ISSN 1000-1026

CN 32-1180/TP

+Advanced Search 中文版
    • Online First
    • |
    • Current Issue
    • |
    • Special Articles
    • |
    • Discussion Corner
    • |
    • Archive
      Select All
      • QIN Chuan, DING Pengfei, LIU Bo, JU Ping

        Available online:October 26, 2021  DOI: 10.7500/AEPS20210524009

        Abstract:The summer peak-load is the focus of the power sectors, and temperature cumulative effect has an important impact on the summer peak-load forecasting. Therefore, a feature decoupling peak-load forecasting model considering meteorological cumulative effect is proposed. The deep neural network model with three input branches is established to realize the decoupling of input features from the structure, which is called feature decoupling model. The three branches take the time feature, load feature and meteorological feature as the input respectively. Branches for the load and meteorology apply the long short-term memory (LSTM) network hidden layer, they process the load and meteorological sequences based on the time sequence processing ability of LSTM network to reflect cumulative effect, and then apply it to the peak-load forecasting. Finally, through case analysis and comparison with conventional methods such as temperature correction method, it is verified that the feature decoupling model is more suitable for peak-load forecasting considering cumulative effect.

      • YU Chen, LI Shangxuan, XIE Yunyun, LIU Shaofeng, CHANG Kang

        Available online:October 26, 2021  DOI: 10.7500/AEPS20210726005

        Abstract:Typhoon disasters often result in multiple physical failures in the distribution network. A reasonable rush repair strategy can reduce the time of outrage and improve the reliability of the distribution network. The existing research mainly focuses on the dispatch of repair resources, including repair crews, materials and the operation of the distribution network. However, the current repair strategy fails to consider the impact of traffic condition, which would affect the implementation effect of the strategy. With the improvement of power grid information collection capabilities, power grid has the ability to obtain traffic information. Therefore, this paper proposes a method to optimize the rush repair strategy for the distribution network after typhoon, which considers the information integration of the traffic network and the distribution network. Based on the construction of an information integration framework, a two-layer optimization model for rush repair strategies considering the interruption of the traffic network is established to coordinate and optimize repair crews, repair materials, mobile power sources, network reconfiguration and other resources to reduce the economic losses. Finally, the effectiveness of the proposed method is verified by example simulation.

      • XU Chaoran, XU Xiaoyuan, YAN Zheng, LI Hengjie

        Available online:October 26, 2021  DOI: 10.7500/AEPS20210413003

        Abstract:A power system dispatch method based on data-driven distributionally robust optimization (DRO) is proposed to deal with the operation problem of power system considering the uncertainty of wind power. Firstly, the statistical information of wind power data is mined, and a construction method of probability distribution ambiguity set of wind power based on principal component analysis and kernel density estimation is proposed to describe the randomness of wind power and the spatial correlation between the outputs of different wind turbines. Secondly, aiming at the dispatch problem with wind power, a two-stage DRO problem considering probability distribution ambiguity set is established. Thirdly, the DRO problem is transformed into its equivalent solvable form, and the affine strategy and duality principle are used to transform it into a linear programming problem for solving. The range parameter selection strategy of probability distribution ambiguity set based on out-of-sample test is proposed to ensure the reliability and economy of the dispatch scheme. Finally, the 6-bus and IEEE 118-bus systems are used for simulation analysis, and the proposed DRO method is compared with the DRO method without considering the correlation of random variables and the traditional stochastic and robust optimization methods to verify the effectiveness of proposed method.

      • ZHU Lin, ZHAO Xueshen, GUO Li, LI Xialin, DENG Zhanfeng, WANG Chengshan

        Available online:October 26, 2021  DOI: 10.7500/AEPS20210622004

        Abstract:Due to the dynamic interaction between converters, the design of control parameters of medium-voltage DC (MVDC) distribution system is much more complicated than that of single-converter situation. First, for the current control loop, the voltage control loop and droop control loop of the converter in the system, the open-loop and closed-loop transfer functions considering the dynamic interaction between the converters are established respectively. In addition, the number of poles (or zeros) of the transfer function considering the dynamic interaction between converters is consistent with the single-converter situation. Then, a control parameter design method for MVDC distribution system considering the dynamic interaction between converters is proposed. In this design method, the control parameters of all converters are designed qualitatively and quantitatively from the perspective of system dynamic stability. Through the mutual cancellation of some zeros and poles, the dynamic characteristics of the system can be designed by this method with a pair of conjugate poles. Moreover, this design method is effective in a wide oscillation frequency range (such as 10~50 Hz). Finally, the proposed transfer function and control parameter design method are simulated and verified in the power electronics simulation software PLECS.

      • ZHU Renlong, ZHOU Xiaoping, LUO An, HONG Lerong, XIA Haitao, DENG Lingfeng

        Available online:October 26, 2021  DOI: 10.7500/AEPS20210324005

        Abstract:Advanced firing control is an important method to reduce the probability of commutation failure in high-voltage direct curren (HVDC) transmition ystem. Accurate advanced firing is the key to mitigate commutation failure. According to the theory of commutation voltage-time area, a commutation failure mitigation method based on commutation voltage-time area prediction is proposed. This method not only takes into account the change of DC current after fault, but also indirectly takes into account the change of commutation angle and equivalent commutation reactance after fault through the measurement of commutation area. At the same time, this method has clear physical significance and does not need to improve the commutation failure mitigation performance through complex parameter tuning. Finally, taking the CIGRE standard model in PSCAD/EMTDC as the test model, the effectiveness of the proposed method is verified under different faults.

      • CAI Xu, YANG Renxin, ZHOU Jianqiao, FANG Zixi, YANG Mingyang, SHI Xianqiang, CHEN Qing

        Available online:October 21, 2021  DOI: 10.7500/AEPS20210909007

        Abstract:With the rapid development of deep and distant offshore wind power, the transmission and the grid integration of wind power via HVDC systems have become technical hotspots. Focusing on the several vital technologies such as the AC collection-based DC transmission, the low-cost DC transmission, the multi-terminal DC transmission and the multi-voltage-level DC transmission, the paper comprehensively discusses the current status, existing problems, research hotspots and development trends of the offshore wind power via DC transmission technologies in terms of system topology, equipment, control and protection. It is pointed out that the transmission and grid integration of offshore wind farms via the centralized HVDC system is the mainstream scheme in the near future. For such a system, the broadband oscillation is an urgent problem to be solved, and to make the whole system perform the characteristics of dominant power sources is the focus of attention. In terms of cost reduction, the technical route of the diode rectifier based HVDC transmission system has good expectations. However, to use pure diode rectifiers as the sending-end converter requires the improvement of wind turbines. For the multi-terminal DC grid with wind power integration, the development depends on the progress of low-cost DC circuit breaker and transient control and protection technology. And the offshore wind power transmission and grid integration via multi-voltage-level DC grid still lacks the support of the DC transformer and other key equipment.

      • ZHOU Jianli, WU Yunna, DONG Haoxin, HE Jiaming, XU Chuanbo, GAO Jianwei

        Available online:October 20, 2021  DOI: 10.7500/AEPS20210614005

        Abstract:The development potential of the integrated energy system coupled with renewable energy power generation and hydrogen energy storage is huge, which is an effective way to achieve clean and low-carbon energy consumption. Taking the natural gas-wind-photovoltaic-hydrogen integrated energy system as the planning object, the stochastic charging demand of electric vehicles in the radiation area is considered. Gaussian kernel density estimation and K-means clustering algorithm are used to characterize the uncertainty of wind-photovoltaic output and electric vehicle charging demand, and to generate typical probability scenes. The capacity allocation of the system is optimized based on mixed-integer linear programming with the minimum annual comprehensive cost and annual carbon emission as the optimization objectives. The weight coefficients are introduced to characterize the optimization preferences of decision-makers. Capacity configuration of system is optimized based on mixed integer linear programming. The case study is based on annual time-to-time data to verify the validity and reasonableness of the model. Compared with the traditional sub-distribution system and interconnected case, the system benefits from the cooperation of wind-photovoltaic power generation and hydrogen energy storage, showing considerable environmental benefits. It is economical and feasible for the system to provide charging service, and charging electric vehicles in the system can significantly reduce indirect carbon emissions.

      • WU Chengxin, SHEN Haijun, WANG Zhihua, HUANG Ting, FAN Shuai, HE Guangyu

        Available online:October 20, 2021  DOI: 10.7500/AEPS20210405003

        Abstract:Accurately identifying the parameters of the air-conditioning load model is an important basis for tapping its energy saving and demand response potential. Current researches mostly adapt offline identification methods with poor accuracy. For this, based on the idea of data-driven, a online identification method of model parameters for air-conditioning is proposed. Firstly, the framework of data-driven online identification forparameters of 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 parameter update discrimination mechanism based on event-driven significant changes in parameters, and 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 greatly improves the calculation speed and accuracy, and can meet the needs of online applications.

      Select All

      Volume 45,2021 Issue 20

        >Cyber Energy Systems
      • CAI Ye, LIU Fang, CAO Yijia, CHEN Yang, LIU Ying

        2021,45(20):1-8, DOI: 10.7500/AEPS20210318002

        Abstract:The deep integration of the cyber layer and the physical layer is an important feature of a smart grid. Attacks carried out using cross-domain relationships bring new risks to the stable operation of the power grid. It is of great significance to study the high-risk attack strategy for improving the ability of power grids to resist attacks considering the attack cost factors. Firstly, based on the coupling relationship between cyber layer and physical layer, an index of line attack cost is defined by combining the structure characteristics and operation characteristics, so as to identify the low-attack-cost lines with weak protection measures in the power grid, which are vulnerable to attacks. Then, the multi-stage attacks on low-attack-cost lines are carried out to induce blackouts, and the damage amplification mechanism of the multi-stage attack under the interaction of cyber and physical systems is analyzed. Finally, an IEEE 118-bus system and a provincial power system are taken as examples for simulation analysis. Simulation results show that multi-stage attacks on low-attack-cost lines can cause blackouts,and the attack cost is obviously lower; the increase of failure probability of information nodes will improve the effectiveness of the attack strategy, and information network plays an important role in preventing blackouts.

      • WANG Ziyu, WANG Jingyu, XIE Jun, WANG Yingying, SHI Dongyuan, DUAN Xianzhong

        2021,45(20):9-17, DOI: 10.7500/AEPS20210108003

        Abstract:The regular operation of the power communication system is the basis of ensuring the reliable transmission of measurement data and control commands, and maintaining the safe and stable operation in power systems. The conventional static models of power communication systems ignore the differences in the functions of different buses, making it hard to characterize the multi-layer connection status of physical and logical links simultaneously. Moreover, it is also inapplicable to precisely describing the changing situation of information flow under various cyber-physical contingencies. Regarding these issues, a static layered modeling method of power communication systems for data availability is proposed. From the physical layer to the session layer, the proposed method employs direct/indirect adjacency matrices and the cross-layer dependencies to model the transmission and processing of the information from bottom to top in detail. The yielded model can more comprehensively analyze the changes of the information flow under cyber-physical contingencies, and thus lay the foundation for more accurate interaction impact analysis and risk assessment in cyber-physical power systems. The effectiveness of the proposed method is verified through the analysis of cyber-physical co-simulation in the IEEE 9-bus system and IEEE 118-bus system under complex contingencies.

      • >Basic Research
      • LAN Puzhe, HAN Dong, XU Xiaoyuan, YAN Zheng, LIU Xutao

        2021,45(20):18-28, DOI: 10.7500/AEPS20201106004

        Abstract:With the development of the national energy revolution strategy and the power system reform in China, the integrated energy system has gradually emerged as a new energy development form that integrates various energy attributes to maximize the energy utilization efficiency. However, the current integrated energy system has the problems of low redundancy of measurement data, large measurement error of measurement equipment, and inconsistent unit time scale of data acquisition of the measurement equipment in the power grid and gas network. These problems bring severe challenges to the state estimation problem of the electricity-gas coupled integrated energy system. Considering the high portability of data-driven methods and the ability to extract and summarize different information, a Bayesian state estimation model for electricity-gas coupled integrated energy systems is established based on the long short-term memory. The complete measurement data is generated by Monte Carlo sampling using Bayesian learning to obtain the probability and statistical characteristics of measurement data. And the rationality of the generated data is verified by the power flow of the electricity-gas coupled integrated energy system. Thus, the training sample set of the deep learning network based on the long short-term memory is obtained. The evaluation standard of root mean square error is used to train the deep learning network of the long short-term memory, and the accuracy of state estimation in the electricity-gas coupled integrated energy system is effectively improved. Compared with the classical model-driven state estimation method, the simulation verifies the effectiveness and robustness of the proposed data-driven state estimation method.

      • HUANG Yu, XU Qingshan, XIA Yuanxing, YUE Dong, DOU Chunxia

        2021,45(20):29-37, DOI: 10.7500/AEPS20210126008

        Abstract:Large-scale wind power integration increases the probability of transmission congestion in power grids. In order to overcome the disadvantages of deterministic methods for congestion management when facing system random perturbations, based on chance-constrained programming conditions considering congestion risk, this paper proposes an adaptive approach for quantifying transmission reserves by taking the wind power uncertainty into account. A bi-level stochastic scheduling model with adaptive transmission reserves is developed. This model incorporates the line transmission reserve into the upper-level unit commitment (UC) model. Aiming at the economic rescheduling problem at the lower level, the improved point estimation method is used to obtain the statistical distribution characteristics of the real-time power flow of the line, and the result is returned to the upper model to dynamically adjust the size of the reserve demand. At the same time, based on the Karush-Kuhn-Tucher (KKT) optimal condition of the underlying model, the bi-level optimal structure is merged into a single layer, and finally transformed into a mixed-integer programming problem that is easy to solve. Finally, the IEEE RTS 24-bus testing case verifies that the proposed model and method can effectively alleviate transmission congestion and improve wind power consumption and reserve availability.

      • MA Lan, XIE Lirong, YE Lin, LU Peng, WANG Kaifeng

        2021,45(20):38-48, DOI: 10.7500/AEPS20200904002

        Abstract:Aiming at the problems of serious wind abandonment, large prediction error and large fluctuation rate of wind power, a cooperative operation strategy of dual energy storage considering multi-objective and multiple operation conditions is proposed. According to different operation conditions, the strategy of energy storage for compensating the negative prediction error within the time scale of 15 minutes is formulated, and the strategy of energy storage for suppressing the wind power fluctuations within the time scale of 5 minutes is set. Two operation modes are designed according to the proposed strategies. On this basis, considering the influence of the cooperative operation strategy of the dual energy storage system on capacity allocation, the optimal allocation model of energy storage capacity is established with the objective function of maximizing the compensation of negative prediction error and suppressing wind power fluctuation. The particle swarm optimization algorithm combined with genetic characteristics is used to solve the problem. A 50 MW wind farm in Xinjiang Uygur Autonomony Region of China is used to verify the rationality of the proposed strategy and the effectiveness of the model solving method.

      • SHI Jinkai, BAO Yan, CHEN Zhen, JIANG Zhenyu, ZHANG Weige

        2021,45(20):49-58, DOI: 10.7500/AEPS20210302005

        Abstract:Aiming at the characteristics of the fluctuation and uncertainty of the charging load in the charging station, a robust optimization configuration method for energy storage of the charging station considering the charging load uncertainty is proposed. Firstly, a polyhedral uncertain set is established to describe the fluctuation and uncertainty of charging load of electric vehicles. Then, the optimal configuration model of energy storage system capacity is established. At the same time, the variable life model of the energy storage battery is used in the optimal configuration model to minimize the comprehensive cost of the energy storage system under multiple constraints, so that the charging station can cope with the random fluctuation of load with less comprehensive cost. Finally, based on the robust optimization theory, the uncertain model is transformed from the stochastic parameter optimization model to the deterministic parameter model, and the optimal configuration parameters of the energy storage system are obtained by combining the mesh adaptive direct search algorithm. Compared with the deterministic algorithm, the feasibility of the proposed method is verified by a case.

      • ZHU Yongqiang, CAI Qinqin, GONG Ping, XIA Chunni, YU Zicong, XIA Ruihua

        2021,45(20):59-66, DOI: 10.7500/AEPS20210331004

        Abstract:A universal mathematical description and a unified energy flow calculation method suitable for different energy forms are the theoretical basis for the planning and operation of multi-energy networks. On the basis of constructing the three-dimensional topological diagram for the multi-energy network, this paper proposes a unified description for energy transfer, and constructs the energy transfer equations of branches in the multi-energy network by borrowing the concepts of intensity quantity and extensive quantity in thermodynamics. Taking the most common cylindrical transmission pipeline as an example, this paper derives the energy transfer processes of branches of the AC and DC power transmission, incompressible viscous fluid, compressible gas, heat conduction and fluid heat transfer, and then establishes a network equation describing extensive quantity conservation of nodes. The steady-state energy flows of the independent energy sub-networks and multi-energy network are solved by the Newton-Raphson method, and the calculation examples verify the feasibility of the proposed method and model.

      • ZHANG Xihong, ZHAO Qi, CHEN Zhong, TIAN Jiang, DU Puliang, XING Qiang

        2021,45(20):67-75, DOI: 10.7500/AEPS20201028004

        Abstract:Delay, packet loss and error codes of data transmission in information systems will cause the information collected by the distribution network control center to deviate from the true value. Therefore, a multi-scenario two-stage stochastic control method is proposed, which considers the information failure of the distribution network in the cyber-physical system. Firstly, the data and predicted values of each sampling point in the trend optimization cycle are discretized into different scenario sets based on the information failure conditions that may occur. With the goal of minimizing the total cost, the stochastic model predictive control is used to perform real-time rolling optimization of each scenario in the cycle and coordinate the output of various distributed energy sources. Secondly, in a shorter real-time correction cycle, according to the fine coordination of dispatchable abilities, the impact of information failure on the distribution network can be further reduced. Finally, in the modified IEEE 33-node distribution network, it is verified that the proposed method can effectively improve the safety and economy of the coordination and optimization of the distribution network when information failure is considered, and the difference and sensitivity of different information failure situations are also analyzed.

      • XIAO Jun, BAO Zhenyu, SONG Chenhui, ZU Guoqiang

        2021,45(20):76-83, DOI: 10.7500/AEPS20200803005

        Abstract:When modeling the distribution system security region (DSSR), total supply capability and security analysis of distribution network, and in order to solve the problem that it lacks the quantitative basis for the selection of DC power flow models, a method for power flow model selection based on capacity boundary and voltage boundary is proposed. Firstly, the concepts of capacity boundary and voltage boundary are proposed on the basis of DSSR of AC power flow model. Secondly, based on the relationship between the two constraints when the length of the feeder varies, the critical length is found, that is, the dividing point dominated by the capacity constraint or the voltage constraint. If the length of feeder is less than the critical length, the capacity constraint is stricter, and the DC power flow model can be adopted. On the contrary, the voltage constraint is stricter, and the AC power flow model should be adopted. Finally, for the urban, town and rural distribution networks in China, the reference values of the critical length for different wire types, power factors and voltage regulation measures are calculated. At the same time, some suggestions are given to select the power flow models of the five types of power supply areas stipulated in the planning guidelines.

      • TANG Jiajun, ZHANG Zhi, YANG Li, CUI Jinrui, CHEN Jiageng, LIN Zhemin

        2021,45(20):84-92, DOI: 10.7500/AEPS20201001001

        Abstract:In the medium- and long-term electricity market, the formation of alliances among incremental distribution and retail companies can reduce the risk of deviation assessment and increase the operating revenue. On this background, the alliance strategy for incremental distribution and retail companies considering the geographical advantage and cooperation cost is proposed. Firstly, according to the physical properties of the distribution network and the partition-wall power sale policy, a depth-first search algorithm is adopted to obtain a complete set of potential alliances that meet the conditions of geographic proximity. Secondly, an alliance revenue model based on the transaction cost theory is built, and three indicators, which are revenue contribution rate, resource scarcity, and an alliance dependence, are proposed to evaluate the revenue allocation right of members, as well as an improved Shapley value method based on the criteria importance through intercriteria correlation (CRITIC) method and analytic hierarchy process (AHP) method is proposed. Finally, a forbearance index is proposed to restrict the members to choose the priority of the alliances, and the alliance optimization algorithm of incremental distribution and retail companies based on the forbearance is proposed. The analysis results of calculation examples show that the incremental distribution and retail companies can effectively increase the revenue through the alliance, the geographical advantage directly affects the revenues of companies, and the cooperation cost affects the size of the alliance and the choices of each member in the alliance.

      • OUYANG Jinxin, YE Junjun, ZHANG Zhen, XIAO Chao

        2021,45(20):93-102, DOI: 10.7500/AEPS20210401008

        Abstract:In the multi-infeed line commutated converter based high-voltage direct current (LCC-HVDC) transmission system, the complex coupling between AC/DC systems and different LCC-HVDC causes the change of commutation failure modes, which makes the control and protection system of LCC-HVDC face challenges. At present, commutation failures of multi-circuit LCC-HVDC directly caused by power grid faults have attracted much attention, but the interaction between LCC-HVDC in the process of commutation failure has been ignored. The state and output characteristics of LCC-HVDC change sharply after commutation failure. Through the coupling of AC system, it is bound to have a direct impact on adjacent LCC-HVDC. Therefore, this paper studies the influence of LCC-HVDC commutation failure on adjacent normal LCC-HVDC, analyzes the generation mechanism of successive commutation failure in multi-infeed LCC-HVDC system under weak power grid fault at the receiving end, and the dynamic reactive power characteristics of inverter stations under the operation of control system during commutation failure. The reactive power expression of inverter station considering the response of control system is deduced; the reactive power exchange characteristics of adjacent LCC-HVDC are analyzed; and the characteristics and influencing factors of successive commutation failure are analyzed. Finally, the correctness of the theoretical analysis is verified through the standard model.

      • DU Yixing, HU Zhijian, LI Ben, CHEN Jinpeng, WENG Changhong

        2021,45(20):103-112, DOI: 10.7500/AEPS20210118001

        Abstract:The conventional machine learning models are weak in overall perception of time series when applied to power system transient stability assessment, so it is difficult to mine the dynamic information contained in the electrical response trajectory, and the reliability of critical sample prediction results is low. Focusing on the aforementioned problems, a two-stage transient stability assessment method based on the bi-directional gated recurrent unit (BiGRU) is proposed. The method takes the dynamic trajectories of the underlying measurement data after disturbance as inputs, first screens out credible samples through continuous dynamic assessment, and then predicts the fault severity of uncertain samples and credible stable samples through the regression model. BiGRU classifier is improved by introducing truncation function and weight coefficient into the loss function, which strengthens the study strength of the model for difficult samples and unstable samples. The experimental results on the modified New England 10-machie 39-bus system show that the proposed method not only significantly reduces the misjudgment of unstable samples, but also improves the recognition ability of stable samples.

      • >Application Research
      • LUO Kui, SHI Wenhui

        2021,45(20):113-120, DOI: 10.7500/AEPS20200908006

        Abstract:The large-scale integration of wind power brings great uncertainty to the operation and planning of power grids, making the system operation mode more variable. The extraction of extreme operation scenarios is of great significance for analyzing the high-risk and weak points in the power grid operation, while the traditional empirical scenario extraction methods have difficulty in dealing with the dual uncertainties of wind power and load. Aiming at the security and stability assessment problem of the power system at the planning level, a method for extracting extreme operation scenarios in the power grid based on data mining and machine learning algorithms is proposed. First, through machine learning, the scenario variables that have a greater impact on the transient stability are identified and sorted according to their importance. At the same time, the entropy-weighted method is used to reflect the degree of contribution of the discreteness of the scenario variable itself to the extreme operation scenarios. Then, the weighted clustering algorithm is used to pick the typical operation scenarios that can represent the transient angle stability level in most scenarios. Furthermore, the outlier and extreme edge operating points far from the clustering center are extracted as extreme scenarios. Finally, an IEEE 39-bus test system is used for the transient simulation analysis, which verifies the effectiveness and rationality of the method that combining data mining and specific problems in extracting extreme scenarios. The proposed method improves the level and efficiency of grid integration planning and stability analysis of wind power.

      • YANG Xiangzhen, YANG Qiuqiang, DU Yan, SHI Yong

        2021,45(20):121-130, DOI: 10.7500/AEPS20201007003

        Abstract:Due to the access of commercial converters with various unknown models, the frequent switching of power supply and load equipment, and the changes in network structure, the microgrid system model has the characteristics of uncertainty and time-varying. The secondary frequency regulation for microgrid based on traditional proportional-integral (PI) control and model predictive control with parameterized prediction model has the problems such as the complexity of system mechanism modeling, the difficulty in online updating, the difficulty in fixed-order non-mechanical modeling to meet the modeling requirements of full working conditions and the complexity of solving the parameterized prediction model. In order to solve the above problems, a secondary frequency regulation strategy with dynamic matrix control (DMC) based on adaptive online model identification is proposed. The strategy consists of two parts. The first part is the adaptive model identification which combines the order optimization method based on Akaike information criterion (AIC) and the forgetting factor recursive least squares method to realize the optimal equivalent model identification of the microgrid. In the second part, the secondary frequency regulation strategy with DMC uses the obtained optimal equivalent model of the microgrid to update the predictive model of the DMC second frequency regulation algorithm in real time, and quickly recovers the system frequency through rolling optimization control. Finally, a simulation example of the microgrid built on the MATLAB/Simulink platform shows the correctness and effectiveness of the proposed identification method and the frequency control strategy.

      • WU Zhangxi, GUO Chunyi, ZHAO Wei, YE Quan, YANG Shuo

        2021,45(20):131-139, DOI: 10.7500/AEPS20210330001

        Abstract:The hybrid cascaded ultra-high voltage DC (HC-UHVDC) system based on the line commutated converter (LCC) and the modular multilevel converter (MMC) has received extensive attention from engineering and academic circles. This paper establishes a model for the HC-UHVDC system in which the rectifier adopts a dual 12-pluse LCC and the inverter is composed of one LCC and three paralleled MMCs in series. And the generation mechanism of the DC overcurrent due to the commutation failure of LCC caused by the AC fault at the inverter station is analyzed. Then this paper proposes an overcurrent suppression method for HC-UHVDC system based on fuzzy clustering and identification. In this suppression method, the fuzzy clustering of multiple electric parameters at the rectifier station is carried out through the simulation in the case of AC fault on the inverter side. According to the characteristics of different transient stages at the inverter station identified by the clustering result, the staged firing angle instruction value is designd in advance. When an AC fault occurs in the system, the DC voltage of the rectifier station is adjusted in time based on the local information of the rectifier station, to quickly suppress the DC overcurrent. The results of the detailed electromagnetic transient simulation in PSCAD/EMTDC show that, under the three-phase and single-phase metallic short-circuit fault conditions on the inverter side, the proposed approach can suppress the DC overcurrent and overvoltage after the commutation failure of LCC at the inverter station to some extent, and can significantly improve the dynamic characteristics of the HC-UHVDC system.

      • JIANG Shunping, DING Yong, SHI Xiangjian, LIU Weiqun, TAO Zheng, SHI Yong

        2021,45(20):140-147, DOI: 10.7500/AEPS20200928006

        Abstract:Flexible power source has the functions of neutral point displacement overvoltage suppression and ground fault arc-suppression, which are essential for safe and stable operation of distribution networks. However, when the displacement overvoltage suppression is switched online to fault arc-suppression, the flexible power source has the problem of output overcurrent. Aiming at the overcurrent problem at the moment of switching, a control method of amplitude difference over-limit active blocking is proposed to realize the reliable switching of the flexible power source operation mode. First, the operation mechanism of flexible power source is introduced, and a closed-loop control method is proposed which shares the current inner control loop. Then, the equivalent circuit models are established in two operation modes, and the principle of overcurrent during the switching process is analyzed. The switching state is determined by whether the quantity of voltage sudden change exceeds the limit in the displacement overvoltage suppression mode. And then the AC output is blocked until the upper arc-suppression command is received, and the flexible power source switches to the ground fault arc-suppression mode. The simulation analyzes the rationality of the two operation modes and their switching control methods. Finally, a prototype is developed and applied in a 66 kV power grid of China, and the field test results verify the reliability and advancement of the proposed method.

      • ZHANG Rui, SONG Yankan, YU Zhitong, CHEN Ying, ZHANG Di, ZHU Tong

        2021,45(20):148-156, DOI: 10.7500/AEPS20201217003

        Abstract:Half-bridge type voltage source converter (VSC) is widely used in the modern power system. When the traditional electromagnetic transients program (EMTP) is used to simulate a large-scale VSC network, there are problems of high time-consuming and low efficiency. Taking the half-bridge sub-circuit as the basic unit for the switch state judgment, by analyzing the freewheeling and turning-off process of the diode when the switch states change, a synchronous switch prediction method that is generally suitable for the half-bridge-type VSC is obtained. This method can directly obtain the stable switch state combination through logic judgment at the current time step, while eliminating the iterative calculation. A series of fast simulation models of the half-bridge type VSC are constructed by combining the fast prediction method of synchronous switch and nested node elimination method. Comparative simulation results verify that the proposed fast simulation models have the same simulation accuracy as the full-detailed models, and can effectively reduce the simulation time and improve the simulation efficiency. The simulation on an 80-module solid-state transformer using the fast simulation model can be accelerated by 20 times compared to the full-detailed model.

      • YOU Zhenfei, ZHAO Jian, WANG Xiaoyu, BIAN Xiaoyan, XU Xianghai, HOU Weihong

        2021,45(20):157-165, DOI: 10.7500/AEPS20200726001

        Abstract:A state recognition method for disconnectors based on semantic segmentation and connected component labeling is proposed. Firstly, a method of disconnector pixel extraction based on semantic segmentation algorithm is proposed. The images are classified according to the semantic category, and the pixels of the disconnector arm are extracted. Secondly, a semantic segmentation image labeling method based on the region growing algorithm is proposed to label the connected component of disconnectors, and an area-sorting statistical method is proposed to optimize the area threshold and eliminate the non disconnector arm region. Finally, the state of the disconnector is judged according to the number of connected components of the disconnector arm. In order to improve the accuracy of disconnector location and segmentation in defaced image, the style transfer algorithm is introduced to generate the stylized images of disconnectors to enhance the training set. Experimental results show that the proposed method can accurately locate and segment disconnectors, and identify the state of disconnectors.

      • >Egineering Application
      • SANG Maosheng, DING Yi, WANG Zhengfeng, LI Zhu, BAO Minglei, SHANG Nan, WU Xu

        2021,45(20):166-174, DOI: 10.7500/AEPS20210105002

        Abstract:With the continuous increase of medium- and long-term transaction electricity, the transaction results keep squeezing the scheduling space of the power system, threatening the safe operation of the system. Due to the operation constraints of power grids, a large number of transaction results are reduced, which tends to cause market members to question the market fairness. Therefore, exploring the scale restriction lifting of the medium- and long-term transaction electricity in the background of coexistence of planning and market will not only help to clarify the evolution trend of power markets, but also help to realize the overall optimization of operation safety and economic efficiency. In this paper, a method for calculating the scale restriction lifting of the medium- and long-term transaction electricity is proposed. This method takes into account both the market-oriented electricity clearing optimization and the power grid operation safety, as well as constructs a two-level optimization model considering market economic benefits and power grid operation constraints. The upper level is the market-oriented electricity optimization clearing model aiming at the maximum social welfare, and the lower level is the market-oriented electricity checking model considering the transmission capacity of transmission lines, heat demand of the system, unit maintenance arrangement, the minimum startup mode of power plants and other grid operation constraints. For the two-level optimization model, the upper and lower iterative algorithm is proposed to solve the problem, and the constraints of market-oriented electricity are relaxed to ensure the solvability of the model. Through the analysis of an example of a provincial power grid of China, the validity and practicability of the proposed model and algorithm are verified.

      • SHI Puxin, SHI Peiran, WANG Peiwen, WANG Yang, CHEN Shunan

        2021,45(20):175-184, DOI: 10.7500/AEPS20201025003

        Abstract:Large-scale grid integration of renewable energy brings new challenges to electricity market design and power system operation, especially power peak-shaving. Firstly, the necessity of developing North China power peak-shaving ancillary service market with China's power source structure dominated by coal is put forward. Then, the mechanism design of North China power peak-shaving market is analyzed from the aspects of peak-shaving service declaration, clearing mechanism and settlement mechanism. Finally, with the actual operation of North China power peak-shaving market in 2020, the operation performance evaluation index system of peak-shaving market is constructed from three aspects: market power and bidding behavior of market entities, market clearing and operation, and market social benefits, which comprehensively reflects the peak-shaving market power, market supply and demand relationship, market operation effectiveness, and operation level of market operation institutions, and provides a useful reference for the construction of China's electricity market system.

      • >Survey
      • GUO Xiaoqiang, WEI Yupeng, WAN Yanming, ZHOU You, KONG Lingguo, ZHANG Zhe, LU Zhigang, HUA Changchun, Mariusz MALINOWSKI

        2021,45(20):185-199, DOI: 10.7500/AEPS20201101004

        Abstract:The combination of hydrogen energy and renewable energy generation is one of the important ways to achieve the goals of carbon neutrality and the peak of carbon dioxide emissions, where power electronic converters are widely used. In order to realize the efficient operation of the electrolysis hydrogen production system, the power electronic converter is required to have the characteristics of low voltage and high current capability, low output current ripple, high step-down ratio, and high reliability. Firstly, this paper analyzes the electrolyzer model, and studies the characteristics and technical indicators of the existing commercial electrolyzer. Secondly, this paper sorts and classifies the power electronic converters applied in various electrolysis hydrogen production systems including the alkaline electrolyzer, the polymer exchange membrane electrolyzer and the solid oxide electrolyzer, and various topologies are analyzed and compared. Meanwhile, the potential applications, existing problems and development trends of power electronic converters for producing hydrogen from renewable energy sources are discussed and summarized.TFoundation:his work is supported by Science and Technology Innovation Project of CHN Energy (No. GJNY-19-136), National Key R&D Program of China (No. 2018YFB1503100), and Science Fund for Creative Research Groups of Hebei Provincial Natural Science Foundation of China (No. F2020203013).