Power grid dispatching model construction method, power grid task generation method and related devices

Abstract

The invention discloses a power grid dispatching model construction method, a power grid task generation method and a related device, the power grid dispatching model construction method comprises the following steps: obtaining an equipment node relation and dispatching data of a power grid, the dispatching data comprising equipment parameters, line parameters and actual task planning data; an adjacency matrix is constructed based on the equipment node relation, the adjacency matrix comprises information of nodes and edges, the nodes are power grid equipment, the edges are the connection relation between the power grid equipment, and a graph neural network model suitable for power grid task generation is constructed; and training the graph neural network model based on the adjacency matrix and the scheduling data to obtain a power grid scheduling model. The power grid dispatching model obtained through training can perform dispatching planning based on the connection relation between the power grid devices, the device parameters and the line parameters to obtain dispatching task planning data, and the actual task planning data can serve as the inspection standard of the model quality, so that the automatically generated task planning data can meet the actual requirements of the power grid. And the feasibility is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid dispatching, and in particular to a power grid dispatching model construction method, a power grid task generation method and related devices. Background Art

[0002] With the rapid development of technology in the power industry and the increasing needs of users, the importance of dispatching automation systems in power grid operation and control has become increasingly prominent. The master station system can centrally manage and control various equipment and resources on the power grid, which plays a very important role in the efficient, safe and reliable operation of the power grid and in ensuring the power supply of users. However, at the same time, with the increasing complexity and intelligence of the master station system, the risks it faces are also increasing. For example, system attacks, malware, power information security and other issues may have a serious impact on the power industry. Therefore, power grid dispatching brings a heavy workload to the dispatchers of the power grid and has low work efficiency. Summary of the invention

[0003] In order to solve the above problems, the present invention provides a power grid dispatching model construction method, a power grid task generation method and related devices.

[0004] In a first aspect, the present invention provides a method for constructing a power grid dispatching model, comprising:

[0005] Obtaining equipment node relationships and dispatching data of the power grid, wherein the dispatching data includes equipment parameters, line parameters, and actual task planning data;

[0006] Constructing an adjacency matrix based on the device node relationship, the adjacency matrix including node and edge information, the nodes are power grid devices, and the edges are connection relationships between power grid devices;

[0007] Build a graph neural network model suitable for power grid task generation;

[0008] The graph neural network model is trained based on the adjacency matrix and the scheduling data to obtain a power grid scheduling model.

[0009] In a second aspect, the present invention provides a method for generating a power grid task, comprising:

[0010] Obtain real-time grid status data;

[0011] The power grid state data is input into a trained power grid dispatching model to obtain task planning data. The power grid dispatching model is trained using the power grid dispatching model construction method described in the first aspect.

[0012] In a third aspect, the present invention provides a power grid dispatching model construction device, comprising:

[0013] A data acquisition module, used to acquire equipment node relationships and dispatching data of the power grid, wherein the dispatching data includes equipment parameters, line parameters and actual task planning data;

[0014] An adjacency matrix construction module, used to construct an adjacency matrix based on the device node relationship, the adjacency matrix including node and edge information, the node being a power grid device, and the edge being a connection relationship between power grid devices;

[0015] An initial model building module, used to build a graph neural network model suitable for power grid task generation;

[0016] A model training module is used to train the graph neural network model based on the adjacency matrix and the scheduling data to obtain a power grid scheduling model.

[0017] In a fourth aspect, the present invention provides a power grid task generation device, comprising:

[0018] A power grid status data acquisition module, used to acquire real-time power grid status data;

[0019] The task planning data generation module is used to input the power grid state data into the trained power grid dispatching model to obtain the task planning data.

[0020] The power grid dispatching model is trained using the power grid dispatching model construction method described in the first aspect.

[0021] In a fifth aspect, the present invention provides an electronic device, the electronic device comprising:

[0022] at least one processor; and

[0023] a memory communicatively connected to the at least one processor; wherein,

[0024] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the power grid dispatching model construction method described in the first aspect of the present invention, and / or the power grid task generation method described in the second aspect.

[0025] In a sixth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the power grid dispatching model construction method described in the first aspect of the present invention, and / or the power grid task generation method described in the second aspect when executed.

[0026] In the above-mentioned method for constructing a power grid dispatching model, the equipment node relationship and dispatching data of the power grid are obtained, and the dispatching data includes equipment parameters, line parameters and actual task planning data; an adjacency matrix is ​​constructed based on the equipment node relationship, and the adjacency matrix includes node and edge information, where the node is the power grid equipment and the edge is the connection relationship between the power grid equipment, and a graph neural network model suitable for power grid task generation is constructed; the graph neural network model is trained based on the adjacency matrix and dispatching data to obtain a power grid dispatching model. The trained power grid dispatching model can be used to perform dispatching planning based on the connection relationship, equipment parameters and line parameters between the power grid equipment to obtain dispatching task planning data. The actual task planning data can be used as a test standard for the quality of the model, so that the task planning data generated by the power grid dispatching model can meet the actual needs of the power grid, making it feasible to automatically generate task planning data.

[0027] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 is a flow chart of a method for constructing a power grid dispatching model provided in the first embodiment of the present invention;

[0030] Figure 2 It is a flow chart of a method for training a graph neural network model provided in Embodiment 1 of the present invention;

[0031] Figure 3 is a flow chart of a method for generating power grid tasks provided by Embodiment 2 of the present invention;

[0032] Figure 3A A flowchart of a power grid dispatching process provided in Embodiment 2 of the present invention;

[0033] Figure 4 It is a structural schematic diagram of a power grid dispatching model building device provided in Embodiment 3 of the present invention;

[0034] Figure 5 is a structural diagram of a power grid task generating device provided by a fourth embodiment of the present invention;

[0035] Figure 6It is a schematic diagram of the structure of an electronic device provided in Embodiment 5 of the present invention. DETAILED DESCRIPTION

[0036] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0037] The solution of the present invention is applied to Graph Neural Network (GNN), which is a type of deep learning model for processing graph structured data. Entities in graph structured data are represented in the form of nodes, and the relationships between entities are represented in the form of edges. The goal of GNN is to learn useful representations from graph structured data and use these representations for various tasks, such as node classification, graph classification, link prediction, etc.

[0038] The core idea of ​​GNN is to aggregate the neighbor information of nodes through a message-passing mechanism and update the feature representation of nodes. This process usually goes through multiple rounds of iterations to capture information at farther distances in the graph. Ultimately, the feature representation of each node will contain information about its neighbors and more distant nodes; the basic components of graph neural networks include:

[0039] Node feature matrix: used to represent the initial features of each node in the graph.

[0040] Adjacency matrix: used to represent the connection relationship between nodes in a graph.

[0041] Graph convolutional layer: used to aggregate the information of neighbor nodes and update node features.

[0042] Output layer: The output layer designed according to task requirements is used to output prediction results.

[0043] Embodiment 1

[0044] Figure 1 This is a flowchart of a method for constructing a power grid dispatch model provided in the first embodiment of the present invention. This embodiment is applicable to the case of constructing a power grid dispatch model. The method can be executed by a power grid dispatch model construction device. The power grid dispatch model construction device can be implemented in the form of hardware and / or software. The power grid dispatch model construction device can be configured in an electronic device. Figure 1 As shown, the power grid dispatching model construction method includes:

[0045] S11. Obtain equipment node relationships and dispatching data of the power grid, where the dispatching data includes equipment parameters, line parameters, and actual task planning data.

[0046] The device node relationship of the power grid mainly refers to devices with associated relationships, mainly referring to devices with power / signal transmission or relationships, such as power signal transmission equipment, power line transmitters, intelligent power transmitters, etc. Different devices are connected by wires and cables or wirelessly.

[0047] Equipment parameters are the information of equipment extracted from the equipment database, mainly including: equipment name, rated power, operating status (normal status or fault status) and load information. The load information includes real-time load data and load type (industrial type, commercial type or residential type).

[0048] Line parameters can indicate the operating status and load conditions of the line, and mainly include: line length, rated current, and impedance value.

[0049] The actual task planning data can be extracted from the dispatching command and control system and the distribution network management platform. The actual task planning data may include maintenance plans and operation tasks.

[0050] Maintenance plan: planned time, power outage equipment, and maintenance duration.

[0051] Operation tasks: specific operation steps, operators, safety measures, power supply transfer plan, reclosing switch operation plan, etc. These tasks include specific maintenance tasks, such as equipment maintenance, line inspection, etc.; operation steps, such as power off, power on, load switching, etc.; as well as safety measures and notes, to ensure that all operations comply with safety regulations and record the specific requirements and precautions of each step in detail.

[0052] Optionally, after obtaining the scheduling data, the extracted data may be cleaned and standardized to remove noise, process missing values, and construct a formatted training data set.

[0053] S12. Construct an adjacency matrix based on the device node relationship. The adjacency matrix includes node and edge information. The node is the power grid device, and the edge is the connection relationship between the power grid devices.

[0054] As mentioned in S11, there are associations between power grid devices, so an adjacency matrix can be constructed based on the device node relationship.

[0055] The adjacency matrix is ​​equivalent to the topological structure of the power grid. The logical structure of the adjacency matrix is ​​divided into two parts: V and E sets, where V is the vertex and E is the edge. Therefore, a one-dimensional array is used to store all the vertex data in the graph; a two-dimensional array is used to store the data of the relationship between vertices (edges or arcs). This two-dimensional array is called the adjacency matrix. The adjacency matrix is ​​a matrix that represents the adjacent relationship between vertices. Let G = (V, E) be a graph, where V = {v1, v2, ..., vn}. Using the adjacency matrix to represent the graph, it is easy to determine whether any two vertices in the graph are connected by an edge.

[0056] Specifically, the construction of the adjacency matrix based on the device node relationship includes: taking all power grid devices in the power grid as matrix nodes of the adjacency matrix; determining any two of the matrix nodes with a connection relationship based on the device node relationship; and using edges to connect any two of the matrix nodes with a connection relationship to obtain an adjacency matrix.

[0057] S13. Construct a graph neural network model suitable for power grid task generation.

[0058] The graph neural network model in this step is an initial, untrained power grid dispatching model.

[0059] S14. Train the graph neural network model based on the adjacency matrix and scheduling data to obtain a power grid scheduling model.

[0060] The adjacency matrix can represent the relationship between power grid equipment. The trained power grid dispatching model can perform dispatching planning based on the connection relationship, equipment parameters, and line parameters between power grid equipment to obtain dispatching task planning data. The actual task planning data can be used as a test standard for model quality, so that the task planning data generated by the power grid dispatching model can meet the actual needs of the power grid, making the automatic generation of task planning data feasible.

[0061] Figure 2 This is a flow chart of the method for training the graph neural network model, such as Figure 2 As shown, the graph neural network model includes a weight matrix, and the graph neural network model is trained based on the adjacency matrix and the scheduling data to obtain a power grid scheduling model, including the following steps:

[0062] S141. For each matrix node in the adjacency matrix, extract node features of the matrix node from corresponding device parameters and line parameters.

[0063] That is, feature vectors are extracted from device parameters and line parameters respectively, and the feature vectors of the two are used as node features of corresponding matrix nodes.

[0064] S142. Based on the weight matrix in the graph neural network model, the node features of adjacent matrix nodes are weighted and summed to obtain aggregate information.

[0065] This step is equivalent to performing an aggregation operation on the node features. For a node feature, through the aggregation operation, the node features related to it are summed to obtain new node information. The one-layer sum can include the features of adjacent nodes. Update iteration When this operation is performed multiple times, the node can include the information of its adjacent nodes, that is, second-order aggregation is achieved. Repeated operations can include more-order adjacent point information, efficiently aggregate the features of each node, and comprehensively consider the node devices in the adjacency matrix, that is, when performing task planning, the equipment parameters and line parameters of all power grid equipment can be comprehensively considered, making the task planning reasonable and feasible.

[0066] Specifically, the formula for the aggregation operation is: The information of these nodes is aggregated by taking the weighted sum of the feature vectors of adjacent nodes.

[0067] By introducing GNN and utilizing the power grid topology and real-time status data, fast and accurate operation task generation can be achieved, reducing the burden on dispatchers.

[0068] The weight matrix is ​​equivalent to the model parameters of the graph neural network model. The model parameters affect the output effect of the model. In this step, the weight matrix in the initial graph neural network model can be set in advance. In the subsequent process, the weight matrix may also be updated.

[0069] S143. Input the aggregated information into the graph neural network model to output the predicted task planning data.

[0070] That is, the aggregated information is used as the input of the graph neural network model, and the graph neural network model outputs the result, which is the prediction task planning data.

[0071] S144. Obtain the information difference between the predicted task planning data and the actual planning data.

[0072] The information difference can be the inverse of the data similarity between the predicted task planning data and the actual planning data, that is, the smaller the data similarity between the two, the larger the information difference, and the larger the similarity, the smaller the information difference.

[0073] Optionally, obtaining the information difference between the predicted task planning data and the actual planning data includes: obtaining the task steps and the number of task steps of the predicted task planning data and the actual planning data respectively; and calculating the information difference between the predicted task planning data and the actual planning data based on the similarity of the task steps and the similarity of the number of task steps. By limiting the size of the information difference, the model can automatically and efficiently generate power grid dispatching operation tasks, thereby improving the intelligence and accuracy of power grid management.

[0074] Optionally, the information difference is inversely proportional to the similarity of the task steps and the similarity of the number of task steps.

[0075] S145. Determine whether the graph neural network model is trained through information difference.

[0076] If yes, execute S146, if no, execute S147.

[0077] Specifically, a certain information difference threshold can be set. When the information difference is greater than the information difference threshold, it is considered that the output result of the graph neural network model is consistent with the actual result and achieves the desired effect, then the model training is considered to be completed; otherwise, the model training is considered to be incomplete.

[0078] S146. Use the trained graph neural network model as a power grid dispatching model.

[0079] When the graph neural network model training is completed, the trained graph neural network model is used as the power grid dispatching model.

[0080] S147. Update the weight matrix based on the information difference between the predicted task planning data and the actual planning data.

[0081] The weight matrix is ​​equivalent to the model parameters of the graph neural network model. The model parameters affect the output effect of the model. The weight matrix W is trainable. After training, it will adapt to the importance of different node features. Through the back propagation algorithm, the model can adjust these weights so that the final output can better reflect the actual state of the power grid.

[0082] When the training of the graph neural network model is not completed, it means that the model parameters affecting the output effect of the model have not met the requirements. At this time, the weight matrix of the graph neural network model needs to be updated.

[0083] After executing S147, the process returns to executing S142.

[0084] Optionally, after updating the weight matrix based on the information difference between the predicted task planning data and the actual planning data, the method further includes: iterating and updating the node features of the matrix nodes through an aggregation operation.

[0085] Specifically, the node features are updated using the following formula:

[0086]

[0087] in, Represents the node features (feature vector) of the jth node in the lth layer. The node features contain various attribute information of the node, such as device status, type, etc.

[0088] The adjacency matrix A represents the connection relationship between nodes in the graph.

[0089] N(i) represents the set of all nodes adjacent to node i, that is, it includes all nodes directly connected to node i. During the aggregation process, information about the adjacent nodes of node i is collected.

[0090] As described above, the aggregation operation can be calculated and updated multiple times. Therefore, after updating the model parameters, the node features of the matrix nodes can also be updated and iterated through the aggregation operation, which can maintain the independence and reusability of each matrix node and prevent the model from falling into the local optimum during the training process.

[0091] Embodiment 2

[0092] Figure 3 This is a flowchart of a power grid task generation method provided in the second embodiment of the present invention. This embodiment can be applied to the case where tasks are automatically generated when power grid equipment is dispatched. The method can be executed by a power grid task generation device. The power grid task generation device can be implemented in the form of hardware and / or software. The power grid task generation device can be configured in an electronic device. Figure 3 As shown, the power grid task generation method includes:

[0093] S31, obtaining real-time power grid status data;

[0094] S32: Input the power grid status data into a trained power grid dispatching model to obtain task planning data.

[0095] Optionally, the power grid task generation method also includes: obtaining a rule base for power grid dispatching tasks, the rule base consisting of standardized task planning data; determining whether there is task planning data in the rule base that is identical to the task planning data generated by the power grid dispatching model; if so, determining that the task planning data generated by the power grid dispatching model is the task data to be executed; if not, returning to the step of inputting the power grid state data into the trained power grid dispatching model to obtain the task planning data. By establishing a complete rule base and automatic comparison mechanism, the rationality and compliance of the generated tasks are ensured, and operational risks are reduced.

[0096] Among them, rule base construction: establish a set of rule bases containing standard operating procedures. The rule base consists of a set of standardized operating tasks, each of which includes detailed operating steps, execution conditions, safety measures and related notes. These standard operating tasks are based on the best practices and safety regulations for power grid management, and are formulated and regularly updated by power grid staff. After generating task planning data, the system compares these tasks with the standard tasks in the rule base. The purpose of the comparison is to check the similarity of the generated tasks with the standard tasks to ensure that they meet the predetermined safety and operating specifications.

[0097] The similarity calculation formula is as follows:

[0098]

[0099] Among them, Tgenerated: generated operation task. Tstandard: standard task in the rule base. CommonSteps: the number of common steps between the generated task and the standard task. TotalSteps: the total number of steps in the standard task. By calculating the similarity Sim(Tgenerated, Tstandard), the similarity between the generated task and the standard task can be quantified. The value of similarity is between 0 and 1, and the larger the value, the more similar the two tasks are.

[0100] Error correction suggestion generation: If the similarity is lower than the set threshold, the system generates corresponding error correction suggestions, including modification steps and supplementary information.

[0101] The above-mentioned power grid dispatching model is trained by the power grid dispatching model construction method in the first embodiment, and can automatically generate task planning data that meets the actual needs of the power grid. After the task planning data is generated, it can be sent to the terminal where the human is located to execute the task. The above functions can also be integrated into a software system. The system supports real-time data input, task generation and self-checking, and has a friendly user interface, which is convenient for dispatchers to operate.

[0102] Figure 3A is a flow chart of a power grid dispatching process, such as Figure 3A As shown, in combination with Embodiment 1 and Embodiment 2, the process of power grid dispatching is as follows:

[0103] S3A1, data extraction and preprocessing;

[0104] The data includes power grid equipment parameters, line parameters, etc.

[0105] S3A2, build graph structure;

[0106] That is, building the structure of the graph neural network model, including building the adjacency matrix.

[0107] S3A3, training graph neural network model;

[0108] S3A4, real-time data input and feature extraction;

[0109] S3A5, intelligent task generation algorithm: generate operation steps (task data);

[0110] Specifically including maintenance and operation steps.

[0111] S3A6. Self-checking and error correction suggestions.

[0112] That is, task comparison and feedback.

[0113] S3A7, task execution and feedback.

[0114] Embodiment 3

[0115] Figure 4 This is a schematic diagram of the structure of a power grid dispatching model building device provided in the third embodiment of the present invention. Figure 4 As shown, the power grid dispatching model building device includes:

[0116] The data acquisition module 401 is used to acquire the equipment node relationship and scheduling data of the power grid, wherein the scheduling data includes equipment parameters, line parameters and actual task planning data;

[0117] An adjacency matrix construction module 402 is used to construct an adjacency matrix based on the device node relationship, wherein the adjacency matrix includes node and edge information, wherein the node is a power grid device and the edge is a connection relationship between power grid devices;

[0118] An initial model building module 403 is used to build a graph neural network model suitable for power grid task generation;

[0119] The model training module 404 is used to train the graph neural network model based on the adjacency matrix and the scheduling data to obtain a power grid scheduling model.

[0120] Optionally, the adjacency matrix construction module 402 includes:

[0121] A matrix node determination submodule is used to use all power grid devices in the power grid as matrix nodes of the adjacency matrix;

[0122] A connection relationship determination submodule, used for determining any two matrix nodes having a connection relationship based on the device node relationship;

[0123] The adjacency matrix determination submodule is used to connect any two matrix nodes having a connection relationship using an edge to obtain an adjacency matrix.

[0124] Optionally, the model training module 404 includes:

[0125] A node feature extraction submodule, used for extracting the node feature of each matrix node in the adjacency matrix from the corresponding device parameters and line parameters;

[0126] An aggregate information determination submodule, used for performing weighted summation on the node features of the adjacent matrix nodes based on the weight matrix in the graph neural network model to obtain aggregate information;

[0127] A prediction task planning data output submodule, used for inputting the aggregated information into the graph neural network model to output prediction task planning data;

[0128] An information difference acquisition submodule, used to acquire the information difference between the predicted task planning data and the actual planning data;

[0129] A judgment submodule, used to determine whether the graph neural network model has been trained through the information difference;

[0130] A power grid dispatching model determination submodule, used to use the trained graph neural network model as a power grid dispatching model if the graph neural network model training is completed;

[0131] The weight matrix updating submodule is used to update the weight matrix based on the information difference between the predicted task planning data and the actual planning data if the training of the graph neural network model is not completed; and return the content of the execution aggregation information determination submodule.

[0132] Optionally, the information difference acquisition submodule includes:

[0133] An information difference comparison item acquisition unit, used to respectively acquire the predicted task planning data, the task steps and the number of task steps of the actual planning data;

[0134] An information difference acquisition unit is used to calculate the information difference between the predicted task planning data and the actual planning data based on the similarity of the task steps and the similarity of the number of task steps.

[0135] Optionally, the information difference is inversely proportional to the similarity of the task steps and the similarity of the number of task steps.

[0136] Optionally, the model training module 404 further includes:

[0137] The node feature updating module is used to iteratively update the node features of the matrix nodes through aggregation operations.

[0138] The power grid dispatch model construction device provided in the embodiment of the present invention can execute the power grid dispatch model construction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0139] Embodiment 4

[0140] Figure 5 This is a schematic diagram of the structure of a power grid task generation device provided by the fourth embodiment of the present invention. Figure 5 As shown, the power grid task generating device includes:

[0141] A power grid status data acquisition module 501 is used to acquire real-time power grid status data;

[0142] The task planning data generating module 502 is used to input the power grid state data into the trained power grid dispatching model to obtain the task planning data.

[0143] The power grid dispatching model is trained by using the power grid dispatching model construction method of the present invention.

[0144] Optionally, the power grid task generating device further includes:

[0145] A rule base acquisition module is used to acquire a rule base for power grid dispatching tasks, wherein the rule base is composed of standardized task planning data;

[0146] The same data search module is used to determine whether there is task planning data in the rule base that is the same as the task planning data generated by the power grid dispatching model; if so, the content of the task data determination module to be executed is executed; if not, the content of the task planning data generation module is executed;

[0147] The module for determining the task data to be executed is used to determine that the task planning data generated by the power grid dispatching model is the task data to be executed.

[0148] The power grid task generation device provided in the embodiment of the present invention can execute the power grid task generation method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0149] Embodiment 5

[0150] Figure 6 A schematic diagram of the structure of an electronic device 40 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0151] like Figure 6As shown, the electronic device 40 includes at least one processor 41, and a memory connected to the at least one processor 41, such as a read-only memory (ROM) 42, a random access memory (RAM) 43, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 41 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 42 or the computer program loaded from the storage unit 48 to the random access memory (RAM) 43. In the RAM 43, various programs and data required for the operation of the electronic device 40 can also be stored. The processor 41, the ROM 42, and the RAM 43 are connected to each other through a bus 44. An input / output (I / O) interface 45 is also connected to the bus 44.

[0152] A number of components in the electronic device 40 are connected to the I / O interface 45, including: an input unit 46, such as a keyboard, a mouse, etc.; an output unit 47, such as various types of displays, speakers, etc.; a storage unit 48, such as a disk, an optical disk, etc.; and a communication unit 49, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 49 allows the electronic device 40 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0153] The processor 41 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 41 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The processor 41 executes the various methods and processes described above, such as a power grid dispatch model construction method, and / or a power grid task generation method.

[0154] In some embodiments, the power grid dispatch model construction method, and / or the power grid task generation method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 48. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 40 via the ROM 42 and / or the communication unit 49. When the computer program is loaded into the RAM 43 and executed by the processor 41, one or more steps of the power grid dispatch model construction method, and / or the power grid task generation method described above may be executed. Alternatively, in other embodiments, the processor 41 may be configured to execute the power grid dispatch model construction method, and / or the power grid task generation method, in any other appropriate manner (e.g., by means of firmware).

[0155] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0156] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0157] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0158] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0159] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0160] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0161] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0162] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for constructing a power grid dispatching model, characterized in that: include: Obtaining equipment node relationships and dispatching data of the power grid, wherein the dispatching data includes equipment parameters, line parameters, and actual task planning data; Constructing an adjacency matrix based on the device node relationship, the adjacency matrix including node and edge information, the nodes are power grid devices, and the edges are connection relationships between power grid devices; Build a graph neural network model suitable for power grid task generation; The graph neural network model is trained based on the adjacency matrix and the scheduling data to obtain a power grid scheduling model.

2. The method for constructing a power grid dispatching model according to claim 1, characterized in that: The step of constructing an adjacency matrix based on the device node relationship includes: All power grid devices in the power grid are regarded as matrix nodes of the adjacency matrix; Determine any two of the matrix nodes having a connection relationship based on the device node relationship; Any two matrix nodes having a connection relationship are connected by edges to obtain an adjacency matrix.

3. The method for constructing a power grid dispatching model according to claim 1, characterized in that: The graph neural network model includes a weight matrix, and the graph neural network model is trained based on the adjacency matrix and the scheduling data to obtain a power grid scheduling model, including: For each matrix node in the adjacency matrix, extracting a node feature of the matrix node from the corresponding device parameters and line parameters; Performing weighted summation on the node features of the adjacent matrix nodes based on the weight matrix in the graph neural network model to obtain aggregate information; Inputting the aggregated information into the graph neural network model to output prediction task planning data; Obtaining information difference between the predicted task planning data and the actual planning data; Determining whether the graph neural network model has been trained through the information difference; If so, the trained graph neural network model is used as a power grid dispatching model; If not, update the weight matrix based on the information difference between the predicted task planning data and the actual planning data; and return to execute the weighted summation of the node features of the adjacent matrix nodes based on the weight matrix in the graph neural network model to obtain aggregate information.

4. The method for constructing a power grid dispatching model according to claim 3, characterized in that: The obtaining of the information difference between the predicted task planning data and the actual planning data comprises: Respectively obtain the predicted task planning data, the task steps of the actual planning data and the number of task steps; The information difference between the predicted task planning data and the actual planning data is calculated based on the similarity of the task steps and the similarity of the number of task steps.

5. The method for constructing a power grid dispatching model according to claim 4, characterized in that: The information difference is inversely proportional to the similarity of the task steps and the similarity of the number of task steps.

6. The method for constructing a power grid dispatching model according to claim 3, characterized in that: After the weight matrix is ​​updated based on the information difference between the predicted task planning data and the actual planning data, the method further includes: The node features of the matrix nodes are updated iteratively through aggregation operations.

7. A method for generating power grid tasks, characterized in that: include: Obtain real-time grid status data; The power grid state data is input into a trained power grid dispatching model to obtain task planning data, and the power grid dispatching model is trained using the power grid dispatching model construction method according to any one of claims 1 to 6.

8. The method for generating power grid tasks according to claim 7, characterized in that: Also includes: Acquire a rule base for power grid dispatching tasks, wherein the rule base is composed of standardized task planning data; Determining whether there is task planning data in the rule base that is identical to the task planning data generated by the power grid dispatching model; If so, determining that the task planning data generated by the power grid dispatching model is task data to be executed; If not, return to the step of inputting the power grid status data into the trained power grid dispatching model to obtain task planning data.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the power grid dispatching model construction method described in any one of claims 1-6, and / or the power grid task generation method described in any one of claims 7-8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the power grid dispatching model construction method described in any one of claims 1-6 and / or the power grid task generation method described in any one of claims 7-8 when executed.

Images