Method and device for generating a new equipment start-up scheme in a power system, computer device, readable storage medium and program product

Abstract

The application relates to a method and device for generating a new equipment starting scheme in a power system, equipment, a storage medium and a program product, and relates to the technical field of power control. The method can improve the efficiency and accuracy of power grid equipment starting scheme writing. The method comprises the following steps: training starting steps and condition statements according to historical operation data and operation-description data sets to obtain a starting step model and a starting condition model; using a graph attention neural network to perform feature extraction in a power grid topology model to obtain topology features, generating a corpus according to a device name library and a starting operation segmented text library; generating an initial starting scheme of a new equipment in a current power system based on the starting step model and the starting condition model, combining the topology features and the corpus; replacing the descriptions of the same operations in the initial starting scheme with standard descriptions, integrating the initial starting scheme according to artificial intervention supplementary information to obtain a target starting scheme of the new equipment in the current power system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power control, in particular to a method and device for generating a new device start-up scheme in a power system, computer equipment, a computer readable storage medium and a computer program product. BACKGROUND

[0002] A start-up scheme is an important guarantee for the safe operation of a power system, and especially before a new device is put into use, a detailed start-up scheme must be prepared to verify whether the device can operate safely. With the rapid development of the economy and the rapid growth of power demand, a large number of infrastructure projects are put into production every year, which puts forward higher requirements for the efficiency and quality of the start-up scheme preparation.

[0003] The start-up scheme preparation method in the traditional technology mainly relies on a manual operation mode, which is prone to low efficiency and accuracy of scheme preparation due to human factors, and also has the problem of low intelligent level, which makes it difficult to meet the development needs of large-scale power grids and the safe and stable operation target. SUMMARY

[0004] Therefore, it is necessary to provide a method and device for generating a new device start-up scheme in a power system, computer equipment, a computer readable storage medium and a computer program product in view of the above technical problems.

[0005] In a first aspect, the present application provides a method for generating a new device start-up scheme in a power system, comprising:

[0006] Obtaining historical operation data of power systems in multiple different regions, identifying different description contents of the same operation in the different regions in the historical operation data, and obtaining an operation-description data set;

[0007] Training a start-up step model according to the historical operation data and the operation-description data set, and training a start-up condition model according to the historical operation data and the operation-description data set;

[0008] Using a preset graph attention neural network to perform feature extraction in a power grid topology model to obtain topology features, and generating a corpus of start-up condition texts according to a device name library and a start-up operation segmented text library;

[0009] Generating an initial start-up scheme for a new device in a current power system based on the start-up step model and the start-up condition model, in combination with the topology features and the corpus;

[0010] replace the description of the same operation in the initial start-up scheme with a standard description, and integrate the initial start-up scheme according to the artificial intervention supplementary information to obtain a target start-up scheme of a new device in the current power system.

[0011] In one of the embodiments, the training of the start-up step according to the historical operation data and the operation-description data set to obtain a start-up step model comprises:

[0012] The standard operation step in the historical operation data is converted into sequence data containing command identification and command description; the command identification is taken as the input of the start-up step model, and the command description is taken as the output of the start-up step model, the start-up step is trained to obtain an initial start-up step model; a first optimizer and a first loss function are generated based on the operation-description data set, and the initial start-up step model is updated according to the first optimizer and the first loss function to obtain the start-up step model.

[0013] In one of the embodiments, the training of the conditional statement according to the historical operation data and the operation-description data set to obtain a start-up condition model comprises:

[0014] The standard conditional statement in the historical operation data is obtained, the conditional statement is trained based on the standard conditional statement and a full connection layer to obtain an initial start-up condition model; a second optimizer and a second loss function are generated based on the operation-description data set, and the initial start-up condition model is updated according to the second optimizer and the second loss function to obtain the start-up condition model.

[0015] In one of the embodiments, before the replacement of the description of the same operation in the initial start-up scheme with a standard description, the method further comprises:

[0016] The different description contents of the same operation in different regions are analyzed for similarity by using a preset similarity calculation method to obtain a comprehensive similarity coefficient; in the different description contents, the description content with the highest comprehensive similarity coefficient is identified as the standard description.

[0017] In one of the embodiments, before the integration of the initial start-up scheme according to the artificial intervention supplementary information, the method further comprises:

[0018] An artificial intervention interface is arranged, which is used to adjust the command sequence, edit the command content, and add a safety check step; a current command sequence, a current command content, and a current safety check step obtained through the artificial intervention interface are acquired, and the artificial intervention supplementary information is generated according to the current command sequence, the current command content, and the current safety check step.

[0019] In one of the embodiments, the initial start-up scheme of the new device in the current power system is generated based on the start-up step model and the start-up condition model in combination with the topology feature and the corpus, which includes:

[0020] The start-up step and the start-up condition of the new device are obtained by prediction based on the start-up step model and the start-up condition model; the start-up step and the start-up condition are adjusted in combination with the topology feature and the corpus, and the initial start-up scheme is generated according to the adjusted start-up step and start-up condition.

[0021] In the second aspect, the application further provides a device for generating a start-up scheme of a new device in a power system, which includes:

[0022] A data recognition module is configured to acquire historical operation data of power systems in multiple different regions, and identify different description contents of the same operation in the historical operation data to obtain an operation-description data set;

[0023] A model training module is configured to train a start-up step based on the historical operation data and the operation-description data set to obtain a start-up step model, and train a condition statement based on the historical operation data and the operation-description data set to obtain a start-up condition model;

[0024] A feature extraction module is configured to perform feature extraction in a power grid topology model by using a preset graph attention neural network to obtain a topology feature, and generate a corpus of start-up condition texts based on a device name library and a start-up operation segmented text library;

[0025] A scheme generation module is configured to generate an initial start-up scheme of a new device in a current power system based on the start-up step model and the start-up condition model in combination with the topology feature and the corpus;

[0026] A scheme integration module is configured to replace the description of the same operation in the initial start-up scheme with a standard description, and integrate the initial start-up scheme according to artificial intervention supplementary information to obtain a target start-up scheme of the new device in the current power system.

[0027] In a third aspect, the present application also provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0028] The historical operation data of power systems in multiple different regions are acquired, different description contents of the same operation in the different regions are identified in the historical operation data to obtain an operation-description data set; the start step is trained according to the historical operation data and the operation-description data set to obtain a start step model, and the conditional statement is trained according to the historical operation data and the operation-description data set to obtain a start condition model; the preset graph attention neural network is used to perform feature extraction in a power grid topology model to obtain topology features, and a start condition text corpus is generated according to a device name library and a start operation segmented text library; the initial start scheme of a new device in a current power system is generated based on the start step model and the start condition model, in combination with the topology features and the corpus; the description of the same operation in the initial start scheme is replaced with a standard description, and the initial start scheme is integrated according to artificial intervention supplementary information to obtain the target start scheme of the new device in the current power system.

[0029] In a fourth aspect, the present application also provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the following steps:

[0030] The historical operation data of power systems in multiple different regions are acquired, different description contents of the same operation in the different regions are identified in the historical operation data to obtain an operation-description data set; the start step is trained according to the historical operation data and the operation-description data set to obtain a start step model, and the conditional statement is trained according to the historical operation data and the operation-description data set to obtain a start condition model; the preset graph attention neural network is used to perform feature extraction in a power grid topology model to obtain topology features, and a start condition text corpus is generated according to a device name library and a start operation segmented text library; the initial start scheme of a new device in a current power system is generated based on the start step model and the start condition model, in combination with the topology features and the corpus; the description of the same operation in the initial start scheme is replaced with a standard description, and the initial start scheme is integrated according to artificial intervention supplementary information to obtain the target start scheme of the new device in the current power system.

[0031] In a fifth aspect, the present application also provides a computer program product comprising a computer program, wherein the computer program is executed by a processor to implement the following steps:

[0032] The historical operation data of a plurality of different regional power systems is acquired, in which the different regions are identified for different descriptions of the same operation to obtain an operation-description data set; a start step model is obtained by training the start step according to the historical operation data and the operation-description data set, and a start condition model is obtained by training the conditional statement according to the historical operation data and the operation-description data set; a topology feature is obtained by performing feature extraction in a power grid topology model by using a preset graph attention neural network, and a corpus of start condition texts is generated according to a device name library and a start operation segmented text library; an initial start scheme of a new device in a current power system is generated based on the start step model and the start condition model, in combination with the topology feature and the corpus; a description of the same operation in the initial start scheme is replaced with a standard description, and the initial start scheme is integrated according to artificial intervention supplementary information to obtain a target start scheme of the new device in the current power system.

[0033] The method, device, computer device, computer readable storage medium and computer program product for generating a start scheme of a new device in a power system described above automatically learn the mode of start steps and conditional statements from historical operation data and operation-description data by using artificial intelligence technology, perform feature extraction in a power grid topology model by using a preset graph attention neural network to obtain a topology feature, generate a corpus of start condition texts according to a device name library and a start operation segmented text library, generate an initial start scheme of a new device in a current power system based on a start step model and a start condition model in combination with a topology feature and a corpus, and integrate the initial start scheme according to artificial intervention supplementary information to obtain a target start scheme of the new device in the current power system. The entire technical solution is summarized as an intelligent compilation method for a start scheme of a new device in a power grid, which improves the efficiency and accuracy of the compilation of a start scheme of a power grid device by designing the compilation, review and operation process of a new device, avoids manual operation, effectively reduces the workload of an operator, improves the intelligent level, and meets the development needs of a large-scale power grid and the operation target of safety and stability. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related art, the drawings needed to be used in the description of the embodiments of the present application or the related art will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other related drawings can be obtained by those skilled in the art without creative labor.

[0035] Figure 1An application environment diagram of the method for generating a new device startup scheme in a power system in an embodiment;

[0036] Figure 2 A flowchart of the method for generating a new device startup scheme in a power system in an embodiment;

[0037] Figure 3 A flowchart of the training step in an embodiment;

[0038] Figure 4 A flowchart of the method for generating a new device startup scheme in a power system in an embodiment;

[0039] Figure 5 A structural block diagram of the device for generating a new device startup scheme in a power system in an embodiment;

[0040] Figure 6 An internal structure diagram of a computer device in an embodiment. DETAILED DESCRIPTION

[0041] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0042] The method for generating a new device startup scheme in a power system provided by the embodiments of the present application can be applied in an application environment as shown in the accompanying drawings. Figure 1 In the application environment, a terminal communicates with a server through a network. A data storage system can store data required to be processed by the server. The data storage system can be integrated on the server, or placed on a cloud or other network server.

[0043] Specifically, the method for generating a new device startup scheme in a power system provided by the embodiments of the present application can be executed by the server.

[0044] Exemplarily, the server acquires historical operation data of power systems in multiple different regions, identifies different descriptions of the same operation in different regions in the historical operation data, and obtains an operation-description data set; the server trains a start step model according to the historical operation data and the operation-description data set, and trains a start condition model according to the historical operation data and the operation-description data set; the server performs feature extraction in a power grid topology model by using a preset graph attention neural network, obtains topology features, and generates a corpus of start condition texts according to a device name library and a start operation segmented text library; the server generates an initial start scheme of a new device in a current power system based on the start step model and the start condition model, in combination with the topology features and the corpus; and the server replaces descriptions of the same operation in the initial start scheme with standard descriptions, and integrates the initial start scheme according to artificial intervention supplementary information to obtain a target start scheme of the new device in the current power system.

[0045] In the application environment as shown in Figure 1 , the terminal can be, but is not limited to, various personal computers, notebook computers, smart phones and tablet computers. The server can be implemented by an independent server or a server cluster composed of multiple servers.

[0046] In one embodiment, as shown in Figure 2 , a method for generating a start scheme of a new device in a power system is provided. Taking the server in Figure 1 as an example, the method includes the following steps:

[0047] Step S201, acquiring historical operation data of power systems in multiple different regions, identifying different descriptions of the same operation in different regions in the historical operation data, and obtaining an operation-description data set.

[0048] The historical operation data includes, but is not limited to, historical commissioning schemes, operation manuals and power grid operation records.

[0049] Specifically, the server collects historical commissioning schemes, operation manuals and power grid operation records of power systems in different regions, identifies different descriptions of the same operation in different regions, obtains an operation-description data set, and stores the operation-description data set into a database.

[0050] Step S202, training a start step model according to the historical operation data and the operation-description data set, and training a start condition model according to the historical operation data and the operation-description data set.

[0051] Specifically, the server, in response to the model training instruction, trains the start-up steps and the conditional statements according to the historical operation data and the operation-description data set, and by using the sequence and non-sequence deep learning model, to obtain a start-up step model and a start-up condition model.

[0052] In step S203, the preset graph attention neural network is used to perform feature extraction in the power grid topology model to obtain topology features, and a corpus of start-up condition texts is generated according to the device name library and the start-up operation segmented text library.

[0053] The graph attention neural network is a neural network architecture for graph data, which combines the advantages of graph neural networks and attention mechanisms.

[0054] Specifically, the server uses the preset graph attention neural network to perform feature extraction in the power grid topology model to obtain topology features, and forms a corpus of start-up condition texts according to the device name library and the start-up operation segmented text library, which will be used for training and prediction of the deep learning model.

[0055] In step S204, based on the start-up step model and the start-up condition model, the topology features and the corpus are combined to generate an initial start-up scheme for the new device in the current power system.

[0056] The topology features can represent the connection structure between devices in the power system.

[0057] Specifically, based on the start-up step model and the start-up condition model, the server combines the extracted topology features and the corpus, and the model can automatically predict the steps and conditions for the new device to be put into operation, and then integrates the predicted start-up steps and conditions to generate a complete start-up scheme, which includes the sequence of device start-up, the conditions and safety measures that need to be met before each step, etc.

[0058] In step S205, the descriptions of the same operations in the initial start-up scheme are replaced with standard descriptions, and the initial start-up scheme is integrated according to the artificial intervention supplementary information to obtain the target start-up scheme for the new device in the current power system.

[0059] Specifically, the server performs similarity analysis on different descriptions of the same operation in different regions, and then generates a standard description based on the obtained comprehensive similarity coefficient, replaces the descriptions of the same operation in the initial start-up scheme with the standard description, and integrates the initial start-up scheme according to the artificial intervention supplementary information to obtain the target start-up scheme for the new device in the current power system.

[0060] In the method for generating a new device startup scheme in the power system, the mode of startup steps and conditional statements is learned from historical operation data and operation-description data automatically by using artificial intelligence technology, feature extraction is performed in a power grid topology model by using a preset graph attention neural network to obtain topology features, a startup condition text corpus is generated according to a device name library and a startup operation segmented text library, an initial startup scheme of a new device in the current power system is generated based on a startup step model and a startup condition model in combination with the topology features and the corpus, and finally the initial startup scheme is integrated according to artificial intervention supplementary information to obtain a target startup scheme of the new device in the current power system. The entire technical solution proposes an intelligent compilation method for a power grid new device startup scheme, improves the efficiency and accuracy of power grid device startup scheme compilation by designing the compilation, review and operation process of the new device, avoids manual operation, effectively reduces the workload of operators, improves the intelligent level, and meets the development needs of large-scale power grids and the safe and stable operation target.

[0061] In one embodiment, as shown in Figure 3 In step S202, the startup step is trained according to the historical operation data and the operation-description data set to obtain a startup step model, which includes the following steps:

[0062] In step S301, the standard operation steps in the historical operation data are obtained, and the standard operation steps are converted into sequence data containing command identifiers and command descriptions.

[0063] In step S302, the command identifiers are taken as the input of the startup step model, and the command descriptions are taken as the output of the startup step model, the startup step is trained to obtain an initial startup step model.

[0064] In step S303, a first optimizer and a first loss function are generated based on the operation-description data set, and the initial startup step model is updated based on the first optimizer and the first loss function to obtain the startup step model.

[0065] Specifically, the startup step model training process is as follows:

[0066] S1: Collect the standard operation steps in the historical operation scheme and convert them into sequence data, each operation step corresponds to a command id and a command description, and these data will be used as the input and output of the model;

[0067] S2: Use LSTM (Long Short-Term Memory) units to build the model, as LSTM can handle long-term dependencies in sequential data, the input of the model is the command description of the previous one or more time steps, and the output is the command description of the current time step;

[0068] S3: Train the model using historical data, update the model parameters through an optimizer (such as Adam) and a loss function (such as sparse categorical cross-entropy), during the training process, the model learns how to predict the next operation step according to the previous steps;

[0069] S4: Evaluate the performance of the model using the validation set, ensure that the model can accurately predict the operation step.

[0070] In one embodiment, the above step S202, according to the historical operation data and operation-description data set to train the conditional statement, get the start condition model, including the following steps:

[0071] Get the standard conditional statement in the historical operation data, train the conditional statement based on the standard conditional statement and the fully connected layer, get the initial start condition model; based on the operation-description data set, generate the second optimizer and the second loss function, and update the initial start condition model according to the second optimizer and the second loss function, get the start condition model.

[0072] Specifically, the start condition model training process is as follows:

[0073] S1: Collect the conditional statements in the historical operation scheme, these statements describe the conditions that must be met before a specific operation;

[0074] S2: Use Dense (Dense) to build the model, as the conditional statement does not involve time dependence of sequential data;

[0075] S3: Train the model using historical data, update the model parameters through an optimizer (such as SGD) and a loss function (such as categorical cross-entropy), during the training process, the model learns how to identify the conditions that meet the specific operation;

[0076] S4: Evaluate the performance of the model using the validation set, ensure that the model can accurately identify the conditional statement.

[0077] Through the combination of the start step model and the start condition model, a new device operation scheme can be automatically generated, where the start step model is responsible for generating ordered operation steps, and the start condition model is responsible for identifying the conditions required for these steps, such combination ensures that the generated scheme not only conforms to the operation logic, but also meets the actual operation constraints.

[0078] The above two models are used to automatically generate the start-up scheme information, specifically as follows:

[0079] The graph attention neural network (GAT) is used to extract features from the power grid topology model, including the connection relationship between devices and the topology structure of the power grid.

[0080] The start-up step and start-up condition text corpus are formed according to the device name library and the start-up operation segmented text library. This corpus will be used for training and prediction of the deep learning model.

[0081] Using the trained start-up step model and start-up condition model, combined with the extracted topology features and the corpus, the model can automatically predict the steps and conditions of the new device commissioning.

[0082] The predicted start-up steps and conditions are integrated to generate a complete start-up scheme. This scheme includes the sequence of device start-up, the conditions that need to be met before each step, safety measures, etc.

[0083] In one embodiment, in step S204, based on the start-up step model and the start-up condition model, combined with the topology features and the corpus, an initial start-up scheme for the new device in the current power system is generated, specifically including the following steps:

[0084] Based on the start-up step model and the start-up condition model, the commissioning steps and conditions of the new device are predicted; combined with the topology features and the corpus, the commissioning steps and conditions are adjusted, and the initial start-up scheme is generated according to the adjusted commissioning steps and conditions.

[0085] Specifically, based on similarity analysis, the standard description is generated, and then the standard description is used to replace the description of the same operation in the start-up scheme; the generation steps of the standard description are as follows:

[0086] The similarity analysis is performed on different descriptions of the same operation in different regions, using methods such as cosine similarity or Jaccard similarity coefficient, etc. The purpose is to generate the similarity q of a certain description to other descriptions in the description of the same operation, and then obtain the comprehensive similarity coefficient of each description to other descriptions The calculation process of the comprehensive similarity coefficient is as follows:

[0087]

[0088] In the above formula, q i represents the similarity of the current description to the i-th description, and n is the total number of collected descriptions.

[0089] In this embodiment, the similarity can be obtained in the manner of cosine similarity or Jaccard similarity coefficient, and the specific manners are as follows.

[0090] 1. Cosine similarity is used for calculation:

[0091]

[0092] In the above formula, is the dot product of vectors A and B, is the Euclidean norm (or length) of vector A; is the Euclidean norm (or length) of vector B, the range of cosine similarity is from -1 to 1, the value of 1 indicates that the vectors are completely the same, the value of -1 indicates that the vectors are completely opposite, and the value of 0 indicates that the vectors are orthogonal (no correlation).

[0093] 2. Jaccard similarity coefficient is used for calculation:

[0094]

[0095] In the above formula, |A∩B| is the size of the intersection of sets A and B, that is, the number of common elements in the two sets; |A∪B| is the size of the union of sets A and B, that is, the number of all elements in the two sets, and the repeated elements are counted only once; the range of Jaccard similarity coefficient is from 0 to 1, the value of 1 indicates that the two sets are completely the same, and the value of 0 indicates that the two sets have no intersection.

[0096] Through clustering analysis and similarity calculation, common points are extracted, and a comprehensive and suitable standard description for different regions is generated, which is the description with the highest comprehensive similarity coefficient.

[0097] In this embodiment, cosine similarity and Jaccard similarity coefficient can be used to compare the similarity between documents or sentences. Cosine similarity is not sensitive to the length of the document, while Jaccard similarity coefficient is sensitive to the size of the set. According to different application scenarios, appropriate similarity calculation methods can be selected, which can be used independently or the average value of the normalized results of the two can be used.

[0098] In one of the embodiments, before the initial starting scheme is integrated according to the human intervention supplementary information, the method of the application further includes the following steps:

[0099] Setting a human intervention interface; the human intervention interface is used to adjust the command sequence, edit the command content, and add a safety check step; obtaining the current command sequence, the current command content, and the current safety check step obtained through the human intervention interface, and generating human intervention supplementary information according to the current command sequence, the current command content, and the current safety check step.

[0100] Specifically, the server sets up a manual intervention interface for manually adjusting the order of commands and editing the content of commands to ensure the rationality and safety of the scheme, i.e. the standard description generated automatically is reviewed and adjusted by professionals. Manual intervention is a key step based on the automatically generated start-up scheme to ensure the rationality, safety and adaptability of the scheme. Here are a few specific cases of manual intervention and examples of results:

[0101] Case one: adjustment of command order, for example:

[0102] Automatically generated scheme:

[0103] 1. Check the power connection of device A.

[0104] 2. Start device B.

[0105] 3. Verify the operating status of device A.

[0106] In actual operation, it may be necessary to verify the operating status of device A first to ensure that it is in a safe state before starting device B to avoid potential safety risks; therefore, the scheme after manual intervention is as follows:

[0107] 1. Check the power connection of device A.

[0108] 2. Verify the operating status of device A.

[0109] 3. Start device B.

[0110] Case two: editing of command content, for example:

[0111] Automatically generated scheme:

[0112] 1. Turn on the power switch of device C.

[0113] Since the automatically generated command may be too general, manual intervention can provide more detailed operation steps to ensure that the operator can accurately perform and verify the operation results. Therefore, after manual intervention, the following can be supplemented:

[0114] 1. Turn on the main power switch of device C and check if the power indicator light is on.

[0115] Case three: adding safety check steps, for example:

[0116] Automatically generated scheme:

[0117] 1. Connect the control line of device D.

[0118] 2. Start device D.

[0119] To ensure the safety of the operation, it is necessary to add a check of the grounding condition before starting the device, which is an important safety step that may be missed in the automatic generation scheme. Therefore, after manual intervention:

[0120] 1. Connect the control line of device D.

[0121] 2. Check the grounding condition of device D.

[0122] 3. Start device D.

[0123] Case four: adapt to specific regional regulations, for example:

[0124] Automatic generation scheme:

[0125] 1. Start device E according to the standard procedure.

[0126] Since different regions may have different operation regulations and safety standards, manual intervention can ensure that the scheme complies with the laws and regulations of the specific region. Therefore, after manual intervention:

[0127] 1. Adjust the start-up parameters of device E according to the special regulations of region X.

[0128] Case five: handle abnormal situations, for example:

[0129] Automatic generation scheme:

[0130] 1. Start device F according to the normal flow.

[0131] Since the automatically generated scheme may not take into account all the abnormal situations that may be encountered in actual operation, manual intervention can adjust the scheme according to the actual situation to ensure the safety and effectiveness of the operation. Therefore, after manual intervention:

[0132] 1. Before starting device F, check the weather conditions, and if there is a thunderstorm, delay the start.

[0133] Through these examples of manual intervention in different situations, we can see that manual intervention is mainly to make up for the shortcomings of the automatically generated scheme, such as lack of details, neglect of safety checks, and non-compliance with specific regional regulations. Manual intervention can improve the practicality and reliability of the start-up scheme, ensuring the smooth operation of the new device.

[0134] Then, the server will integrate the automatically generated information and the information supplemented by manual intervention to generate a perfect new device start-up scheme, that is, based on the automatically generated information, if it is determined to introduce manual intervention, replace the manually intervened part with the originally automatically generated part, and finally implement these schemes in actual operation, and further adjust and perfect the rule base according to feedback.

[0135] In one embodiment, asFigure 4 As shown, a method for generating a new device startup scheme in a power system in a specific embodiment is provided, which specifically includes the following steps:

[0136] In step S401, historical operation data of power systems in multiple different regions is obtained, and different descriptions of the same operation in different regions are identified in the historical operation data to obtain an operation-description data set.

[0137] In step S402, standard operation steps in the historical operation data are obtained, and the standard operation steps are converted into sequence data containing command identifiers and command descriptions; the command identifiers are taken as inputs of a startup step model, and the command descriptions are taken as outputs of the startup step model, the startup steps are trained to obtain an initial startup step model; based on the operation-description data set, a first optimizer and a first loss function are generated, and the initial startup step model is updated according to the first optimizer and the first loss function to obtain a startup step model.

[0138] In step S403, standard condition sentences in the historical operation data are obtained, and the condition sentences are trained based on the standard condition sentences and a full connection layer to obtain an initial startup condition model; based on the operation-description data set, a second optimizer and a second loss function are generated, and the initial startup condition model is updated according to the second optimizer and the second loss function to obtain a startup condition model.

[0139] In step S404, a preset graph attention neural network is used to extract features in a power grid topology model to obtain topology features, and a startup condition text corpus is generated based on a device name library and a startup operation segmented text library; the startup step model and the startup condition model are used for prediction to obtain commissioning steps and commissioning conditions of a new device; the commissioning steps and the commissioning conditions are adjusted in combination with the topology features and the corpus, and an initial startup scheme is generated according to the adjusted commissioning steps and the commissioning conditions.

[0140] In step S405, a preset similarity calculation method is used to analyze the similarity of different descriptions of the same operation in different regions to obtain a comprehensive similarity coefficient; in the different descriptions, the description with the highest comprehensive similarity coefficient is identified as a standard description.

[0141] In step S406, a manual intervention interface is set; the manual intervention interface is used to adjust the order of commands, edit the content of commands, and add safety check steps; the current order of commands, the current content of commands, and the current safety check steps obtained through the manual intervention interface are obtained, and manual intervention supplementary information is generated according to the current order of commands, the current content of commands, and the current safety check steps.

[0142] Step S407, replace the description of the same operation in the initial start-up scheme with the standard description, and integrate the initial start-up scheme according to the artificial intervention supplementary information to obtain the target start-up scheme of the new device in the current power system.

[0143] The beneficial effects brought by the above embodiments are as follows:

[0144] The present application automatically learns the mode of the start-up steps and the conditional statements from the historical data through deep learning technology, generates a start-up step model and a start-up condition model, and thus automatically compiles the start-up scheme. This method significantly improves the efficiency and accuracy of the power grid device start-up scheme compilation, reduces human errors, and ensures the consistency and reliability of the scheme. Moreover, the present application constructs an intelligent compilation system for the start-up scheme of the new device in the power grid, adopts a B / S architecture system, designs the compilation, review process, and commissioning and checking functions of the new device commissioning, effectively improves the intelligent level, and meets the needs of the development of large power grids. In summary, the implementation of the present application can effectively reduce the workload of the operating personnel, improve the risk management and control of the new device commissioning, and plays an important role in ensuring the safe and stable operation of the power system. Through the intelligent compilation system, potential risk points can be identified and avoided in advance to ensure that each step of the new device commissioning is strictly checked and reviewed. This systematic risk management capability is unmatched by traditional manual compilation methods, and it improves the safety and stability of the power system through intelligent means.

[0145] It should be understood that although each step in the flowchart involved in each embodiment as described above is displayed in sequence according to the arrow, these steps are not necessarily executed in the order indicated by the arrow. Unless otherwise specified herein, there is no strict order limitation for the execution of these steps, and these steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in each embodiment as described above can include multiple steps or stages, which 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 sequential, but can be executed in rotation or alternation with at least part of other steps or steps or stages in other steps.

[0146] Based on the same inventive concept, the present application also provides a device for generating a start-up scheme for a new device in a power system, which is used to implement the method for generating a start-up scheme for a new device in a power system as described above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme described in the above method, and therefore the specific limitations in one or more device embodiments for generating a start-up scheme for a new device in a power system provided below can refer to the limitations of the method for generating a start-up scheme for a new device in a power system described above, which will not be repeated here.

[0147] In one example embodiment, as shown in Figure 5 A device startup scheme generation apparatus in a power system is provided, comprising:

[0148] A data recognition module 501 is configured to obtain historical operation data of power systems in multiple different regions, and identify different descriptions of the same operation in the historical operation data to obtain an operation-description data set.

[0149] A model training module 502 is configured to train startup steps based on the historical operation data and the operation-description data set to obtain a startup step model, and train conditional statements based on the historical operation data and the operation-description data set to obtain a startup condition model.

[0150] A feature extraction module 503 is configured to perform feature extraction in a power grid topology model using a preset graph attention neural network to obtain topology features, and generate a corpus of startup condition texts based on a device name library and a startup operation segmented text library.

[0151] A scheme generation module 504 is configured to generate an initial startup scheme for a new device in a current power system based on the startup step model and the startup condition model, in combination with the topology features and the corpus.

[0152] A scheme integration module 505 is configured to replace descriptions of the same operation in the initial startup scheme with standard descriptions, and integrate the initial startup scheme based on artificial intervention supplementary information to obtain a target startup scheme for the new device in the current power system.

[0153] In one embodiment, the model training module 502 is further configured to obtain standard operation steps in the historical operation data, convert the standard operation steps into sequence data containing command identifiers and command descriptions, train the startup steps by taking the command identifiers as inputs of the startup step model and the command descriptions as outputs of the startup step model to obtain an initial startup step model, generate a first optimizer and a first loss function based on the operation-description data set, and update the initial startup step model based on the first optimizer and the first loss function to obtain the startup step model.

[0154] In one embodiment, the model training module 502 is further configured to obtain standard conditional statements in the historical operation data, train the conditional statements based on the standard conditional statements and a full connection layer to obtain an initial startup condition model, generate a second optimizer and a second loss function based on the operation-description data set, and update the initial startup condition model based on the second optimizer and the second loss function to obtain the startup condition model.

[0155] In one embodiment, the device for generating a new equipment start-up scheme in a power system further comprises a similarity analysis module configured to perform similarity analysis on different descriptions of the same operation in different regions by using a preset similarity calculation method to obtain a comprehensive similarity coefficient; and identify the description with the highest comprehensive similarity coefficient in the different descriptions as a standard description.

[0156] In one embodiment, the device for generating a new equipment start-up scheme in a power system further comprises an information supplement module configured to set up a manual intervention interface; the manual intervention interface is configured to adjust the order of commands, edit the content of the commands, and add a safety check step; obtain the current order of commands, the current content of the commands, and the current safety check step through the manual intervention interface, and generate manual intervention supplement information based on the current order of commands, the current content of the commands, and the current safety check step.

[0157] In one embodiment, the scheme generation module 504 is configured to perform prediction based on the start-up step model and the start-up condition model to obtain the commissioning step and the commissioning condition of the new equipment; adjust the commissioning step and the commissioning condition in combination with the topological features and the corpus, and generate an initial start-up scheme based on the adjusted commissioning step and the commissioning condition.

[0158] The modules in the device for generating a new equipment start-up scheme in a power system described above can be implemented in whole or in part by software, hardware, or a combination thereof. The modules described above can be embedded in or independent of a processor in a computer device in hardware form, or stored in a memory in a computer device in software form, so as to be called and executed by a processor to perform the operations corresponding to the modules.

[0159] In one exemplary embodiment, a computer device is provided, which can be a server, and an internal structure diagram of the computer device can be as shown in Figure 6 The computer device includes a processor, a memory, an input / output interface (I / O), and a communication interface. 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. The processor of the computer device is configured 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 operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is configured to store historical operation data of a power system. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with external terminals through a network connection. The computer program is executed by the processor to implement a method for generating a new equipment start-up scheme in a power system.

[0160] Those skilled in the art can understand that, Figure 6 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0161] In one embodiment, a computer device is also provided, including a memory and a processor, the memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0162] In one embodiment, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0163] In one embodiment, a computer program product is provided, which includes a computer program, and the computer program is executed by a processor to implement the steps in the above method embodiments.

[0164] 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 for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0165] 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. The computer program can be stored in a non-volatile computer readable storage medium, and when executed, can include the processes of the above-mentioned embodiment methods. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. The non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, 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. The volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, the RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., without being limited thereto.

[0166] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present application.

[0167] The above-described embodiments are merely illustrative of several embodiments of the present application, and the description is relatively specific and detailed, but should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A method for generating a startup scheme for new equipment in a power system, characterized in that, The method includes: Historical operation data of power systems in multiple different regions are acquired, and different descriptions of the same operation in different regions are identified in the historical operation data to obtain an operation-description dataset. The system acquires standard operation steps from the historical operation data and converts these steps into sequence data containing command identifiers and command descriptions. It then uses the command identifiers as input to a startup step model and the command descriptions as output to train the startup step model, obtaining an initial startup step model. Based on the operation-description dataset, it generates a first optimizer and a first loss function, and updates the initial startup step model according to the first optimizer and the first loss function, obtaining the startup step model. Additionally, it acquires standard conditional statements from the historical operation data, trains these conditional statements using the standard conditional statements and fully connected layers, obtaining an initial startup condition model. Based on the operation-description dataset, it generates a second optimizer and a second loss function, and updates the initial startup condition model according to the second optimizer and the second loss function, obtaining a startup condition model. Using a pre-defined graph attention neural network, feature extraction is performed in the power grid topology model to obtain topological features, and a corpus of startup condition texts is generated based on the equipment name library and the startup operation word segmentation text library. Based on the startup step model and the startup condition model, combined with the topological features and the corpus, an initial startup scheme for new equipment in the current power system is generated. The initial startup scheme is modified by replacing the descriptions of identical operations with standard descriptions and integrating the initial startup scheme with supplementary information based on manual intervention, thereby obtaining the target startup scheme for new equipment in the current power system.

2. The method according to claim 1, characterized in that, Before replacing the description of the same operation in the initial startup scheme with a standard description, the method further includes: Using a preset similarity calculation method, a similarity analysis is performed on different descriptions of the same operation in different regions to obtain a comprehensive similarity coefficient; Among the different descriptive contents, the descriptive content with the highest comprehensive similarity coefficient is identified and used as the standard description.

3. The method according to claim 2, characterized in that, Before integrating the initial startup scheme based on supplementary information obtained through human intervention, the process also includes: A manual intervention interface is set up; the manual intervention interface is used to adjust the command sequence, edit the command content, and add security check steps; Obtain the current command sequence, current command content, and current security check steps obtained through the manual intervention interface, and generate the supplementary information for manual intervention based on the current command sequence, current command content, and current security check steps.

4. The method according to any one of claims 1 to 3, characterized in that, The process of generating an initial startup scheme for new equipment in the current power system based on the startup step model and the startup condition model, combined with the topological features and the corpus, includes: Based on the startup step model and the startup condition model, the commissioning steps and commissioning conditions of the new equipment are obtained through prediction. By combining the topological features and the corpus, the commissioning steps and conditions are adjusted, and the initial startup scheme is generated based on the adjusted commissioning steps and conditions.

5. A device for generating a startup scheme for new equipment in a power system, characterized in that, The device includes: The data identification module is used to acquire historical operation data of power systems in multiple different regions, identify different descriptions of the same operation in the historical operation data, and obtain an operation-description dataset. The model training module is used to acquire standard operation steps from the historical running data, convert the standard operation steps into sequence data containing command identifiers and command descriptions; train the startup steps using the command identifiers as input to the startup step model and the command descriptions as output to obtain an initial startup step model; generate a first optimizer and a first loss function based on the operation-description dataset, and update the initial startup step model according to the first optimizer and the first loss function to obtain the startup step model; acquire standard condition statements from the historical running data, train the condition statements based on the standard condition statements and fully connected layers to obtain an initial startup condition model; generate a second optimizer and a second loss function based on the operation-description dataset, and update the initial startup condition model according to the second optimizer and the second loss function to obtain a startup condition model; The feature extraction module is used to extract features from the power grid topology model using a preset graph attention neural network to obtain topological features, and to generate a corpus of startup condition texts based on the equipment name library and the startup operation word segmentation text library. The scheme generation module is used to generate an initial startup scheme for new equipment in the current power system based on the startup step model and the startup condition model, combined with the topology features and the corpus. The scheme integration module is used to replace the descriptions of the same operations in the initial startup scheme with standard descriptions, and to integrate the initial startup scheme based on supplementary information from manual intervention, so as to obtain the target startup scheme for the new equipment in the current power system.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images