English Version Journal Login
EN
最新刊期
卷 49 ， 期 20 ， 2025
- 摘要：Enhancing demand-side coordination capability is one of the key initiatives to accelerate the construction of new power system. This facilitates the development of demand-side flexibility, improves the retail-side market framework, and diversifies grid regulation methods. Among various typical models emerging in demand-side management, the virtual power plant (VPP) functions as a novel “power source” combining flexible regulation attributes with market trading capabilities. Through multi-dimensional interactions with the power system, VPPs provide multi-temporal balancing regulation capabilities, thereby obtaining market value. However, existing VPP technologies still lack a dedicated systematic framework and analytical theory. The operation characteristics, interaction mechanisms, and dynamic aggregation problems of massive distributed resources have yet been fully addressed, which leads to significant obstacles to large-scale application and routine operation of VPPs for sustainable development. To this end, this paper first conducts an in-depth analysis of the technical essence of VPPs from first principles perspective, organizes the technical requirements imposed by differentiated market trading varieties for various types of VPPs, and proposes the objectives for constructing the VPP technical system. Then, a key technology system for multi-type VPPs across four dimensions is established: fundamental commonality, engineering application, key equipment, and characteristic research. Finally, this paper summarizes project implementation, operational status, and market effectiveness of VPPs in China. Forward-looking recommendations to address limitations in existing technical system are proposed, thereby providing theoretical support for routine operation and standardized application of VPPs.关键词：virtual power plant;new power system;demand response;technical system;foundational commonality;engineering application;key equipment;first principles3|2|0更新时间：2026-02-28
- 摘要：As an innovative mode for the efficient management of distributed energy resource (DER), the virtual power plant (VPP) has become an important solution for the large-scale grid connection of DER through extensive interconnection and flexible scheduling of energy. While improving the energy utilization efficiency, the VPP faces serious challenge of privacy data leakage due to its massive, multi-source data and the broad and open “cloud-pipeline-edge-terminal” communication architecture. This paper thoroughly analyzes the privacy risks and preservation requirements of the VPP, systematically reviews the full-process privacy preservation technologies from data collection and sharing, trading, scheduling to revenue settlement, and summarizes the shortcomings of existing privacy preservation schemes in terms of computational cost, dynamic adaptability, and the balance between privacy and data availability. In response to the above shortcomings,a hierarchical and collaborative overall privacy preservation system of “cloud-pipeline-edge-terminal” is proposed, and the feasibility and technical challenges faced by deep learning and generative artificial intelligence in overcoming above deficiencies through behavior feature mining, dynamic game strategies, and semantic perception are prospected, providing new ideas for building a safe and intelligent VPP.关键词：virtual power plant;distributed energy resource (DER);privacy preservation;cyber attack;security analysis;artificial intelligence2|3|0更新时间：2026-02-28
- 摘要：Virtual power plants (VPPs) provide power ancillary services by aggregating distributed user resources, delivering significant safety and economic benefits. Since VPP operators do not directly own these resources, incentive mechanisms become crucial for achieving friendly interactions with users. This paper constructs a VPP incentive mechanism in the form of power ancillary service packages. Compared with other methods, the form of packages can better balance the demands of multiple parties and attract massive user to participate. First, the market rules and operation schemes related to VPPs in China are reviewed, clarifying the policy direction and analyzing the profitability conditions of VPPs based on this foundation. Secondly, the connotations of VPP incentive mechanisms, along with their engineering practices, classifications, and specific manifestations, are summarized, and the rationality and effectiveness of establishing power ancillary service incentive mechanisms through packages are analyzed. On this basis, the overall implementation framework for customizing power ancillary service packages in VPPs is proposed, focusing on three key aspects: customized design, parameter decision-making, and dynamic adjustment. Key issues and challenges within this framework are examined. Finally, relevant research progress regarding these three aspects is reviewed, and specific solutions are proposed, offering a feasible path for innovation in VPP business models.关键词：virtual power plant;incentive mechanism;customized package;ancillary service;user resource;new power system2|1|0更新时间：2026-02-28
- 摘要：Quantitative assessment of interactive regulation capability for a virtual power plant (VPP) serves as critical foundation for evaluating its regulation performance and optimal operation under diverse spatiotemporal conditions. To achieve a refined assessment of the regulation capability for a VPP, this paper proposes a multi-dimensional quantitative assessment method for VPP based on hierarchical aggregation. Firstly, a data-driven method is employed to address the requirements of market transaction decisions across various time scales. By applying the k-means algorithm to extract typical user load profiles, a quantitative assessment model is obtained for the regulation capability of single flexible resource, which incorporates an incentive price-based correction to account for realizable potential. Secondly, considering factors such as regulation capacity, ramp rate, response time, and response cost, a quantitative assessment model is developed for the aggregated interactive regulation capability of heterogeneous VPP resources and solved through a multi-objective optimization approach that combines the particle swarm optimization algorithm and the comprehensive membership degree method. Finally, simulation results demonstrate that the proposed method can accurately quantify the potential boundaries of VPPs in participating in ancillary services, providing a decision-making basis for operators to formulate optimal strategies and maximize their service value.关键词：virtual power plant (VPP);distributed energy resource (DER);regulation capability;multi-dimensional quantitative assessment;resource aggregation;ancillary service;multi-objective optimization;incentive price3|5|0更新时间：2026-02-28
- 摘要：With the rapid expansion of virtual power plants (VPPs) in scale, the existing quantification results of regulation capability for VPPs under multiple uncertainties fail to meet the requirements of risk-aware dispatch of power systems and bidding in electricity markets. Building VPP, as an important form of VPPs, possess the characteristics of rapid regulation capability and high reliability by aggregating building load resources. Among these resources, as the primary regulation equipment in building VPPs, the quantification of credible regulation capability for air conditioning is of significant importance for the reliable operation of building VPPs. To this end, the concept and a quantification method for the credible regulation capability of building VPPs are proposed by taking air conditioning as an example. Firstly, the quantification framework of credible regulation capability for building VPPs is constructed based on hybrid data-physical-driven model, and the concept of credible regulation capability is proposed, and the connection of the credible regulation capability interval is clarified. Secondly, a probabilistic prediction model for the energy consumption baseline is developed based on the Informer architecture and conformal quantile regression (Informer-CQR), By extracting temporal features, asymmetric residual sets are constructed to generate reliable baseline probability intervals. Furthermore, based on a thermal transfer model of building air conditioning, the upward and downward regulation capabilities of building air conditioning are quantified at different confidence levels. Finally, simulation analysis is are conducted with actual data from a public building in Macau, China. The results demonstrate that the proposed method can effectively quantify the credible regulation capability of building air conditioning within predefined confidence intervals, and reveal the dynamic fluctuation behavior of the regulation capability of air conditioning through quantification intervals.关键词：virtual power plant (VPP);building;air conditioning;credible regulation capability;quantification;probability prediction2|4|0更新时间：2026-02-28
- 摘要：As a vital component of flexible regulation resources in power grids, feeder loads exhibit significant regulation potential and can serve as a critical component of the virtual power plant (VPP), playing an important role in the power grid security, stability control, and demand-side response. This paper focuses on the assessment of adjustable capacity of feeder loads, aiming to provide capacity constraints for their participation in grid optimization control in scenarios such as non-routine regulation and emergency balancing. Firstly, the voltage-power coupling characteristics of feeder loads are validated through engineering tests. On this basis, a time-varying characteristic-based model for conservation voltage reduction (CVR) coefficient is established, and a corresponding parameter identification method is proposed. Secondly, on the premise of using on-load tap changers (OLTCs) as the primary regulation means and shunt reactive power regulation devices as auxiliary means, a dynamic optimization model for assessing the adjustable capacity of feeder loads is proposed based on analytical voltage sensitivity. This model maximizes the voltage regulation boundaries while reducing the total regulation of reactive power. And combined with the CVR coefficient model, the regulation capability of the feeder cluster is updated every minute on a rolling basis. Finally, an modified IEEE 33-bus model is built on the MATLAB simulation platform to verify the effectiveness of the proposed method.关键词：virtual power plant (VPP);feeder load;voltage regulation;voltage sensitivity;adjustable capacity;dynamic assessment3|3|0更新时间：2026-02-28
- 摘要：Electric vehicle virtual power plants (EV-VPPs) can aggregate large-scale electric vehicles (EVs) to participate in green electricity and spot market trading. Green electricity contract is a key link between the green electricity medium- and long-term market and the spot market. Its decomposition modes may directly influence the bidding strategies, market settlement costs, and operation profits of EV-VPPs. To this end, considering the influences of decomposition modes of green electricity contracts, demand response mechanisms, and multiple uncertainties such as day-ahead electricity prices, real-time electricity prices, and charging loads of EVs, the classification modeling of regulation capabilities for various EV types is conducted. Based on the two-stage stochastic optimization and conditional value at risk theory, the optimal day-ahead market bidding model for EV-VPPs is established. The case study results show that the proposed model can smooth EV loads, allow the bidding strategies of EV-VPPs to be developed in line with their risk preference, reduce the trading risks of EV-VPPs, and improve operating profits.关键词：virtual power plant (VPP);electric vehicle (EV);spot market;green electricity;bidding model3|1|0更新时间：2026-02-28
- 摘要：Multiple virtual power plants (VPPs) participating in the joint market as a cluster can enhance their market competitiveness. To solve the resource allocation of the cluster in the multi-market coupling decision-making and balance the electric power energy and the frequency regulation revenues, this paper proposes a bidding strategy of VPP clusters participating in the electric power energy market and frequency regulation market. Firstly, an internal transaction mode between the VPP and the cluster operator is proposed, and a pricing strategy for the cluster operator and a method for refining the response behavior of the VPP when participating in the external joint market are designed. Secondly, an operation cost model of VPP considering the electric power energy and frequency regulation capacity transaction is established to realize multi-product transactions in the cluster, providing a foundation for the operator to formulate bidding strategies for the joint market. Then, the price quota curve is adopted to characterize the relationship between clearing results of the frequency regulation market and the bidding strategy, and a risk-averse optimal bidding strategy is proposed with the objective of maximizing cluster revenues. Finally, a simulation analysis is conducted on a cluster consisting of three VPPs. The case study results demonstrate that the proposed joint bidding strategy stimulates the response willingness of each VPP and improves the economic revenue of the cluster.关键词：virtual power plant (VPP);bidding strategy;electric power energy market;risk management;frequency regulation;ancillary service2|3|0更新时间：2026-02-28
- 摘要：The linkage uncertainty among distributed resources in a virtual power plant (VPP) poses significant challenges to the accuracy of the response of the VPP to power grid scheduling. To address this issue, this paper proposes an optimal scheduling strategy considering the linkage uncertainty of distributed resources in a VPP, aiming to enhance the economic efficiency of VPP scheduling and the accuracy of response. Firstly, this paper portrays the linkage uncertainty of distributed resource power variations through seasonal trend decomposition and cross-correlation analysis, and constructs a linkage matrix. Secondly, a comprehensive assessment method is proposed. This method optimizes the recovery process of distributed resources after responding to scheduling to minimize the economic cost of scheduling, which holistically considers the economics and linkage uncertainty. Then, a dynamic scheduling strategy based on the Gaussian Copula function is introduced. By establishing a correlation model among resources, this strategy dynamically adjusts the scheduling process to improve both accuracy and economic efficiency. Finally, the real data from 247 charging stations in Shenzhen, China, are used for simulation validation. Simulation results demonstrate that, compared with existing methods, the proposed scheduling strategy of distributed resources can reduce the costs by 10.09%.关键词：virtual power plant (VPP);distributed resource;linkage;uncertainty;optimal scheduling;response capability recovery;Copula function2|2|0更新时间：2026-02-28
- 摘要：The disorderly integration of a large number of electric vehicles poses unprecedented challenges to the optimization scheduling of the power system. The virtual power plant (VPP), as a solution that aggregates distributed resources such as electric vehicles and photovoltaics at the edge of the cloud for participation in the scheduling of the transmission network, has great application potential. However, the large-scale homogeneous electric vehicles have consistency in parameters such as charging and discharging power, efficiency, and battery capacity, resulting in the same edge optimization strategies for responding the control signals of the VPP. During the collaborative scheduling of the VPP, problems such as non-convergence of iterations and even oscillations are prone to occur. Therefore, this paper constructs a cloud-edge collaborative optimization scheduling framework and model for VPPs based on Lagrangian relaxation, reveals the mechanism of iteration oscillation caused by the consistency of parameters of large-scale electric vehicles, and proposes a convergence acceleration algorithm based on the disturbing function. By breaking the parameter consistency in the condition of maintaining the optimal solution unchanged, it achieves efficient collaborative optimization of a large number of electric vehicles and the VPP. Through parameter sensitivity analysis, the closed form of the disturbance limit is theoretically proved, and the effectiveness and efficiency of the proposed method are verified based on the case of the users with various scales in Shenzhen, China关键词：virtual power plant (VPP);electric vehicle;Lagrangian relaxation;disturbing function;cloud;edge2|2|0更新时间：2026-02-28
- 摘要：The virtual power plant (VPP) aggregates massive distributed resources to form a large-scale market entity interacting with the power grid. However, the energy management of large-scale distributed resources faces a contradiction between global optimization and local demands. Centralized optimization and regulation struggle to balance electricity costs and comfort of the users, while local user energy management lacks coordinated control over the overall operational state of the VPP. Therefore, a lightweight user energy management strategy for cloud-edge collaborative VPPs is proposed. Firstly, a hierarchical decision-making framework for energy management of VPPs is constructed to shift unidirectional control to bidirectional interaction. Secondly, considering the limited computational resources of low-cost edge devices, a local decision-making algorithm based on sparse reinforcement learning is proposed, which uses adaptive topology evolution to reduce the number of trainable parameters in neural networks. Finally, case studies conducted on real-world datasets validate the effectiveness and efficiency of the proposed lightweight strategy.关键词：virtual power plant (VPP);cloud;edge;energy management;distributed resource;sparse reinforcement learning2|2|0更新时间：2026-02-28
- 摘要：To fully leverage the characteristics of various types of heterogeneous distributed energy resources (DERs) and utilize the adjustable capacity of the distribution network to provide superior frequency regulation auxiliary services, this paper proposes a primary frequency control strategy of distribution-level virtual power plants (DLVPPs). First, models suitable for aggregated control are developed for grid-following and grid-forming heterogeneous DERs. Next, leveraging the derived aggregation conditions and considering factors such as the response time, system topology, and frequency regulation capacity of each DER, an adaptive disaggregation allocation strategy is employed to decompose the total expected characteristics into each DER. Then, a local controller is designed based on robust model predictive control (RMPC) to enhance the system robustness in the presence of external disturbances and model uncertainties. Meanwhile, on the premise of meeting the device-level transient constraints, each device can precisely match the local expected characteristics. Finally, simulation results demonstrate that the proposed strategy significantly improves the primary frequency regulation characteristics while considering the local dynamic characteristics, and achieves unified aggregated frequency control of heterogeneous DERs, providing an effective solution for the distribution network participating in frequency regulation auxiliary services关键词：virtual power plant (VPP);distribution network;frequency regulation;ancillary service;predictive control;primary frequency control;distributed energy resource3|2|0更新时间：2026-02-28
- 摘要：The automatic generation control (AGC) for virtual power plants (VPPs) can meet the secondary frequency regulation requirements in the power grid with high renewable energy penetration through coordinated control of multiple resources. To address the issues of inaccurate convergence and slow convergence of traditional distributed AGC methods, this paper proposes an improved distributed algorithm that incorporates line power flow and convergence acceleration. Firstly, a linearized model for the active power flow of the VPP is established. Accordingly, an optimal AGC command decomposition problem for the VPP is formulated to enhance frequency regulation performance and economic efficiency while ensuring the safe operation of the power grid. Secondly, an improved exact diffusion algorithm (EDA) based on the Nesterov accelerated gradient (NAG) method is proposed to further improve the solution speed and accuracy of the optimization model. Case studies based on the IEEE 33-bus system and a VPP demonstration base at Linhai Toumen Port in Taizhou of Zhejiang Province, China demonstrate that the proposed algorithm significantly outperforms other algorithms in both response accuracy and convergence speed. Furthermore, the AGC system based on this algorithm achieves significant improvements in both frequency regulation performance and economic efficiency.关键词：virtual power plant (VPP);frequency regulation;automatic generation control (AGC);power flow;command decomposition;distributed optimization;exact diffusion algorithm (EDA);Nesterov accelerated gradient2|2|0更新时间：2026-02-28
- 摘要：As the coordinated operation of main grids and distribution networks in new power systems continues to deepen, the response demands of virtual power plants (VPPs) in multi-level power grids are showing a multi-dimensional development trend, making the dynamic allocation of limited aggregated capacity pivotal. To address the problems of high resource occupancy rates, low utilization efficiency, and difficulties in multi-time-series coordination caused by traditional proportional allocation methods, this paper takes VPPs integrated with aggregated photovoltaic-energy storage battery swapping station (BSS) as its research object. A bottom-up capacity allocation method for heterogeneous resources is introduced in the dynamic construction to achieve a “low-resource, high-efficiency” response strategy. First, this paper constructs a linear modeling framework for a photovoltaic-energy storage BSS, considering the dynamic boundaries of aggregated capacity. Then, based on Minkowski sum and approximation techniques, the normative extraction of feasibility regions for heterogeneous resources is realized. Finally, with the objective of maximizing the utilization of individual feasibility regions, a bi-level planning dynamic allocation model is established based on the affine transformations of feasibility regions. The original non-deterministic polynomial problem is reformulated into a linear optimization problem by employing Karush-Kuhn-Tucker (KKT) conditions and the McCormick convex envelope relaxation method, which significantly improves the solution efficiency of the VPP construction process. On this basis, considering the high-frequency volatility of regulation commands and the time-varying operational parameters of swapping stations, a dynamic regulation strategy for real-time secondary frequency regulation is proposed to achieve a closed-loop linkage from “offline capacity allocation” to “online dynamic regulation”. A validation scenario is established based on Hubei power grid in China to verify the effectiveness and feasibility of the proposed strategy in engineering practice.关键词：virtual power plant (VPP);battery swapping station (BSS);new power system;feasibility region;capacity allocation;demonstration2|1|0更新时间：2026-02-28
Special Session on Virtual Power Plant with Enormous Flexible Resources
- 摘要：The distribution network with high-penetration distributed photovoltaics has problems of mismatched operation of regulating equipment and frequent voltage violations, which pose serious challenges to the operation and control of the distribution network. Therefore, a two-stage centralized-distributed voltage control method for distribution networks based on deep reinforcement learning is proposed. This method transforms the switching operation of on-load tap changers into a Markov decision process on a long time scale, aiming to comprehensively optimize the action of discrete regulating equipment and voltage deviation, and establishes a day-ahead centralized optimization model based on discrete equipment switching plans. This method also transforms the coordinated control of multiple photovoltaic inverters into a Markov game process on a short time scale, with the goal of minimizing node voltage deviation and line loss, and establishes a multi-agent based intra-day distributed control model. Considering the regulation characteristics and interactive effects between various equipment, the discrete-continuous mixed actions between various equipment are coordinated through day-ahead optimization and intra-day real-time control to mitigate node voltage fluctuations. Finally, the simulation analysis is conducted based on the modified IEEE 33-bus distribution network and a high-penetration photovoltaic distribution network in Southern China to verify the effectiveness and scalability of the proposed method.关键词：distribution network;distributed photovoltaic;deep reinforcement learning;time scale;agent;voltage control5|2|0更新时间：2026-02-28
- 摘要：The low three-phase voltage imbalance on the low-voltage side of distribution transformers in low-voltage distribution station areas and the presence of vacant households and light-load customers make it difficult to achieve accurate identification of phase-customer relationship using single electrical measurement data. This paper proposes a two-step identification method for phase-customer relationship in low-voltage distribution station areas by integrating voltage and current monitoring data. Firstly, trend turning points are extracted from customer voltage monitoring data. A common trend transition point set is defined and used to filter and arrange the original voltage sequences, followed by similarity matrix calculation and k-means clustering to partition user clusters. Secondly, the LASSO regression algorithm is employed to compute the regression coefficient matrix for the phase-customer relationship, thereby obtaining the optimal phase-customer relationship matching results and achieving accurate identification of phase-customer relationships in low-voltage distribution station areas. Finally, the effectiveness and superiority of the proposed algorithm are validated through actual case study analysis.关键词：distribution network;low-voltage distribution area;data-driven;phase-customer relationship;trend transition point;cluster analysis;LASSO regression algorithm2|0|0更新时间：2026-02-28
- 摘要：To address the modeling challenge of doubly-fed induction generator (DFIG) under “black/gray box” models, this paper proposes an efficient identification method for control parameters based on frequency-domain response sensitivity weight. First, the frequency-domain admittance model is derived covering the entire operating area of the DFIG, and the frequency-domain response sensitivity of control parameters is further derived. On this basis, sensitivity weights are introduced to improve the differential evolution algorithm, which adaptively adjusts mutation and crossover operations, and achieves dynamic optimization of parameter search direction and amplitude. The above method combines theoretical interpretability and computational efficiency, significantly improving identification accuracy. The identification results based on a certain actual black box model under various typical working conditions show that the proposed method outperforms traditional methods in terms of accuracy, applicability, and computational efficiency. This method is easy to obtain data, has clear physical meanings, and can effectively improve the identification accuracy of DFIGs.关键词：doubly-fed induction generator;parameter identification;response sensitivity;differential evolution algorithm;black box2|2|0更新时间：2026-02-28
- 摘要：As a critical component of energy storage systems, the accurate estimation of the state of charge (SOC) and state of health (SOH) of lithium-ion battery is essential for improving battery usage efficiency and safety. To achieve precise estimation of SOC and SOH of lithium-ion batteries, this paper proposes a joint estimation method that combines light gradient boosting machine (LightGBM), convolutional neural network (CNN), and bidirectional long short-term memory (BiLSTM) network. Firstly, based on parameters such as charging voltage, charging time, and surface temperature of lithium-ion batteries, a health factor extraction method using the Kendall correlation coefficient is proposed. Secondly, a temporal SOH estimation approach integrating CNN and BiLSTM networks is introduced, which enhances estimation accuracy through the combination of local feature extraction and bidirectional temporal modeling. Furthermore, the estimated SOH values are used as input features for SOC estimation. To address the need for processing large-scale data of lithium-ion batteries in energy storage systems, the LightGBM algorithm, based on a gradient boosting framework, is employed to achieve joint estimation of SOC and SOH, thereby improving computational efficiency and response speed. Finally, the proposed joint estimation model is validated by real battery datasets, and the results demonstrate its high computational accuracy and fast computation speed.关键词：lithium-ion battery;state of charge (SOC);state of health (SOH);state estimation;light gradient boosting machine (LightGBM);convolutional neural network (CNN);long short-term memory network2|1|0更新时间：2026-02-28
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