English Version Journal Login
EN
最新刊期
卷 46 ， 期 1 ， 2022
- 摘要：Electricity theft in the power distribution system is the main cause of non-technical loss of power grids, and it is a chronic problem in operation and management of power utilities. The electricity information acquisition system collects massive user data, which makes it possible to carry out data-driven abnormal electricity detection and accurately pinpoint electricity theft consumers. Affected by the diversity of electricity consumption behaviors of users, the false positive rate of data-driven based electricity theft detection method is still difficult to meet the practical demands in some scenarios, which seriously restricts the engineering application of this method. Firstly, this paper describes the implementation measures of electricity theft, and then sorts out the basic ideas and limitations of the electricity theft detection methods applied in engineering practice and the data-driven based electricity theft detection methods. On this basis, combining with different requirements of engineering application for the evaluation index of electricity theft detection, it is pointed out that the lack of useful extracted information, the low sensitivity and reliability of the characteristic index items are the major reasons that hinder the data-driven based electricity theft detection methods from being practical in engineering. Finally, the electricity theft detection with low false positive rate is prospected from different levels such as algorithm design, state space subdivision, and design and selection of characteristic index items.关键词：electricity theft detection;low false positive rate;data-driven;feature engineering;state space3|1|0更新时间：2025-06-16
- 摘要：With the expansion of AC/DC hybrid power system and the large-scale grid-connection of power electronic equipment, the dynamic characteristics of novel power system dominated by renewable energy become more and more complex. The mechanism interpretability of the physical model is highly complementary to the characteristic fitting ability of the data model. How to leverage the construction of fusion model from qualitative analysis to quantitative analysis needs in-depth research. Based on the characteristics of data methods and physical methods in the power system, the corresponding application scenarios of four typical data-physical fusion models are analyzed. Taking parallel mode as the research object, the generalization errors of parallel mode, physical only model and data only model are compared and analyzed respectively, and the selection method of fusion model parameters is proposed. The generalization error upper bound of the fusion model in parallel mode is derived, and the feasible suggestions to improve the performance of the fusion model are put forward. Finally, combined with the transient power angle stability problem, the rationality of the proposed assumptions and conclusions is verified.关键词：novel power system;data-driven;data-physical fusion model;generalization error;transient power angle3|0|0更新时间：2025-06-16
- 摘要：Synchrophasor measurement units (PMUs) can provide real-time measurement data for power system monitoring and control. However, continuous bad data of PMU can be highly similar to the disturbance data, which may cause the control center to make wrong decisions. Aiming at the problem that continuous bad data of PMU is difficult to be distinguished from the disturbance data, a continuous bad data detection algorithm is proposed based on the local outlier factor (LOF). Through analysis of a large amount of measured data, it is concluded that the continuous bad data has poor spatial similarity, and the disturbance data has strong spatial similarity. According to this conclusion, a method for spatial similarity evaluation is proposed based on dynamic time warping (DTW). The LOF value of each PMU is calculated by evaluating the spatial similarity between different PMUs. Further, a method for determining the threshold value based on the box-plot is proposed. By checking whether the LOF value of each PMU in the current window exceeds the threshold, the continuous bad data can be detected online. Simulation and testing results illustrate that the proposed method can achieve the efficient continuous bad data identification and detection, and the disturbance data can be distinguished.关键词：synchrophasor measurement unit;continuous bad data detection;dynamic time warping;local outlier factor3|0|0更新时间：2025-06-16
- 摘要：With the wide application of advanced communication and information technology, the false data injection (FDI) attack has become one of the important factors which threat the security of automatic generation control (AGC) system. The detection of cyber-attacks is a priority for defense. This paper proposes a method of FDI attack detection for AGC system based on the set-membership filtering. Firstly, this paper analyzes the influence of the FDI attacks in AGC systems, and establishes a model of AGC system of interconnected power grids and a model of FDI attacks. Secondly, based on the real-time control commands and measurement data, the prediction update and measurement update of the AGC system ellipsoid set are performed. By determining whether there is an intersection between the prediction update ellipsoid set and the measurement update ellipsoid set, the possible FDI attacks in the data transmission of the system are detected. Finally, the effectiveness of the proposed method is verified in the IEEE dual-area interconnected grid.关键词：false data injection attack;automatic generation control;set-membership filtering;cyber-attack detection4|0|0更新时间：2025-06-16
- 摘要：The power grid and communication network are highly coupled in the distribution Internet of Things (DIoT). The anomalies of a single network interact with another network, which may lead to the expansion of anomaly range. However, it is difficult to comprehensively and accurately identify the types and locations of the anomaly source in the DIoT by using the information of the power grid or the communication network alone. Therefore, this paper proposes an anomaly type identification and location method for DIoT based on the three-dimensional convolutional neural network (3D-CNN). Firstly, the communication flow characteristics of DIoT are analyzed and an interactive simulation model of DIoT based on Simulink and OPNET is built. Secondly, a 3D-CNN oriented sample construction method is proposed, in which the electric parameters and communication flow information of each node in DIoT are composed into a feature sub-pixel, and the state of DIoT at each moment is represented as a feature frame, thus forming a cubic sample matrix that contains the anomaly process of DIoT. Thirdly, a deep learning model is built, which includes the three-dimensional feature extraction network and the hierarchical softmax classifier. By extracting and identifying the abnormal information hidden in the cubic sample matrix, the type and location of anomalies in DIoT could be determined simultaneously. Finally, the model is tested by using abnormal data of the IEEE 33-node DIoT. The results show that the proposed method can precisely classify and locate the short-circuit fault, communication interruption fault, and protection maloperation and rejection caused by abnormal communication data.关键词：distribution Internet of Things (DIoT);deep learning;interactive simulation;anomaly identification and localization;three-dimensional convolutional neural network (3D-CNN)3|0|0更新时间：2025-06-16
- 摘要：The increase of the penetration of renewable energy brings continuous challenges to the safe and stable operation of the power system. It becomes more and more difficult to analyze the system stability and control the stable operation of the power system by traditional methods. To solve this problem, a power system optimal operation framework with embedded security and stability constraints and an oblique regression tree and its ensemble algorithm for extracting power system security and stability rules are proposed. The algorithm first optimizes the oblique split coefficient to train a single oblique regression tree, then uses the boosting idea to integrate the oblique regression tree, and uses the regularization method to ensure the sparsity of the tree and enhance the interpretability of the algorithm. Compared with the black box model such as neural network, the proposed method can extract explicit security and stability rules, which lays a foundation for the optimal operation of the power system with embedded security and stability constraints. Finally, the static voltage stability problem is taken as an example to verify the effectiveness of the algorithm. The results show that the algorithm has good interpretability, strong representation ability and high ensemble efficiency.关键词：static voltage stability;high proportion of renewable energy;ensemble learning;oblique regression tree;eXtreme Gradient Boosting (XGBoost)3|0|0更新时间：2025-06-16
- 摘要：While the flexibility of power system operation can be improved by topology optimization, the dimension of system-level discrete decision variables, including the connections of lines and substation busbars, is prohibitively high. Thus, the topology optimization problem of power systems can hardly be solved by the conventional mixed-integer optimization method. Aiming at this problem, a reinforcement learning based method is proposed combining asynchronous advantage Actor-Critic (A3C) and power system domain knowledge, which transfers the computational burden of online optimization to the offline agent training stage. The defined reward function is adopted to minimize the violations of power transmission line flow limits. Forced constraints verification is employed to reduce the searching space and improve the efficiency of the reinforcement learning. The fast computation of the topology optimization of power system operation is realized,and the operation security of power systems is enhanced. The effectiveness of the proposed method is validated by simulation testing results.关键词：power system operation;deep reinforcement learning;topology optimization;power output adjustment3|1|0更新时间：2025-06-16
- 摘要：When the transmission section limit is formulated by adopting the traditional operation modes, the worst operation modes of power grids are mainly considered to ensure the safety, which can not adapt to the complex and variable characteristics of the operation modes of the power grid with high proportion of renewable energy. There are the problems such as insufficient operation mode scenarios and too large safety margins, etc. Therefore, a clustering analysis method for the transmission section modes of the power grid with renewable energy is proposed. Firstly, the key sections of the power grid are discovered based on the historical data statistics, and the relevant factors that affect the transmission power of the key sections of the power grid are discovered by correlation analysis. Secondly, the key factors are sorted according to the correlation index, and the operation modes of the power grid are clustered according to the key factors of the top ranking. Thirdly, according to the clustering results, the transmission section limits at each level of the boundary modes are optimized to increase the transmission capacity of the power grid. Finally, the effectiveness and practicability of the proposed method are verified by the simulation analysis of Northwest China power grid.关键词：power grid;operation mode;clustering;section limit optimization4|1|0更新时间：2025-06-16
- 摘要：In the scenarios requiring large-scale repetitive power flow calculations, such as N
- 摘要：As a representative deep learning algorithm, the convolutional neural network (CNN) has been used in the transient voltage stability assessment of the power system. However, the construction method and rationality verification for the input features of CNN have not been thoroughly studied. For the transient voltage stability assessment of an AC/DC system, a CNN based construction method for input features is proposed. Firstly, a two-stage partition scheme is used to reduce the dimensionality and redundancy of the input features, i.e., the initial partition result is first given according to the system topology and geographic location constraints, then aggregated by the similarity of transient voltage characteristics to obtain the best partition scheme with low dimension and redundancy. Secondly, based on the partition result, the key factors affecting the transient voltage stability of the AC/DC system are investigated and the input features that take in account both the steady-state characteristics and multi-dimensional fault information are constructed. Finally, the constructed input features are applied to the CNN transient voltage assessment model and verified by actual power grid data. Simulation results demonstrate that the proposed method has higher accuracy than the traditional feature selection methods.关键词：transient voltage;stability assessment;convolutional neural network;power grid partition;input feature3|2|0更新时间：2025-06-16
- 摘要：The ambient data under the natural excitation of the system contains rich electromechanical characteristic information. Accurate identification of small signal stability characteristic parameters from ambient data is of great practical significance for guiding the safe and stable operation of the power system. An on-line identification algorithm of small signal stability characteristic parameters driven by ambient data based on subspace optimal mode decomposition (Sub-OpMD) is proposed. The algorithm obtains the singular sub-matrix by the matrix linear transformation of the input data based on the orthogonal projection, and the conjugate gradient algorithm is used to iteratively solve the optimal low-dimensional orthogonal space to achieve the optimal low-dimensional approximation of the high-dimensional mapping matrix between singular sub-matrices. According to the eigenvalue decomposition result of the optimal low-dimensional matrix, the small signal stability characteristic parameters of the system can be accurately obtained, i.e., the oscillation frequency, damping ratio and mode. The matrix linear transformation based on orthogonal projection and the introduction of conjugate gradient method make the dynamic mode decomposition (DMD) method better adapt to the ambient data. The calculation and analysis of IEEE 16-generator 68-bus system and actual system measured data verify the effectiveness and accuracy of the proposed algorithm.关键词：ambient data;small signal stability characteristic parameter;subspace optimal mode decomposition (Sub-OpMD);orthogonal projection;conjugate gradient algorithm3|5|0更新时间：2025-06-16
- 摘要：Aiming at the problem of low coverage of real-time measurements and uncertainty of probability distribution of measurement errors, an interval state estimation method based on a pseudo measurement modeling method using deep neural networks (DNN) is proposed for regional AC/DC distribution network. Firstly, DNN is trained offline in this method. Then the real-time measurement data and the variable values controlled by VSC are used as the input features of DNN to establish a pseudo-measurement model. Secondly, the trained DNN is used to generate the pseudo measurements quickly when the real-time measurements are updated. Finally, the uncertainty of the pseudo-measurement and the real-time measurement is modeled in the interval form and the interval state estimation is carried out in order to accurately monitor the states of the AC/DC distribution system. The simulation results of the calculation example show that the proposed method can avoid the assumptions about the probability distribution of the measurement errors, and it can obtain the accurate upper and lower bounds of the state variables in the case of low real-time measurement redundancy or insufficient configuration.关键词：AC/DC distribution network;interval state estimation;deep neural network;pseudo measurement;uncertainty3|1|0更新时间：2025-06-16
- 摘要：Alongside the wide proliferation of distributed energy resources at the residential sector, how to meet the needs of real-time autonomous energy management while considering the heterogeneous operating characteristics of these resources so as to maximize the utility for residential end-users deserves significant research attention. In this area, conventional model-based optimization methods are generally burdened with inaccurate system modeling and inability to efficiently deal with uncertainties stemmed from multiple sources. In order to address these challenges, this paper proposes a model-free method based on deep reinforcement learning to achieve real-time autonomous energy management optimization. First, the user's resources are classified into different categories, their operating characteristics are then described using a unified 3-element tuple, and the associated energy management actions are also identified. Next, the long short-term memory neural network is employed to extract the future trends of multi-source sequential data from the environment states. Then, based on the proximal policy optimization algorithm,it enables efficient learning of the optimal energy management policies in the multi-dimensional continuous-discrete mixed action space, which can adaptively adjust to system uncertainties towards the user's electricity cost minimization objective. Finally, the effectiveness of the proposed method is verified by benchmarking its performance against several existing methods through case studies on an actual scenario.关键词：real-time autonomous energy management optimization;uncertainties;continuous-discrete mixed actions;long short-term memory neural network;deep reinforcement learning3|3|0更新时间：2025-06-16
- 摘要：Accurately identifying the parameters of the air-conditioning load model is an important basis for tapping its energy saving and demand response potential. Current studies mostly adopt offline identification methods with low accuracy. Therefore, a data-driven online identification method of model parameters for air-conditioning is proposed. Firstly, the framework of data-driven online identification for parameters of the air-conditioning load model is established. Secondly, based on the air-conditioning load model, the data-driven online identification mechanism and method are proposed. Among them, the data-driven online identification mechanism is designed as the parameter update discrimination mechanism based on event-driven significant changes in parameters, and the dynamic threshold setting mechanism for parameters based on fluctuation ranges of historical parameters. Under this mechanism, a fast online identification method is established through the particle swarm optimization algorithm. Finally, the effectiveness of the proposed online identification method is verified in a practical experimental environment. Compared with offline identification methods, the proposed method improves the calculation speed and accuracy more greatly, and can meet the needs of online applications.关键词：inverter air-conditioning;online identification;data-driven mechanism;second-order equivalent thermal parameter model3|0|0更新时间：2025-06-16
- 摘要：A detection method of hidden danger for power utilization of low-voltage users based on local outlier factor (LOF)-Boosting is proposed by using voltage measurement data of the electricity information acquisition system. First, a method of voltage information reconstruction based on information entropy is used to expand the difference of voltage information. Second, a voltage data sparse coding method based on K-singular value decomposition (K-SVD) is presented to deal with the redundancy problem caused by the excessively high dimension of the original load characteristics. Then, the accuracy and the generalization ability of hidden danger detection for power utilization of low-voltage users are improved by optimizing the multiple LOF combination models. Finally, a case study of a certain low-voltage distribution area in Zhejiang Province of China demonstrates that the proposed method can achieve a higher detection accuracy than the traditional LOF algorithm.关键词：low-voltage distribution area;hidden danger detection;K-singular value decomposition (K-SVD);sparse coding;local outlier factor-Boosting3|0|0更新时间：2025-06-16
- 摘要：The traditional harmonic state estimation based on the least-square method is limited by few measuring devices, difficulty in obtaining accurate harmonic impedance, complex network topology and variation of grid operation mode, which results in the problems such as underdetermined measurement equations, non-global observation of the power system and difficulty in accurately extracting coupling relationships between buses. This paper proposes a harmonic state estimation method based on the improved generative adversarial network. The method is based on the pix2pix harmonic state estimation network to fit the coupling relationship between the monitored buses and the target buses. And the collected historical harmonic data is used to batch train the model. The harmonic current and harmonic voltage amplitude of the target buses are estimated by the generated network after training to achieve data-driven harmonic state estimation. The model is tested in an environment with noise, and the simulation results verify the effectiveness of the proposed method.关键词：harmonic state estimation;deep learning;generative adversarial network;pix2pix;convolutional neural network3|1|0更新时间：2025-06-16
Special Topic on Data-Driven Methods for Modernizing Power Systems
-
Optimization Method of Resource Combination for Virtual Power Plant Based on Modern Portfolio Theory
摘要：For the optimal combination of heterogeneous resources in virtual power plants, a response characteristic model including demand response, flexible controllable load, and distributed power generation is proposed. Based on the control authority of the virtual power plant over interactive resources, the interactive resource response power is divided into risk-free and risk assets. Then the resource response allocation problem is mapped to the asset weight allocation problem in the modern portfolio theory. The portfolio risk is used to measure the uncertainty of the user’s response to electricity, and the optimal combination model is established with the goal of maximizing the expected return. Based on the decentralization of interactive resources, the virtual power plant is involved in congestion management, and the proposed method is verified on the CIGRE 20 kV medium-voltage distribution network. Simulation results verify the practicability of the proposed model.关键词：virtual power plant;congestion management;resource combination;distributed energy resources3|2|0更新时间：2025-06-16 - 摘要：In the future power system where renewable energy is predominant, the energy management of multiple virtual power plants (VPPs) will be more complicated and challenging. Considering the characteristics of blockchain technology including information security, transparent mechanism, and regional co-governance, a blockchain consisting of commercial chain, technical chain and asset chain is designed inspired by the functional characteristics of VPPs. Then, comprehensively considering the association and conflict of the interest of the multiple decision-making entities formed by multiple VPPs, a Stackelberg game based optimal operation method is introduced with the network security constraints. On this basis, the optimal operation mode of multiple VPPs based on the blockchain technology is analyzed and designed from the aspects including the network level framework, interaction mechanism, smart contracts, and incentive mechanism. Finally, case studies are conducted to verify the effectiveness of the proposed method.关键词：multiple virtual power plants;Stackelberg game;network security constraint;blockchain technology;operation mode4|2|0更新时间：2025-06-16
- 摘要：The power generation of upstream hydropower stations directly affects the power generation of downstream power stations through the upstream and downstream hydraulic connection. This hydraulic connection increases the difficulty in establishing and solving the spot market clearing model. Therefore, a clearing model for the cross-regional transaction spot market coupled with complex hydraulic connection is proposed. Firstly, the nonlinear constraints are linearized in this model. Secondly, the cascaded hydropower family is introduced, and the complex hydraulic connection is transformed into the maximum and minimum power generation of cascaded hydropower stations. Furthermore, the hydropower units participate in the spot market clearing by means of the equivalent thermal power units which meet the maximum and minimum power generation constraints of the hydropower station groups, so the complex hydraulic connection is eliminated. Taking the hydropower transmission from Yunnan Province to Guangdong Province in China as an example for verification, the results demonstrate that the clearing results can meet various operation constraints of the units, and can reflect the market price signals truthfully; the congestion components of the power generation of cascaded hydropower stations can reasonably represent the congestion cost caused by the basin inflow.关键词：spot market;clearing model;locational marginal price (LMP);cascaded hydropower family;Lagrangian duality multiplier3|1|0更新时间：2025-06-16
- 摘要：In view of the unclear quantification of the substitution ability of battery energy storage systems with the same power for the frequency regulation by thermal power units, this paper analyzes the principle of frequency regulation of battery energy storage systems and thermal power units, which determines the frequency regulation advantages of battery energy storage systems in terms of frequency response stability and rapidity. Starting from the difference coefficients, the influence of energy storage on the frequency regulation system of high-proportion renewable energy is analyzed, and the advantages of energy storage with the same power in improving the frequency regulation effect by thermal power units are determined. Then, the power and energy substitution indices for energy storage frequency regulation are defined, and a substitution index for energy storage to increase the nadir of the frequency is proposed. Finally, according to the related quantitative indices, the regional power system in the east of China is taken as an example to verify the high substitution ability of battery energy storage systems for thermal power units by analyzing the substitution ability of energy storage with multiple time scales in different disturbance scenarios.关键词：battery energy storage;thermal power units;frequency regulation;substitution ability;difference coefficients3|3|0更新时间：2025-06-16
- 摘要：In recent years, the level and scale of power grid interconnection have developed rapidly, integration the uncertainty and the regulation and control requirement of the power system keep increasing with the large-scale integration of new energy and power electronic equipment. The current power grid regulation and control method has the problems such as dependence on the abnormal triggering of the experience characteristic quantity and the lack of the regulation and control means for proactivity and predictability based on the out-of-limit of the tie line. In order to solve the problems, a method of risk assessment and proactive regulation and control is proposed for power grid operation safety based on probability prediction. Firstly, a rolling probability prediction model based on long short-term memory (LSTM) network and support vector machine (SVM) is constructed. Then, the severity function of common risk events is established from the perspective of sufficiency to achieve the over-limit probability prediction of key elements. Also, the quantitative risk is calculated to form a trigger mechanism so that the proactive regulation and control of the power system for risk events is realized. Finally, a simulation is carried out on the IEEE 39-bus system combined with the actual load data of a certain provincial power grid in China. The calculation results verify that the proposed method and model can realize the proactive regulation and control in advance and effectively avoid the risks of safe operation.关键词：probabilistic prediction;risk pre-warning;proactive regulation and control;long short-term memory (LSTM) network;support vector machine (SVM);time series prediction;machine learning;data-driven3|3|0更新时间：2025-06-16
Basic Research
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0