Power grid adjustment method and device for joint operation of main and power distribution networks, computer equipment and storage medium

Abstract

The invention relates to a power grid adjustment method and device for joint operation of a main power distribution network, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring key operation data of a to-be-analyzed power grid; performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data; inputting the feature vector into a pre-trained current load prediction model to obtain a target prediction load of the to-be-analyzed power grid; obtaining power supply information of the to-be-analyzed power grid, and generating a power grid node adjustment instruction for the to-be-analyzed power grid according to the target prediction load and the power supply information of the to-be-analyzed power grid; and adjusting the grid of the power grid to be analyzed according to the grid node adjustment instruction of the power grid. By adopting the method, the adjustment accuracy of the grid of the power grid can be improved.

Description

Technical Field

[0001] The present application relates to the field of power grid technology, and in particular to a power grid adjustment method, device, computer equipment, computer-readable storage medium and computer program product for joint operation of a main distribution network. Background Art

[0002] In the field of power grids, how to adjust the power grid is crucial to optimizing the grid structure and improving power supply reliability.

[0003] In traditional technology, manual planning is generally used to adjust the power grid; however, this manual planning method has subjective factors and is prone to errors, resulting in low accuracy in adjusting the power grid. Summary of the invention

[0004] Based on this, it is necessary to provide a power grid adjustment method, device, computer equipment, computer-readable storage medium and computer program product for joint operation of main and distribution networks, which can improve the adjustment accuracy of the power grid, in order to address the above technical problems.

[0005] In a first aspect, the present application provides a power grid adjustment method for joint operation of a main distribution network, comprising:

[0006] Obtain key operating data of the power grid to be analyzed;

[0007] Performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data;

[0008] Inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed;

[0009] Acquire power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed;

[0010] The power grid to be analyzed is adjusted according to the power grid node adjustment instruction.

[0011] In one of the embodiments, the power grid to be analyzed includes a plurality of power grid nodes;

[0012] The step of obtaining key operation data of the power grid to be analyzed includes:

[0013] Determining the importance of each of the power grid nodes;

[0014] From each of the power grid nodes, select a power grid node whose corresponding importance is greater than a preset importance as a key power grid node;

[0015] The operation data of the key power grid nodes are used as the key operation data of the power grid to be analyzed.

[0016] In one embodiment, the obtaining of power supply information of the power grid to be analyzed includes:

[0017] Determining power supply nodes in the power grid to be analyzed;

[0018] The location information, expected power generation and load demand information within a preset range of the power node are all used as the power information of the power grid to be analyzed.

[0019] In one embodiment, the performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data includes:

[0020] Preprocessing the key operation data to obtain preprocessed key operation data;

[0021] Identifying the data type of the pre-processed key operation data;

[0022] According to the data type of the preprocessed key operation data, query the correspondence between the data type and the feature extraction model to obtain the feature extraction model corresponding to the preprocessed key operation data;

[0023] The preprocessed key operation data is input into a feature extraction model corresponding to the preprocessed key operation data to perform feature extraction processing, and a feature vector corresponding to the preprocessed key operation data is obtained as the feature vector corresponding to the key operation data.

[0024] In one embodiment, the step of inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed includes:

[0025] determining a historical load forecasting model associated with the pre-trained current load forecasting model;

[0026] Inputting the feature vector into the pre-trained current load prediction model to obtain a first predicted load of the power grid to be analyzed, and inputting the feature vector into the historical load prediction model to obtain a second predicted load of the power grid to be analyzed;

[0027] The first predicted load and the second predicted load of the power grid to be analyzed are fused to obtain a target predicted load of the power grid to be analyzed.

[0028] In one embodiment, the pre-trained current load prediction model is trained in the following manner:

[0029] Obtaining sample key operating data of a sample power grid;

[0030] Performing feature extraction processing on the sample key operation data to obtain a sample feature vector corresponding to the sample key operation data;

[0031] Inputting the sample feature vector into a current load prediction model to be trained to obtain a predicted load of the sample power grid;

[0032] According to the difference between the predicted load and the actual load of the sample power grid, the current load prediction model to be trained is iteratively trained to obtain the pre-trained current load prediction model.

[0033] In a second aspect, the present application further provides a power grid adjustment device for joint operation of a main distribution network, comprising:

[0034] A data acquisition module, used to acquire key operation data of the power grid to be analyzed;

[0035] A feature extraction module, used to perform feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data;

[0036] A load prediction module, used for inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed;

[0037] An instruction generation module, used to obtain power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed;

[0038] A grid adjustment module is used to adjust the power grid to be analyzed according to the power grid node adjustment instruction.

[0039] In a third aspect, the present application further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0040] Obtain key operating data of the power grid to be analyzed;

[0041] Performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data;

[0042] Inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed;

[0043] Acquire power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed;

[0044] The power grid to be analyzed is adjusted according to the power grid node adjustment instruction.

[0045] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the following steps are implemented:

[0046] Obtain key operating data of the power grid to be analyzed;

[0047] Performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data;

[0048] Inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed;

[0049] Acquire power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed;

[0050] The power grid to be analyzed is adjusted according to the power grid node adjustment instruction.

[0051] In a fifth aspect, the present application further provides a computer program product, including a computer program, which implements the following steps when executed by a processor:

[0052] Obtain key operating data of the power grid to be analyzed;

[0053] Performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data;

[0054] Inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed;

[0055] Acquire power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed;

[0056] The power grid to be analyzed is adjusted according to the power grid node adjustment instruction.

[0057] The above-mentioned grid adjustment method, device, computer equipment, storage medium and computer program product for the joint operation of the main distribution network first obtain the key operation data of the grid to be analyzed, then perform feature extraction processing on the key operation data to obtain the feature vector corresponding to the key operation data, then input the feature vector into the pre-trained current load prediction model to obtain the target predicted load of the grid to be analyzed, then obtain the power supply information of the grid to be analyzed, and generate the grid node adjustment instruction for the grid to be analyzed according to the target predicted load and power supply information of the grid to be analyzed, and finally, adjust the grid to be analyzed according to the grid node adjustment instruction. In this way, when adjusting the grid, the grid to be analyzed is adjusted according to the target predicted load and power supply information of the grid to be analyzed, so as to better match the power supply and demand, reduce the problem of voltage instability caused by load overload or insufficient power supply, and help improve the reliability of grid operation; moreover, the method does not require manual intervention, avoids the defects of subjective factors in the manual planning method, which is prone to errors and leads to low adjustment accuracy of the grid, thereby improving the adjustment accuracy of the grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0059] Figure 1 A schematic flow chart of a power grid adjustment method for joint operation of a main distribution network in one embodiment;

[0060] Figure 2 A schematic flow chart of steps for obtaining key operation data of a power grid to be analyzed in one embodiment;

[0061] Figure 3 A schematic flow chart of the steps of obtaining a feature vector corresponding to key operation data in one embodiment;

[0062] Figure 4It is a flowchart of a power grid adjustment method for joint operation of a main distribution network in another embodiment;

[0063] Figure 5 A structural block diagram of a power grid adjustment device for joint operation of a main distribution network in one embodiment;

[0064] Figure 6 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0066] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0067] In an exemplary embodiment, Figure 1 As shown, a grid adjustment method for the joint operation of the main distribution network is provided. This embodiment uses the method applied to the server as an example for illustration; it can be understood that the method can also be applied to the terminal, and can also be applied to the system including the terminal and the server, and is implemented through the interaction between the terminal and the server. Among them, the terminal can be but not limited to various personal computers, laptops, smart phones and tablets; the server can be implemented with an independent server or a server cluster composed of multiple servers. In this embodiment, the method includes the following steps:

[0068] Step S101, obtaining key operation data of the power grid to be analyzed.

[0069] The grid to be analyzed refers to the grid that needs to be analyzed. The grid refers to the network structure formed by interconnecting power facilities such as transmission lines and substations in the power system.

[0070] Among them, key operation data refers to operation data with a greater degree of importance, including voltage data, current data, frequency data, power data, temperature data, state data, environmental data, etc. of the power grid to be analyzed.

[0071] Exemplarily, the server obtains the operating data of the power grid to be analyzed from the database; then, the server determines the importance corresponding to the operating data of the power grid to be analyzed; then, the server filters out the operating data with a corresponding importance greater than a preset importance from the operating data of the power grid to be analyzed as the key operating data of the power grid to be analyzed.

[0072] Step S102: performing feature extraction processing on the key operation data to obtain feature vectors corresponding to the key operation data.

[0073] The feature extraction process refers to a process of obtaining a feature vector corresponding to the key operating data based on the key operating data.

[0074] Among them, the feature vector refers to the characterization vector corresponding to the key operating data.

[0075] Exemplarily, the server inputs the key operation data into the feature extraction model, performs feature extraction processing on the key operation data through the feature extraction model, and obtains a feature vector corresponding to the key operation data.

[0076] Step S103, inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed.

[0077] The current load prediction model refers to the load prediction model for the current time period. The load prediction model refers to a network model that can use the feature vector to obtain the target predicted load corresponding to the power grid to be analyzed, such as a machine learning model.

[0078] The target predicted load refers to the estimated value of the power load that the power grid to be analyzed is expected to bear in a certain period of time in the future (such as the next day, week, month, etc.).

[0079] Exemplarily, the server inputs the feature vector into a pre-trained current load prediction model, and obtains the predicted load corresponding to the feature vector through the current load prediction model; then, the server uses the predicted load corresponding to the feature vector as the target predicted load of the power grid to be analyzed.

[0080] Step S104, obtaining power source information of the power grid to be analyzed, and generating a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power source information of the power grid to be analyzed.

[0081] The power supply information refers to the information related to the power supply in the power grid to be analyzed.

[0082] The power grid node adjustment instruction is used to represent instruction information corresponding to adjusting the power grid nodes in the power grid to be analyzed.

[0083] Exemplarily, the server obtains information related to power supply in the power grid to be analyzed from the database as the power supply information of the power grid to be analyzed; then, the server determines the power grid node adjustment instructions corresponding to the target predicted load and power supply information of the power grid to be analyzed through the control algorithm, according to the target predicted load and power supply information of the power grid to be analyzed, as the power grid node adjustment instructions for the power grid to be analyzed. It should be noted that the control algorithm includes fuzzy control algorithm, PID (Proportional-Integral-Derivative Control) control algorithm, etc.

[0084] Further, the server determines the load-power matching situation of the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed, and determines the power grid node adjustment instruction corresponding to the load-power matching situation as the power grid node adjustment instruction for the power grid to be analyzed; for example, when the load data corresponding to the target predicted load is greater than the power supply data corresponding to the power supply information, it means that the predicted load exceeds the power generation capacity, and the instruction corresponding to the power generation amount is increased (such as increasing the number of power stations in the power grid to be analyzed) as the power grid node adjustment instruction for the power grid to be analyzed; when the load data corresponding to the target predicted load is less than the power supply data corresponding to the power supply information, it means that the predicted load is lower than the power generation capacity, and the instruction corresponding to the power generation amount is reduced (such as reducing the number of power stations in the power grid to be analyzed) as the power grid node adjustment instruction for the power grid to be analyzed; when the load data corresponding to the target predicted load is equal to the power supply data corresponding to the power supply information, it means that the predicted load is equal to the power generation capacity, and the instruction corresponding to the power generation amount is maintained (such as keeping the number of power stations in the power grid to be analyzed unchanged) as the power grid node adjustment instruction for the power grid to be analyzed.

[0085] Step S105, adjusting the power grid to be analyzed according to the power grid node adjustment instruction.

[0086] Exemplarily, the server adjusts the power grid to be analyzed according to the power grid node adjustment instruction to obtain the adjusted power grid to be analyzed; then, the server uses the adjusted power grid to be analyzed as the new power grid to be analyzed, and jumps to the step of obtaining the key operating data of the power grid to be analyzed, until the key operating data of the power grid to be analyzed meets the preset conditions (for example, the key operating data is within the preset data threshold range), and then uses the adjusted power grid to be analyzed as the target power grid.

[0087] In the above-mentioned grid adjustment method for the joint operation of the main distribution network, the key operation data of the grid to be analyzed is first obtained, and then the key operation data is subjected to feature extraction processing to obtain the feature vector corresponding to the key operation data. Then, the feature vector is input into the pre-trained current load prediction model to obtain the target predicted load of the grid to be analyzed. Then, the power supply information of the grid to be analyzed is obtained, and according to the target predicted load and power supply information of the grid to be analyzed, a grid node adjustment instruction for the grid to be analyzed is generated. Finally, according to the grid node adjustment instruction, the grid to be analyzed is adjusted. In this way, when the grid is adjusted, the grid to be analyzed is adjusted according to the target predicted load and power supply information of the grid to be analyzed, so that the power supply and demand can be better matched, the problem of voltage instability caused by load overload or insufficient power supply can be reduced, and the reliability of grid operation can be improved. Moreover, the method does not require manual intervention, avoids the defects of subjective factors in the manual planning method, which is prone to errors and leads to low adjustment accuracy of the grid, thereby improving the adjustment accuracy of the grid.

[0088] In an exemplary embodiment, the power grid to be analyzed includes a plurality of power grid nodes.

[0089] So, if Figure 2 As shown, the above step S101, obtaining key operation data of the power grid to be analyzed, specifically includes the following steps:

[0090] Step S201, determining the importance of each power grid node.

[0091] Step S202 : Filter out, from among the grid nodes, grid nodes whose corresponding importance is greater than a preset importance as key grid nodes.

[0092] Step S203: using the operation data of the key power grid nodes as the key operation data of the power grid to be analyzed.

[0093] Among them, the power grid node refers to the connection point or intersection point with specific functions in the power grid to be analyzed, such as distributed power access node, substation node, load access node, communication node, etc.

[0094] The preset importance level refers to a preset importance level threshold value, which is used to judge the importance level corresponding to each power grid node. It should be noted that the preset importance level can be determined according to the circumstances.

[0095] Among them, the key power grid node refers to a power grid node whose corresponding importance is greater than a preset importance.

[0096] Exemplarily, the server determines the node type of each power grid node; then, the server queries the correspondence between the node type and the importance according to the node type of each power grid node, and obtains the corresponding importance of each power grid node; then, the server selects the power grid nodes whose corresponding importance is greater than the preset importance from each power grid node, and uses these power grid nodes as key power grid nodes; then, the server obtains the operating data of the key power grid nodes through sensors installed at the key power grid nodes, and uses the operating data of the key power grid nodes as the key operating data of the power grid to be analyzed.

[0097] In this embodiment, by using the operating data of key power grid nodes as the key operating data of the power grid to be analyzed, operation and maintenance personnel can conduct focused monitoring without having to apply equal effort to all nodes, thereby greatly improving monitoring efficiency and providing a data basis for the follow-up.

[0098] In an exemplary embodiment, the above step S104, obtaining the power supply information of the power grid to be analyzed, specifically includes the following contents: determining the power supply nodes in the power grid to be analyzed; using the location information, expected power generation and load demand information within a preset range of the power supply nodes as the power supply information of the power grid to be analyzed.

[0099] The power source node refers to the distributed power source access node in the power grid to be analyzed.

[0100] The location information is used to characterize the specific geographical location of the power node, including the coordinate information of the power node.

[0101] Among them, the expected power generation refers to the estimated value of the total amount of electric energy that the power node is expected to generate in a specific period of time in the future.

[0102] The load demand information within the preset range refers to the power demand of users or power-consuming equipment within a certain period of time in a specific area covered by the power node.

[0103] Exemplarily, the server determines the power node in the power grid to be analyzed; then, the server obtains the location information of the power node, and obtains the location information of other nodes associated with the power node (such as the surrounding power grid infrastructure); then, the server determines the distance information between the power node and the other nodes based on the location information of the power node and the location information of the other nodes; then, the server determines the target node whose distance information to the power node is within a preset distance range based on the distance information between the power node and the other nodes; then, the server obtains the load demand information of the target node, and uses the load demand information of the target node as the load demand information within a preset range of the power node; then, the server obtains the historical power generation of the power node, and inputs the historical power generation of the power node into a power generation prediction model to obtain the expected power generation of the power node; then, the server splices the location information, expected power generation and load demand information within the preset range of the power node to obtain the spliced ​​information corresponding to the power node; then, the server uses the spliced ​​information corresponding to the power node as the power information of the power grid to be analyzed.

[0104] In this embodiment, by clarifying the location information of the power supply node, the layout and line direction of the power grid can be planned more reasonably, transmission losses can be reduced, and the overall efficiency of the power grid can be improved; moreover, by determining the expected power generation of the power supply node and the load demand information within a preset range, power allocation and reserves can be made in advance to ensure stable power supply during peak power consumption or emergencies.

[0105] In an exemplary embodiment, Figure 3 As shown, the above step S102, performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data, specifically includes the following steps:

[0106] Step S301, preprocessing the key operation data to obtain preprocessed key operation data.

[0107] Step S302, identifying the data type of the pre-processed key operation data.

[0108] Step S303, according to the data type of the preprocessed key operation data, query the corresponding relationship between the data type and the feature extraction model to obtain the feature extraction model corresponding to the preprocessed key operation data.

[0109] Step S304, input the preprocessed key operation data into a feature extraction model corresponding to the preprocessed key operation data to perform feature extraction processing, and obtain a feature vector corresponding to the preprocessed key operation data as the feature vector corresponding to the key operation data.

[0110] Among them, preprocessing includes cleaning, formatting, and outlier processing.

[0111] The preprocessed key operation data refers to the preprocessed key operation data.

[0112] The data type refers to the type corresponding to the sensor data, such as numeric type, character type, etc.

[0113] The correspondence between the data type and the feature extraction model is used to characterize the association information between the data type and the feature extraction model. For example, data type A corresponds to feature extraction model a, data type B corresponds to feature extraction model b, and data type C corresponds to feature extraction model c.

[0114] Among them, the feature extraction model refers to the network model that can perform feature extraction processing on data, such as convolutional neural network, recurrent neural network, etc.

[0115] Exemplarily, the server preprocesses the key operating data to obtain the preprocessed key operating data as the preprocessed key operating data; for example, the server identifies abnormal data in the key operating data, and deletes the abnormal data in the key operating data to obtain the preprocessed key operating data; then, the server identifies the data type of the preprocessed key operating data; then, the server queries the correspondence between the data type and the feature extraction model according to the data type of the preprocessed key operating data, and obtains the feature extraction model corresponding to the data type of the preprocessed key operating data as the feature extraction model corresponding to the preprocessed key operating data; then, the server inputs the preprocessed key operating data into the feature extraction model corresponding to the preprocessed key operating data for feature extraction processing, and obtains the feature vector corresponding to the preprocessed key operating data as the feature vector corresponding to the key operating data.

[0116] In this embodiment, the feature extraction model corresponding to the key operating data after preprocessing is determined based on the correspondence between the data type and the feature extraction model, so that special feature extraction models can be matched according to different data types, thereby improving the accuracy and adaptability of feature extraction, which is conducive to improving the accuracy of determining the feature vector corresponding to the key operating data.

[0117] In an exemplary embodiment, the above step S103, inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed, specifically includes the following contents: determining a historical load prediction model associated with the pre-trained current load prediction model; inputting the feature vector into the pre-trained current load prediction model to obtain a first predicted load of the power grid to be analyzed, and inputting the feature vector into the historical load prediction model to obtain a second predicted load of the power grid to be analyzed; fusing the first predicted load and the second predicted load of the power grid to be analyzed to obtain the target predicted load of the power grid to be analyzed.

[0118] The historical load forecasting model refers to the load forecasting model for the past time period.

[0119] The first predicted load refers to the predicted load of the power grid to be analyzed output by the current load prediction model.

[0120] The second predicted load refers to the predicted load of the power grid to be analyzed output by the historical load prediction model.

[0121] The fusion processing may refer to weighted summation processing.

[0122] Exemplarily, the server determines a historical load prediction model associated with a pre-trained current load prediction model; then, the server inputs the feature vector into the pre-trained current load prediction model to obtain the predicted load of the power grid to be analyzed output by the current load prediction model as the first predicted load; then, the server inputs the feature vector into the historical load prediction model to obtain the predicted load of the power grid to be analyzed output by the historical load prediction model as the second predicted load; then, the server obtains the model weight corresponding to the current load prediction model and the model weight corresponding to the historical load prediction model; then, the server uses the model weight corresponding to the current load prediction model as the first weight corresponding to the first predicted load, and uses the model weight corresponding to the historical load prediction model as the second weight corresponding to the second predicted load; finally, the server fuses the first predicted load and the second predicted load of the power grid to be analyzed according to the first weight and the second weight, and obtains the fused predicted load as the target predicted load of the power grid to be analyzed; for example, the server performs weighted summation on the first predicted load and the second predicted load of the power grid to be analyzed according to the first weight and the second weight, and obtains the weighted summed predicted load as the target predicted load of the power grid to be analyzed.

[0123] In this embodiment, by combining the predicted loads of the current load prediction model and the historical load prediction model for fusion processing, the advantages and characteristics of different models can be comprehensively utilized, which is beneficial to improving the accuracy of the load prediction of the power grid to be analyzed, and further improving the accuracy of determining the target predicted load of the power grid to be analyzed.

[0124] In an exemplary embodiment, the power grid adjustment method for joint operation of the main distribution network provided in the present application also includes a training step of a pre-trained current load prediction model, which specifically includes the following contents: obtaining sample key operating data of the sample power grid; performing feature extraction processing on the sample key operating data to obtain a sample feature vector corresponding to the sample key operating data; inputting the sample feature vector into the current load prediction model to be trained to obtain the predicted load of the sample power grid; and iteratively training the current load prediction model to be trained according to the difference between the predicted load and the actual load of the sample power grid to obtain a pre-trained current load prediction model.

[0125] The sample power grid refers to the power grid corresponding to the current load prediction model used for training.

[0126] Among them, the sample key operating data refers to the key operating data of the sample power grid.

[0127] The sample feature vector refers to the feature vector corresponding to the key operating data of the sample.

[0128] The predicted load of the sample power grid refers to the predicted load value of the sample power grid.

[0129] The actual load refers to the actual predicted value of the sample power grid.

[0130] Exemplarily, in response to a model training instruction for a current load prediction model to be trained, the server obtains key operating data of a sample power grid from a database as sample key operating data; then, the server inputs the sample key operating data into a feature extraction model, performs feature extraction processing on the sample key operating data through the feature extraction model, and obtains a feature vector corresponding to the sample key operating data as a sample feature vector; then, the server inputs the sample feature vector into the current load prediction model to be trained to obtain a predicted load of the sample power grid; then, the server obtains a loss value based on the difference between the predicted load and the actual load of the sample power grid; then, the server adjusts the model parameters of the current load prediction model to be trained based on the loss value; then, the server retrains the current load prediction model after the model parameters are adjusted until the loss value obtained by the trained current load prediction model is less than the loss value threshold, then stops the training, and uses the trained current load prediction model as the pre-trained current load prediction model.

[0131] In this embodiment, by pre-training the current load prediction model, it is convenient to predict the target predicted load of the power grid to be analyzed after obtaining the key operating data of the power grid to be analyzed in actual applications; moreover, the current load prediction model receives new data in each round of iteration, and performs internal improvements and optimizations on the model, so that predictions can be made more effectively, which is beneficial to improving the prediction accuracy of the current load prediction model.

[0132] In an exemplary embodiment, Figure 4 As shown, another grid adjustment method for joint operation of the main distribution network is provided, and the method is applied to a server for example, including the following steps:

[0133] Step S401, determine the importance of each grid node in the grid to be analyzed; select grid nodes whose corresponding importance is greater than a preset importance from each grid node as key grid nodes; and use the operating data of the key grid nodes as the key operating data of the grid to be analyzed.

[0134] Step S402, preprocessing the key operating data to obtain preprocessed key operating data; identifying the data type of the preprocessed key operating data; according to the data type of the preprocessed key operating data, querying the correspondence between the data type and the feature extraction model to obtain the feature extraction model corresponding to the preprocessed key operating data.

[0135] Step S403, input the preprocessed key operation data into a feature extraction model corresponding to the preprocessed key operation data to perform feature extraction processing, and obtain a feature vector corresponding to the preprocessed key operation data as the feature vector corresponding to the key operation data.

[0136] Step S404: determine a historical load prediction model associated with the pre-trained current load prediction model.

[0137] Step S405, inputting the feature vector into a pre-trained current load prediction model to obtain a first predicted load of the power grid to be analyzed, and inputting the feature vector into a historical load prediction model to obtain a second predicted load of the power grid to be analyzed.

[0138] Step S406 , fusing the first predicted load and the second predicted load of the power grid to be analyzed to obtain a target predicted load of the power grid to be analyzed.

[0139] Step S407, determining the power supply nodes in the power grid to be analyzed; taking the location information, expected power generation and load demand information within a preset range of the power supply nodes as the power supply information of the power grid to be analyzed.

[0140] Step S408: generating a grid node adjustment instruction for the grid to be analyzed based on the target predicted load and power source information of the grid to be analyzed.

[0141] Step S409: adjusting the power grid to be analyzed according to the power grid node adjustment instruction.

[0142] In the above-mentioned grid adjustment method for the joint operation of the main distribution network, when adjusting the grid, the grid to be analyzed is adjusted according to the target predicted load and power supply information of the grid to be analyzed, so as to better match the power supply and demand, reduce the problem of voltage instability caused by load overload or insufficient power supply, and help improve the reliability of grid operation; moreover, the method does not require human intervention, avoids the subjective factors in manual planning, prone to errors, and leads to the defect of low adjustment accuracy of the grid, thereby improving the adjustment accuracy of the grid.

[0143] In an exemplary embodiment, in order to more clearly illustrate the power grid adjustment method for the joint operation of the main distribution network provided by the embodiment of the present application, the power grid adjustment method for the joint operation of the main distribution network is specifically described below with a specific embodiment. In one embodiment, the present application also provides a dynamic grid optimization system and method for the joint operation of the main distribution network. When adjusting the power grid, the key operating data of the power grid to be analyzed is first obtained, and then the key operating data is subjected to feature extraction processing to obtain the feature vector corresponding to the key operating data. Then, the feature vector is input into the pre-trained current load prediction model to obtain the target predicted load of the power grid to be analyzed. Then, the power supply information of the power grid to be analyzed is obtained, and based on the target predicted load and power supply information of the power grid to be analyzed, a power grid node adjustment instruction for the power grid to be analyzed is generated. Finally, according to the power grid node adjustment instruction, the power grid to be analyzed is adjusted. Specifically including the following contents:

[0144] In the current power management system, we often face the challenges of dynamic adjustment of grid structure and optimization of power distribution, especially when load fluctuates and resource allocation requirements change rapidly. Traditional methods rely on manual monitoring and adjustment, which is not only time-consuming but also often unable to respond to changes in grid status in real time, thus affecting the efficiency and safety of the grid.

[0145] In order to solve these problems, the present invention proposes an adaptive network reconstruction algorithm, which collects real-time data by installing sensors at key nodes of the power grid and uses machine learning technology to predict short-term and long-term load fluctuations, thereby dynamically adjusting the grid structure. In addition, the present invention also designs a grid management system for distributed resources, which combines geographic information system (GIS) and power grid data analysis to optimize the access points of distributed power sources, dynamically monitor the real-time status of the power grid, and automatically calculate the maximum safe access capacity of each potential access point. Through these technologies, the present invention can optimize energy distribution and grid configuration strategies, improve the response speed and operating efficiency of the power grid, and ensure the stable operation of the power grid.

[0146] Load forecasting provides key information about the current and future state of the power grid, allowing the system to predict when and where load peaks or troughs will occur; while the optimization of power access points uses these forecast data to dynamically adjust the access points and access amounts of distributed power sources to ensure the optimal allocation and use of power resources. By combining load forecasting with power access optimization, the present invention can improve the response speed and operating efficiency of the power grid while coping with fluctuations in new energy sources, thereby ensuring the stable operation of the power grid.

[0147] The embodiment of the present invention provides a main-coordinated dynamic grid optimization system and method thereof, including but not limited to the following steps S100, S200 and S300.

[0148] Step S100, adaptive network reconstruction algorithm;

[0149] The adaptive network reconstruction algorithm first collects real-time data through sensors installed at key nodes of the power grid, and then makes decisions based on the collected data to evaluate the power status and potential risks of each grid in the current power grid. Finally, the results of all operations and adjustments are fed back to the central control system.

[0150] Step S200, grid management of distributed resources;

[0151] In the grid management of distributed resources, an advanced automation system is first designed to optimize the access of distributed power sources, and then GIS tools are used to analyze their relative position relationship with the surrounding power grid infrastructure. Finally, the real-time status of the power grid is dynamically monitored to predict short-term and long-term power demand fluctuations.

[0152] Step S300, optimization algorithm and implementation strategy.

[0153] Model and algorithm design:

[0154] Grid Optimization Configuration Model: The present invention constructs a model that includes mathematical representations of all grid components (e.g., power stations, substations, transmission lines, and load centers). The operating range (e.g., generation, load demand, line capacity) and operating costs (e.g., fuel costs, power loss costs) of each component are part of the model input.

[0155] Output of the model:

[0156] Optimal power distribution plan: Based on the operating range and cost data of the grid components input by the model, the model outputs the optimal power distribution plan. This includes the distribution of power generation of each power station, the power demand of each load center, and the flow path of power in the transmission line, aiming to maximize the efficiency of grid operation, reduce transmission losses, and ensure the balance of power supply and demand.

[0157] Real-time adjustment suggestions: Using real-time data and load forecasts, the model provides real-time adjustment suggestions, including adjusting the output power of power plants, changing the power transmission path, or switching backup power sources to cope with sudden load fluctuations or power shortages to ensure the stable operation of the power grid.

[0158] Cost optimization results: The model outputs the cost optimization results of power grid operation, including fuel cost, power loss cost and maintenance cost, etc. By optimizing these costs, the overall operating cost can be reduced and the economic efficiency of the power grid can be improved.

[0159] Grid stability assessment: The model provides assessment results of grid stability. This includes voltage stability, frequency stability, and load balance at each node. These assessment results help grid managers to promptly identify and resolve potential stability issues and ensure safe operation of the grid.

[0160] Nonlinear programming to handle complex constraints: Nonlinear programming is used to handle nonlinear problems in power grid operation, such as power generation efficiency curves, random fluctuations in renewable energy output, etc.

[0161] Integration of Machine Learning:

[0162] Prediction and Adaptive Adjustment: Integrate machine learning models to predict grid demand and renewable energy supply fluctuations. Use these predictions to adjust grid configuration, optimize power distribution, and reduce unnecessary energy storage or purchases.

[0163] Real-time data-driven decision-making: Grid operation data is collected in real time, and the machine learning model continuously updates and optimizes algorithm parameters based on real-time data to adapt to the current grid status and predict future changes.

[0164] Based on the above steps S100 to S300 , a specific implementation scheme is described in detail below.

[0165] In step S100, the target power station is provided with a plurality of power-consuming devices (these power-consuming devices may be power-consuming devices of the same type or power-consuming devices of different types). Then, when collecting power consumption patterns, the following steps may be followed:

[0166] Step S110, data preprocessing;

[0167] Step S120, algorithm decision making;

[0168] Step S130: Feedback to the central control system.

[0169] In specific implementation, the module first collects real-time data through sensors installed at key nodes of the power grid. The data is pre-processed, including cleaning, formatting, and outlier processing to ensure its quality and consistency. The processed data is fed into a prediction model that uses machine learning technology to predict short-term and long-term load fluctuations based on historical data trends and real-time input from the field. The prediction results will be used for decision-making, and optimization algorithms will be used to determine how to adjust the grid most effectively.

[0170] Voltage data, current data, frequency data, power data, load data, temperature data, status data, environmental data, etc.

[0171] The decision-making process of the adaptive network reconstruction algorithm includes evaluating the power status and potential risks of each grid of the current power grid, such as overload or energy shortage. Based on this evaluation, the algorithm decides whether to expand certain grids to increase their capacity, merge adjacent grids to optimize resource sharing, or split a grid to reduce load or improve response speed. These dynamic adjustments are automatically performed by the control system, but also allow operators to intervene manually to respond to complex or urgent grid management situations.

[0172] Ultimately, the results of all operations and adjustments are fed back to the central control system for performance monitoring and subsequent system optimization. The system's performance is evaluated based on key indicators such as grid operation efficiency, energy utilization, and power supply reliability, ensuring that the algorithm is continuously iterated and improved to adapt to the ever-changing grid operation and new market demands.

[0173] Step S200 may specifically include the following steps:

[0174] Step S210, optimizing access to distributed power sources;

[0175] Step S220, resource aggregation strategy;

[0176] Step S230: real-time adjustment and response.

[0177] In this invention, we have designed an advanced automation system for the optimized access of distributed power sources, which combines geographic information system (GIS) and implements grid data analysis to automatically identify and optimize the access points of distributed power sources. The core innovation of this system is that it can comprehensively consider the geographical location of the power source, the expected power generation and the nearby load demand, so as to achieve the optimal configuration of the power access point.

[0178] The system uses GIS tools to analyze the relative position of distributed power sources (such as wind turbines and photovoltaic panels) and the surrounding power grid infrastructure based on their specific geographical locations. In addition, the system also takes into account the expected power generation of these power sources, as well as the power demand and load conditions in the nearby areas. Through the comprehensive analysis of this data, the system can automatically determine the best power access point, optimize energy distribution, and reduce losses in energy transmission.

[0179] The relative position relationship here refers to the geographical distance and spatial relationship between distributed power sources (such as wind turbines and photovoltaic panels) and surrounding power grid infrastructure (such as substations, transmission lines and load centers). By analyzing these relative positions, the optimal location of distributed power access points can be determined to reduce energy transmission losses and improve power supply efficiency.

[0180] In addition, the system takes into account the expected power generation of these power sources, as well as the power demand and load conditions in the nearby area.

[0181] Expected power generation is determined by a variety of methods, including the following:

[0182] Historical Generation Data: Analyze historical generation records to identify generation patterns and trends.

[0183] Environmental conditions: Evaluate current and future environmental conditions (such as sunshine, wind speed, etc.) and predict the power generation of wind turbines and photovoltaic panels.

[0184] Equipment parameters: Consider the rated capacity and performance parameters of the power generation equipment, and make adjustments based on actual operating data.

[0185] Machine Learning Model: Uses machine learning techniques to combine historical data and real-time inputs to predict future power generation.

[0186] Through these methods, the system can more accurately predict the power generation of distributed power sources and optimize the operation of the power grid.

[0187] Determination of the load condition includes the following steps:

[0188] Real-time data collection: Collect current load data, including voltage, current, power, etc.

[0189] Historical data analysis: Analyze historical load data to identify load patterns and changing trends.

[0190] Environmental and social factors: Consider external factors that affect load, such as weather, holidays, etc.

[0191] Machine Learning Forecasting: Use machine learning techniques to predict short-term and long-term load fluctuations based on the above data and factors.

[0192] The determination of load conditions is closely related to the load fluctuation prediction in step S100. The load prediction model in step S100 provides dynamic prediction results of load changes, which are used to optimize the access points of distributed power sources and the operation strategy of the power grid to ensure supply and demand balance and power grid stability.

[0193] Through comprehensive analysis of this data, the system can automatically determine the best power access point, optimize energy distribution, and reduce losses in energy transmission.

[0194] Determining the best power access point is closely related to grid optimization. The specific links are as follows:

[0195] Supply and demand matching: Determining the best access point can ensure that the distributed generation is located close to the load center, reduce transmission losses, and achieve the best match between supply and demand.

[0196] Grid flexibility: By optimizing access points, the flexibility and adaptability of the grid are enhanced, which can better cope with load fluctuations and the uncertainty of distributed power generation.

[0197] Improved stability: Reasonable access point optimization reduces the operating pressure of the power grid and improves the stability and reliability of the power grid.

[0198] Cost optimization: By reducing the transmission distance and optimizing power distribution, the operating cost of the power grid can be reduced and the economic benefits can be improved.

[0199] By combining load forecasting and access point optimization, the system can achieve dynamic optimization of the grid and improve the operating efficiency and stability of the overall power grid.

[0200] The system dynamically monitors the real-time status of the power grid, including the existing grid load and capacity, and uses advanced forecasting models to predict short-term and long-term fluctuations in power demand. Based on this information, the system's algorithm automatically calculates the maximum safe access capacity of each potential access point, ensuring that the newly connected distributed power generation does not exceed the grid's carrying capacity, thereby avoiding grid stability issues caused by overload.

[0201] The optimized access of distributed power sources is achieved through two core technologies: automated access point identification and access capacity management, which aims to improve the overall operating efficiency and stability of the power grid. By using geographic information systems (GIS) and real-time power grid data, the system can accurately identify the most suitable locations for accessing distributed power sources and evaluate the power demand and power generation potential of these locations. In addition, by real-time monitoring of grid load and capacity, the system can dynamically calculate the maximum safe access capacity of each access point to effectively prevent grid overload. The application of these two technologies not only optimizes resource allocation and reduces energy losses, but also improves the grid's ability to adapt to fluctuations in new energy, ensuring efficient and reliable operation of the grid, while promoting the development of the power system towards a low-carbon and sustainable direction.

[0202] In this invention, the resource aggregation strategy is to optimize the management and utilization of distributed resources and improve the overall efficiency and responsiveness of the power grid. The core of this strategy is to develop a dynamic resource aggregation algorithm that integrates scattered small-scale power generation resources into a unified and powerful virtual power plant through virtual power plant (VPP) technology. The construction and implementation process of the algorithm is described in detail below.

[0203] The dynamic resource aggregation algorithm is based on real-time data analysis and optimization theory, and aims to achieve the optimal configuration and management of distributed power sources. The core of the algorithm design is to aggregate each independently operated distributed power source (such as wind power, solar power, small hydropower, etc.) into a virtual power plant through intelligent control technology, thereby improving the integrated efficiency of the power source and the stability of power output.

[0204] The algorithm first collects real-time data of each distributed generation unit, including but not limited to power generation, operating status, environmental parameters, etc. The collected data is standardized, including data cleaning and normalization, to ensure the quality and consistency of the input data.

[0205] The implementation of the aggregation strategy mainly includes two core parts: the aggregation decision model and the virtual power plant implementation.

[0206] The aggregate decision model is responsible for determining the output and dispatch of generation units based on the grid’s demand and availability. We employ a constraint-based optimization model to solve this problem.

[0207] Model objective function: maximize the benefits of the virtual power plant while meeting the costs and losses of the power grid.

[0208]

[0209] in:

[0210] P i Represents the power generation of the ith power generation unit in megawatts (MW).

[0211] C i It represents the electricity price of the i generating unit, in units of price per megawatt-hour ($ / MWh).

[0212] Q j Represents the power demand of the jth load unit in megawatts (MW).

[0213] L j represents the electricity loss cost of the j load unit, in units of price per megawatt-hour ($ / MWh).

[0214] Constraints:

[0215] Power balance constraint: This constraint ensures that the total power generation is equal to the total load demand.

[0216]

[0217] Output limit of power generation unit: The output of each power generation unit shall not exceed its maximum capacity

[0218]

[0219] Voltage safety constraint: The voltage of all nodes must be maintained at a minimum value V min and maximum value V max Between voltages.

[0220]

[0221] Dispatch cost constraint: The total power generation cost must not exceed budget B.

[0222]

[0223] In the "real-time adjustment and response" module of the present invention, the system is specially designed to monitor load changes and power generation fluctuations in the power grid in real time, and automatically perform necessary adjustments to ensure the stable operation of the power grid. The module analyzes data from various nodes in real time through an integrated control system, including power generation, load demand, and power grid status. Once a significant load deviation or power generation fluctuation is detected, the system automatically adjusts the output of distributed resources to ensure load balance of the entire power grid. In addition, the system also has a high fault response capability. In the event of a local fault or abnormal situation, the control system can quickly identify and isolate the affected area, while dynamically reconfiguring the surrounding power sources and loads to minimize the impact of the fault on the entire power grid and quickly restore normal power supply. This real-time adjustment and fault response mechanism not only improves the operating efficiency of the power grid, but also significantly enhances the system's adaptability to uncertain events and the reliability of the overall power grid. Through this integrated and automated management, power grid operators can more effectively handle the dynamic changes of the power grid, optimize resource allocation, and maintain the continuity and safety of power grid operation.

[0224] Power generation adjustment:

[0225] Real-time adjustment of the power generation of each distributed power source (such as wind turbines, photovoltaic panels, etc.) to match the current load demand of the power grid. For example, increase or decrease the power generation of a wind farm or photovoltaic power station.

[0226] Backup power start / stop:

[0227] Start or stop backup power sources (such as energy storage systems or backup generators) to balance the grid load. For example, start the energy storage system to discharge during peak load periods and charge during low load periods.

[0228] Power generation unit start and stop control:

[0229] Control the start or stop of each distributed power generation unit, and dynamically adjust the operating status of the power generation unit according to real-time load demand and power generation conditions.

[0230] Power Scheduling:

[0231] Dynamically dispatch the output power of different types of distributed power sources, adjust the power output of different types of power generation sources such as wind power, photovoltaics, and energy storage according to real-time demand, and ensure the stability of the power grid and the balance of supply and demand.

[0232] The implementation of dynamic adjustment mechanisms relies on advanced machine learning techniques that allow the system to monitor the state of the grid in real time and make rapid adjustment decisions. Machine learning plays a central role in this process, especially when dealing with complex problems in the grid caused by fluctuations and changes in renewable energy and distributed power sources.

[0233] 1. Real-time monitoring and data processing.

[0234] Data collection: The system collects key operating data such as voltage, current, frequency and power in real time through sensors and smart meters installed at key nodes of the power grid. These data provide the necessary input for machine learning models to support subsequent predictions and decisions.

[0235] Data stream processing: Using data stream processing technology, real-time data is quickly transmitted and processed to ensure the timeliness and accuracy of the data and provide real-time dynamic input for machine learning models.

[0236] 2. Forecasting and decision-making.

[0237] Building forecasting models: Using time series forecasting models and machine learning algorithms such as neural networks, models are trained to predict short-term and long-term load demand on the grid and the supply of renewable energy. These forecasts take into account seasonal changes, weather conditions, historical data trends, and possible market changes.

[0238] Optimization decision: Based on the output of the predictive model, the optimization algorithm (combining linear and nonlinear programming techniques) calculates the optimal energy distribution and grid configuration strategy. This calculation takes into account the safe operation criteria of the power grid, such as voltage and frequency limits, while trying to minimize operating costs and energy losses.

[0239] 3. Adaptive adjustment.

[0240] Decision execution: The control system receives adjustment suggestions from machine learning models and optimization algorithms and automatically executes related adjustment operations, such as adjusting power generation, switching transmission lines, and reconfiguring loads.

[0241] Feedback loop: The results of the execution are fed back to the machine learning model for retraining and optimization. This continuous learning process enables the model to adapt to new data patterns and grid operating environments, improving the accuracy of predictions and the effectiveness of operations.

[0242] In the above-mentioned embodiment, when adjusting the power grid, the power grid to be analyzed is adjusted according to the target predicted load and power supply information of the power grid to be analyzed, so as to better match the power supply and demand, reduce the problem of voltage instability caused by load overload or insufficient power supply, and help improve the reliability of power grid operation; moreover, the method does not require human intervention, avoids the defect of low adjustment accuracy of the power grid caused by subjective factors and prone to errors in manual planning, thereby improving the adjustment accuracy of the power grid.

[0243] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0244] Based on the same inventive concept, the embodiment of the present application also provides a power grid adjustment device for joint operation of the main distribution network for implementing the power grid adjustment method for joint operation of the main distribution network involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more embodiments of the power grid adjustment device for joint operation of the main distribution network provided below can refer to the limitations of the power grid adjustment method for joint operation of the main distribution network above, and will not be repeated here.

[0245] In an exemplary embodiment, Figure 5 As shown, a power grid adjustment device for joint operation of a main distribution network is provided, comprising: a data acquisition module 501, a feature extraction module 502, a load prediction module 503, an instruction generation module 504 and a grid adjustment module 505, wherein:

[0246] The data acquisition module 501 is used to acquire key operation data of the power grid to be analyzed.

[0247] The feature extraction module 502 is used to perform feature extraction processing on the key operation data to obtain feature vectors corresponding to the key operation data.

[0248] The load prediction module 503 is used to input the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed.

[0249] The instruction generation module 504 is used to obtain the power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed.

[0250] The grid adjustment module 505 is used to adjust the power grid to be analyzed according to the power grid node adjustment instruction.

[0251] In an exemplary embodiment, the data acquisition module 501 is also used to determine the importance of each power grid node; from each power grid node, the power grid nodes whose corresponding importance is greater than the preset importance are screened out as key power grid nodes; and the operating data of the key power grid nodes are used as the key operating data of the power grid to be analyzed.

[0252] In an exemplary embodiment, the instruction generation module 504 is also used to determine the power supply nodes in the power grid to be analyzed; the location information, expected power generation and load demand information within a preset range of the power supply nodes are used as the power supply information of the power grid to be analyzed.

[0253] In an exemplary embodiment, the feature extraction module 502 is also used to preprocess the key operating data to obtain preprocessed key operating data; identify the data type of the preprocessed key operating data; query the correspondence between the data type and the feature extraction model according to the data type of the preprocessed key operating data, and obtain the feature extraction model corresponding to the preprocessed key operating data; input the preprocessed key operating data into the feature extraction model corresponding to the preprocessed key operating data for feature extraction processing, and obtain the feature vector corresponding to the preprocessed key operating data as the feature vector corresponding to the key operating data.

[0254] In an exemplary embodiment, the load prediction module 503 is also used to determine a historical load prediction model associated with a pre-trained current load prediction model; input the feature vector into the pre-trained current load prediction model to obtain a first predicted load of the power grid to be analyzed, and input the feature vector into the historical load prediction model to obtain a second predicted load of the power grid to be analyzed; and fuse the first predicted load and the second predicted load of the power grid to be analyzed to obtain a target predicted load of the power grid to be analyzed.

[0255] In an exemplary embodiment, the power grid adjustment device for joint operation of the main distribution network also includes a model training module, which is used to obtain sample key operating data of the sample power grid; perform feature extraction processing on the sample key operating data to obtain a sample feature vector corresponding to the sample key operating data; input the sample feature vector into a current load prediction model to be trained to obtain a predicted load of the sample power grid; and iteratively train the current load prediction model to be trained according to the difference between the predicted load and the actual load of the sample power grid to obtain a pre-trained current load prediction model.

[0256] Each module in the above-mentioned power grid adjustment device for joint operation of the main distribution network can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0257] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 6 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as target predicted load and power supply information. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a power grid adjustment method for the joint operation of the main distribution network is implemented.

[0258] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0259] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0260] In an exemplary embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0261] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0262] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0263] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0264] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A grid adjustment method for joint operation of main distribution networks, It is characterized in that The method comprises: Obtain key operating data of the power grid to be analyzed; Performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data; Inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed; Acquire power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed; The power grid to be analyzed is adjusted according to the power grid node adjustment instruction.

2. The method according to claim 1, It is characterized in that The power grid to be analyzed includes a plurality of power grid nodes; The step of obtaining key operation data of the power grid to be analyzed includes: Determining the importance of each of the power grid nodes; From each of the power grid nodes, select a power grid node whose corresponding importance is greater than a preset importance as a key power grid node; The operation data of the key power grid nodes are used as the key operation data of the power grid to be analyzed.

3. The method according to claim 1, It is characterized in that The obtaining of power supply information of the power grid to be analyzed includes: Determining power supply nodes in the power grid to be analyzed; The location information, expected power generation and load demand information within a preset range of the power node are all used as the power information of the power grid to be analyzed.

4. The method according to claim 1, It is characterized in that The performing feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data includes: Preprocessing the key operation data to obtain preprocessed key operation data; Identifying the data type of the pre-processed key operation data; According to the data type of the preprocessed key operation data, query the correspondence between the data type and the feature extraction model to obtain the feature extraction model corresponding to the preprocessed key operation data; The preprocessed key operation data is input into a feature extraction model corresponding to the preprocessed key operation data to perform feature extraction processing, and a feature vector corresponding to the preprocessed key operation data is obtained as the feature vector corresponding to the key operation data.

5. The method according to claim 1, It is characterized in that The step of inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed includes: determining a historical load forecasting model associated with the pre-trained current load forecasting model; Inputting the feature vector into the pre-trained current load prediction model to obtain a first predicted load of the power grid to be analyzed, and inputting the feature vector into the historical load prediction model to obtain a second predicted load of the power grid to be analyzed; The first predicted load and the second predicted load of the power grid to be analyzed are fused to obtain a target predicted load of the power grid to be analyzed.

6. The method according to any one of claims 1 to 5, It is characterized in that The pre-trained current load prediction model is trained in the following way: Obtaining sample key operating data of a sample power grid; Performing feature extraction processing on the sample key operation data to obtain a sample feature vector corresponding to the sample key operation data; Inputting the sample feature vector into a current load prediction model to be trained to obtain a predicted load of the sample power grid; According to the difference between the predicted load and the actual load of the sample power grid, the current load prediction model to be trained is iteratively trained to obtain the pre-trained current load prediction model.

7. A power grid adjustment device for joint operation of the main distribution network, It is characterized in that The device comprises: A data acquisition module, used to acquire key operation data of the power grid to be analyzed; A feature extraction module, used to perform feature extraction processing on the key operation data to obtain a feature vector corresponding to the key operation data; A load prediction module, used for inputting the feature vector into a pre-trained current load prediction model to obtain a target predicted load of the power grid to be analyzed; An instruction generation module, used to obtain power supply information of the power grid to be analyzed, and generate a power grid node adjustment instruction for the power grid to be analyzed according to the target predicted load and power supply information of the power grid to be analyzed; A grid adjustment module is used to adjust the power grid to be analyzed according to the power grid node adjustment instruction.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, It is characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images