Power grid operation ticket processing method and system based on artificial intelligence

Abstract

The invention discloses a power grid operation order processing method and system based on artificial intelligence. The method comprises the steps of performing key field information extraction on at least one power grid maintenance plan application data uploaded by a user through an RPA intelligent agent to generate an application form information table including at least one piece of application form information; according to key field information of the application form information, classifying the application form information in the application form information table to obtain at least one application form set; according to the operation task corresponding to each application form set and the current state and the target state of the power grid equipment in the target power system, performing execution strategy prediction through a reasoning model, and obtaining the strategy information of the security measure required to be executed by each application form set to generate a power grid operation ticket of the corresponding application form set, power grid equipment in a target power system is indicated to be adjusted to a target state; according to the invention, the power grid operation order is automatically generated, the generation efficiency and accuracy of the operation order are improved, and stable operation of a power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of power distribution network dispatching management, and in particular to a power grid operation ticket processing method and system based on artificial intelligence. Background Art

[0002] In power systems, grid equipment requires regular maintenance, dispatching, testing, replacement, addition, and emergency repairs. Operation tickets are typically used to guide and regulate grid equipment operators in performing distribution network dispatching operations. Accurate descriptions and completion of operation tickets are directly related to the stable operation of the power system.

[0003] At present, operation tickets are filled out manually by staff according to work tasks. The filling speed of manually filled operation tickets is slow and it is easy to make mistakes due to human factors, which affects the stable operation of the power system and easily causes safety accidents of power grid equipment. Summary of the Invention

[0004] In response to the problems existing in the existing technology, an embodiment of the present invention provides an artificial intelligence-based power grid operation ticket processing method and system, which can realize the automatic generation of power grid operation tickets, effectively improve the efficiency and accuracy of operation ticket generation, thereby ensuring the stable operation of the power system and reducing the occurrence of power grid equipment safety accidents.

[0005] In a first aspect, an embodiment of the present invention provides a method for processing power grid operation tickets based on artificial intelligence, comprising: Extract key field information from at least one power grid maintenance plan application data uploaded by the user through an RPA agent, and generate at least one application form information based on the key field information of at least one power grid maintenance plan application data to form an application form information table; Classifying the application form information in the application form information table according to key field information corresponding to each application form information in the application form information table to obtain at least one application form set; wherein one application form set corresponds to the same job type and operation task; Based on the operation tasks corresponding to each of the application form sets, the current state and target state of the power grid equipment in the target power system, an execution strategy prediction is performed through an inference model to obtain strategy information of the security measures required to be executed for each of the application form sets; Based on the policy information of each of the application form sets, a power grid operation ticket for the corresponding application form set is generated; wherein the power operation ticket is used to instruct the target power system to adjust the operating status of at least one power grid node according to the corresponding policy information, so as to adjust the power grid equipment in the target power system to the target state.

[0006] As an improvement to the above solution, the RPA agent extracts key field information from at least one power grid maintenance plan application data uploaded by the user, including: Performing semantic analysis on at least one power grid maintenance plan application data uploaded by the user through an RPA agent built based on a recurrent neural network, and extracting key field information of at least one of the power grid maintenance plan application data; Among them, the key field information includes: application number, operation type, operation time, operation task, and power outage scope.

[0007] As an improvement to the above solution, the application form information in the application form information table is classified according to the key field information corresponding to each application form information in the application form information table to obtain at least one application form set, including: Extracting a first key field indicating a job type from each application form information in the application form information table; Classifying the application form information in the application form information table according to the first key field of the application form information to obtain at least one application form information sub-table; wherein each application form information sub-table corresponds to a job type; Extracting the second key field indicating the operation task and the power outage scope from each application form information in each application form information sub-table; The application form information of the corresponding application form information subtable is classified according to the second key field of the application form information to obtain at least one application form set; wherein each application form set corresponds to the same operation task under the same power outage scope.

[0008] As an improvement to the above solution, the execution strategy prediction is performed through the inference model based on the operation tasks corresponding to each application set, the current state and target state of the power grid equipment in the target power system, and the strategy information of the security measures required to be executed for each application set is obtained, including: Obtain the grid topology model, power flow status data, and operation mode of the target power system; The operation tasks corresponding to each application form set, the current state and target state of the power grid equipment, the power grid topology model, the flow state data, and the operating mode are input into the reasoning model to obtain the policy information of the security measures required to be executed for each application form set.

[0009] As an improvement to the above solution, the method further includes: Performing a first verification on the entity name in the power grid operation ticket; performing a second verification on the operation terms in the power grid operation ticket; Performing a third verification on the power system operation mode in the power grid operation ticket; Performing a fourth check on the grid equipment status in the grid operation ticket; When the first check, the second check, the third check and the fourth check are passed, the power grid operation ticket is sent to the user end; When any one of the first check, the second check, the third check and the fourth check fails, the grid operation ticket is corrected and the corrected grid operation ticket is sent to the user end.

[0010] As an improvement to the above solution, the method further includes the following training process of the inference model: Collect grid topology information, historical flow status data, historical operation mode, and historical grid equipment status information of multiple power systems; Collecting a plurality of historical application forms of the power system and their corresponding historical grid operation tickets; Time-align the historical power flow state data, the historical operation mode, and the historical power grid equipment state information of each power system to obtain the operation state time series data of the corresponding power system; The operation tasks in the historical application form and the time series data of the operating status of the power system are used as the input of the long short-term memory network, and the strategy information in the historical power grid operation ticket is used as the output of the long short-term memory network. The long short-term memory network is trained through self-supervised learning to obtain a trained inference model.

[0011] As an improvement to the above solution, the first verification of the entity name in the power grid operation ticket includes: Performing entity recognition on the power grid operation ticket using a pre-built entity recognition model to obtain entity names of power grid equipment in the power grid operation ticket; Matching the entity name with the device name in the standard naming library; When the entity name matches any device name in the standard naming library, determining that the first verification is passed; When the entity name does not match any device name in the standard naming library, it is determined that the first verification has failed.

[0012] As an improvement to the above solution, the third verification of the power system operation mode in the power grid operation ticket includes: obtaining a current operating mode of the power system; Comparing and analyzing the operation mode in the grid operation ticket with the current operation mode of the power system; When the operation mode in the grid operation ticket is consistent with the current operation mode of the power system, determining that the third verification is passed; When the operation mode in the grid operation ticket is inconsistent with the current operation mode of the power system, it is determined that the third verification has failed.

[0013] As an improvement to the above solution, the correction of the grid operation ticket includes: When the first verification or the second verification fails, standardizing the entity names or operation terms in the power grid operation ticket to obtain a corrected power grid operation ticket; When the third verification or the fourth verification fails, the inference model is re-adopted to perform execution strategy prediction based on the current operating mode of the target power system and the latest status of the grid equipment to obtain updated strategy information, and the grid operation ticket is regenerated based on the updated strategy information as a corrected grid operation ticket.

[0014] In a second aspect, an embodiment of the present invention provides an artificial intelligence-based power grid operation ticket processing system, comprising: A key field information extraction module is configured to extract key field information from at least one power grid maintenance plan application data uploaded by a user through an RPA agent, and generate at least one application form information based on the key field information of at least one power grid maintenance plan application data to form an application form information table; an application form information classification module, configured to classify the application form information in the application form information table according to key field information corresponding to each application form information in the application form information table, to obtain at least one application form set; wherein one application form set corresponds to the same job type and operation task; A policy information reasoning module is used to obtain policy information of security measures required to be executed for each application set based on the operation tasks corresponding to each application set, the power grid topology model, the current state and target state of the power grid equipment through a pre-built reasoning model; The power grid operation ticket generation module is used to generate the power grid operation ticket of the corresponding application form set according to the key field information of each application form in each application form set, the job type, operation task and strategy information of each application form set.

[0015] Compared with the prior art, an embodiment of the present invention provides an artificial intelligence-based power grid operation ticket processing method and system, which extracts key field information of at least one power grid maintenance plan application data uploaded by the user based on the RPA agent, and generates application form information according to the key field information of each power grid maintenance plan application data to form an application form information table; then, according to the key field information corresponding to each application form information in the application form information table, the application form information in the application form information table is classified to obtain at least one application form set; wherein, one application form set corresponds to the same job type and operation task; then, according to the operation tasks corresponding to each application form set, the current state and target state of the power grid equipment in the target power system, the execution strategy is predicted through the inference model to obtain the strategy information of the safety measures required to be executed for each application form set; finally, based on the According to the policy information of each application form set, a power grid operation ticket of the corresponding application form set is generated; wherein, the power operation ticket is used to instruct the target power system to adjust the operating status of at least one power grid node according to the corresponding policy information, so as to adjust the power grid equipment in the target power system to the target state; the embodiment of the present invention adopts the natural language processing technology of RPA to mine key application form information, and mines the execution strategy required for the operation task indicated by the application form information under the operating status of the target power system and power grid equipment based on the inference model, so as to realize the automatic generation of power grid operation tickets, effectively improve the generation efficiency and accuracy of operation tickets, and through the policy information of the power grid operation ticket, the power grid equipment in the target power system can be adjusted to the target state, so as to ensure the stable operation of the power system when the operating personnel perform the operation task, and reduce the occurrence of human safety accidents and power grid equipment safety accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings used in the implementation methods. Obviously, the drawings described below are only some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 This is a flow chart of a method for processing power grid operation tickets based on artificial intelligence provided by an embodiment of the present invention; Figure 2 This is a schematic diagram of the process of aggregating application form information tables provided by an embodiment of the present invention; Figure 3 This is a structural block diagram of an artificial intelligence-based power grid operation ticket processing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0019] It is understood that the various numbers used in the embodiments of the present invention are merely for ease of description and are not intended to limit the scope of this application. The order of execution of each process does not necessarily imply a specific order of execution. The order of execution of each process should be determined by its function and inherent logic.

[0020] In embodiments of the present invention, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. The terms "include", "comprises", or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further restrictions, the elements defined by the statement "includes..." do not exclude the presence of additional identical elements in the process, method, article, or device comprising the elements. The term "plurality or several" refers to two or more.

[0021] See Figure 1 , Figure 1 This is a flow chart of a method for processing power grid operation tickets based on artificial intelligence provided by an embodiment of the present invention. The method for processing power grid operation tickets based on artificial intelligence specifically includes: S11: extracting key field information from at least one power grid maintenance plan application data uploaded by the user through an RPA agent, and generating at least one application form information based on the key field information of at least one power grid maintenance plan application data to form an application form information table; S12: Classifying the application form information in the application form information table according to key field information corresponding to each application form information in the application form information table to obtain at least one application form set; wherein one application form set corresponds to the same job type and operation task; S13: Based on the operation tasks corresponding to each application form set, the current state and target state of the power grid equipment in the target power system, an execution strategy prediction is performed using a reasoning model to obtain strategy information of the security measures required to be executed for each application form set; S14: Generate a power grid operation ticket for the corresponding application form set based on the policy information of each application form set; wherein the power operation ticket is used to instruct the target power system to adjust the operating status of at least one power grid node according to the corresponding policy information, so as to adjust the power grid equipment in the target power system to the target state.

[0022] It should be noted that the embodiments of the present invention can be executed by intelligent terminal devices such as servers and computers. The embodiments of the present invention construct an RPA agent based on natural language processing technology, such as a recurrent neural network model and a bag-of-words model. The RPA agent then guides users to upload power grid maintenance plan application data and extracts key field information from the uploaded power grid maintenance plan application data. It should be noted that the training process of large language models and bag-of-words models is prior art and will not be described in detail here. Specifically, the RPA agent connects to the distribution network control and management system. The user wakes up the RPA agent and enters the login information for the distribution network control and management system. The RPA agent automatically logs in to the distribution network control and management system according to preset process rules and locates the target area in the system where maintenance plans are stored. The maintenance plan to be processed is then downloaded from the target area. The user can upload the downloaded maintenance plan as power grid maintenance plan application data to the RPA agent's interactive dialog box, and further extract key field information using the bag-of-words model in the RPA agent.

[0023] Specifically, extracting key field information from at least one power grid maintenance plan application data uploaded by the user through the RPA agent includes: The RPA agent built based on the recurrent neural network performs semantic analysis on at least one power grid maintenance plan application data uploaded by the user, and extracts key field information of at least one of the power grid maintenance plan application data.

[0024] Among them, key field information includes but is not limited to: application number (can also be described as application form code, such as XXX company 202208XXX), operation type (including power outage operation, live operation, etc.), operation time (that is, the power outage time applied by the operator), operation task (the specific work content of the operator, such as disconnecting the leads on both sides of the XXX switch on pole A of line A and removing the leads of the lightning arrester), power outage scope, power supply station, etc.

[0025] When a user uploads data for a power grid maintenance plan application, the RPA agent directly extracts key field information from the data, generates application form information, and forms an application form information table based on the application form information. When a user uploads multiple power grid maintenance plan application data, the RPA agent separates the key field information of the multiple power grid maintenance plan application data, generates respective application form information based on the key field information of each power grid maintenance plan application data, and then aggregates the multiple application form information into the same application form information table, such as Figure 2 shown.

[0026] The embodiment of the present invention uses RPA's natural language processing technology to automatically download power grid maintenance plan application data and mine key field information to generate application form information. Compared with the manual filling method, it can automate the download and generation process of power grid maintenance application forms, effectively shorten the download and generation time of power grid maintenance application forms, and improve the download and generation efficiency of power grid maintenance application forms.

[0027] For the above-mentioned application form information table, the embodiment of the present invention classifies all application form information in the application form information table based on the key field information of each application form information, so as to integrate the application form information of the same operation type and operation task into the same application form set; subsequently, according to the operation tasks corresponding to each application form set, the current state and target state of the power grid equipment in the target power system, the execution strategy is predicted through the reasoning model to obtain the strategy information of the safety measures required to be executed for each application form set, which is used to generate the power grid operation ticket of the corresponding application form set, realize the automatic generation of the power grid operation ticket, and effectively improve the generation efficiency and accuracy of the operation ticket. Based on the reasoning model, the embodiment of the present invention can mine the relationship between the specific operation task indicated by the application form information and the required execution strategy under the operating state of the target power system and power grid equipment, so as to ensure that after the target power system adjusts the power grid equipment to the target state according to the strategy information indicated by the power operation ticket, the power system can operate stably without responding to the normal operation task execution of the operator, thereby reducing the occurrence of human safety accidents and power grid equipment safety accidents.

[0028] Specifically, the application form information in the application form information table is classified according to the key field information corresponding to each application form information in the application form information table to obtain at least one application form set, including: Extracting a first key field indicating a job type from each application form information in the application form information table; Classifying the application form information in the application form information table according to the first key field of the application form information to obtain at least one application form information sub-table; wherein each application form information sub-table corresponds to a job type; Extracting the second key field indicating the operation task and the power outage scope from each application form information in each application form information sub-table; The application form information of the corresponding application form information subtable is classified according to the second key field of the application form information to obtain at least one application form set; wherein each application form set corresponds to the same operation task under the same power outage scope.

[0029] For example, after the RPA agent completes the generation of the application form information table, the key fields of each application form information in the application form information table can be screened through a feature extraction project. For example, based on the first key field indicating the job type in the key field information, the application form information of the live job type is screened out from the application form information table to generate an application form information sub-table; then, based on the second key field indicating the job time, power outage range, and operation task in the key field information, the application form information with the same job time, power outage range, and operation task is screened out from the application form information sub-table to form an application form set. The application form set generated by two rounds of key field screening in the embodiment of the present invention has more refined data and stronger correlation, which greatly improves the pertinence and efficiency of data processing, and lays the foundation for the subsequent analysis and adaptation of safety measures execution strategies for specific job type applications.

[0030] Specifically, the execution strategy prediction is performed through the inference model based on the operation tasks corresponding to each application form set, the current state and target state of the power grid equipment in the target power system, and the strategy information of the security measures required to be executed for each application form set is obtained, including: Obtain the grid topology model, power flow status data, and operation mode of the target power system; The operation tasks corresponding to each application form set, the current state and target state of the power grid equipment, the power grid topology model, the flow state data, and the operating mode are input into the reasoning model to obtain the policy information of the security measures required to be executed for each application form set.

[0031] It should be noted that the power grid topology model describes the connections between various power grid devices in the power system. Nodes in the power grid topology model represent the connection points of power grid devices, such as busbar nodes at substations and intersections of transmission lines. These nodes are interconnected by edges (i.e., electrical connections such as transmission lines and transformer windings), forming the power grid topology. Power flow state data primarily includes node data (e.g., the electrical state of each node in the power system at a specific steady-state moment, such as the voltage amplitude at each node) and line data (e.g., the power distribution of each line). Operational modes include main connection modes (including various operating modes of the power system at different times and conditions, such as single busbar connection, double busbar connection, and bridge connection) and power source distribution (which involves the geographical distribution of various power generation sources in the power system, such as thermal power, hydropower, wind power, and solar power, and their respective power output contributions).

[0032] Furthermore, the training process of the inference model includes: Collect grid topology information, historical flow status data, historical operation mode, and historical grid equipment status information of multiple power systems; Collecting a plurality of historical application forms of the power system and their corresponding historical grid operation tickets; Time-align the historical power flow state data, the historical operation mode, and the historical power grid equipment state information of each power system to obtain the operation state time series data of the corresponding power system; The operation tasks in the historical application form and the time series data of the operating status of the power system are used as the input of the long short-term memory network, and the strategy information in the historical power grid operation ticket is used as the output of the long short-term memory network. The long short-term memory network is trained through self-supervised learning to obtain a trained inference model.

[0033] The embodiment of the present invention collects grid topology information (i.e., grid topology models), historical flow status data, and historical operating modes of multiple different power systems; simultaneously collects historical application forms and their corresponding historical grid operation tickets of multiple power systems, as well as historical grid equipment status information (including pre-execution status and post-execution status) before and after the corresponding power systems execute the historical grid operation tickets; then time-aligns the historical flow status data, historical operating modes, and historical grid equipment status information with time attributes to obtain operating status time series data of the corresponding power systems; and uses the operating status time series data, operating status time series data, and historical application forms of the corresponding power systems as input items, and the policy information of the safety measures indicated by the historical grid operation tickets as output items, and performs self-supervised learning training on the long short-term memory network until the model loss converges to obtain a trained inference model.

[0034] The embodiment of the present invention uses a long short-term memory network to mine potential correlation patterns between different data, extract operation tasks from historical application forms, and combine them with the corresponding power system operating status to explore under what operating conditions, without affecting the stable operation of the power system, the safety measures strategy adopted for specific operation tasks; at the same time, combined with self-supervised learning training, the trained inference model has a strong generalization ability, which can not only accurately infer safety measures for operation tasks under working conditions similar to the training data, but also adapt to new and incomplete operating states and operation scenarios to a certain extent.

[0035] After the inference model is trained, for each request set, the model extracts the current and target states of the grid equipment in the target power system, the operational tasks, and the grid topology, operational mode, and power flow data as inputs. The inference model then performs execution strategy reasoning and prediction to obtain policy information for the safety measures required for the corresponding request set. This policy information includes the execution strategy steps for the nodes / devices in the power system that must be regulated / controlled when the operator performs a specific operational task, such as changing a switch from hot standby to operation or changing a recloser from tripping to signaling.

[0036] The corresponding grid operation ticket can be generated by integrating this strategy information. Among them, a grid operation ticket of the live working type is shown in the following table: The embodiment of the present invention combines RPA intelligent agents and neural network models to automatically generate power grid operation tickets throughout the entire process. On the one hand, it uses natural language technology to extract and aggregate operation tasks and uses neural networks trained based on historical data to infer appropriate strategy information. This can improve the accuracy of operation ticket generation, ensure the stable operation of the power system, and reduce the occurrence of human safety accidents and power grid equipment safety accidents; on the other hand, it can greatly shorten the time for operation ticket generation and improve the efficiency of operation ticket generation.

[0037] Furthermore, the method further comprises: Performing a first verification on the entity name in the power grid operation ticket; performing a second verification on the operation terms in the power grid operation ticket; Performing a third verification on the power system operation mode in the power grid operation ticket; Performing a fourth check on the grid equipment status in the grid operation ticket; When the first check, the second check, the third check and the fourth check are passed, the power grid operation ticket is sent to the user end; When any one of the first check, the second check, the third check and the fourth check fails, the grid operation ticket is corrected and the corrected grid operation ticket is sent to the user end.

[0038] Specifically, the first verification of the entity name in the power grid operation ticket includes: Performing entity recognition on the power grid operation ticket using a pre-built entity recognition model to obtain entity names of power grid equipment in the power grid operation ticket; Matching the entity name with the device name in the standard naming library; When the entity name matches any device name in the standard naming library, determining that the first verification is passed; When the entity name does not match any device name in the standard naming library, it is determined that the first verification has failed.

[0039] Specifically, the second verification of the operation terms in the power grid operation ticket includes: Extracting operational terms of the strategy information in the power grid operation ticket; matching the operational terms with operational terms in a standard operational term library; When the operation term matches any operation term in the standard operation term library, determining that the second verification is passed; When the operation term does not match any operation term in the standard naming library, it is determined that the second verification has failed.

[0040] Specifically, the third verification of the power system operation mode in the power grid operation ticket includes: obtaining a current operating mode of the power system; Comparing and analyzing the operation mode in the grid operation ticket with the current operation mode of the power system; When the operation mode in the grid operation ticket is consistent with the current operation mode of the power system, determining that the third verification is passed; When the operation mode in the grid operation ticket is inconsistent with the current operation mode of the power system, it is determined that the third verification has failed.

[0041] Specifically, the fourth verification of the grid equipment status in the grid operation ticket includes: Obtaining the current state of power grid equipment in the power system; Comparing and analyzing the grid equipment status in the grid operation ticket with the current grid equipment status of the power system; When the grid device state in the grid operation ticket is consistent with the current grid device state of the power system, determining that the fourth verification is passed; When the grid device status in the grid operation ticket is inconsistent with the current grid device status of the power system, it is determined that the fourth verification has failed.

[0042] Specifically, the correcting the power grid operation ticket includes: When the first verification or the second verification fails, standardizing the entity names or operation terms in the power grid operation ticket to obtain a corrected power grid operation ticket; For example, the entity names or operation terms in the power grid operation ticket are unified into the set standard entity names or standard operation terms.

[0043] When the third verification or the fourth verification fails, the inference model is re-adopted to perform execution strategy prediction based on the current operating mode of the target power system and the latest status of the grid equipment to obtain updated strategy information, and the grid operation ticket is regenerated based on the updated strategy information as a corrected grid operation ticket.

[0044] The embodiment of the present invention can further improve the accuracy of the operation ticket by checking and correcting the entity name, operation terminology, operation mode, grid equipment status and other information in the grid operation ticket, and can avoid the occurrence of wrong tickets caused by changes in wiring methods and changes in grid equipment status.

[0045] Compared with the existing technology, the embodiment of the present invention uses RPA's natural language processing technology to mine key application form information, and based on the inference model, it mines the execution strategy required for the operation task indicated by the application form information under the operating status of the target power system and power grid equipment, thereby realizing the automatic generation of power grid operation tickets, effectively improving the efficiency and accuracy of operation ticket generation, and ensuring that operators adjust the power grid equipment to the target state according to the strategy information indicated in the power grid operation ticket, thereby ensuring the stable operation of the power system and reducing the occurrence of power grid equipment safety accidents.

[0046] See also Figure 3 , Figure 3 The embodiment of the present invention provides a structural block diagram of an artificial intelligence-based power grid operation ticket processing system, which includes: A key field information extraction module 11 is configured to extract key field information from at least one power grid maintenance plan application data uploaded by a user through an RPA agent, and generate at least one application form information based on the key field information of at least one power grid maintenance plan application data to form an application form information table; The application form information classification module 12 is configured to classify the application form information in the application form information table according to the key field information corresponding to each application form information in the application form information table to obtain at least one application form set; wherein one application form set corresponds to the same job type and operation task; A policy information reasoning module 13 is configured to obtain policy information of security measures required to be executed for each application set based on the operation tasks, power grid topology model, current state and target state of power grid equipment corresponding to each application set through a pre-built reasoning model; The power grid operation ticket generating module 14 is configured to generate a power grid operation ticket for a corresponding application form set according to key field information of each application form in each application form set, job type, operation task and strategy information of each application form set.

[0047] In an optional embodiment, the key field information extraction module 11 includes: A semantic analysis unit, configured to perform semantic analysis on at least one power grid maintenance plan application data uploaded by a user through an RPA agent constructed based on a recurrent neural network, and extract key field information of at least one of the power grid maintenance plan application data; Among them, the key field information includes: application number, operation type, operation time, operation task, and power outage scope.

[0048] In an optional embodiment, the application form information classification module 12 includes: A first key field extraction unit, configured to extract a first key field indicating a job type from each application form information in the application form information table; a first classification unit, configured to classify the application form information in the application form information table according to the first key field of the application form information to obtain at least one application form information sub-table; wherein each application form information sub-table corresponds to a job type; A second key field extraction unit, configured to extract a second key field indicating an operation task and a power outage range from each application form information in each application form information sub-table; The second classification unit is used to classify the application form information of the corresponding application form information subtable according to the second key field of the application form information to obtain at least one application form set; wherein each application form set corresponds to the same operation task under the same power outage range.

[0049] In an optional embodiment, the policy information reasoning module 13 includes: An information acquisition unit, used to obtain the grid topology model, flow status data, and operation mode of the target power system; The strategy reasoning unit is used to input the operation tasks corresponding to each application form set, the current state and target state of the power grid equipment, the power grid topology model, the flow state data, and the operating mode into the reasoning model to obtain the strategy information of the security measures that need to be executed for each application form set.

[0050] In an optional embodiment, the system further includes: A first verification module, configured to perform a first verification on the entity name in the power grid operation ticket; A second verification module, configured to perform a second verification on the operation terms in the power grid operation ticket; A third verification module, configured to perform a third verification on the power system operation mode in the power grid operation ticket; a fourth verification module, configured to perform a fourth verification on the grid equipment status in the grid operation ticket; a sending module, configured to send the grid operation ticket to a user terminal when the first check, the second check, the third check, and the fourth check are passed; A correction module is used to correct the grid operation ticket when any one of the first check, the second check, the third check and the fourth check fails, and send the corrected grid operation ticket to the user end.

[0051] In an optional embodiment, the system further includes: A first historical information collection module is used to collect grid topology information, historical flow status data, historical operation mode, and historical grid equipment status information of multiple power systems; A second historical information collection module is used to collect a plurality of historical application forms of the power system and their corresponding historical grid operation tickets; A time alignment module, configured to time-align the historical power flow state data, the historical operation mode, and the historical power grid equipment state information of each power system to obtain the operation state time series data of the corresponding power system; The self-supervised learning training module is used to use the operation tasks in the historical application form and the operating status time series data of the power system as the input of the long short-term memory network, and use the strategy information in the historical power grid operation ticket as the output of the long short-term memory network, perform self-supervised learning training on the long short-term memory network, and obtain a trained inference model.

[0052] In an optional embodiment, the first verification module includes: An entity recognition unit, configured to perform entity recognition on the power grid operation ticket using a pre-built entity recognition model, and obtain entity names of power grid devices in the power grid operation ticket; A matching unit, configured to match the entity name with the device name in a standard naming library; When the entity name matches any device name in the standard naming library, it is determined that the first verification has been passed; when the entity name does not match any device name in the standard naming library, it is determined that the first verification has not been passed.

[0053] In an optional embodiment, the third verification module includes: a current operation mode acquisition unit, configured to acquire the current operation mode of the power system; a comparison and analysis unit, configured to compare and analyze the operation mode in the grid operation ticket with the current operation mode of the power system; Among them, when the operating mode in the grid operation ticket is consistent with the current operating mode of the power system, it is determined that the third verification has been passed; when the operating mode in the grid operation ticket is inconsistent with the current operating mode of the power system, it is determined that the third verification has not been passed.

[0054] In an optional embodiment, the correction module includes: a standardization processing unit, configured to, when the first verification or the second verification fails, standardize the entity names or operation terms in the power grid operation ticket to obtain a corrected power grid operation ticket; Among them, when the third verification or the fourth verification fails, the inference model is re-adopted to perform execution strategy prediction based on the current operating mode of the target power system and the latest status of the grid equipment to obtain updated strategy information, and the grid operation ticket is regenerated based on the updated strategy information as a corrected grid operation ticket.

[0055] It should be noted that the working process of each module in the artificial intelligence-based power grid operation ticket processing system described in the embodiment of the present invention can refer to the working process of the artificial intelligence-based power grid operation ticket processing method described in the above embodiment, and the technical effect achieved is also the same as the artificial intelligence-based power grid operation ticket processing method described in the above embodiment, which will not be repeated here.

[0056] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, various improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for processing power grid operation tickets based on artificial intelligence, characterized in that: include: Extract key field information from at least one power grid maintenance plan application data uploaded by the user through an RPA agent, and generate at least one application form information based on the key field information of at least one power grid maintenance plan application data to form an application form information table; Classify the application form information in the application form information table according to key field information corresponding to each application form information in the application form information table to obtain at least one application form set; wherein one application form set corresponds to the same job type and operation task; Based on the operation tasks corresponding to each of the application form sets, the current state and target state of the power grid equipment in the target power system, an execution strategy prediction is performed through an inference model to obtain strategy information of the security measures required to be executed for each of the application form sets; Based on the policy information of each of the application form sets, a power grid operation ticket for the corresponding application form set is generated; wherein the power operation ticket is used to instruct the target power system to adjust the operating status of at least one power grid node according to the corresponding policy information, so as to adjust the power grid equipment in the target power system to the target state.

2. The method for processing power grid operation tickets based on artificial intelligence according to claim 1, characterized in that: The extracting key field information from at least one power grid maintenance plan application data uploaded by the user through the RPA agent includes: Performing semantic analysis on at least one power grid maintenance plan application data uploaded by the user through an RPA agent built based on a recurrent neural network, and extracting key field information of at least one of the power grid maintenance plan application data; Among them, the key field information includes: application number, operation type, operation time, operation task, and power outage scope.

3. The method for processing power grid operation tickets based on artificial intelligence according to claim 2, characterized in that: The application form information in the application form information table is classified according to the key field information corresponding to each application form information in the application form information table to obtain at least one application form set, including: Extracting a first key field indicating a job type from each application form information in the application form information table; Classifying the application form information in the application form information table according to the first key field of the application form information to obtain at least one application form information sub-table; wherein each application form information sub-table corresponds to a job type; Extracting the second key field indicating the operation task and the power outage scope from each application form information in each application form information sub-table; The application form information of the corresponding application form information subtable is classified according to the second key field of the application form information to obtain at least one application form set; wherein each application form set corresponds to the same operation task under the same power outage scope.

4. The method for processing power grid operation tickets based on artificial intelligence according to claim 1, characterized in that: The method of performing execution strategy prediction based on the operation tasks corresponding to each application form set, the current state and target state of the power grid equipment in the target power system through the inference model to obtain strategy information of the security measures required to be executed for each application form set includes: Obtain the grid topology model, power flow status data, and operation mode of the target power system; The operation tasks corresponding to each application form set, the current state and target state of the power grid equipment, the power grid topology model, the flow state data, and the operating mode are input into the reasoning model to obtain the policy information of the security measures required to be executed for each application form set.

5. The method for processing power grid operation tickets based on artificial intelligence according to claim 4, characterized in that: The method further comprises: Performing a first verification on the entity name in the power grid operation ticket; performing a second verification on the operation terms in the power grid operation ticket; Performing a third verification on the power system operation mode in the power grid operation ticket; Performing a fourth check on the grid equipment status in the grid operation ticket; When the first check, the second check, the third check and the fourth check are passed, the power grid operation ticket is sent to the user end; When any one of the first check, the second check, the third check and the fourth check fails, the grid operation ticket is corrected and the corrected grid operation ticket is sent to the user end.

6. The method for processing power grid operation tickets based on artificial intelligence according to claim 4, characterized in that: The method further includes the following training process of the inference model: Collect grid topology information, historical flow status data, historical operation mode, and historical grid equipment status information of multiple power systems; Collecting a plurality of historical application forms of the power system and their corresponding historical grid operation tickets; Time-align the historical power flow state data, the historical operation mode, and the historical power grid equipment state information of each power system to obtain the operation state time series data of the corresponding power system; The operation tasks in the historical application form and the time series data of the operating status of the power system are used as the input of the long short-term memory network, and the strategy information in the historical power grid operation ticket is used as the output of the long short-term memory network. The long short-term memory network is trained through self-supervised learning to obtain a trained inference model.

7. The method for processing power grid operation tickets based on artificial intelligence according to claim 5, characterized in that: The first verification of the entity name in the power grid operation ticket includes: Performing entity recognition on the power grid operation ticket using a pre-built entity recognition model to obtain entity names of power grid equipment in the power grid operation ticket; Matching the entity name with the device name in the standard naming library; When the entity name matches any device name in the standard naming library, determining that the first verification is passed; When the entity name does not match any device name in the standard naming library, it is determined that the first verification has failed.

8. The method for processing power grid operation tickets based on artificial intelligence according to claim 5, characterized in that: The third verification of the power system operation mode in the power grid operation ticket includes: obtaining a current operating mode of the power system; Comparing and analyzing the operation mode in the grid operation ticket with the current operation mode of the power system; When the operation mode in the grid operation ticket is consistent with the current operation mode of the power system, determining that the third verification is passed; When the operation mode in the grid operation ticket is inconsistent with the current operation mode of the power system, it is determined that the third verification has failed.

9. The method for processing power grid operation tickets based on artificial intelligence according to claim 5, characterized in that: The correcting of the power grid operation ticket includes: When the first verification or the second verification fails, standardizing the entity names or operation terms in the power grid operation ticket to obtain a corrected power grid operation ticket; When the third verification or the fourth verification fails, the inference model is re-adopted to perform execution strategy prediction based on the current operating mode of the target power system and the latest status of the grid equipment to obtain updated strategy information, and the grid operation ticket is regenerated based on the updated strategy information as a corrected grid operation ticket.

10. An artificial intelligence-based power grid operation ticket processing system, characterized in that: include: A key field information extraction module is configured to extract key field information from at least one power grid maintenance plan application data uploaded by a user through an RPA agent, and generate at least one application form information based on the key field information of at least one power grid maintenance plan application data to form an application form information table; an application form information classification module, configured to classify the application form information in the application form information table according to key field information corresponding to each application form information in the application form information table, to obtain at least one application form set; wherein one application form set corresponds to the same job type and operation task; A policy information reasoning module is used to obtain policy information of security measures required to be executed for each application set based on the operation tasks corresponding to each application set, the power grid topology model, the current state and target state of the power grid equipment through a pre-built reasoning model; The power grid operation ticket generation module is used to generate the power grid operation ticket of the corresponding application form set according to the key field information of each application form in each application form set, the job type, operation task and strategy information of each application form set.

Images