Distributed energy aggregation control method, device, equipment, medium and chip

Abstract

This disclosure relates to the field of power distribution network technology, specifically to a distributed energy aggregation and management method, device, equipment, medium, and chip. The distributed energy aggregation and management method includes: when a first resource regulation condition is met, the group layer selects the most suitable subgroup from its subordinate subgroup layers and sends a first regulation command to it; after receiving the first regulation command, or when a second resource regulation condition is met, the most suitable subgroup selects the most suitable terminal from its subordinate terminal layer and sends a second regulation command to it; the most suitable terminal performs resource scheduling according to the received second regulation command. This technical solution achieves efficient and flexible global-local collaborative optimization through dynamic selection of the "most suitable subgroup" and "most suitable terminal" for hierarchical and progressive resource regulation, providing effective technical support for the stable operation of energy systems with a high proportion of renewable energy access.

Description

Technical Field

[0001] This disclosure relates to the field of power distribution network technology, specifically to a method, apparatus, equipment, medium, and chip for the aggregation and management of distributed energy resources. Background Art

[0002] In recent years, with the steady advancement of low-carbon energy transformation and electricity market reform, the personalized development of user-side demand, the mature application of distributed energy technologies, and the significant deepening of distribution network application scenarios have driven the increasing penetration rate of flexible resources in the power system. Promoting the participation of flexible resources in power system operation and electricity market transactions is a catalyst for achieving flexible and stable power system operation, industrialization of distributed resources, optimization of energy supply structure, and improvement of comprehensive energy efficiency. Currently, efforts are being made to build a new type of power system with new energy sources as the main body, characterized by extensive interconnection, intelligent interaction, flexibility, and security controllability. Aggregate architecture technology for the new power system can provide intelligent interconnection, flexible clustering, flexible regulation, and reliable collaboration technologies for massive, small-scale, multi-energy heterogeneous, geographically dispersed, and demand-diverse flexible resources, promoting the technological upgrade of the power grid to the energy internet.

[0003] Distributed energy aggregation and management methods refer to the use of advanced technologies to unify and optimize the scheduling of dispersed distributed energy resources, thereby achieving efficient energy utilization and supply-demand balance. This method is of great significance for improving the reliability of energy systems, reducing energy costs, and promoting the integration and consumption of renewable energy. The diversity and complexity of distributed energy technologies increase the difficulty of aggregation and management. Different types of distributed energy differ in their power generation principles, operating characteristics, and grid connection requirements, making unified management and optimization more complex. Distributed energy aggregation and management methods require the coordinated optimization and scheduling of multiple distributed energy resources. However, the decentralized and uncertain nature of distributed energy increases the difficulty of system optimization and scheduling. Existing optimization algorithms and scheduling strategies may not be fully adapted to the characteristics of distributed energy, leading to problems such as reduced system efficiency and energy waste. Summary of the Invention

[0004] To address the problems in related technologies, embodiments of this disclosure provide a distributed energy aggregation and management method, apparatus, device, medium, and chip.

[0005] In a first aspect, this disclosure provides a distributed energy aggregation and control method applicable to a distributed energy aggregation and control system. The distributed energy aggregation and control system includes a group layer, a subgroup layer, and a terminal layer. The group layer is configured with a group control center for collective decision-making at the regional distribution network level. The subgroup layer consists of edge decision-making devices for decomposing the control tasks of the group layer. The terminal layer includes a distribution network field monitoring unit and a distribution network field control unit. The method includes:

[0006] When the first resource regulation condition is met, the group layer selects the best-fit subgroup from its subordinate subgroup layers and sends the first regulation instruction to it;

[0007] After receiving the first control instruction or when the second resource control condition is met, the most suitable subgroup selects the most suitable terminal from its subordinate terminal layer and sends the second control instruction to it.

[0008] The most suitable terminal performs resource scheduling based on the received second control instruction.

[0009] In one embodiment of this disclosure, the condition of satisfying the first resource regulation condition includes: if a user demand instruction is received, then it is determined that the first resource regulation condition is satisfied; or, if the distributed energy supply within the group cannot meet the load requirements, then it is determined that the first resource regulation condition is satisfied; and / or,

[0010] The conditions under which the second resource regulation condition is met include: if the distributed energy supply within the subgroup cannot meet the load requirements, then the second resource regulation condition is determined to be met.

[0011] In one embodiment of this disclosure, the group layer selects the best-fit subgroup from its subordinate subgroup layers and sends a first control command to it, including:

[0012] The group layer obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers;

[0013] First decision information is obtained by extracting the first feature and configuring the first weight on the resource information of the subgroup nodes.

[0014] Based on the first decision information, the most suitable subgroup is selected and a first control instruction is sent to it;

[0015] The subgroup node resource information is obtained by integrating the terminal layer resource information uploaded by the terminal layer under the subgroup layer; the terminal layer resource information includes the operation information and resource supply status information of each terminal device.

[0016] In one embodiment of this disclosure, obtaining first decision information by performing first feature extraction and first weight configuration on the resource information of the subgroup nodes includes:

[0017] The first input vector is obtained by extracting the first feature from the resource information of the subgroup nodes;

[0018] The first input vector is input into the first feature attention mechanism to configure the first sub-weights, and the first feature vector is obtained.

[0019] The first feature vector is input into the temporal attention mechanism to configure the second sub-weights, resulting in the second feature vector;

[0020] The first decision information is obtained based on the second feature vector.

[0021] In one embodiment of this disclosure, selecting the most suitable terminal from its subordinate terminal layer and sending a second control command to it includes:

[0022] The second decision information is obtained by performing second feature extraction and second weight configuration on the terminal layer resource information;

[0023] Based on the second decision information, the most suitable terminal is selected and a second control command is sent to it.

[0024] In one embodiment of this disclosure, the step of obtaining second decision information by performing second feature extraction and second weight configuration on the terminal layer resource information includes:

[0025] The second input vector is obtained by performing a second feature extraction on the terminal layer resource information;

[0026] The second input vector is input into the second feature attention mechanism to configure the second weights, resulting in the third feature vector;

[0027] The second decision information is obtained based on the third feature vector.

[0028] In one embodiment of this disclosure, it further includes:

[0029] The second input vector is uploaded to the group layer as subgroup node resource information and stored in the subgroup node resource information database.

[0030] The group layer obtains the subgroup node resource information uploaded by its subordinate subgroup layers, including:

[0031] Obtain the subgroup node resource information uploaded by its subordinate subgroup layers from the subgroup node resource information database.

[0032] In one embodiment of this disclosure, the first decision information is a subgroup adaptation score calculated by weighted summation of the feature values ​​of the features configured with the first weight and the corresponding first weight.

[0033] The step of selecting the most suitable subgroup based on the first decision information and sending a first control command to it includes:

[0034] The calculated subgroup fit scores are compared, and the subgroup corresponding to the highest subgroup fit score is taken as the best-fit subgroup and a first control command is sent to it.

[0035] In one embodiment of this disclosure, the second decision information is a terminal adaptation score calculated by weighted summation of the feature values ​​of the features configured with the second weight and the corresponding second weight;

[0036] The step of selecting the most suitable terminal based on the second decision information and sending a second control command to it includes:

[0037] The calculated terminal adaptation scores are compared, and the terminal with the highest terminal adaptation score is selected as the most adapted terminal and a second control command is sent to it.

[0038] In one embodiment of this disclosure, it further includes:

[0039] If the distributed energy supply within the terminal layer cannot meet the load requirements, the terminal layer will perform resource scheduling on its subordinate terminal equipment.

[0040] Secondly, this disclosure provides a distributed energy aggregation and management device, comprising:

[0041] The first control module is configured to select the most suitable subgroup from its subordinate subgroup layers and send the first control instruction to it when the first resource control condition is met.

[0042] The second control module is configured to select the most suitable terminal from its subordinate terminal layer and send the second control instruction to the most suitable subgroup after it receives the first control instruction or when the second resource control conditions are met.

[0043] The resource scheduling module is configured to best suit the terminal to perform resource scheduling according to the received second control instruction;

[0044] The group layer is configured with a group control center to realize collective decision-making at the regional distribution network level. The subgroup layer is composed of edge decision-making devices to decompose the control tasks of the group layer. The terminal layer includes a distribution network field monitoring unit and a distribution network field control unit.

[0045] In one embodiment of this disclosure, the condition of satisfying the first resource regulation condition includes: if a user demand instruction is received, then it is determined that the first resource regulation condition is satisfied; or, if the distributed energy supply within the group cannot meet the load requirements, then it is determined that the first resource regulation condition is satisfied; and / or,

[0046] The conditions under which the second resource regulation condition is met include: if the distributed energy supply within the subgroup cannot meet the load requirements, then the second resource regulation condition is determined to be met.

[0047] In one embodiment of this disclosure, the part of the first control module that selects the most suitable subgroup from its subordinate subgroup layers and sends a first control command to it is configured as follows:

[0048] The group layer obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers;

[0049] First decision information is obtained by extracting the first feature and configuring the first weight on the resource information of the subgroup nodes.

[0050] Based on the first decision information, the most suitable subgroup is selected and a first control instruction is sent to it;

[0051] The subgroup node resource information is obtained by integrating the terminal layer resource information uploaded by the terminal layer under the subgroup layer; the terminal layer resource information includes the operation information and resource supply status information of each terminal device.

[0052] In one embodiment of this disclosure, the portion of the first control module that obtains the first decision information after performing first feature extraction and first weight configuration on the resource information of the subgroup nodes is configured as follows:

[0053] The first input vector is obtained by extracting the first feature from the resource information of the subgroup nodes;

[0054] The first input vector is input into the first feature attention mechanism to configure the first sub-weights, and the first feature vector is obtained.

[0055] The first feature vector is input into the temporal attention mechanism to configure the second sub-weights, resulting in the second feature vector;

[0056] The first decision information is obtained based on the second feature vector.

[0057] In one embodiment of this disclosure, the part of the second control module that selects the most suitable terminal from its subordinate terminal layer and sends a second control command to it is configured as follows:

[0058] The second decision information is obtained by performing second feature extraction and second weight configuration on the terminal layer resource information;

[0059] Based on the second decision information, the most suitable terminal is selected and a second control command is sent to it.

[0060] In one embodiment of this disclosure, the portion of the second decision information obtained by the second control module after performing second feature extraction and second weight configuration on the terminal layer resource information is configured as follows:

[0061] The second input vector is obtained by performing a second feature extraction on the terminal layer resource information;

[0062] The second input vector is input into the second feature attention mechanism to configure the second weights, resulting in the third feature vector;

[0063] The second decision information is obtained based on the third feature vector.

[0064] In one embodiment of this disclosure, it further includes:

[0065] The storage module is configured to upload the second input vector as subgroup node resource information to the group layer and store it in the subgroup node resource information database;

[0066] The part of the first control module that obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers from the group layer is configured as follows:

[0067] Obtain the subgroup node resource information uploaded by its subordinate subgroup layers from the subgroup node resource information database.

[0068] In one embodiment of this disclosure, the first decision information is a subgroup adaptation score calculated by weighted summation of the feature values ​​of the features configured with the first weight and the corresponding first weight.

[0069] The part of the first control module that selects the most suitable subgroup based on the first decision information and sends the first control command to it is configured as follows:

[0070] The calculated subgroup fit scores are compared, and the subgroup corresponding to the highest subgroup fit score is taken as the best-fit subgroup and a first control command is sent to it.

[0071] In one embodiment of this disclosure, the second decision information is a terminal adaptation score calculated by weighted summation of the feature values ​​of the features configured with the second weight and the corresponding second weight;

[0072] The part of the second control module that selects the most suitable terminal based on the second decision information and sends the second control command to it is configured as follows:

[0073] The calculated terminal adaptation scores are compared, and the terminal with the highest terminal adaptation score is selected as the most adapted terminal and a second control command is sent to it.

[0074] In one embodiment of this disclosure, it further includes:

[0075] The terminal layer scheduling module is configured to perform resource scheduling on its subordinate terminal devices if the distributed energy supply within the terminal layer cannot meet the load requirements.

[0076] Thirdly, embodiments of this disclosure provide an electronic device including a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method as described in any of the first aspects.

[0077] Fourthly, this disclosure provides a computer-readable storage medium having computer instructions stored thereon, which, when executed by a processor, implement the method as described in any of the first aspects.

[0078] Fifthly, embodiments of this disclosure provide a chip, which includes a processor for calling a computer program in memory to execute the above-described distributed energy aggregation and management method.

[0079] The technical effects provided by the embodiments of this disclosure may include the following beneficial effects:

[0080] According to the technical solution provided in this disclosure, a distributed energy aggregation and control method is applicable to a distributed energy aggregation and control system. The distributed energy aggregation and control system includes a group layer, a subgroup layer, and a terminal layer. The group layer is configured with a group control center for collective decision-making at the regional distribution network level. The subgroup layer consists of edge decision-making devices for decomposing the control tasks of the group layer. The terminal layer includes a distribution network field monitoring unit and a distribution network field control unit. The method includes: when a first resource control condition is met, the group layer selects the most suitable subgroup from its subordinate subgroup layers and sends a first control command to it; after receiving the first control command, or when a second resource control condition is met, the most suitable subgroup selects the most suitable terminal from its subordinate terminal layers and sends a second control command to it; the most suitable terminal performs resource scheduling according to the received second control command. The above technical solution performs hierarchical and progressive resource regulation by dynamically selecting the "most suitable subgroup" and "most suitable terminal". It decomposes the global problem of distributed resource scheduling into local optimization tasks, significantly reducing the computational complexity of a single decision and shortening the response time. While ensuring real-time performance, it achieves efficient and flexible global-local collaborative optimization, providing effective technical support for the stable operation of energy systems with a high proportion of renewable energy access.

[0081] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. Attached Figure Description

[0082] Figure 1 A flowchart is shown for a distributed energy aggregation and management method according to an embodiment of the present disclosure.

[0083] Figure 2 A flowchart illustrating a distributed energy aggregation and management method according to a specific embodiment of the present disclosure is shown.

[0084] Figure 3 A structural block diagram of a distributed energy aggregation and management device according to an embodiment of the present disclosure is shown.

[0085] Figure 4 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0086] Figure 5 A schematic diagram of the structure of a computer system suitable for implementing the method according to embodiments of the present disclosure is shown. Detailed Implementation

[0087] In the following, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings to enable those skilled in the art to readily implement them. Furthermore, for clarity, portions unrelated to the description of exemplary embodiments have been omitted from the drawings.

[0088] In this disclosure, it should be understood that terms such as “comprising” or “having” are intended to indicate the presence of features, figures, steps, behaviors, components, parts or combinations thereof disclosed in this specification, and are not intended to exclude the possibility of the presence or addition of one or more other features, figures, steps, behaviors, components, parts or combinations thereof.

[0089] It should also be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0090] Distributed energy aggregation and management methods refer to the use of advanced technologies to unify and optimize the scheduling of dispersed distributed energy resources, thereby achieving efficient energy utilization and supply-demand balance. This method is of great significance for improving the reliability of energy systems, reducing energy costs, and promoting the integration and consumption of renewable energy. The diversity and complexity of distributed energy technologies increase the difficulty of aggregation and management. Different types of distributed energy differ in their power generation principles, operating characteristics, and grid connection requirements, making unified management and optimization more complex. Distributed energy aggregation and management methods require the coordinated optimization and scheduling of multiple distributed energy resources. However, the decentralized and uncertain nature of distributed energy increases the difficulty of system optimization and scheduling. Existing optimization algorithms and scheduling strategies may not be fully adapted to the characteristics of distributed energy, leading to problems such as reduced system efficiency and energy waste.

[0091] In view of the above-mentioned shortcomings, the distributed energy aggregation and management method provided in this disclosure performs hierarchical and progressive resource regulation by dynamically selecting the "most suitable subgroup" and "most suitable terminal". It decomposes the global problem of distributed resource scheduling into local optimization tasks, significantly reduces the computational complexity of a single decision, shortens the response time, and achieves efficient and flexible global-local collaborative optimization while ensuring real-time performance. This provides effective technical support for the stable operation of energy systems with a high proportion of renewable energy access.

[0092] Figure 1 A flowchart is shown for a distributed energy aggregation and management method according to an embodiment of the present disclosure.

[0093] like Figure 1 As shown, the distributed energy aggregation and management method includes the following steps S110-S130:

[0094] In step S110, when the first resource regulation condition is met, the group layer selects the most suitable subgroup from its subordinate subgroup layers and sends the first regulation instruction to it.

[0095] In step S120, after receiving the first control instruction or when the second resource control condition is met, the most suitable subgroup selects the most suitable terminal from its subordinate terminal layer and sends the second control instruction to it.

[0096] In step S130, the most suitable terminal performs resource scheduling according to the received second control instruction.

[0097] The distributed energy aggregation and management method disclosed herein is applied to a distributed energy distribution network scheduling architecture consisting of a group layer, a subgroup layer, and a terminal layer. This distribution network scheduling architecture is a hierarchical interactive scheduling framework. Through a multi-level interactive energy scheduling model, the system operation is optimized iteratively by optimizing the operation mode of each subgroup layer and utilizing the adjustable multiple interactions between the subgroup layer and the group layer.

[0098] The group layer is configured with a group control center, such as a server device. Considering that the master station is the top layer of the distribution network dispatch architecture, the group control center is located on the master station side, and sends dispatch tasks to multiple subgroup layers below. It coordinates the subgroup layers to jointly complete the group decision through control, incentive and other instructions, so as to realize collective decision-making at the regional distribution network level.

[0099] The subgroup layer consists of edge decision-making devices. It is divided statically or dynamically. For groups composed of heterogeneous resources, static grouping is performed considering resource type, geographical distribution, affiliated stakeholders, and hierarchical relationships. For groups composed of homogeneous resources, which generally belong to the same stakeholder and do not involve interest conflicts, dynamic subgrouping is performed based on spatiotemporal characteristics such as resource geographical distribution, operational status, and adjustment capabilities. The subgroup layer's role is to further decompose the control tasks of the group layer, break down control center instructions, and optimize execution plans through distributed collaboration among groups.

[0100] The terminal layer includes distribution network field monitoring units and distribution network field control units, specifically comprising individual resources located at the very end of the power grid, such as distributed photovoltaics, electric vehicles, wind turbines, energy storage, and air conditioners. The terminal layer is regulated by the subgroup layer and also has a self-decision-making function. Self-decision-making instructions have lower priority than upper-layer instructions, achieving optimal response.

[0101] In one embodiment of this disclosure, the condition of satisfying the first resource regulation condition includes: if a user demand instruction is received, then it is determined that the first resource regulation condition is satisfied; or, if the distributed energy supply within the group cannot meet the load requirements, then it is determined that the first resource regulation condition is satisfied; and / or, the condition of satisfying the second resource regulation condition includes: if the distributed energy supply within the subgroup cannot meet the load requirements, then it is determined that the second resource regulation condition is satisfied.

[0102] In this disclosed method, when any of the above-mentioned first resource control conditions are met, the group layer is triggered to select the most suitable subgroup from the subgroup layer and send a first control instruction. After receiving the first control instruction, the most suitable subgroup is triggered to select the most suitable terminal from the terminal layer and send a second control instruction. Alternatively, when the second resource control conditions are met, the most suitable subgroup can also be triggered to select the most suitable terminal from the terminal layer and send a second control instruction. In this case, the subgroup layer does not make decisions based on the instructions from the group layer, but rather independently decides whether to issue a control instruction based on the distributed resources and load conditions within the subgroup.

[0103] In this disclosure, user demand instructions can be demand response instructions, such as responding to instructions from the power grid company to meet peak shaving and valley filling demands, or emergency control instructions, such as instructions to increase the grid frequency, etc. Situations where distributed energy supply within a group cannot meet load requirements and / or where distributed energy supply within a subgroup cannot meet load requirements can be events such as unexpected load surges, events where special weather causes a sudden drop in renewable energy output, or events where equipment failure leads to power shortages, etc.

[0104] In this disclosed method, when the group layer and subgroup layer make decisions, they select the most suitable subgroup / most suitable terminal to execute the control command. The number of the most suitable subgroup is usually one, thus achieving single-objective focus, avoiding the complexity of multi-resource coordination, and improving response speed. The number of the most suitable terminal can be one, two, or more, thus facilitating the leveraging of complementary advantages of different resources. Unselected subgroups / terminals can maintain their original operating state and do not participate in the current control cycle, achieving resource isolation.

[0105] In this disclosed method, the first control command issued from the group level to the most suitable subgroup can carry instructions with global control objectives, such as power adjustment amounts; the second control command issued from the most suitable subgroup to the most suitable terminal can carry instructions with specific equipment operation instructions, such as energy storage charging and discharging power settings, photovoltaic load reduction ratios, etc. After receiving the second control command, the most suitable terminal performs resource scheduling, specifically by parsing the second command and generating control equipment parameters, such as inverter power and energy storage SOC threshold. The most suitable terminal also monitors the execution results in real time, provides feedback on deviations, and dynamically adjusts parameters to optimize resource scheduling results.

[0106] In one embodiment of this disclosure, the group layer selects the best-fit subgroup from its subordinate subgroup layers and sends a first control command to it, including:

[0107] The group layer obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers;

[0108] First decision information is obtained by extracting the first feature and configuring the first weight on the resource information of the subgroup nodes.

[0109] Based on the first decision information, the most suitable subgroup is selected and a first control instruction is sent to it;

[0110] The subgroup node resource information is obtained by integrating the terminal layer resource information uploaded by the terminal layer under the subgroup layer; the terminal layer resource information includes the operation information and resource supply status information of each terminal device.

[0111] In this disclosed method, terminal devices collect and perceive information data through methods such as sensor information acquisition and image recognition, thereby enhancing communication and collaboration capabilities between devices. The operational information of the terminal devices can be power system operational information such as current, voltage, and frequency. Resource supply status information can be decision-making information based on their own reasoning, including information on whether there is surplus power supply or whether additional power supply is needed. The operational information of each terminal device and the corresponding resource supply status information constitute the terminal layer resource information. The terminal layer aggregates the terminal layer resource information and uploads it to the subgroup layer, storing it in the subgroup layer's terminal layer resource information database. The subgroup layer extracts features and integrates resources from the aggregated terminal layer resource information to form subgroup node resource information, which is then uploaded to the group layer and stored in the group layer's subgroup node resource information database.

[0112] In this disclosed method, the resource information of the subgroup nodes is first extracted using a first feature extraction technique, such as principal component analysis (PCA) or long short-term memory network (LSTM), to extract key features from the original data, thereby reducing data dimensionality and retaining important information.

[0113] After the first feature is extracted, the first weight can be configured based on the current control objective. The first weight can be adjusted. For example, in the control scenario of emergency frequency regulation, the response speed feature can be given a high weight, while in the economic scheduling scenario, the control cost feature can be given a high weight. Then, the feature values ​​of each feature can be weighted and summed to calculate the subgroup fit score. The subgroup fit score is used as the first decision information, and the best-fitting subgroup is selected based on the score value. Then, the first control command is sent to the best-fitting subgroup.

[0114] In one embodiment of this disclosure, obtaining first decision information by performing first feature extraction and first weight configuration on the resource information of the subgroup nodes includes:

[0115] The first input vector is obtained by extracting the first feature from the resource information of the subgroup nodes;

[0116] The first input vector is input into the first feature attention mechanism to configure the first sub-weights, and the first feature vector is obtained.

[0117] The first feature vector is input into the temporal attention mechanism to configure the second sub-weights, resulting in the second feature vector;

[0118] The first decision information is obtained based on the second feature vector.

[0119] In this disclosure, to highlight the more critical influencing features of the subgroup node resource information, a first feature attention mechanism is introduced, enabling the model to adaptively weight the input features during intelligent regulation of the group layer. The model in this disclosure can be a reinforcement learning strategy model with an embedded attention layer, as described in the prior art, and will not be elaborated upon here.

[0120] Specifically, the first weight is calculated first. Since two attention mechanisms are introduced, in this implementation, the first weight includes a first sub-weight related to the first feature attention mechanism and a second sub-weight related to the temporal attention mechanism: the feature after first feature extraction is set as... As the first input vector of the model, N is the number of samples, T is the number of subgroups, and n f Let e ​​be the number of terminal devices. The quantization process for the weights corresponding to each feature is: e = σ(XW) e +b e e = (e1, e2, ..., e T ) represents the first unnormalized sub-weight. The coefficient matrix is ​​trainable. Here, σ is the bias parameter, and σ is the Sigmoid activation function. Next, the first sub-weights are normalized: to ensure the first sub-weights satisfy a probability distribution with a sum of 1, e is normalized using the softmax function to obtain the normalized first sub-weights α. j Recalculate the intermediate semantic vector: Calculate the normalized first sub-weight α. j With the corresponding eigenvector x j Multiplication can enhance or weaken x. j The purpose of this expression is to obtain the first feature vector for adaptive optimization: X ATT =(α1x1,α2x2,...,α T x T ).

[0121] Since different types of distributed energy and load absorption vary greatly in terms of timing, a timing attention mechanism is further introduced into the group layer.

[0122] Specifically, firstly, the first feature vector X containing the feature association relationship is used. ATT As input to the temporal attention mechanism, the weights at time t during the iteration process are quantized: f t =σ([X ATT W f +b f ), where f t =(f t 1 ,f t 2,...,f t T ) represents the unnormalized second sub-weight. The coefficient matrix is ​​trainable. These are the bias parameters. Next, the second sub-weights are normalized: the softmax function is used to adjust f. t Normalization is performed to obtain the normalized second sub-weights. To obtain the implicit time information at time t again, we will... With α t x t We perform a weighted summation to obtain the second eigenvector. It contains features and temporal correlation information.

[0123] In this disclosure, the first decision information is the subgroup adaptation score obtained by weighted summation of the feature values ​​of the features configured with the first weight and the corresponding first weight;

[0124] The step of selecting the most suitable subgroup based on the first decision information and sending a first control command to it includes:

[0125] The calculated subgroup fit scores are compared, and the subgroup corresponding to the highest subgroup fit score is taken as the best-fit subgroup and a first control command is sent to it.

[0126] In one embodiment of this disclosure, selecting the most suitable terminal from its subordinate terminal layer and sending a second control command to it includes:

[0127] The second decision information is obtained by performing second feature extraction and second weight configuration on the terminal layer resource information;

[0128] Based on the second decision information, the most suitable terminal is selected and a second control command is sent to it.

[0129] In this disclosure, the step of obtaining the second decision information by performing second feature extraction and second weight configuration on the terminal layer resource information includes:

[0130] The second input vector is obtained by performing a second feature extraction on the terminal layer resource information;

[0131] The second input vector is input into the second feature attention mechanism to configure the second weights, resulting in the third feature vector;

[0132] The second decision information is obtained based on the third feature vector.

[0133] The processes of second feature extraction and second weight configuration in this disclosure can refer to the processes of first feature extraction and first weight configuration in the above scheme, and will not be repeated here.

[0134] In this disclosure, the second decision information is the terminal adaptation score obtained by weighted summation of the feature values ​​of the features configured with the second weight and the corresponding second weight;

[0135] The step of selecting the most suitable terminal based on the second decision information and sending a second control command to it includes:

[0136] The calculated terminal adaptation scores are compared, and the terminal with the highest terminal adaptation score is selected as the most adapted terminal and a second control command is sent to it.

[0137] In one embodiment of this disclosure, it further includes:

[0138] The second input vector is uploaded to the group layer as subgroup node resource information and stored in the subgroup node resource information database.

[0139] The group layer obtains the subgroup node resource information uploaded by its subordinate subgroup layers, including:

[0140] Obtain the subgroup node resource information uploaded by its subordinate subgroup layers from the subgroup node resource information database.

[0141] In this disclosed method, the second input vector is useful data for subgroup decision-making obtained by feature extraction of terminal layer resource information. This useful data is then uploaded to the group layer as subgroup node resource information, enabling the group layer to make further decisions based on the useful data, thereby improving decision-making efficiency.

[0142] In one embodiment of this disclosure, it further includes:

[0143] If the distributed energy supply within the terminal layer cannot meet the load requirements, the terminal layer will perform resource scheduling on its subordinate terminal equipment.

[0144] In this disclosed method, the inability of distributed energy supply within the terminal layer to meet load requirements can be due to events such as unexpected load surges, sudden drops in renewable energy output caused by severe weather, or power shortages caused by equipment failures. The terminal layer has a self-decision-making function; in the event of the above situations, it directly schedules resources for the corresponding terminal equipment based on local control strategies, enabling better adaptation to local dynamic changes and improved response speed.

[0145] Figure 2 A flowchart illustrating a distributed energy aggregation and management method according to a specific embodiment of the present disclosure is shown.

[0146] like Figure 2 As shown, the specific process of the distributed energy aggregation and management method is as follows:

[0147] Step 1: Each terminal device 1 through n in the terminal layer uploads system operation information and its own inferred terminal layer self-decision information (i.e., resource supply status information) to its respective subgroup in the subgroup layer. Each subgroup aggregates the uploaded information and stores it in the terminal layer resource information database. The subgroup layer performs a second feature extraction on the information in the terminal layer resource information database to obtain the subgroup node resource information, and then uploads it to the group layer. The group layer aggregates the uploaded information and stores it in the subgroup node resource information database.

[0148] Step 2: After analyzing the resource information of the subgroup nodes, the group layer decides whether to issue the first control command based on the distributed resources and load within the group. If not, it waits for the subgroup node resource information to iterate; if so, it extracts the first feature from the subgroup node resource information. The extracted information is then used for group-level control decision analysis through a first feature attention mechanism and a time-series attention mechanism to select the most suitable subgroup and send the first control command. Furthermore, if a user demand command is received, group-level control decision analysis is performed directly to select the most suitable subgroup and send the first control command.

[0149] Step 3: After receiving the first control instruction, the subgroup layer performs second feature extraction on the terminal layer resource information. The extracted information is then processed through the second feature attention mechanism for subgroup layer control decision analysis. The most suitable terminal device is selected and the second control instruction is sent (not shown in the figure). After receiving the second control instruction, the execution unit of the terminal device performs resource scheduling.

[0150] Step 4: The subgroup layer can also independently analyze the resource information of all terminal layers within its region. Based on the distributed resources and load within the subgroup, it decides whether to issue a second control command. If not, it waits for the terminal layer resource information to iterate; if so, it extracts the second feature from the terminal layer resource information. The extracted information is then used for subgroup layer control decision analysis through a second feature attention mechanism to select the most suitable terminal device and send the second control command. The execution unit of the terminal device performs resource scheduling upon receiving the second control command.

[0151] Figure 3 A structural block diagram of a distributed energy aggregation and management device according to an embodiment of the present disclosure is shown. This device can be implemented as part or all of an electronic device through software, hardware, or a combination of both.

[0152] like Figure 3 As shown, the distributed energy aggregation and management device 300 includes:

[0153] The first control module 310 is configured to select the most suitable subgroup from its subordinate subgroup layers and send the first control instruction when the first resource control condition is met.

[0154] The second control module 320 is configured to select the most suitable terminal from its subordinate terminal layer and send the second control instruction after the most suitable subgroup receives the first control instruction or when the second resource control condition is met.

[0155] Resource scheduling module 330 is configured to perform resource scheduling after the terminal receives the second control instruction.

[0156] The group layer is configured with a group control center to realize collective decision-making at the regional distribution network level. The subgroup layer is composed of edge decision-making devices to decompose the control tasks of the group layer. The terminal layer includes a distribution network field monitoring unit and a distribution network field control unit.

[0157] According to the technical solution provided in this disclosure, by dynamically selecting the "most suitable subgroup" and the "most suitable terminal" for hierarchical and progressive resource regulation, the global problem of distributed resource scheduling is decomposed into local optimization tasks, which significantly reduces the computational complexity of a single decision and shortens the response time. While ensuring real-time performance, it achieves efficient and flexible global-local collaborative optimization, providing effective technical support for the stable operation of energy systems with a high proportion of renewable energy access.

[0158] In one embodiment of this disclosure, the condition of satisfying the first resource regulation condition includes: if a user demand instruction is received, then it is determined that the first resource regulation condition is satisfied; or, if the distributed energy supply within the group cannot meet the load requirements, then it is determined that the first resource regulation condition is satisfied; and / or,

[0159] The conditions under which the second resource regulation condition is met include: if the distributed energy supply within the subgroup cannot meet the load requirements, then the second resource regulation condition is determined to be met.

[0160] In one embodiment of this disclosure, the part of the first control module that selects the most suitable subgroup from its subordinate subgroup layers and sends a first control command to it is configured as follows:

[0161] The group layer obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers;

[0162] First decision information is obtained by extracting the first feature and configuring the first weight on the resource information of the subgroup nodes.

[0163] Based on the first decision information, the most suitable subgroup is selected and a first control instruction is sent to it;

[0164] The subgroup node resource information is obtained by integrating the terminal layer resource information uploaded by the terminal layer under the subgroup layer; the terminal layer resource information includes the operation information and resource supply status information of each terminal device.

[0165] In one embodiment of this disclosure, the portion of the first control module that obtains the first decision information after performing first feature extraction and first weight configuration on the resource information of the subgroup nodes is configured as follows:

[0166] The first input vector is obtained by extracting the first feature from the resource information of the subgroup nodes;

[0167] The first input vector is input into the first feature attention mechanism to configure the first sub-weights, and the first feature vector is obtained.

[0168] The first feature vector is input into the temporal attention mechanism to configure the second sub-weights, resulting in the second feature vector;

[0169] The first decision information is obtained based on the second feature vector.

[0170] In one embodiment of this disclosure, the part of the second control module that selects the most suitable terminal from its subordinate terminal layer and sends a second control command to it is configured as follows:

[0171] The second decision information is obtained by performing second feature extraction and second weight configuration on the terminal layer resource information;

[0172] Based on the second decision information, the most suitable terminal is selected and a second control command is sent to it.

[0173] In one embodiment of this disclosure, the portion of the second decision information obtained by the second control module after performing second feature extraction and second weight configuration on the terminal layer resource information is configured as follows:

[0174] The second input vector is obtained by performing a second feature extraction on the terminal layer resource information;

[0175] The second input vector is input into the second feature attention mechanism to configure the second weights, resulting in the third feature vector;

[0176] The second decision information is obtained based on the third feature vector.

[0177] In one embodiment of this disclosure, it further includes:

[0178] The storage module is configured to upload the second input vector as subgroup node resource information to the group layer and store it in the subgroup node resource information database;

[0179] The part of the first control module that obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers from the group layer is configured as follows:

[0180] Obtain the subgroup node resource information uploaded by its subordinate subgroup layers from the subgroup node resource information database.

[0181] In one embodiment of this disclosure, the first decision information is a subgroup adaptation score calculated by weighted summation of the feature values ​​of the features configured with the first weight and the corresponding first weight.

[0182] The part of the first control module that selects the most suitable subgroup based on the first decision information and sends the first control command to it is configured as follows:

[0183] The calculated subgroup fit scores are compared, and the subgroup corresponding to the highest subgroup fit score is taken as the best-fit subgroup and a first control command is sent to it.

[0184] In one embodiment of this disclosure, the second decision information is a terminal adaptation score calculated by weighted summation of the feature values ​​of the features configured with the second weight and the corresponding second weight;

[0185] The part of the second control module that selects the most suitable terminal based on the second decision information and sends the second control command to it is configured as follows:

[0186] The calculated terminal adaptation scores are compared, and the terminal with the highest terminal adaptation score is selected as the most adapted terminal and a second control command is sent to it.

[0187] In one embodiment of this disclosure, it further includes:

[0188] The terminal layer scheduling module is configured to perform resource scheduling on its subordinate terminal devices if the distributed energy supply within the terminal layer cannot meet the load requirements.

[0189] This disclosure also discloses an electronic device. Figure 4 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0190] like Figure 4 As shown, the electronic device includes a memory and a processor, wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method according to embodiments of the present disclosure.

[0191] The distributed energy aggregation and control method is applicable to a distributed energy aggregation and control system, wherein the distributed energy aggregation and control system includes: a group layer, a subgroup layer, and a terminal layer. The group layer is configured with a group control center for collective decision-making at the regional distribution network level. The subgroup layer consists of edge decision-making devices for decomposing the control tasks of the group layer. The terminal layer includes a distribution network field monitoring unit and a distribution network field control unit. The method includes:

[0192] When the first resource regulation condition is met, the group layer selects the best-fit subgroup from its subordinate subgroup layers and sends the first regulation instruction to it;

[0193] After receiving the first control instruction or when the second resource control condition is met, the most suitable subgroup selects the most suitable terminal from its subordinate terminal layer and sends the second control instruction to it.

[0194] The most suitable terminal performs resource scheduling based on the received second control instruction.

[0195] In one embodiment of this disclosure, the condition of satisfying the first resource regulation condition includes: if a user demand instruction is received, then it is determined that the first resource regulation condition is satisfied; or, if the distributed energy supply within the group cannot meet the load requirements, then it is determined that the first resource regulation condition is satisfied; and / or,

[0196] The conditions under which the second resource regulation condition is met include: if the distributed energy supply within the subgroup cannot meet the load requirements, then the second resource regulation condition is determined to be met.

[0197] In one embodiment of this disclosure, the group layer selects the best-fit subgroup from its subordinate subgroup layers and sends a first control command to it, including:

[0198] The group layer obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers;

[0199] First decision information is obtained by extracting the first feature and configuring the first weight on the resource information of the subgroup nodes.

[0200] Based on the first decision information, the most suitable subgroup is selected and a first control instruction is sent to it;

[0201] The subgroup node resource information is obtained by integrating the terminal layer resource information uploaded by the terminal layer under the subgroup layer; the terminal layer resource information includes the operation information and resource supply status information of each terminal device.

[0202] In one embodiment of this disclosure, obtaining first decision information by performing first feature extraction and first weight configuration on the resource information of the subgroup nodes includes:

[0203] The first input vector is obtained by extracting the first feature from the resource information of the subgroup nodes;

[0204] The first input vector is input into the first feature attention mechanism to configure the first sub-weights, and the first feature vector is obtained.

[0205] The first feature vector is input into the temporal attention mechanism to configure the second sub-weights, resulting in the second feature vector;

[0206] The first decision information is obtained based on the second feature vector.

[0207] In one embodiment of this disclosure, selecting the most suitable terminal from its subordinate terminal layer and sending a second control command to it includes:

[0208] The second decision information is obtained by performing second feature extraction and second weight configuration on the terminal layer resource information;

[0209] Based on the second decision information, the most suitable terminal is selected and a second control command is sent to it.

[0210] In one embodiment of this disclosure, the step of obtaining second decision information by performing second feature extraction and second weight configuration on the terminal layer resource information includes:

[0211] The second input vector is obtained by performing a second feature extraction on the terminal layer resource information;

[0212] The second input vector is input into the second feature attention mechanism to configure the second weights, resulting in the third feature vector;

[0213] The second decision information is obtained based on the third feature vector.

[0214] In one embodiment of this disclosure, it further includes:

[0215] The second input vector is uploaded to the group layer as subgroup node resource information and stored in the subgroup node resource information database.

[0216] The group layer obtains the subgroup node resource information uploaded by its subordinate subgroup layers, including:

[0217] Obtain the subgroup node resource information uploaded by its subordinate subgroup layers from the subgroup node resource information database.

[0218] In one embodiment of this disclosure, the first decision information is a subgroup adaptation score calculated by weighted summation of the feature values ​​of the features configured with the first weight and the corresponding first weight.

[0219] The step of selecting the most suitable subgroup based on the first decision information and sending a first control command to it includes:

[0220] The calculated subgroup fit scores are compared, and the subgroup corresponding to the highest subgroup fit score is taken as the best-fit subgroup and a first control command is sent to it.

[0221] In one embodiment of this disclosure, the second decision information is a terminal adaptation score calculated by weighted summation of the feature values ​​of the features configured with the second weight and the corresponding second weight;

[0222] The step of selecting the most suitable terminal based on the second decision information and sending a second control command to it includes:

[0223] The calculated terminal adaptation scores are compared, and the terminal with the highest terminal adaptation score is selected as the most adapted terminal and a second control command is sent to it.

[0224] In one embodiment of this disclosure, it further includes:

[0225] If the distributed energy supply within the terminal layer cannot meet the load requirements, the terminal layer will perform resource scheduling on its subordinate terminal equipment.

[0226] Figure 5 A schematic diagram of the structure of a computer system suitable for implementing the method according to embodiments of the present disclosure is shown.

[0227] like Figure 5 As shown, the computer system includes a processing unit that can execute various methods described above based on a program stored in a read-only memory (ROM) or a program loaded from a storage portion into a random access memory (RAM). The RAM also stores various programs and data required for the operation of the computer system. The processing unit, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0228] The following components are connected to the I / O interface: input sections including keyboards, mice, etc.; output sections including cathode ray tubes (CRTs), liquid crystal displays (LCDs), and speakers; storage sections including hard disks, etc.; and communication sections including network interface cards such as LAN cards and modems. The communication section performs communication processes via a network such as the Internet. Drives are also connected to the I / O interface as needed. Removable media, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on the drive as needed so that computer programs read from them can be installed into the storage section as needed. The processing unit can be implemented as a CPU, GPU, TPU, FPGA, NPU, etc.

[0229] In particular, according to embodiments of this disclosure, the methods described above can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program containing program code for performing the methods described above. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium.

[0230] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0231] The units or modules described in the embodiments of this disclosure can be implemented in software or programmable hardware. The described units or modules can also be located in a processor, and the names of these units or modules do not necessarily constitute a limitation on the unit or module itself.

[0232] In another aspect, this disclosure also provides a chip including at least one processor that can be used to implement the methods involved in the above system embodiments.

[0233] In one possible design, the chip also includes a memory for storing program instructions and data, which may be located inside or outside the processor.

[0234] In another aspect, this disclosure also provides a computer-readable storage medium, which may be a computer-readable storage medium included in the electronic device or computer system described above; or it may be a standalone computer-readable storage medium not assembled into a device. The computer-readable storage medium stores one or more programs, which are used by one or more processors to perform the methods described in this disclosure.

[0235] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

Claims

1. A distributed energy aggregation and management method, applicable to a distributed energy aggregation and management system, characterized in that, The distributed energy aggregation and management system includes: a group layer, a subgroup layer, and a terminal layer. The group layer is configured with a group control center for collective decision-making at the regional distribution network level. The subgroup layer consists of edge decision-making devices for decomposing the control tasks of the group layer. The terminal layer includes a distribution network field monitoring unit and a distribution network field control unit. The method includes: When the first resource regulation condition is met, the group layer selects the most suitable subgroup from its subordinate subgroup layers and sends a first regulation instruction to it, including: the group layer obtaining subgroup node resource information uploaded by its subordinate subgroup layers; performing first feature extraction and first weight configuration on the subgroup node resource information to obtain first decision information; selecting the most suitable subgroup based on the first decision information and sending a first regulation instruction to it; wherein, the subgroup node resource information is obtained by integrating terminal layer resource information uploaded by the terminal layers under the subgroup layer; the terminal layer resource information includes the operating information and resource supply status information of each terminal device; After receiving the first control instruction or when the second resource control condition is met, the most suitable subgroup selects the most suitable terminal from its subordinate terminal layer and sends the second control instruction to it, including: performing second feature extraction and second weight configuration on the terminal layer resource information to obtain second decision information; selecting the most suitable terminal based on the second decision information and sending the second control instruction to it; The most suitable terminal performs resource scheduling based on the received second control instruction.

2. The distributed energy aggregation and management method according to claim 1, characterized in that, The conditions under which the first resource regulation condition is met include: if a user demand instruction is received, then it is determined that the first resource regulation condition is met; or, if the distributed energy supply within the group cannot meet the load requirements, then it is determined that the first resource regulation condition is met; and / or, The conditions under which the second resource regulation condition is met include: if the distributed energy supply within the subgroup cannot meet the load requirements, then the second resource regulation condition is determined to be met.

3. The distributed energy aggregation and management method according to claim 1, characterized in that, The first decision information obtained after extracting the first feature and configuring the first weight of the subgroup node resource information includes: The first input vector is obtained by extracting the first feature from the resource information of the subgroup nodes; The first input vector is input into the first feature attention mechanism to configure the first sub-weights, and the first feature vector is obtained. The first feature vector is input into the temporal attention mechanism to configure the second sub-weights, resulting in the second feature vector; The first decision information is obtained based on the second feature vector.

4. The distributed energy aggregation and management method according to claim 1, characterized in that, The second decision information obtained after performing second feature extraction and second weight configuration on the terminal layer resource information includes: The second input vector is obtained by performing a second feature extraction on the terminal layer resource information; The second input vector is input into the second feature attention mechanism to configure the second weights, resulting in the third feature vector; The second decision information is obtained based on the third feature vector.

5. The distributed energy aggregation and management method according to claim 4, characterized in that, Also includes: The second input vector is uploaded to the group layer as subgroup node resource information and stored in the subgroup node resource information database. The group layer obtains the subgroup node resource information uploaded by its subordinate subgroup layers, including: Obtain the subgroup node resource information uploaded by its subordinate subgroup layers from the subgroup node resource information database.

6. The distributed energy aggregation and management method according to claim 1, characterized in that, The first decision information is the subgroup adaptation score calculated by weighted summation of the feature values ​​of the features configured with the first weight and the corresponding first weight; The step of selecting the most suitable subgroup based on the first decision information and sending a first control command to it includes: The calculated subgroup fit scores are compared, and the subgroup corresponding to the highest subgroup fit score is taken as the best-fit subgroup and a first control command is sent to it.

7. The distributed energy aggregation and management method according to claim 1, characterized in that, The second decision information is the terminal adaptation score calculated by weighted summation of the feature values ​​of the features configured with the second weight and the corresponding second weight; The step of selecting the most suitable terminal based on the second decision information and sending a second control command to it includes: The calculated terminal adaptation scores are compared, and the terminal with the highest terminal adaptation score is selected as the most adapted terminal and a second control command is sent to it.

8. The distributed energy aggregation and management method according to claim 1, characterized in that, Also includes: If the distributed energy supply within the terminal layer cannot meet the load requirements, the terminal layer will perform resource scheduling on its subordinate terminal equipment.

9. A distributed energy aggregation and management device, characterized in that, include: The first control module is configured to select the most suitable subgroup from its subordinate subgroup layers and send the first control instruction to it when the first resource control condition is met. The second control module is configured to select the most suitable terminal from its subordinate terminal layer and send the second control instruction to the most suitable subgroup after it receives the first control instruction or when the second resource control conditions are met. The resource scheduling module is configured to best suit the terminal to perform resource scheduling according to the received second control instruction; The group layer is configured with a group control center to realize collective decision-making at the regional distribution network level. The subgroup layer is composed of edge decision-making devices to decompose the control tasks of the group layer. The terminal layer includes a distribution network field monitoring unit and a distribution network field control unit. The part of the first control module that selects the most suitable subgroup from its subordinate subgroup layers and sends the first control command to it is configured as follows: The group layer obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers; First decision information is obtained by extracting the first feature and configuring the first weight on the resource information of the subgroup nodes. Based on the first decision information, the most suitable subgroup is selected and a first control instruction is sent to it; The subgroup node resource information is obtained by integrating the terminal layer resource information uploaded by the terminal layer under the subgroup layer; the terminal layer resource information includes the operation information and resource supply status information of each terminal device. The part of the second control module that selects the most suitable terminal from its subordinate terminal layer and sends the second control command to it is configured as follows: The second decision information is obtained by performing second feature extraction and second weight configuration on the terminal layer resource information; Based on the second decision information, the most suitable terminal is selected and a second control command is sent to it.

10. The distributed energy aggregation and management device according to claim 9, characterized in that, The conditions under which the first resource regulation condition is met include: if a user demand instruction is received, then it is determined that the first resource regulation condition is met; or, if the distributed energy supply within the group cannot meet the load requirements, then it is determined that the first resource regulation condition is met; and / or, The conditions under which the second resource regulation condition is met include: if the distributed energy supply within the subgroup cannot meet the load requirements, then the second resource regulation condition is determined to be met.

11. The distributed energy aggregation and management device according to claim 9, characterized in that, The portion of the first decision information obtained by performing first feature extraction and first weight configuration on the resource information of the subgroup nodes in the first regulation module is configured as follows: The first input vector is obtained by extracting the first feature from the resource information of the subgroup nodes; The first input vector is input into the first feature attention mechanism to configure the first sub-weights, and the first feature vector is obtained. The first feature vector is input into the temporal attention mechanism to configure the second sub-weights, resulting in the second feature vector; The first decision information is obtained based on the second feature vector.

12. The distributed energy aggregation and management device according to claim 9, characterized in that, The portion of the second decision information obtained by the second control module after performing second feature extraction and second weight configuration on the terminal layer resource information is configured as follows: The second input vector is obtained by performing a second feature extraction on the terminal layer resource information; The second input vector is input into the second feature attention mechanism to configure the second weights, resulting in the third feature vector; The second decision information is obtained based on the third feature vector.

13. The distributed energy aggregation and management device according to claim 12, characterized in that, Also includes: The storage module is configured to upload the second input vector as subgroup node resource information to the group layer and store it in the subgroup node resource information database; The part of the first control module that obtains resource information of subgroup nodes uploaded by its subordinate subgroup layers from the group layer is configured as follows: Obtain the subgroup node resource information uploaded by its subordinate subgroup layers from the subgroup node resource information database.

14. The distributed energy aggregation and management device according to claim 9, characterized in that, The first decision information is the subgroup adaptation score calculated by weighted summation of the feature values ​​of the features configured with the first weight and the corresponding first weight; The part of the first control module that selects the most suitable subgroup based on the first decision information and sends the first control command to it is configured as follows: The calculated subgroup fit scores are compared, and the subgroup corresponding to the highest subgroup fit score is taken as the best-fit subgroup and a first control command is sent to it.

15. The distributed energy aggregation and management device according to claim 9, characterized in that, The second decision information is the terminal adaptation score calculated by weighted summation of the feature values ​​of the features configured with the second weight and the corresponding second weight; The part of the second control module that selects the most suitable terminal based on the second decision information and sends the second control command to it is configured as follows: The calculated terminal adaptation scores are compared, and the terminal with the highest terminal adaptation score is selected as the most adapted terminal and a second control command is sent to it.

16. The distributed energy aggregation and management device according to claim 9, characterized in that, Also includes: The terminal layer scheduling module is configured to perform resource scheduling on its subordinate terminal devices if the distributed energy supply within the terminal layer cannot meet the load requirements.

17. An electronic device, characterized in that, The method includes a memory and a processor; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method according to any one of claims 1-8.

18. A computer-readable storage medium storing computer instructions thereon, characterized in that, When executed by a processor, the computer instructions implement the method described in any one of claims 1-8.

19. A chip, characterized in that, include: At least one processor is configured to implement the method as described in any one of claims 1-8.

Images