection, adaptive threshold adjustment, and sliding window cumulative detection strategies to achieve rapid attack identification and classification. The Kalman filter handles system state estimation and generates a residual sequence, while the attack detector performs statistical hypothesis testing based on the residual sequence, identifying attacks by monitoring abnormal residual variations. Finally, the Kalman filter parameters are optimized using a simulated annealing algorithm to enhance the system's frequency stability and state estimation accuracy under attack conditions. Simulation results demonstrate the effectiveness and superiority of the proposed method in improving CPPS hybrid attack detection.
Risk identification and load transfer calculation method for distribution network framework considering new energy access
Ji SU, Tao CHEN, Yu DUAN, Zhantao GAO, Wei YANG, and Yuqiao OU
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.008
Driven by the "dual carbon" target, it has become a trend for an extremely high proportion of new energy to be integrated into the distribution network. The random "bidirectional flow" of power flow in the distribution network has a great impact on its operational safety. It is urgent to identify the network framework risks and load transfer methods of the distribution network. A risk identification and load transfer calculation method for the distribution network framework considering the integration of new energy is proposed. First, a risk identification method for the distribution network with a high proportion of new energy access is proposed, and its risk quantification indicators are given from the aspects of network structure, operating load rate, power flow factor, power flow transfer entropy, and their comprehensive aspects. Second, considering cascading conditions across multiple high-risk lines in the distribution network, a model for calculating cascading load transfer is proposed. Finally, a case study on an actual power grid is conducted to validate the proposed method, and the results demonstrate its effectiveness.
A risk assessment method for three-phase unbalance in distribution transformer areas considering zero-sequence current
Mo SHI, Yingting LUO, Xin LI, Bin ZHANG, Bowei WEI, Shenzhou ZHOU, Qin YAN, and Rui MA
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.009
With the significant growth of seasonal loads such as agriculture and tourism, power equipment failures and power outages caused by the three-phase unbalance problem in distribution transformer areas become major challenges in the operation and maintenance of distribution networks. Therefore, it is urgent to strengthen the assessment of three-phase unbalance in distribution transformer areas. To address the problems of few evaluation indices and single weight distribution methods for three-phase unbalance in existing distribution transformer areas, a risk assessment method for three-phase unbalance in distribution transformer areas considering zero-sequence current is proposed based on the traditional three-phase unbalance evaluation method for distribution networks. First, the correlation of the severity of three-phase current unbalance with both the zero-sequence current and the three-phase current unbalance degree of the distribution network is fully considered. Then, an evaluation model for three-phase unbalance in distribution transformer areas considering zero-sequence current is established by using the improved utility theory (IUT). Based on the analytic hierarchy process and the anti-entropy weight method, a combined weighting method is proposed to strengthen the comprehensive analysis of three-phase unbalance and zero-sequence current indices. Finally, an example is used to verify the method. The research shows that traditional single evaluation indices cannot effectively and comprehensively evaluate the three-phase unbalance degree in distribution transformer areas, while the proposed risk assessment method can better reflect the severity of three-phase unbalance in each distribution transformer area, achieving more accurate risk assessment.
Identification method for household-phase-transformer relationship in distribution station area based on FDDTW and K-median clustering
Qiangsheng BU, Ye GUO, Jiangping JING, Wenfei YI, Pengpeng LYU, and Shuyi ZHUANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.010
To study the hierarchical relationship of branch nodes involving the identification of distributed power generation access and effectively perform the topology identification of distribution station areas, a topology identification method combining the fast derivative dynamic time-regularization algorithm (FDDTW) and K-median clustering is proposed for low-voltage active distribution station areas. Firstly, under the premise of not changing the data characteristics, the input original voltage data of the station area are normalized by utilizing the characteristics of removing the mean and standardizing the variance of the Z-score standardization method. Secondly, the distance value is calculated by adopting the FDDTW algorithm; the clustering number is set according to the number of distribution transformers divided in the station area, and the household-transformer relationship is clustered by utilizing K-median clustering. Thirdly, based on the clustering result of the household-transformer relationship, the distance value inside the station area is calculated again by using the FDDTW algorithm, and the phase relationship is clustered. Finally, the method is compared and analyzed by adopting experimental simulation. The research results show that the calculation time of the FDDTW algorithm is shortened by 28.6% compared with that of DTW, and the accuracy rate of the FDDTW algorithm is improved by 23.8 percentage points compared with that of DTW.
Identification method for high-resistance ground faults in distribution networks based on traveling wave fault feature difference and RMT model
Junjie SHI, Daoyi GU, Feng DENG, Pengyu QI, Junwen LUO, Ruijun LI, and Chang TANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.011
When a high-resistance grounding fault occurs in a distribution network, the fault signal is weak, and traditional methods based on single-feature extraction are prone to causing protection maloperation. In response, this paper introduces an effective identification method for high-resistance ground faults in distribution networks based on differential features of traveling wave voltage signals and a "retentive network meets vision transformer" (RMT) model. Initially, the traveling wave voltage signal is extracted to implement wavelet packet time-frequency features, and the differences in the time-frequency domain responses between high-resistance ground faults and normal disturbance conditions are visualized. Then, a composite RMT model integrating retentive network (RetNet) and vision transformer (ViT) is constructed, utilizing RetNet's memory mechanism and incorporating temporal priors into ViT to enhance the memory and attention to rare events and thereby address the issues of limited sample data and their imbalanced category distribution in distribution networks. Subsequently, the panoramic waveform of traveling wave is input into the RMT model, and hyperparameters are optimized to improve the model's discriminative performance, so as to achieve precise identification of high-resistance ground faults. Finally, a simulation test is conducted to verify the feasibility and effectiveness of the proposed method. Simulation results confirm that this approach overcomes the limitations of traditional methods that rely on a single feature quantity, maintaining high classification accuracy even with small, imbalanced datasets, and accurately identifying high-resistance ground faults under various fault conditions.
Cable fault identification and precise location method based on SSA-BP and ICEEMDAN-NTEO algorithms
Fating YUAN, Haoyue LI, Huqiang LI, Yi YANG, Shengkai JIAN, Yuqing JIANG, and Bo TANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.012
Existing power cable traveling wave ranging methods rely on the accurate identification of the initial traveling wave and suffer from the problem of inaccurate fault location. Based on the electromagnetic transient simulation software ATP-EMTP, this paper establishes a 10 kV power cable transmission line model and proposes a double-ended traveling wave location method based on back propagation (BP) neural network optimized by the sparrow search algorithm (SSA) and novel Teager energy operator (NTEO). Firstly, the power cable transmission line model is established, the fault current waveforms under different working conditions are studied, and the SSA-BP algorithm is utilized to identify the cable fault types. The prediction results of the training set and the test set show that the SSA-BP algorithm is able to accurately and quickly identify the power cable fault types. Then, through the phase-mode transformation of the three-phase current of the cable, appropriate fault components are selected for fault location of the cable according to the different fault types of power cables. The improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) algorithm is used to decompose the fault waveform, filter out the noise interference in the fault signal, enhance the initial traveling wave head characteristics by the NTEO algorithm, and accurately determine the time when the initial traveling wave arrives at the detector, so as to realize the accurate location of power cable faults. The method proposed in this paper achieves the fault identification accuracy of 98.3 % with the consideration of different short-circuit faults, grounding resistance, fault distance, etc. The fault location accuracy of the complete ensemble empirical mode decomposition (CEEMD)-NTEO and wavelet transform algorithms is 99.83 % and 99.67 %, respectively, while that of the proposed method is 99.88 %. The research results provide an important theoretical basis for the accurate identification and location of cable faults.
Low-error carbon monitoring method for industrial users based on equipment sta tus identification
Yutao XU, Zongyi WANG, Zhuk ui TAN, Yun ZHAO, Qihui FENG, and Ziwen CAI
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.013
A low-error carbon monitoring method base d on equipment status identification is proposed to solve the problem of insufficient accuracy in existing carbon monitoring methods for industrial users. Firstly, an equipment status identification model based on temporal convolutional network-gated recurrent unit (TCN-GRU) is built to accurately identify the operating status of key carbon-emitting equipment for industrial users. Secondly, genetic algorithms (GAs) are introduced to dynamically optimize the parameters of the fully connected layer in the model, enhancing the classifier's identification capability for equipment with high carbon emissions. Finally, low-error carbon emission monitoring is achieved based on the optimized status identification results. Experiments conducted on the industrial dataset IAID demonstrate that the proposed method significantly reduces carbon monitoring errors, with the root mean square error (RMSE) decreasing by approximately 13%. Additionally, it outperforms existing methods in key metrics such as RMSE and R², effectively improving the accuracy and reliability of carbon emission monitoring for industrial users.
Game bidding strategy for electric thermal gas coupled virtual power plants participating in main and auxiliary joint markets
Chunyan ZHANG, Zhenlan DOU, NG Jihang ZHA, Lingling WANG, and Chuanwen JIANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.014
As integrated energy systems develop, multi-energy coupled virtual power plants have become an effective way to manage integrated energy systems. However, due to their numerous internal energy types, complex coupling relationships, and inability to obtain competitor information when participating in the market, it is difficult to solve their bidding strategies by adopting traditional mathematical methods. A game bidding strategy for electric thermal gas coupled virtual power plants participating in main and auxiliary joint markets based on constraint-aware multi-agent reinforcement learning is proposed. Firstly, a mathematical model of a multi-energy coupled virtual power plant is built and the energy flow relationship of its internal electric thermal gas multi-energy coupling is described. Secondly, a game model of multi-energy coupled virtual power plants is built, in which price setters participate in the incomplete information main and auxiliary joint market, and virtual power plants are allowed to trade electricity and auxiliary services. Finally, constraint-aware reinforcement learning is improved, and the constraint-aware multi-agent reinforcement learning algorithm is proposed for solving the model. By setting a watchdog module in the algorithm, the virtual power plant strategy is effectively ensured to comply with power grid safety constraints, and compared with the theoretical optimal solution to verify the effectiveness of the proposed method.
Clean energy and energy storage
Analysis method for power system ine rtia intervals with coupled uncertainty of new energy output
Xuebin WANG, Guobin FU, Kaix uan YANG, Qixuan WEN, Rui SONG, and Yunfeng WEN
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.015
The large-scale grid connection of new energy leads to the continuous compression of the start-up capacity of synchronous power sources, and the system faces the risk of low-inertia operation. Existing deterministic inertia trend assessment methods ignore the time-varying impact of random new energy fluctuations on unit commitment, which may lead to the misjudgment of inertia adequacy. In this regard, an analysis method for power system inertia intervals is proposed to realize a panoramic portrayal of inertia boundaries by coupling the uncertainty of new energy output. First, a confidence interval is constructed based on probability modeling of wind and solar power prediction errors. Then, an operational constraint system involving unit start-stop, new energy fluctuation, virtual inertia, and power flow security is established. An interval possibility transformation is adopted to convert the uncertain optimization problem into a deterministic inertia extreme value search model, and a multi-time-scale rotational kinetic energy optimization framework is constructed. After the trigger, the quantitative assessment of the trends of system inertia intervals and the risk early warning for the inertia situation are realized. Finally, a simulation analysis and validation are conducted taking a provincial power grid with a high proportion of new energy in China as an example. The results show that the proposed method well quantifies the inertia intervals under the fluctuation of new energy output, providing a decision-making basis for early warnings during periods of weak inertia.
Identification method for control parameters of grid-connected photovoltaic system based on sensitivity characteristic analysis
Yunhe CHEN, Peiqiang LI, and Jiajie XIAO
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.016
Accurate identification of fault control parameters is the basis for establishing an accurate simulation model of a grid-connected photovoltaic system. The dynamic coupling of multiple control links and the low sensitivity of some parameters lead to a low overall identification accuracy of control parameters. To address this problem, an identification method for control parameters of a grid-connected photovoltaic system based on sensitivity characteristic analysis is proposed in this paper. Firstly, a method of injecting disturbances into the measurement signals on the secondary side of the system is proposed to decouple the dynamics of multiple controllers. Secondly, a sensitivity algorithm is used to analyze the dynamic correlation between the control parameters and the system fault transient current, and the optimal observation time window is selected. Thirdly, the sensitivity differences of control parameters under different types of disturbances are used to identify the control parameters step by step, which effectively solves the problem that multiple control parameters are difficult to be identified simultaneously. Finally, a 29-node simulation test system with grid-connected photovoltaics is built on the MATLAB/Simulink platform, and the effectiveness of the proposed method is verified.
Reliability analysis of photovoltaic system based on output prediction
Feilong ZHAI, Xingyong ZHAO, Lilong HAO, Zhiyi ZHANG, Qiang YAN, and Xing WEN
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.017
To study the reliability of photovoltaic (PV) power generation in the context of "double high" (i.e., high penetration of renewable energy plus high penetration of power electronics equipment) of PV systems, this paper proposes a method of analyzing the reliability of PV output based on the combination of output prediction and an improved Monte Carlo method. Firstly, output characteristic analysis and output prediction are performed on the PV historical data to obtain the probability distribution of PV output under different meteorological conditions. After that, a probabilistic model of PV system reliability is established by taking into account the failure rate of PV modules. Then, a Monte Carlo method is established by introducing adaptive sampling to assess the output reliability of the PV system under different meteorological conditions. The data are summarized to obtain the annual PV output reliability. Finally, the rationality and validity of the analysis method are verified by the data from actual PV stations. The results demonstrate that the proposed model and reliability evaluation system can correctly and comprehensively reflect the output reliability of PV power stations.
Research on monitoring strategy for direct power supply of distributed photovoltaic power sources
Shidong CHEN, Shuai YANG, Xing HE, Minqi YU, and Rui HUANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.018
The randomness and intermittency of the output of distributed photovoltaic power sources pose significant challenges to the stable operation and power quality of distribution networks. In this paper, a method based on affinity propagation clustering algorithm (APCA) is proposed to effectively identify the direct power supply situations of photovoltaic power sources. Unlike traditional methods, photovoltaic power sources in the same station area are selected by this algorithm, which reduces the influence of factors such as climate and light intensity and improves the accuracy and operability of the algorithm. First, correlation coefficients are used to screen out photovoltaic power sources with high correlation. Secondly, cluster analysis is conducted on the three-phase grid-connected power curves using the AP clustering algorithm, which further identifies suspected direct power supply and determines possible direct power supply types. Finally, a case analysis is used to verify that the proposed method helps power supply companies better manage distributed photovoltaic power source systems and improves the fairness and transparency of the electricity market.
Multi-dimensional quantitative assessment method for flexibility of hydro-wind-PV multi-energy complementary power generation systems across multiple time scales
Xiangyi CHEN, Quan QING, Chunyi HUANG, Yonglong ZHAO, Ruoyun WANG, Tu XU, and Kai LIAO
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.019
With the gradual transition of the energy system toward clean energy, multi-energy complementary power generation systems with hydro-wind-PV (MCPSs-HWP) have developed rapidly. However, the volatility and uncertainty of wind and PV output, as well as the seasonal variation of hydropower regulation capability, make the dynamic interactions of diversified generation resources across multiple time scales complex, and the key influencing factors of the flexible regulation capability of MCPSs-HWP and their influence mechanisms remain unclear. Therefore, this paper explores the differentiated characteristics of hydropower, wind power, and PV power over long-term and medium- to short-term horizons, and proposes a multi-dimensional quantitative assessment method for the flexibility of MCPSs-HWP across multiple time scales. First, the volatility and uncertainty of wind and PV, as well as the effects of interval inflow, hydraulic relationships, reservoir capacity, and other factors of cascade hydropower on the regulation capability of MCPSs-HWP, are analyzed, and flexibility assessment indices are constructed from three dimensions：regulation range, regulation speed, and regulation direction. Second, based on the balance relationship between power and energy, a flexibility supply-demand balance model across multiple time scales is established, and the long-term and medium- to short-term operational correlation of MCPSs-HWP is constructed. On this basis, a two-stage long-term and medium- to short-term operation model is developed using a chronological operation simulation method. Furthermore, scenario sets of wind and PV output and natural inflow considering volatility and uncertainty are generated based on the Monte Carlo method, and a flexibility index calculation method for MCPSs-HWP under multi-time-scale coupled operation is proposed. Finally, a case study of a river basin in Southwest China is conducted for simulation analysis, and the influence of key factors on the regulation capability of MCPSs-HWP is quantitatively investigated, verifying the effectiveness and practicality of the proposed method.
An evaluation method of carrying capacity of new energy based on two-stage scenario reduction for multi-time-scale time series simulation
Yongming JING, Qiang LI, Linna ZHANG, Hongli LIU, Minru XU, and Weihao WANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.020
To tackle the intricate computat ional challenges arising from the multi-time-scale coupling in evaluation of the carrying capacity of new energy, a novel evaluation methodology is proposed. This approach, grounded in multi-time-scale time series simulation with a two-stage scenario reduction strategy, is designed to achieve precise and efficient evaluation of the carrying capacity of new energy. First, a time-series production simulation model of the power system is formulated, utilizing the minimization of operational costs as its objective function. A penalty mechanism is incorporated to quantitatively constrain photovoltaic curtailment, wind power curtailment, hydropower curtailment, and load shedding. The operational constraints of aggregated thermal power units and pumped-storage hydroelectric units are also fully considered in the model. Subsequently, a two-stage scenario reduction framework based on the general algebraic modeling system-scenario reduction (GAMS-SCENRED) tool is developed. By integrating key metrics such as absolute probability distance, relative probability distance, and marginal relative probability distance, the optimal number of reduced scenarios and their specific configurations are determined. Then, a comprehensive evaluation index system for the carrying capacity of new energy is established. From the perspectives of power supply capability, reliability, and security, this system quantifies the impacts of high-penetration new energy integration on the power grid, thereby providing robust support for grid planning, operation regulation, and decision-making processes. Finally, an actual power grid data from a province in China was used as a case to validate the feasibility and effectiveness of the proposed method. The results show that the proposed method can accurately identify the optimal number of reduced scenarios, significantly enhancing both the evaluation efficiency and accuracy of the carrying capacity of new energy.
Research on multi-time scale fast voltage control strategy for regional power grids with large-scale new energy collection
Zimin ZHU, Xiaoyun WANG, Yu DUAN, Xiaofang WU, Jian MA, and Xiaoyu DENG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.021
To address the problems of frequent voltage fluctuations and significantly increased voltage control complexity caused by high-proportion new energy connected to the power grid, a voltage control strategy considering fast response and fine regulation capabilities is constructed for large-scale new energy collection areas, combined with the requirements of multi-time scale voltage regulation, to realize the joint control between devices with discrete response characteristics (CB) and devices with continuous regulation capabilities (SVC). According to the day-ahead expected results, coarse adjustment of discrete reactive power devices is realized by this strategy in the day-ahead stage; robust reinforcement learning is utilized in the real-time stage to achieve fast adjustment of continuous reactive power devices. First, a non-convex nonlinear stochastic optimization model considering both voltage deviation and economic cost as the objective function is established, and calculation efficiency is improved through model simplification and convex optimization processing. Secondly, in view of the uncertainty of new energy output, an objective function based on the economic cost of continuous device regulation and the improved voltage distribution index is established; a Markov decision process is constructed, and solving is performed by a real-time voltage control model based on robust reinforcement learning to achieve fast response in the real-time stage. Finally, it is verified through an improved case that the proposed strategy effectively improves voltage control accuracy and response speed of the power grid under new energy fluctuation scenarios and possesses strong engineering practicability and promotion value.
Long-term optimal scheduling method for cascade hydro-wind-solar systems based on l2-box ADMM algorithm
Dasong CHEN, Haojie SHI, Yaoxi HE, Bin LIU, Jun YAN, Xiao ZHOU, Yihao GUO, and Rouyu LIN
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.022
The construction of integrated hydro-wind-solar energy bases is a key initiative to promote the development of the new energy industry and achieve energy transition. However, the randomness and intermittency of new energy output pose significant challenges to the long-term optimal scheduling of cascade hydro-wind-solar systems. In response, this paper proposes a coordinated dispatch method for cascade hydro-wind-solar systems that incorporates new energy confidence capacity. First, a definition of new energy confidence capacity incorporating hydropower reserve compensation is proposed and transformed into a mixed-integer linear programming problem. Based on this, the constraints of hydropower units are meticulously modeled. To solve the model, an integer variable relaxation strategy based on the l2-Box ADMM is developed, enabling an equivalent transformation from a discrete-domain model to a continuous-domain model. Finally, the proposed dispatch model and solving algorithm are validated using a complementary system comprising two cascade hydropower stations along with wind and solar power stations. The results demonstrate the high effectiveness and practicality of the proposed method.
Microgrid and integrated energy
Energy optimization management method for microgrids based on safe reinforcement learning
Yunfeng ZHANG, Tao XU, Wen LI, Jingjing LI, Yanhe FAN, and Leijiao GE
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.023
Energy management of microgrids faces the dual challenges of poor adaptability to dynamic environments and insufficient safety in the training process. Traditional model-based energy optimization methods rely heavily on the accurate parameters of microgrids, making it difficult to cope with the dynamic changes of microgrids. A safe reinforcement learning method based on the constrained Markov game is proposed. First, a multi-agent safety boundary constraint including wind turbines, energy storage, and adjustable loads is constructed to limit policy exploration within the preset operation domain; second, an asynchronous safety verification thread is designed to correct the gradient update direction of the policy network in real time; finally, a simulation analysis of the proposed method is conducted using an instance. The research results show that under the premise of ensuring system safety, the proposed method increases the daily profit by 120 yuan compared with other methods, obtains the highest reward value, reduces the wind curtailment volume, and improves the energy storage utilization rate. By decoupling the spatiotemporal correlation between safety constraints and policy optimization, this method provides a scalable safe reinforcement learning paradigm for distributed energy systems.
Active distribution network self-healing method based on microgrid splitting
Zhi ZHANG, Jun HU, Li DING, Fu LI, and Yongwei QIU
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.024
A distribution network can be divided into small-scale microgrids to respond effectively to system disturbances. This flexible structural change not only facilitates fast fault recovery, but also ensures the supply of critical loads. Therefore, a regional microgrid division and adjustable interval self-healing optimization approach for high uncertainty conditions is proposed. First, the existing large-scale distribution network is split into a collection of well-supplied microgrids with sufficient resilience to facilitate the system's self-healing function during faults. Under this framework, the operation of microgrids is decomposed into interconnection and island modes by identifying the combination of heterogeneous renewable generation resources and the configuration of remote control switches. Scheduling optimization for the two modes is carried out separately. After the faults of the system occur, social benefits are maximized through regional segmentation, load reconstruction, power rescheduling, and necessary load reduction operations. Secondly, reliability and resilience needs are considered and a self-healing performance index applicable to interconnection and island modes is proposed to quantify the recovery ability and resilience level of regional microgrids. Furthermore, the self-healing control problem is constructed as a mixed-integer linear programming model, and an adjustable interval optimization strategy is introduced to make the scheduling plan robust under renewable energy prediction errors. The model is solved using a column and constraint generation algorithm to ensure feasibility and calculation efficiency under uncertain conditions. Finally, the effectiveness of the proposed methodology is fully validated through a case study.
Stochastic load disturbance estimation and compensation control method for grid-forming energy storage systems
Xinchi WEI, Chenwei Deng, Chen FANG, Weiyu WANG, and Jinsong LIU
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.025
The grid-forming energy storage system is an effective means to support the independent operation of islanded microgrids. Due to the small overall inertia of the microgrid, its frequency is easily affected by distributed generation and load fluctuations, resulting in a weak disturbance rejection capability of the system. To improve the frequency stability of the microgrid during islanded operation, a stochastic load disturbance estimation and compensation control method for grid-forming converters is proposed. First, an equivalent nonlinear model of the islanded microgrid containing grid-forming energy storage systems is constructed. Then, the Ornstein-Uhlenbeck process is used to describe the stochastic load disturbances in the microgrid. Subsequently, a hybrid observer (HO) is designed by utilizing the active response characteristics of grid-forming energy storage to grid frequency fluctuations and combining with the simplified nonlinear model of the system. This HO is composed of a Luenberger observer and a super-twisting sliding mode observer. Requiring only limited measurement signals, it can quickly estimate the critical states and stochastic load disturbances of the system, thereby enabling the energy storage system to rapidly and accurately compensate for the stochastic load disturbances and ensuring the frequency of the islanded microgrid is within a reasonable range. Finally, the accuracy of the load disturbance observer and the effectiveness of the disturbance compensation controller are verified through time-domain simulations.
Power electronics
A soft-switching high voltage gain DC-DC converter based on coupled inductor
Zhixia BAI, Yushu CHENG, Xiangyuan LYU, Jiayi LIU, and Xuerui ZHANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.026
To solve the problem of voltage mismatch among photovoltaic systems, fuel cells, and high-voltage direct current (DC) buses, a soft-switching high voltage gain DC-DC converter based on a coupled inductor is proposed, which is suitable for non-isolated input-output scenarios. First, the working principle of the converter is analyzed, and the voltage gain of the converter is derived; then, the soft-switching characteristics and voltage stress are analyzed, and a comparison of parameters and characteristics with other converters is provided. Finally, an experimental prototype with an input of 25‒35 V, an output of 200 V, and a power of 500 W is built to verify the effectiveness of the topology and theory. Based on the traditional Boost converter, this converter magnetically couples the Boost inductor with the auxiliary circuit inductor and utilizes the leakage inductance of the coupled inductor and the resonant capacitor to form a resonant tank. It can not only achieve a high voltage gain under non-extreme duty cycle conditions but also effectively reduce the number of magnetic cores and lower the capacitance value. All switches of this converter achieve zero-voltage turn-on, and all diodes achieve zero-current turn-off, which effectively mitigates the reverse recovery problem of the diodes and improves the overall efficiency.
Arc suppression method for grounding fault based on additional energy loop
Ming LI, Chengzhi WEI, Yang LIU, Runhong HUANG, Ruifeng ZHAO, Jiangang LU, Yizhe CHEN, and Zejun HUANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.027
The problems of low utilization and high cost of traditional arc suppression devices can be effectively solved by a multi-function arc suppression device (MF-ASD). However, active power is consumed by the MF-ASD during arc suppression. How to realize the DC-link voltage stability of the active module without an additional power supply unit is worth further discussion. An arc suppression method for grounding faults based on an additional energy loop is proposed. First, the topological structure and working principle of the MF-ASD are briefly described. Second, the influence mechanism of traditional arc suppression methods on the output active power and energy flow direction of the MF-ASD is analyzed, and the internal reason for the change in DC-link capacitor voltage is clarified. Third, an additional energy loop is constructed between the power grid and the MF-ASD to compensate for the active power and reactive power required for the DC-link voltage stability of the MF-ASD during arc suppression, thereby solving the problem that single line-to-ground faults cannot be stably suppressed. Finally, the effectiveness and feasibility of the proposed method are comprehensively verified by simulations and experiments.
Fast identification algorithm for wide-band oscillation parameters based on SFFT
Yongyan CHEN, Liang LUO, I Yue ME, Ying XIA, Ying LU, Zhonghao PAN, and Shuang FENG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.028
As the "dual-carbon" strategy goals are implemented, the oscillations induced by high-penetration power electronic devices are increasingly characterised by the wide frequency domain, strong time-varying and multimodality. The sampling frequency of modern wide-band measurement devices has been substantially increased accordingly to monitor such oscillations. Traditional fast Fourier transform (FFT) parameter identification methods face a contradiction between identification accuracy and identification speed. To this end, a fast identification algorithm for oscillation parameters of multi-modal wide-band signals based on the sparse Fourier transform (SFT) is proposed. Firstly, by utilizing the frequency-domain sparsity of wide-band signals of power systems, multiple non-zero frequency-domain coefficients are hashed into a limited number of buckets to realize computational optimization. Then, multiple iterations are adopted to perform multiple times of spectral rearrangement and localization for wide-band signals. On the premise of maintaining high identification accuracy, the computational efficiency of the identification process is effectively improved. Finally, case studies are employed to verify the proposed algorithm. The results show that the proposed method achieves high precision, fast processing, and improved robustness in identifying multi-modal wide-band signals under the scenarios of high sampling rates.
High voltage and insulation
Identification of frequent partial discharge defects in cables using feature map and multi-attention mechanism-CNN model
Guohua LONG, Qiong LI, Long CHEN, Da XIE, and Song ZENG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.029
The partial discharge (PD) phenomenon is an early indicator of insulation degradation in power equipment, and its frequent occurrence leads to severe failures. Traditional PD detection methods are susceptible to noise interference in complex environments, and their detection accuracy is low. To solve the problem of accurately identifying the PD of typical defects in cable joints, an identification method based on a feature map and a multi-attention mechanism with a convolutional neural network (CNN) is proposed. First, a PD feature matrix map is established, and the features of the time domain, spatial domain, and channel domain are organically combined and visualized, which enables the model to more comprehensively capture the multidimensional information of PD signals. Second, an improved CNN model is constructed by combining the channel attention (CA), spatial attention (SA), and temporal attention (TA) mechanisms. This effectively enhances the perception ability of the model for key features and improves detection accuracy and robustness. Finally, simulation experiments are conducted to analyze the accuracy and effectiveness of the model, and it is compared with other models. The research results indicate that the comprehensive accuracy of the model in PD defect identification reaches 98.89%. The effectiveness of the multi-attention mechanism is verified in an ablation experiment. After the temporal attention mechanism is removed, the identification accuracies of the surface discharge and air gap discharge categories decrease to 97.09% and 91.28%, respectively. Compared with the BP neural network, support vector machine, and random forest models, the performance of this model is more outstanding in all aspects.
Identification of winding deformation and inter-turn short circuit faults in dry-type transformers based on a combined vibration/ultrasound method
Shenjiong YAO, Jialu o CHAI, Yingjing CHEN, Yiyang ZHOU, Lingxuan ZHANG, and Zhousheng ZHANG
Date posted: 4-30-2026
DOI: https://doi.org/10.19781/j.issn.1673-9140.2026.02.030
The insulation structure of dry-type transformers differs from that of oil-immersed transformers, and their oil-free nature makes it impossible to diagnose and forewarn inter-turn insulation faults in real time via chromatographic analysis of gas in oil. To solve this problem, a method based on a combined vibration/ultrasound method is proposed to identify winding deformation and inter-turn short circuit faults in dry-type transformers. First, the ultrasonic and vibration signals generated during winding deformation and inter-turn short circuits in dry-type transformers are collected, and the time-frequency features of the ultrasonic and vibration signals are extracted, including kurtosis, variance, mean, centroid frequency, root mean square frequency, and frequency entropy. Then, combined with the 1DCNN-Transformer, the collection, feature fusion, and classification of the time-frequency features of these two types of signals are achieved to improve the accuracy of fault identification. Finally, the proposed method is verified through simulation analysis. The research results show that the accuracy of the proposed method in identifying winding deformation and inter-turn short circuit faults reaches 98.22%, which is higher than that of traditional identification methods relying solely on vibration signals.
