A method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage

Abstract

The application discloses a kind of adjustable load and energy storage's virtual power plant resource pool construction and regulation method, belong to electric power system dispatching control technical field;The method includes: adjustable load and energy storage resource is accessed virtual power plant operation platform, formulates unified modeling standard and establishes standardization resource model;Model is synchronized to virtual power plant aggregation control subsystem, constructs virtual power plant multidimensional aggregation model, carries out nested data aggregation, forms time-space nested virtual power plant adjustable resource pool and checks;By virtual power plant aggregation control subsystem receives regulation instruction, forms multi-strategy decomposition scheme and is issued to virtual power plant operation platform execution;By virtual power plant operation platform, continuously monitor regulation effect in regulation execution process;The present application realizes the complete closed-loop regulation of heterogeneous resource standardization access, time-space nested aggregation, trusted check, intelligent step-by-step decomposition and self-adapting correction, improves the response accuracy and operating safety of virtual power plant participating in grid frequency modulation.

Description

Technical Field

[0001] This invention relates to the field of power system dispatch and control technology, specifically to a method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage. Background Technology

[0002] With the advancement of the construction of new power systems, the penetration rate of distributed adjustable loads and energy storage resources in distribution networks continues to increase; electric vehicle charging and swapping facilities, distributed energy storage, interruptible industrial loads and other diverse entities have rapid power regulation capabilities, which can theoretically provide high-quality frequency regulation auxiliary services for the power grid; virtual power plants, as an important technical means to aggregate the above-mentioned dispersed resources and achieve unified control, have become a key way to tap the regulation potential of distributed resources and support the safe and economical operation of the power grid.

[0003] However, current virtual power plants still face significant challenges in actual participation in grid automatic generation control (AGC) frequency regulation. On the one hand, adjustable loads and energy storage resources have fundamental differences in response characteristics, physical constraints, and communication interfaces. Existing technologies lack standardized modeling and cross-level fine-grained aggregation methods for heterogeneous resources, making it difficult to efficiently synchronize model data between the municipal resource side and the provincial dispatch side, and making it difficult for the dispatch side to accurately perceive and quantify the resource regulation capabilities. On the other hand, in the decomposition and execution of AGC regulation commands, existing methods mostly adopt a single fixed allocation strategy, which does not fully consider the differences in response speed, regulation cost, and performance coefficients of different resources, making it difficult to adaptively adjust according to the grid operating status and real-time resource conditions. More importantly, existing technologies generally lack continuous monitoring and dynamic correction mechanisms for the actual response effect after the command is issued. When sudden changes in resource conditions or communication delays cause the actual response to deviate from the target value, the system cannot initiate feedback correction in time, resulting in the continuous accumulation of deviations in the aggregation units' following of the regulation target, making it difficult to meet the strict requirements for response speed and control accuracy in high-frequency grid regulation scenarios.

[0004] Therefore, it is essential to design a method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage. Summary of the Invention

[0005] The purpose of this invention is to provide a method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage, in order to solve the problems mentioned in the background art, such as insufficient standardized modeling and cross-level refined aggregation of heterogeneous resources, a single AGC regulation command decomposition strategy lacking adaptive adjustment, and the lack of continuous monitoring and dynamic correction in the regulation execution process, which makes it difficult for the aggregated units to follow the regulation target throughout the entire process.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage, comprising:

[0007] Adjustable loads and energy storage resources are connected to a virtual power plant operation platform. The virtual power plant operation platform formulates a unified modeling standard from five aspects: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost, and establishes a standardized resource model based on the unified modeling standard.

[0008] The virtual power plant operation platform synchronizes the standardized resource model to the virtual power plant aggregation control subsystem. Based on the standardized resource model, the virtual power plant aggregation control subsystem constructs a multi-dimensional aggregation model of the virtual power plant from four dimensions: administrative region, power zone, load type, and response time level, with virtual machine groups as the smallest control unit. The subsystem performs nested data aggregation on the real-time operation data and controllability of the virtual machine groups to form a spatiotemporally nested virtual power plant adjustable resource pool, and verifies the validity of the real-time adjustment capability data of the virtual machine groups in the virtual power plant adjustable resource pool.

[0009] The virtual power plant aggregation control subsystem receives AGC adjustment instructions and, based on the real-time adjustment capability data of the virtual machine groups in the virtual power plant's adjustable resource pool, decomposes the AGC adjustment instructions to the smallest control unit level according to the power grid level, forming a multi-strategy decomposition scheme and sending it to the virtual power plant operation platform for execution.

[0010] The virtual power plant operation platform continuously monitors the regulation effect during the regulation execution process, and initiates a real-time correction process when the actual response fails to achieve the regulation effect.

[0011] As a further technical solution of the present invention, the adjustable load includes electric vehicle charging and swapping facilities and interruptible industrial loads, and the energy storage resources include mobile energy storage vehicles, energy storage warehouses, and tower base station energy storage.

[0012] The physical location of the adjustable load and energy storage resources is the resource side;

[0013] The virtual power plant operation platform is located at the prefecture-level city level, physically close to the resource side; the virtual power plant aggregation control subsystem is located at the provincial level, physically close to the power grid dispatching side.

[0014] The adjustable load and energy storage resources are connected to the virtual power plant operation platform through a three-in-one integrated terminal deployed on the resource side. The three-in-one integrated terminal integrates communication gateway function, vertical encryption authentication function, and AGC remote control function, and uses 5G short slice communication technology for data transmission.

[0015] The resource-side three-in-one converged terminal is equipped with a miniature vertical encryption device. The miniature vertical encryption device, together with the vertical encryption device deployed in the secure access area, forms an encrypted tunnel. Data is transmitted to the production control area where the virtual power plant operation platform is located via the power 5G private network and the encrypted tunnel.

[0016] After being connected, the adjustable load and energy storage resources transmit data to the virtual power plant operation platform in accordance with the unified information transmission standard. The unified information transmission standard specifies the transmission requirements for the following data items: three telemetry data items including grid-connected active power, frequency regulation active power upper limit, and frequency regulation active power lower limit; two tele-signaling data items indicating whether the power plant participates in frequency regulation and whether the power plant controls local activation or deactivation; one remote adjustment data item indicating the frequency regulation active power target value; and one remote control data item indicating whether the power plant remotely controls frequency regulation activation or deactivation.

[0017] As a further technical solution of the present invention, the establishment of the standardized resource model includes:

[0018] The virtual power plant operation platform clearly defines the measurement benchmarks and standardization rules for each of the five dimensions: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost; and collects the original characteristic parameters of the above five dimensions for each adjustable load and energy storage resource.

[0019] The original feature parameters are normalized, wherein the adjustable capacity, response speed, and adjustable duration are normalized using a forward normalization method, and the response delay time and adjustment cost are normalized using a reverse normalization method. The original feature parameters are converted into dimensionless values ​​in the range of 0 to 1, generating a standardized resource model containing the above five dimensions of normalized parameters.

[0020] The standardized resource model is used to construct the subsequent multidimensional aggregation model of the virtual power plant.

[0021] As a further technical solution of the present invention, the standardized resource model is synchronized to the virtual power plant aggregation control subsystem using a combination of full synchronization and incremental synchronization, specifically as follows:

[0022] When there are new adjustable loads and energy storage resources or when the characteristics of existing adjustable loads and energy storage resources change, the virtual power plant operation platform actively sends the changed standardized resource model to the virtual power plant aggregation control subsystem.

[0023] The virtual power plant aggregation control subsystem can proactively request a full set of standardized resource models from the virtual power plant operation platform, and after parsing and confirming the full set of standardized resource models, store them in the local real-time database.

[0024] The virtual power plant aggregation control subsystem constructs a multi-dimensional aggregation model of the virtual power plant based on the synchronized standardized resource model.

[0025] As a further technical solution of the present invention, the specific method for constructing the virtual power plant multidimensional aggregation model is as follows:

[0026] The virtual power plant aggregation control subsystem groups multiple adjustable loads and energy storage resources that have the same administrative region, the same power zone, the same load type, and the same response time level into one group;

[0027] Each group constitutes a virtual machine group, and the four-dimensional label of the virtual machine group is its administrative region, power zone, load type, and response time level.

[0028] The overall adjustability of the virtual machine group is obtained by weighting and summing the adjustable capacities of all adjustable loads and energy storage resources contained in the virtual machine group according to their respective collaborative weights, and the sum of the collaborative weights of all adjustable loads and energy storage resources in the same virtual machine group is 1.

[0029] The virtual power plant aggregation control subsystem uses virtual machine groups as the smallest unit and performs hierarchical nested aggregation according to the order of response time level, load type, power zone, and administrative region to form a five-layer virtual power plant multidimensional aggregation model including virtual machine group layer, response time level layer, load type layer, power zone layer, and administrative region layer.

[0030] The four-dimensional label serves as a unique identifier for the virtual machine group and is used for rapid filtering and location during subsequent instruction decomposition.

[0031] As a further technical solution of the present invention, the specific formation method of the spatiotemporally nested virtual power plant adjustable resource pool is as follows:

[0032] The virtual power plant aggregation control subsystem performs nested aggregation of the real-time operating data and controllability of virtual machine groups at all levels in the virtual power plant multidimensional aggregation model, according to four time scales: second, minute, hour, and day-ahead, and three spatial scales: distribution area, line, and main transformer.

[0033] The spatiotemporal nesting aggregation coefficients of each virtual machine group at different time and space scales are obtained by multiplying the preset time coefficients corresponding to the time scales, the preset space coefficients corresponding to the space scales, and the normalization adjustability of the virtual machine group.

[0034] The adjustable resource pool of the virtual power plant records the spatiotemporal nesting aggregation coefficients, real-time power, upper limit of adjustment capability, lower limit of adjustment capability, and corresponding four-dimensional labels of all levels of virtual machine groups in the multi-dimensional aggregation model of the virtual power plant.

[0035] As a further technical solution of the present invention, the validity verification of the real-time adjustment capability data of the virtual machine group in the adjustable resource pool of the virtual power plant adopts a three-level verification process, specifically as follows:

[0036] The first level of physical boundary verification verifies that the upward adjustment capability reported by the corresponding level virtual machine group in the adjustable resource pool of the virtual power plant does not exceed the difference between the rated power and the real-time power of the corresponding level virtual machine group, and the downward adjustment capability reported does not exceed the real-time power of the corresponding level virtual machine group.

[0037] The second level of availability verification determines whether the corresponding virtual machine group is qualified to participate in regulation based on the controllable status, maintenance status, adjustable time period, and power supply guarantee time period of the corresponding level virtual machine group in the adjustable resource pool of the virtual power plant. The effective regulation capability of the corresponding level virtual machine group that is not qualified to participate in regulation is set to zero.

[0038] The third level of historical deviation verification compares the total adjustment capacity reported by the corresponding level of virtual machine group in the adjustable resource pool of the virtual power plant with the average adjustment capacity of the corresponding level of virtual machine group in similar historical periods. When the deviation ratio exceeds a set threshold, the corresponding data is marked as suspicious data, the automatic control of the corresponding level of virtual machine group is suspended and an alarm is issued.

[0039] As a further technical solution of the present invention, the AGC adjustment command is decomposed level by level to the smallest control unit according to the power grid level, using a three-level decomposition process, specifically as follows:

[0040] The virtual power plant aggregation control subsystem receives AGC regulation commands issued by the power grid dispatch master station, and parses them to obtain the total regulation demand as well as the corresponding power zoning and response time requirements;

[0041] Based on the hierarchical structure of the virtual power plant multidimensional aggregation model, virtual machine groups that meet the requirements of power zoning and response time are quickly selected according to the four-dimensional labels; the total regulation demand is allocated to each main transformer-level aggregation unit.

[0042] Within each of the main transformer-level aggregation units, the adjustment demand is allocated to each line-level aggregation unit;

[0043] Within each of the line-level aggregation units, the adjustment requirements are allocated to each area-level virtual machine group;

[0044] The adjustment amount allocated to each of the aforementioned district-level virtual machine groups is determined according to the proportion of the effective adjustment capacity of the district-level virtual machine group to the total effective adjustment capacity of all eligible virtual machine groups.

[0045] As a further technical solution of the present invention, the formation and distribution of the multi-strategy decomposition scheme includes:

[0046] For each level of the virtual power plant multidimensional aggregation model, the virtual power plant aggregation control subsystem forms at least the following six multi-strategy decomposition schemes: regulation capacity ratio strategy, standardized regulation capacity ratio strategy, regulation capacity priority strategy, regulation cost minimum strategy, regulation performance coefficient priority strategy, and manually set priority strategy.

[0047] Based on the current power grid operation status and regulation requirements, a corresponding multi-strategy decomposition scheme is selected for each level of the virtual power plant multi-dimensional aggregation model;

[0048] The selected multi-strategy decomposition schemes at each level are distributed to the virtual power plant operation platform;

[0049] The virtual power plant operation platform generates a corresponding remote control or remote adjustment control command sequence based on the received multi-strategy decomposition schemes at each level, and sends it to the three-in-one fusion terminal deployed on the resource side for execution;

[0050] The applicable scenarios for the multi-strategy decomposition scheme are as follows: the regulation capacity ratio strategy is suitable for conventional frequency regulation scenarios, the regulation capacity priority strategy is suitable for emergency scenarios with large regulation amounts, the regulation cost minimum strategy is suitable for conventional frequency regulation market scenarios, the regulation performance coefficient priority strategy is suitable for high-precision frequency regulation demand scenarios, and the manually set priority strategy is suitable for special power supply or emergency control scenarios.

[0051] As a further technical solution of the present invention, the real-time correction process includes:

[0052] The virtual power plant operation platform continuously collects the operation data of all adjustable loads and energy storage resources according to the set sampling period, and calculates the adjustment target deviation of each virtual machine group in the multi-dimensional aggregation model of the virtual power plant. The adjustment target deviation is the difference between the target adjustment power and the actual total response power.

[0053] When the adjustment target deviation of any level of virtual machine group in the virtual power plant multidimensional aggregation model exceeds the correction deviation threshold and the duration of the adjustment target deviation exceeds the correction duration threshold, the virtual power plant operation platform starts the real-time correction process of the corresponding level, adopts an incremental PID feedback control algorithm, and calculates the additional adjustment amount by comprehensively considering the current adjustment target deviation, the historical cumulative adjustment target deviation and the rate of change of the adjustment target deviation.

[0054] The virtual power plant operation platform reports the additional regulation amount as a new regulation demand to the virtual power plant aggregation control subsystem. The virtual power plant aggregation control subsystem then performs local decomposition of the corresponding level based on the virtual power plant multidimensional aggregation model and sends it to the virtual power plant operation platform for execution until the regulation target deviation of all virtual machine groups at all levels meets the allowable range.

[0055] After the regulation cycle ends, the virtual power plant operation platform will statistically analyze the actual regulation amount and regulation performance of virtual machine groups at each level in the virtual power plant multidimensional aggregation model, release virtual machine groups whose remaining adjustable capacity is lower than the regulation potential depletion threshold, update the real-time regulation capability data of the virtual power plant adjustable resource pool, and report the regulation execution results and performance evaluation data to the virtual power plant aggregation control subsystem.

[0056] Compared with existing technologies, the beneficial effects of this method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage are:

[0057] By establishing a unified modeling standard based on five dimensions—adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost—and standardizing adjustable loads and energy storage resources accordingly, the dimensional differences and interface barriers caused by the different physical nature of heterogeneous resources are eliminated. This enables various types of resources, such as electric vehicle charging and swapping facilities, interruptible industrial loads, mobile energy storage vehicles, energy storage warehouses, and tower base station energy storage, to be uniformly described and collaboratively managed within the same feature space. Simultaneously, through 5G short-slice communication technology, a three-in-one fusion terminal, and a full-link vertical encryption device, millisecond-level secure transmission from the resource side to the production control area is achieved, meeting the dual requirements of data timeliness and security for real-time power grid regulation.

[0058] A multi-dimensional aggregation model of the virtual power plant is constructed based on four dimensions: administrative region, power zone, load type, and response time level. Data aggregation is then nested at time scales of seconds, minutes, hours, and day-ahead, and spatial scales of distribution area, line, and main transformer levels, forming a spatiotemporally nested adjustable resource pool for the virtual power plant. This resource pool retains the original regulatory characteristics of various resources while possessing the ability to quickly retrieve data by time and space. This allows the dispatching side to simultaneously grasp the hierarchical and zonal regulation capabilities of the virtual power plant at multiple time scales, including second-level primary frequency regulation, minute-level secondary frequency regulation, hourly peak shaving, and day-ahead planned dispatch, achieving refined resource management and rapid response across multiple scenarios.

[0059] The real-time regulation capability data reported by the virtual machine group is verified through a three-level process of physical boundary verification, availability verification, and historical deviation verification. When the historical deviation ratio exceeds the set threshold, the data is marked as suspicious and automatic control is suspended. This effectively avoids interference of erroneous data with the AGC control strategy and significantly improves the operational safety and data reliability of the virtual power plant participating in grid frequency regulation.

[0060] The total regulation demand issued by the dispatching master station is decomposed step by step to the smallest control unit according to the power grid hierarchy of main transformer, line, and distribution area, so that the allocation of regulation commands matches the actual physical topology of the power grid. At the same time, proportional allocation is carried out based on the standardized resource adjustability, which ensures that adjustable loads and energy storage resources share regulation tasks fairly according to their actual capabilities. This avoids the problem that energy storage resources are overloaded due to their fast response speed, while adjustable loads are marginalized due to their slow response speed, and realizes the coordinated participation of the two types of resources.

[0061] By constructing six decomposition strategies—regulation capacity ratio, standardized regulation capacity ratio, regulation capacity priority, lowest regulation cost, regulation performance coefficient priority, and manually set priority—the optimal strategy can be flexibly selected according to different operating scenarios such as conventional frequency regulation, emergency regulation with large regulation volume, high-precision frequency regulation requirements, special power supply protection, or emergency control. This significantly improves the adaptability and regulation effect of the virtual power plant to the complex operating conditions of the power grid.

[0062] By continuously monitoring the deviation of the regulation target during the regulation execution process and using an incremental PID feedback control algorithm to calculate the additional regulation amount, the provincial virtual power plant aggregation control subsystem re-decomposes and distributes the execution, forming a closed-loop control of monitoring, judgment, correction, re-decomposition, and re-execution. This enables the aggregation unit to dynamically follow the AGC regulation target throughout the entire process, keeping the regulation deviation within the allowable range and significantly improving the control accuracy and response quality of the virtual power plant participating in grid frequency regulation.

[0063] The system adopts a two-tier collaborative architecture consisting of a provincial-level virtual power plant aggregation control subsystem and a municipal-level virtual power plant operation platform. The provincial platform is responsible for receiving, aggregating, verifying, and intelligently decomposing AGC commands, while the municipal platform is responsible for resource access, standardized modeling, command execution, and real-time monitoring. This approach not only meets the unified control requirements of the dispatching side but also enables refined operation of the resource side, forming a highly efficient collaborative virtual power plant operation mode between the dispatching and resource sides. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of the method flow of the present invention.

[0065] Figure 2 This is a schematic diagram of the two-level collaborative architecture of the provincial-level virtual power plant aggregation control subsystem and the municipal-level virtual power plant operation platform of the present invention.

[0066] Figure 3 This is a schematic diagram of the virtual power plant resource pool construction and control functional architecture of the present invention.

[0067] Figure 4 This is a schematic diagram of the virtual power plant resource security access and boundary protection architecture of the present invention. Detailed Implementation

[0068] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] Please see Figure 1 The present invention provides an embodiment of a method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage, comprising:

[0070] Adjustable loads and energy storage resources are connected to the virtual power plant operation platform. The virtual power plant operation platform formulates unified modeling standards in five aspects: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost, and establishes a standardized resource model based on the unified modeling standards.

[0071] The virtual power plant operation platform synchronizes the standardized resource model to the virtual power plant aggregation control subsystem. Based on the standardized resource model, the virtual power plant aggregation control subsystem uses virtual machine groups as the smallest control unit to construct a multi-dimensional aggregation model of the virtual power plant from four dimensions: administrative region, power zone, load type, and response time level. It performs nested data aggregation on the real-time operation data and controllability of virtual machine groups to form a spatiotemporally nested virtual power plant adjustable resource pool, and verifies the validity of the real-time adjustment capability data of virtual machine groups in the virtual power plant adjustable resource pool.

[0072] The virtual power plant aggregation control subsystem receives AGC adjustment instructions and, based on the real-time adjustment capability data of virtual machine groups in the virtual power plant's adjustable resource pool, decomposes the AGC adjustment instructions to the smallest control unit level according to the power grid level, forming a multi-strategy decomposition scheme and sending it to the virtual power plant operation platform for execution.

[0073] The virtual power plant operation platform continuously monitors the regulation effect during the regulation execution process, and initiates a real-time correction process when the actual response fails to achieve the regulation effect.

[0074] Furthermore, in another embodiment, the adjustable load includes electric vehicle charging and swapping facilities and interruptible industrial loads, and the energy storage resources include mobile energy storage vehicles, energy storage warehouses, and tower base station energy storage.

[0075] The physical location of adjustable loads and energy storage resources is the resource side;

[0076] The virtual power plant operation platform is located at the prefecture-level city level, which is physically close to the resource side; the virtual power plant aggregation control subsystem is located at the provincial level, which is physically close to the power grid dispatching side.

[0077] Adjustable loads and energy storage resources are connected to the virtual power plant operation platform through a three-in-one integrated terminal deployed on the resource side. The three-in-one integrated terminal integrates communication gateway functions, vertical encryption authentication functions, and AGC remote control functions, and uses 5G short slice communication technology for data transmission.

[0078] The resource-side three-in-one converged terminal is equipped with a miniature vertical encryption device. The miniature vertical encryption device works with the vertical encryption device deployed in the secure access area to form an encrypted tunnel. Data is transmitted to the production control area where the virtual power plant operation platform is located via the power 5G private network and the encrypted tunnel.

[0079] After being connected, the adjustable load and energy storage resources transmit data to the virtual power plant operation platform in accordance with the unified information transmission standard. The unified information transmission standard specifies the transmission requirements for the following data items: three telemetry data items, namely the grid-connected active power, the upper limit of frequency regulation active power, and the lower limit of frequency regulation active power; two tele-signaling data items, namely whether the power plant participates in frequency regulation and whether the power plant controls local control to be put into or taken out of regulation; one remote adjustment data item, namely the target value of frequency regulation active power; and one remote control data item, namely the power plant remote control to put into or take out of frequency regulation.

[0080] Furthermore, in another embodiment, the establishment of the standardized resource model includes:

[0081] The virtual power plant operation platform clearly defines the measurement benchmarks and standardization rules for each of the five dimensions: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost; and collects the original characteristic parameters of the above five dimensions for each adjustable load and energy storage resource.

[0082] The original feature parameters are normalized. The adjustable capacity, response speed, and adjustable duration are normalized using a forward normalization method, while the response delay time and adjustment cost are normalized using a reverse normalization method. The original feature parameters are converted into dimensionless values ​​in the range of 0 to 1, generating a standardized resource model containing the above five dimensions of normalized parameters.

[0083] The standardized resource model is used to construct the subsequent multidimensional aggregation model of the virtual power plant.

[0084] Furthermore, in another embodiment, the standardized resource model is synchronized to the virtual power plant aggregation control subsystem using a combination of full synchronization and incremental synchronization, specifically as follows:

[0085] When there are new adjustable loads and energy storage resources or when the characteristics of existing adjustable loads and energy storage resources change, the virtual power plant operation platform actively sends the changed standardized resource model to the virtual power plant aggregation control subsystem.

[0086] The virtual power plant aggregation control subsystem can proactively request a full set of standardized resource models from the virtual power plant operation platform, and after parsing and confirming the full set of standardized resource models, store them in the local real-time database.

[0087] The virtual power plant aggregation control subsystem constructs a multi-dimensional aggregation model of the virtual power plant based on the synchronized standardized resource model.

[0088] Furthermore, in another embodiment, the virtual power plant multidimensional aggregation model is constructed in the following specific way:

[0089] The virtual power plant aggregation control subsystem groups multiple adjustable loads and energy storage resources that have the same administrative region, the same power zone, the same load type, and the same response time level into one group;

[0090] Each group constitutes a virtual machine group, and the four-dimensional label of the virtual machine group is its administrative region, power zone, load type, and response time level.

[0091] The overall adjustability of a virtual machine group is obtained by weighting and summing the adjustable capacities of all adjustable loads and energy storage resources contained in the virtual machine group according to their respective collaborative weights, and the sum of the collaborative weights of all adjustable loads and energy storage resources within the same virtual machine group is 1.

[0092] The virtual power plant aggregation control subsystem uses virtual machine groups as the smallest unit and performs hierarchical nested aggregation according to the order of response time level, load type, power zone, and administrative region to form a five-layer virtual power plant multidimensional aggregation model including virtual machine group layer, response time level layer, load type layer, power zone layer, and administrative region layer.

[0093] The four-dimensional label serves as a unique identifier for the virtual machine group, used for rapid filtering and location during subsequent instruction decomposition.

[0094] Furthermore, in another embodiment, the specific method for forming the spatiotemporally nested virtual power plant adjustable resource pool is as follows:

[0095] The virtual power plant aggregation control subsystem performs nested aggregation of the real-time operating data and controllability of virtual machine groups at all levels in the virtual power plant multidimensional aggregation model, according to four time scales: second, minute, hour, and day-ahead, and three spatial scales: distribution area, line, and main transformer.

[0096] The spatiotemporal nesting aggregation coefficients of each virtual machine group at different time and space scales are obtained by multiplying the preset time coefficients corresponding to the time scales, the preset space coefficients corresponding to the space scales, and the normalization adjustability of the virtual machine group.

[0097] The adjustable resource pool of the virtual power plant records the spatiotemporal nesting aggregation coefficients, real-time power, upper limit of regulation capability, lower limit of regulation capability, and corresponding four-dimensional labels of all levels of virtual machine groups in the multi-dimensional aggregation model of the virtual power plant.

[0098] Furthermore, in another embodiment, the validity verification of the real-time adjustment capability data of virtual machine groups in the adjustable resource pool of the virtual power plant adopts a three-level verification process, specifically as follows:

[0099] The first level of physical boundary verification verifies that the upward adjustment capability reported by the corresponding level of virtual machine group in the adjustable resource pool of the virtual power plant does not exceed the difference between the rated power and the real-time power of the corresponding level of virtual machine group, and the downward adjustment capability reported does not exceed the real-time power of the corresponding level of virtual machine group.

[0100] The second level of availability verification determines whether the corresponding virtual machine group is qualified to participate in regulation based on the controllable status, maintenance status, adjustable time period, and power supply guarantee time period of the corresponding level virtual machine group in the adjustable resource pool of the virtual power plant. The effective regulation capability of the corresponding level virtual machine group that is not qualified to participate in regulation is set to zero.

[0101] The third level of historical deviation verification compares the total regulation capacity reported by the corresponding level of virtual machine group in the adjustable resource pool of the virtual power plant with the deviation ratio of the average regulation capacity of the corresponding level of virtual machine group in similar historical periods. When the deviation ratio exceeds the set threshold, the corresponding data is marked as suspicious data, the automatic control of the corresponding level of virtual machine group is suspended and an alarm is issued.

[0102] Furthermore, in another embodiment, the AGC adjustment commands are decomposed level by level to the smallest control unit according to the power grid hierarchy, using a three-level decomposition process, specifically as follows:

[0103] The virtual power plant aggregation control subsystem receives AGC regulation commands issued by the power grid dispatch master station, and parses them to obtain the total regulation demand, the corresponding power zoning, and the response time requirements;

[0104] Based on the hierarchical structure of the virtual power plant multidimensional aggregation model, virtual machine groups that meet the requirements of power zoning and response time are quickly selected according to four-dimensional labels; the total regulation demand is allocated to each main transformer-level aggregation unit.

[0105] Within each main transformer-level aggregation unit, the regulation demand is allocated to each line-level aggregation unit;

[0106] Within each line-level aggregation unit, adjustment requirements are allocated to each area-level virtual machine group.

[0107] The adjustment capacity allocated to each area-level virtual machine group is determined according to the proportion of the effective adjustment capacity of the area-level virtual machine group to the total effective adjustment capacity of all eligible virtual machine groups.

[0108] Furthermore, in another embodiment, the formation and distribution of the multi-strategy decomposition scheme includes:

[0109] For each level of the virtual power plant multidimensional aggregation model, the virtual power plant aggregation control subsystem forms at least the following six multi-strategy decomposition schemes: regulation capacity ratio strategy, standardized regulation capacity ratio strategy, regulation capacity priority strategy, regulation cost minimum strategy, regulation performance coefficient priority strategy, and manually set priority strategy.

[0110] Based on the current power grid operation status and regulation requirements, select a corresponding multi-strategy decomposition scheme for each level of the virtual power plant multi-dimensional aggregation model;

[0111] The selected multi-strategy decomposition schemes at each level are distributed to the virtual power plant operation platform;

[0112] The virtual power plant operation platform generates a corresponding remote control or remote adjustment control command sequence based on the received multi-strategy decomposition schemes at each level, and sends it to the three-in-one integrated terminal deployed on the resource side for execution;

[0113] The applicable scenarios for the multi-strategy decomposition scheme are as follows: the regulation capacity ratio strategy is suitable for conventional frequency regulation scenarios, the regulation capacity priority strategy is suitable for emergency scenarios with large regulation amounts, the regulation cost minimum strategy is suitable for conventional frequency regulation market scenarios, the regulation performance coefficient priority strategy is suitable for high-precision frequency regulation demand scenarios, and the manually set priority strategy is suitable for special power supply or emergency control scenarios.

[0114] Furthermore, in another embodiment, the real-time correction process includes:

[0115] The virtual power plant operation platform continuously collects the operation data of all adjustable loads and energy storage resources according to the set sampling period, and calculates the adjustment target deviation of each virtual machine group in the multi-dimensional aggregation model of the virtual power plant. The adjustment target deviation is the difference between the target adjustment power and the actual total response power.

[0116] When the adjustment target deviation of any level of virtual machine group in the virtual power plant multidimensional aggregation model exceeds the correction deviation threshold and the duration of the adjustment target deviation exceeds the correction duration threshold, the virtual power plant operation platform starts the real-time correction process of the corresponding level, adopts the incremental PID feedback control algorithm, and calculates the additional adjustment amount by comprehensively considering the current adjustment target deviation, the historical cumulative adjustment target deviation and the rate of change of the adjustment target deviation.

[0117] The virtual power plant operation platform reports the additional regulation amount as a new regulation demand to the virtual power plant aggregation control subsystem. The virtual power plant aggregation control subsystem then performs local decomposition of the corresponding level based on the virtual power plant multidimensional aggregation model and sends it to the virtual power plant operation platform for execution until the regulation target deviation of all virtual machine groups at all levels meets the allowable range.

[0118] After the control period ends, the virtual power plant operation platform will statistically analyze the actual control volume and control performance of virtual machine groups at each level in the virtual power plant multidimensional aggregation model, release virtual machine groups whose remaining adjustable capacity is lower than the control potential depletion threshold, update the real-time control capability data of the virtual power plant adjustable resource pool, and report the control execution results and performance evaluation data to the virtual power plant aggregation control subsystem.

[0119] Please see Figures 1-4 This invention provides an application example: a method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage, which operates in a two-level collaborative architecture of a virtual power plant aggregation control subsystem and a virtual power plant operation platform. The virtual power plant operation platform is located at the prefecture-level city level, physically close to the resource side, and is responsible for the access of adjustable load and energy storage resources, standardized modeling, physical execution of regulation commands, and real-time monitoring of the regulation execution process. The virtual power plant aggregation control subsystem is located at the provincial level, physically close to the power grid dispatching side, and is responsible for receiving AGC regulation commands issued by the dispatching master station, performing multi-dimensional aggregation, regulation capability effectiveness verification, and intelligent hierarchical decomposition of commands based on the synchronized standardized resource model and real-time regulation capability data, and issuing the decomposed sub-commands to the virtual power plant operation platform for execution.

[0120] Step 1: Adjustable load and energy storage resource integration and standardized modeling:

[0121] 1.1 Resource Filtering:

[0122] Before connecting adjustable loads and energy storage resources to the virtual power plant operation platform, distributed adjustable loads and energy storage resources are screened for access according to the screening quantitative standards of response time not exceeding 10 seconds, adjustable capacity accounting for not less than 60%, and duration not less than 15 minutes. Adjustable loads include electric vehicle charging and swapping facilities and interruptible industrial loads; energy storage resources include mobile energy storage vehicles, energy storage warehouses, and tower base station energy storage.

[0123] 1.2 Communication Access:

[0124] By deploying a three-in-one converged terminal on the resource side and using 5G short-slice communication technology, the screened adjustable loads and energy storage resources are connected to the virtual power plant operation platform through the secure access area. The three-in-one converged terminal integrates communication gateway functions, vertical encryption authentication functions, and AGC remote control functions. The three-in-one converged terminal on the resource side is equipped with a miniature vertical encryption device. The miniature vertical encryption device and the vertical encryption device deployed in the secure access area work together to form an encrypted tunnel. Data is transmitted to the production control area where the virtual power plant operation platform is located via the power 5G private network and the encrypted tunnel. The end-to-end transmission delay is controlled at the millisecond level.

[0125] 1.3 Unified Information Submission Standards:

[0126] After being connected, the adjustable load and energy storage resources transmit data to the virtual power plant operation platform in accordance with the unified information transmission standard. The unified information transmission standard specifies the transmission requirements for the following data items: three telemetry data items, namely, the active power connected to the grid, the upper limit of frequency regulation active power, and the lower limit of frequency regulation active power; two tele-signaling data items, namely, whether the power plant participates in frequency regulation and whether the power plant controls local control to be put into or taken out of regulation; one remote adjustment data item, namely, the target value of frequency regulation active power; and one remote control data item, namely, whether the power plant remotely controls frequency regulation to be put into or taken out of regulation.

[0127] 1.4 Establishment of unified modeling standards:

[0128] The virtual power plant operation platform establishes a unified modeling standard for adjustable loads and energy storage resources in five aspects: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost. The unified modeling standard clarifies the measurement benchmarks and standardization rules for each of the five dimensions: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost.

[0129] 1.5 Establishment of a standardized resource model:

[0130] The virtual power plant operation platform establishes a standardized resource model based on unified modeling standards; let the first... The original feature parameter vector of each resource is:

[0131]

[0132] in: The resource number indicates the first... One adjustable load or energy storage resource; For the first Increase the capacity of each resource; For the first Reduce the capacity of individual resources; For the first The response speed of each resource; For the first Response latency of each resource; For the first The sustainable adjustment time of a resource; For the first The adjustment cost coefficient of each resource;

[0133] The virtual power plant operation platform collects the above five dimensions of original characteristic parameters for each adjustable load and energy storage resource, and performs normalization processing on the original characteristic parameters. Among them, the adjustable capacity, response speed, and adjustable duration are normalized using a forward normalization method, while the response delay time and adjustment cost are normalized using a reverse normalization method. The original characteristic parameters are converted into values ​​in the range of 0 to 1, generating a standardized resource model containing the above five dimensions of normalized parameters.

[0134] For adjustable capacity, response speed, and adjustable duration, a positive range normalization method is used:

[0135]

[0136] For response delay time and adjustment cost, the inverse range normalization method is used:

[0137]

[0138] in: For resource serial number; For dimension number, These correspond to five dimensions: adjustable capacity, response speed, response latency, adjustable duration, and adjustment cost. For the first The resource in the first The original value of the dimension; For the first The maximum original value of all resources in a dimension; For the first The minimum value of the original values ​​of all resources in a dimension;

[0139] After normalization, the obtained normalized parameters are mapped to normalized resource model components according to their dimension indices: Let , , , , , Adopted and The same normalization method is used to calculate independently, resulting in a standardized resource model:

[0140]

[0141] in: For the first Increased capacity after resource standardization; For the first Reduced capacity after resource standardization; For the first Response speed after resource standardization; For the first Response latency after resource normalization; For the first The sustainable adjustment time after resource standardization; For the first Adjustment costs after resource standardization; standardized resource model All components are in Within the range;

[0142] Step 2: Model synchronization and construction of a multi-dimensional aggregated resource pool:

[0143] 2.1 Synchronization of Standardized Resource Models:

[0144] The virtual power plant operation platform synchronizes the standardized resource model to the virtual power plant aggregation control subsystem. The model synchronization adopts a combination of full synchronization and incremental synchronization: when there are new adjustable loads and energy storage resources or when the characteristics of existing adjustable loads and energy storage resources change, the virtual power plant operation platform actively sends the changed standardized resource model to the virtual power plant aggregation control subsystem. The virtual power plant aggregation control subsystem can request the full standardized resource model from the virtual power plant operation platform, parse and confirm the full standardized resource model, and then store it in the local real-time database.

[0145] 2.2 Construction of a Multidimensional Aggregation Model for Virtual Power Plants:

[0146] The virtual power plant aggregation control subsystem constructs a multi-dimensional aggregation model of the virtual power plant based on the synchronized standardized resource model. The virtual power plant aggregation control subsystem divides multiple adjustable loads and energy storage resources with the same administrative region, the same power zone, the same load type, and the same response time level into a group. Each group constitutes a virtual machine group, and the four-dimensional label of the virtual machine group is its administrative region, power zone, load type, and response time level.

[0147] set up For resources The overall standardized adjustable capacity under the current operating conditions is obtained by weighting the standardized upward and downward adjustment capacities according to the current power margin:

[0148]

[0149] in: For resources Comprehensive standardized adjustable capacity; For resource serial number; For the first Increased capacity after resource standardization; For the first Reduced capacity after resource standardization; For resources The percentage of upward adjustment margin is determined by the current operating conditions. , For resources Rated power, For resources Real-time power;

[0150] Set up virtual machine group From resource collection The aggregated standardized adjustable capacity, composed of several adjustable loads and energy storage resources, is calculated using the following formula:

[0151]

[0152] in: For virtual machine groups Adjustable standardization capability after aggregation; This refers to the virtual machine group number; For resource serial number; For virtual machine groups The collection of resources it contains; For resources Comprehensive standardized adjustable capacity; For resources In virtual machine group The collaborative weight coefficient within the range satisfies ;

[0153] The virtual power plant aggregation control subsystem uses virtual machine groups as the smallest unit and performs hierarchical nested aggregation according to the order of response time level, load type, power zone, and administrative region to form a five-layer virtual power plant multidimensional aggregation model including virtual machine group layer, response time level layer, load type layer, power zone layer, and administrative region layer.

[0154] 2.3 Formation of a Spatiotemporally Nested Virtual Power Plant Adjustable Resource Pool:

[0155] The virtual power plant aggregation control subsystem aggregates the real-time operating data and controllability of virtual machine groups at all levels in the virtual power plant multidimensional aggregation model in a nested manner according to four time scales: second, minute, hour, and day-ahead, and three spatial scales: distribution area, line, and main transformer, forming a spatiotemporally nested virtual power plant adjustable resource pool.

[0156] Define spatiotemporal nested aggregation coefficients:

[0157]

[0158] in: For virtual machine groups In time scale and spatial scale Spatiotemporal nested aggregation coefficients; This refers to the virtual machine group number; This is the time scale number. These correspond to the timescales of seconds, minutes, hours, and days, respectively. This is the spatial scale number. These correspond to the distribution area level, line level, and main transformer level, respectively. For the first The preset time coefficients corresponding to each time scale have a value range of [value range missing]. Among them, the second-level data is taken Minute-level retrieval hourly Recently, the level was taken ; For the first The preset spatial coefficients corresponding to each spatial scale have a range of values. Among them, the district-level selection Line-level , main variable level ; For virtual machine groups Standardized adjustability;

[0159] Virtual power plant adjustable resource pool at a specific time scale and spatial scale The total adjustable capacity is:

[0160]

[0161] in: For the virtual power plant's adjustable resource pool on the time scale and spatial scale Total adjustable capacity; This refers to the virtual machine group number; For virtual machine groups The spatiotemporal nested aggregation coefficients;

[0162] The adjustable resource pool of the virtual power plant records the spatiotemporal nesting aggregation coefficients, real-time power, upper limit of regulation capability, lower limit of regulation capability, and corresponding four-dimensional labels of all virtual machine groups at all levels in the multi-dimensional aggregation model of the virtual power plant.

[0163] 2.4 Verification of the validity of data on the real-time adjustment capability of virtual machine groups:

[0164] The virtual power plant aggregation control subsystem verifies the validity of the real-time adjustment capability data of virtual machine groups in the adjustable resource pool of the virtual power plant; the validity verification adopts a three-level verification process.

[0165] The first level is physical boundary verification; (Set up virtual machine group) The reported real-time adjustment capability is Real-time downgrade capability is Real-time power is Rated power is Verify that the reported upward adjustment capacity of the corresponding level virtual machine group in the adjustable resource pool of the virtual power plant does not exceed the difference between the rated power and the real-time power of the corresponding level virtual machine group, and the reported downward adjustment capacity does not exceed the real-time power of the corresponding level virtual machine group.

[0166]

[0167]

[0168] in: This refers to the virtual machine group number; For virtual machine groups The ability to report and adjust in real time; For virtual machine groups The ability to report and adjust in real time; For virtual machine groups Real-time power; For virtual machine groups The rated power; if the reported data exceeds the above physical boundaries, it will be deemed invalid data;

[0169] The second level is availability verification; the virtual power plant aggregation control subsystem determines whether the corresponding level of virtual machine group is qualified to participate in regulation based on the controllable status, maintenance status, adjustable time period, and power supply guarantee time period of the corresponding level of virtual machine group in the adjustable resource pool of the virtual power plant; the effective regulation capability of the corresponding level of virtual machine group that is not qualified to participate in regulation is set to zero;

[0170] The third level is historical deviation verification; assuming the average adjustment capability of this virtual machine group during similar historical periods is... Define the deviation ratio:

[0171]

[0172] in: For virtual machine groups The deviation ratio; This refers to the virtual machine group number; For virtual machine groups Average adjustment capacity during similar historical periods; when the deviation ratio Exceeding the historical deviation threshold When this happens, the virtual power plant aggregation control subsystem will mark the corresponding data as suspicious, suspend automatic control of the corresponding virtual machine group, and issue an alarm; historical deviation threshold Usually taken to ;

[0173] Step 3: Decompose and issue AGC adjustment instructions level by level for execution:

[0174] 3.1 AGC adjustment command reception:

[0175] The virtual power plant aggregation control subsystem receives AGC regulation commands issued by the power grid dispatch master station; the virtual power plant aggregation control subsystem parses the commands to obtain the total regulation demand and the corresponding power zoning and response time requirements.

[0176] 3.2 Three-level decomposition process:

[0177] The virtual power plant aggregation control subsystem is based on the hierarchical structure of the virtual power plant multidimensional aggregation model. It quickly selects virtual machine groups that meet the requirements of power zoning and response time according to four-dimensional labels. It decomposes AGC adjustment commands to the smallest control unit level according to the grid level. The grid level includes three levels: main transformer level, line level, and distribution area level.

[0178] Let the total adjustment demand required by the dispatch master station be... The total effective tolerability of all compliant virtual machine groups is Define virtual machine groups The proportion of standardized adjustable capabilities is:

[0179]

[0180] in: For virtual machine groups The proportion of standardized adjustable capabilities; This is the current virtual machine group number; For virtual machine groups Effective regulatory capacity after verification in step 2.4; Total effective tolerability for all eligible virtual machine groups; The index for traversing all virtual machine groups in the summation operation;

[0181] Assume a primary variable-level aggregation unit Includes virtual machine group collection Then it is allocated to the main variable level aggregation unit. The adjustment requirements are:

[0182]

[0183] in: To be allocated to the main transformer stage aggregation unit The need for regulation; The sequence number of the primary variable-level polymerization unit; To traverse the principal variable level aggregation unit in the summation operation Indexes of internal virtual machine groups; Main variable-level polymerization unit The collection of virtual machine groups included; For virtual machine groups Effective regulatory capacity; Total effective tolerability for all eligible virtual machine groups;

[0184] Set up line-level aggregation unit Includes virtual machine group collection And this line-level aggregation unit Belongs to the main variable level aggregation unit Then it is allocated to the line-level aggregation unit. The adjustment requirements are:

[0185]

[0186] in: To be allocated to line-level aggregation units The need for regulation; This refers to the sequence number of the line-level aggregation unit; The index of the virtual machine group within the corresponding set is used for summation operations; Line-level aggregation unit The collection of virtual machine groups included; To be allocated to the main transformer stage aggregation unit The need for regulation; Main variable-level polymerization unit The collection of virtual machine groups included; For virtual machine groups Effective regulatory capacity;

[0187] In each line-level aggregation unit Internally, adjustment requirements are allocated to each area-level virtual machine group; the adjustment amount allocated to each area-level virtual machine group is determined according to the proportion of the effective adjustment capacity of the area-level virtual machine group to the total effective adjustment capacity of all virtual machine groups within its corresponding line-level aggregation unit.

[0188]

[0189] in: To be assigned to virtual machine group The adjustment sub-instruction; The current virtual machine group number, and ; To traverse line-level aggregation units in the summation operation Indexes of all virtual machine groups within; To be allocated to line-level aggregation units The need for regulation; For virtual machine groups Effective regulatory capacity; Line-level aggregation unit The collection of virtual machine groups included;

[0190] 3.3 Generation of Multi-Strategy Decomposition Schemes:

[0191] For each level of the virtual power plant multidimensional aggregation model, the virtual power plant aggregation control subsystem forms at least the following six multi-strategy decomposition schemes: regulation capacity ratio strategy, standardized regulation capacity ratio strategy, regulation capacity priority strategy, regulation cost minimum strategy, regulation performance coefficient priority strategy, and manually set priority strategy.

[0192] When decomposing adjustment capabilities proportionally, the adjustment sub-instructions undertaken by each virtual machine group are... Sure;

[0193] When decomposing the standardized adjustment capability proportionally, a response time correction coefficient is introduced:

[0194]

[0195] in: To be assigned to virtual machine group The adjustment sub-instruction; The current virtual machine group number, and ; To traverse line-level aggregation units in the summation operation Indexes of all virtual machine groups within; To be allocated to line-level aggregation units The need for regulation; For virtual machine groups The proportion of standardized adjustable capabilities; For virtual machine groups Response latency; The system's baseline response time has a range of values. to The settings shall be determined by those skilled in the art based on the system characteristics; For virtual machine groups The proportion of standardized adjustable capabilities; For virtual machine groups Response latency; Line-level aggregation unit The collection of virtual machine groups included;

[0196] Based on the decomposition of minimum adjustment cost, an optimization model is established with the objective of minimizing adjustment cost:

[0197]

[0198] The constraints are and ;in: To traverse line-level aggregation units in the summation operation Indexes of internal virtual machine groups; For virtual machine groups The adjustment cost coefficient; To be assigned to virtual machine group The adjustment sub-instruction; To be allocated to line-level aggregation units The need for regulation; For virtual machine groups Effective regulatory capacity; Line-level aggregation unit The collection of virtual machine groups included;

[0199] When prioritizing the decomposition of performance coefficients, define virtual machine groups. performance coefficient for:

[0200]

[0201] in: For virtual machine groups Performance coefficient; This is the current virtual machine group number; For the weighting coefficients, satisfying , usually take , , ; For virtual machine groups Response speed; This represents the maximum response speed for the entire virtual machine group. For virtual machine groups Response latency; This represents the maximum response latency for the entire virtual machine group. For virtual machine groups Historical adjustment deviation rate; To allow for adjustment of the deviation rate, it is usually taken as follows: to ;according to Arrange the virtual machines in descending order, prioritizing the allocation of tuning sub-instructions to virtual machine groups with high performance coefficients, until the desired performance is achieved. ,in To traverse line-level aggregation units in the summation operation Indexes of internal virtual machine groups;

[0202] When decomposing based on manually set priorities, the priority weights are used. Proportional allocation:

[0203]

[0204] in: To be assigned to virtual machine group The adjustment sub-instruction; The current virtual machine group number, and ; To traverse line-level aggregation units in the summation operation Indexes of all virtual machine groups within; For virtual machine groups Manually set priority weights; To be allocated to line-level aggregation units The need for regulation; Line-level aggregation unit The collection of virtual machine groups included;

[0205] The virtual power plant aggregation control subsystem selects a corresponding multi-strategy decomposition scheme for each level of the virtual power plant multi-dimensional aggregation model based on the current power grid operation status and regulation requirements. The applicable scenarios for the multi-strategy decomposition scheme are as follows: the regulation capacity ratio strategy is suitable for conventional frequency regulation scenarios, the regulation capacity priority strategy is suitable for emergency scenarios with large regulation volume, the regulation cost minimum strategy is suitable for conventional frequency regulation market scenarios, the regulation performance coefficient priority strategy is suitable for high-precision frequency regulation demand scenarios, and the manually set priority strategy is suitable for special power supply protection or emergency control scenarios.

[0206] 3.4 Decomposition plan issued and executed:

[0207] The virtual power plant aggregation control subsystem distributes the selected multi-strategy decomposition schemes at each level to the virtual power plant operation platform; the virtual power plant operation platform generates the corresponding remote control or remote adjustment control command sequence based on the received multi-strategy decomposition schemes at each level, and distributes it to the three-in-one fusion terminal deployed on the resource side for execution.

[0208] Step 4: Real-time monitoring and adaptive correction:

[0209] 4.1 Continuous monitoring of the adjustment effect:

[0210] The virtual power plant operation platform continuously collects operational data from all adjustable loads and energy storage resources according to a set sampling period, and calculates the adjustment target deviation of each virtual machine group at each level in the virtual power plant multidimensional aggregation model; let the first... The target adjustment power at each sampling time is The actual total response power is Define the target deviation for adjustment:

[0211]

[0212] in: For the first The adjustment target deviation at each sampling time; The sampling time sequence number; For the first Target adjustment power at each sampling time; For the first The actual total response power at each sampling time;

[0213] 4.2 Real-time correction process initiation and execution:

[0214] When the adjustment target deviation of any level of virtual machine group in the virtual power plant multidimensional aggregation model exceeds the correction deviation threshold and the duration of the adjustment target deviation exceeds the correction duration threshold, the virtual power plant operation platform initiates the real-time correction process at the corresponding level; the judgment condition is:

[0215]

[0216] in: For the first The adjustment target deviation at each sampling time; The deviation threshold for correction is usually taken as the adjustment command. to ; The duration of the deviation; The threshold for correction duration is usually set to [value missing]. to ;

[0217] After initiating the real-time correction process, the virtual power plant operation platform employs an incremental PID feedback control algorithm to calculate the additional adjustment amount by comprehensively considering the current adjustment target deviation, the historical cumulative adjustment target deviation, and the rate of change of the adjustment target deviation.

[0218]

[0219] in: For the first Additional adjustment amount at each sampling time; The sampling time sequence number; For the summation variable, let represent the summation from the th... From the sampling time to the... The accumulation of each sampling moment; This is the proportionality coefficient, and its value range is... to ; The integral coefficient has a range of values ​​of 1. to ; These are the differential coefficients, and their range is... to ; For the first The adjustment target deviation at each sampling time; For the first The adjustment target deviation at each sampling time;

[0220] The virtual power plant operation platform will add regulation. As a new regulation requirement, it is reported to the virtual power plant aggregation control subsystem. The virtual power plant aggregation control subsystem then performs local decomposition of the corresponding level based on the virtual power plant multidimensional aggregation model and sends it to the virtual power plant operation platform for execution until the regulation target deviation of all virtual machine groups meets the allowable range.

[0221] 4.3 Handling at the end of the regulation cycle:

[0222] After the control period ends, the virtual power plant operation platform statistically analyzes the actual control volume and performance of virtual machine groups at each level in the virtual power plant multi-dimensional aggregation model, releases virtual machine groups whose remaining adjustable capacity is below the controllable potential depletion threshold, updates the real-time controllable capacity data of the virtual power plant adjustable resource pool, and reports the control execution results and performance evaluation data to the virtual power plant aggregation control subsystem. The controllable potential depletion threshold is typically taken as a percentage of the rated capacity of the virtual machine group. to .

[0223] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage, characterized in that, include: Adjustable loads and energy storage resources are connected to a virtual power plant operation platform. The virtual power plant operation platform formulates a unified modeling standard from five aspects: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost, and establishes a standardized resource model based on the unified modeling standard. The virtual power plant operation platform synchronizes the standardized resource model to the virtual power plant aggregation control subsystem. Based on the standardized resource model, the virtual power plant aggregation control subsystem constructs a multi-dimensional aggregation model of the virtual power plant from four dimensions: administrative region, power zone, load type, and response time level, with virtual machine groups as the smallest control unit. The subsystem performs nested data aggregation on the real-time operation data and controllability of the virtual machine groups to form a spatiotemporally nested virtual power plant adjustable resource pool, and verifies the validity of the real-time adjustment capability data of the virtual machine groups in the virtual power plant adjustable resource pool. The virtual power plant aggregation control subsystem receives AGC adjustment instructions and, based on the real-time adjustment capability data of the virtual machine groups in the virtual power plant's adjustable resource pool, decomposes the AGC adjustment instructions to the smallest control unit level according to the power grid level, forming a multi-strategy decomposition scheme and sending it to the virtual power plant operation platform for execution. The virtual power plant operation platform continuously monitors the regulation effect during the regulation execution process, and initiates a real-time correction process when the actual response fails to achieve the regulation effect.

2. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The adjustable loads include electric vehicle charging and swapping facilities and interruptible industrial loads; the energy storage resources include mobile energy storage vehicles, energy storage warehouses, and tower base station energy storage. The physical location of the adjustable load and energy storage resources is the resource side; The virtual power plant operation platform is located at the prefecture-level city level, physically close to the resource side; the virtual power plant aggregation control subsystem is located at the provincial level, physically close to the power grid dispatching side. The adjustable load and energy storage resources are connected to the virtual power plant operation platform through a three-in-one integrated terminal deployed on the resource side. The three-in-one integrated terminal integrates communication gateway function, vertical encryption authentication function, and AGC remote control function, and uses 5G short slice communication technology for data transmission. The resource-side three-in-one converged terminal is equipped with a miniature vertical encryption device. The miniature vertical encryption device, together with the vertical encryption device deployed in the secure access area, forms an encrypted tunnel. Data is transmitted to the production control area where the virtual power plant operation platform is located via the power 5G private network and the encrypted tunnel. After being connected, the adjustable load and energy storage resources transmit data to the virtual power plant operation platform in accordance with the unified information transmission standard. The unified information transmission standard specifies the transmission requirements for the following data items: three telemetry data items including grid-connected active power, frequency regulation active power upper limit, and frequency regulation active power lower limit; two tele-signaling data items indicating whether the power plant participates in frequency regulation and whether the power plant controls local activation or deactivation; one remote adjustment data item indicating the frequency regulation active power target value; and one remote control data item indicating whether the power plant remotely controls frequency regulation activation or deactivation.

3. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The establishment of the standardized resource model includes: The virtual power plant operation platform clearly defines the measurement benchmarks and standardization rules for each of the five dimensions: adjustable capacity, response speed, response delay time, adjustable duration, and adjustment cost; and collects the original characteristic parameters of the above five dimensions for each adjustable load and energy storage resource. The original feature parameters are normalized, wherein the adjustable capacity, response speed, and adjustable duration are normalized using a forward normalization method, and the response delay time and adjustment cost are normalized using a reverse normalization method. The original feature parameters are converted into dimensionless values ​​in the range of 0 to 1, generating a standardized resource model containing the above five dimensions of normalized parameters. The standardized resource model is used to construct the subsequent multidimensional aggregation model of the virtual power plant.

4. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The standardized resource model is synchronized to the virtual power plant aggregation control subsystem using a combination of full synchronization and incremental synchronization, specifically as follows: When there are new adjustable loads and energy storage resources or when the characteristics of existing adjustable loads and energy storage resources change, the virtual power plant operation platform actively sends the changed standardized resource model to the virtual power plant aggregation control subsystem. The virtual power plant aggregation control subsystem can proactively request a full set of standardized resource models from the virtual power plant operation platform, and after parsing and confirming the full set of standardized resource models, store them in the local real-time database. The virtual power plant aggregation control subsystem constructs a multi-dimensional aggregation model of the virtual power plant based on the synchronized standardized resource model.

5. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The specific method for constructing the virtual power plant multidimensional aggregation model is as follows: The virtual power plant aggregation control subsystem groups multiple adjustable loads and energy storage resources that have the same administrative region, the same power zone, the same load type, and the same response time level into one group; Each group constitutes a virtual machine group, and the four-dimensional label of the virtual machine group is its administrative region, power zone, load type, and response time level. The overall adjustability of the virtual machine group is obtained by weighting and summing the adjustable capacities of all adjustable loads and energy storage resources contained in the virtual machine group according to their respective collaborative weights, and the sum of the collaborative weights of all adjustable loads and energy storage resources in the same virtual machine group is 1. The virtual power plant aggregation control subsystem uses virtual machine groups as the smallest unit and performs hierarchical nested aggregation according to the order of response time level, load type, power zone, and administrative region to form a five-layer virtual power plant multidimensional aggregation model, including virtual machine group layer, response time level layer, load type layer, power zone layer, and administrative region layer.

6. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The specific method for forming the spatiotemporally nested virtual power plant adjustable resource pool is as follows: The virtual power plant aggregation control subsystem performs nested aggregation of the real-time operating data and controllability of virtual machine groups at all levels in the virtual power plant multidimensional aggregation model, according to four time scales: second, minute, hour, and day-ahead, and three spatial scales: distribution area, line, and main transformer. The spatiotemporal nesting aggregation coefficients of each virtual machine group at different time and space scales are obtained by multiplying the preset time coefficients corresponding to the time scales, the preset space coefficients corresponding to the space scales, and the normalization adjustability of the virtual machine group. The adjustable resource pool of the virtual power plant records the spatiotemporal nesting aggregation coefficients, real-time power, upper limit of adjustment capability, lower limit of adjustment capability, and corresponding four-dimensional labels of all levels of virtual machine groups in the multi-dimensional aggregation model of the virtual power plant.

7. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The validity verification of the real-time adjustment capability data of virtual machine groups in the adjustable resource pool of the virtual power plant adopts a three-level verification process, specifically: The first level of physical boundary verification verifies that the upward adjustment capability reported by the corresponding level virtual machine group in the adjustable resource pool of the virtual power plant does not exceed the difference between the rated power and the real-time power of the corresponding level virtual machine group, and the downward adjustment capability reported does not exceed the real-time power of the corresponding level virtual machine group. The second level of availability verification determines whether the corresponding virtual machine group is qualified to participate in regulation based on the controllable status, maintenance status, adjustable time period, and power supply guarantee time period of the corresponding level virtual machine group in the adjustable resource pool of the virtual power plant. The effective regulation capability of the corresponding level virtual machine group that is not qualified to participate in regulation is set to zero. The third level of historical deviation verification compares the total adjustment capacity reported by the corresponding level of virtual machine group in the adjustable resource pool of the virtual power plant with the average adjustment capacity of the corresponding level of virtual machine group in similar historical periods. When the deviation ratio exceeds a set threshold, the corresponding data is marked as suspicious data, the automatic control of the corresponding level of virtual machine group is suspended and an alarm is issued.

8. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The AGC regulation commands are decomposed level by level to the smallest control unit according to the power grid hierarchy, using a three-level decomposition process, specifically: The virtual power plant aggregation control subsystem receives AGC regulation commands issued by the power grid dispatch master station, and parses them to obtain the total regulation demand, the corresponding power zoning, and the response time requirements; Based on the hierarchical structure of the virtual power plant multidimensional aggregation model, virtual machine groups that meet the requirements of power zoning and response time are quickly selected according to the four-dimensional labels; the total regulation demand is allocated to each main transformer-level aggregation unit. Within each of the main transformer-level aggregation units, the adjustment demand is allocated to each line-level aggregation unit; Within each of the line-level aggregation units, the adjustment requirements are allocated to each area-level virtual machine group; The adjustment amount allocated to each of the aforementioned district-level virtual machine groups is determined according to the proportion of the effective adjustment capacity of the district-level virtual machine group to the total effective adjustment capacity of all eligible virtual machine groups.

9. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The formation and distribution of the multi-strategy decomposition scheme includes: For each level of the virtual power plant multidimensional aggregation model, the virtual power plant aggregation control subsystem forms at least the following six multi-strategy decomposition schemes: regulation capacity ratio strategy, standardized regulation capacity ratio strategy, regulation capacity priority strategy, regulation cost minimum strategy, regulation performance coefficient priority strategy, and manually set priority strategy. Based on the current power grid operation status and regulation requirements, a corresponding multi-strategy decomposition scheme is selected for each level of the virtual power plant multi-dimensional aggregation model; The selected multi-strategy decomposition schemes at each level are distributed to the virtual power plant operation platform; The virtual power plant operation platform generates a corresponding remote control or remote adjustment control command sequence based on the received multi-strategy decomposition schemes at each level, and sends it to the three-in-one fusion terminal deployed on the resource side for execution; The applicable scenarios for the multi-strategy decomposition scheme are as follows: the regulation capacity ratio strategy is suitable for conventional frequency regulation scenarios, the regulation capacity priority strategy is suitable for emergency scenarios with large regulation amounts, the regulation cost minimum strategy is suitable for conventional frequency regulation market scenarios, the regulation performance coefficient priority strategy is suitable for high-precision frequency regulation demand scenarios, and the manually set priority strategy is suitable for special power supply or emergency control scenarios.

10. The method for constructing and regulating a virtual power plant resource pool with adjustable load and energy storage according to claim 1, characterized in that, The real-time correction process includes: The virtual power plant operation platform continuously collects the operation data of all adjustable loads and energy storage resources according to the set sampling period, and calculates the adjustment target deviation of each virtual machine group in the multi-dimensional aggregation model of the virtual power plant. The adjustment target deviation is the difference between the target adjustment power and the actual total response power. When the adjustment target deviation of any level of virtual machine group in the virtual power plant multidimensional aggregation model exceeds the correction deviation threshold and the duration of the adjustment target deviation exceeds the correction duration threshold, the virtual power plant operation platform starts the real-time correction process of the corresponding level, adopts an incremental PID feedback control algorithm, and calculates the additional adjustment amount by comprehensively considering the current adjustment target deviation, the historical cumulative adjustment target deviation and the rate of change of the adjustment target deviation. The virtual power plant operation platform reports the additional regulation amount as a new regulation demand to the virtual power plant aggregation control subsystem. The virtual power plant aggregation control subsystem then performs local decomposition of the corresponding level based on the virtual power plant multidimensional aggregation model and sends it to the virtual power plant operation platform for execution until the regulation target deviation of all virtual machine groups at all levels meets the allowable range. After the regulation cycle ends, the virtual power plant operation platform will statistically analyze the actual regulation amount and regulation performance of virtual machine groups at each level in the virtual power plant multidimensional aggregation model, release virtual machine groups whose remaining adjustable capacity is lower than the regulation potential depletion threshold, update the real-time regulation capability data of the virtual power plant adjustable resource pool, and report the regulation execution results and performance evaluation data to the virtual power plant aggregation control subsystem.

Images