Intelligent identification question-answering system for illegal behaviors of power grid information operation

Abstract

The invention discloses a power grid information operation violation behavior intelligent identification question-answering system, and relates to the technical field of power grid operation violation behavior intelligent identification. According to the method, violation behavior judgment and coping rules are extracted from power grid operation records, industry standards and historical violation logs through a natural language processing technology, the intelligent recognition efficiency of violation behaviors is improved, manual intervention is reduced, and operation automation and precision are achieved; by constructing a plurality of sub-databases and combining a version control technology, dynamic update and timely supplement of illegal behavior rules are ensured, so that the system can continuously adapt to industry changes, and the accuracy and timeliness of illegal behavior recognition are improved; besides, the mapping relation between the illegal behavior ID and the coping rule ID is optimized by adopting the graph neural network, so that the automatic association capability of the system is improved, the most suitable coping measure can be recommended according to the illegal behavior type, and the safety and the high efficiency of the power grid operation are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent identification of illegal behaviors in power grid operations, and specifically to an intelligent identification Q&A system for illegal behaviors in power grid information operations. Background Art

[0002] With the expansion of the scale of the power grid and the improvement of the degree of intelligence, the safety and reliability of power grid operations are facing increasing challenges. Traditional identification of illegal behaviors relies on manual review and post-event inspection, which is not only inefficient but also has the risks of missed judgment and misjudgment, and it is difficult to cope with complex and high-frequency operation environments.

[0003] Although some existing intelligent identification technologies for illegal behaviors in power grid operations have applied technologies such as sensors and image recognition for on-site monitoring, most of them have not fully integrated the standards and specifications related to power grid operations and historical illegal records, resulting in a low level of intelligence in the determination of illegal behaviors and response measures. In addition, the update and management of illegal behavior rules and response measures usually rely on manual operations, lacking an effective dynamic update mechanism and being difficult to adapt to the rapid changes in power grid industry standards.

[0004] In view of the above problems, it is necessary to propose an intelligent identification Q&A system for illegal behaviors in power grid information operations. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems existing in the background art and propose an intelligent identification Q&A system for illegal behaviors in power grid information operations.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An intelligent identification Q&A system for illegal behaviors in power grid information operations includes a data integration module, a knowledge base generation and management module, an illegal behavior identification module, an intelligent Q&A module, and a progress tracking module.

[0008] The data integration module is responsible for accessing industry standard specification documents, operation manuals, power grid operation records, and historical illegal logs, and extracting the machine language of the illegal behavior determination rules and illegal behavior response rules therein.

[0009] Through natural language processing technology, keywords including equipment names, equipment parameters, operation requirements, environmental requirements, numerical ranges, and personnel configurations in industry standard specification documents and operation manuals are located, identified, and relationship-extracted, and are transformed into illegal behavior event determination rules with a standard format, and an illegal behavior ID is assigned to each illegal behavior event determination rule.

[0010] The standard format of the illegal behavior event determination rule is specifically: illegal behavior ID + judgment principle + judgment result.

[0011] As a preferred embodiment of the present invention, through natural language processing technology, keywords including rectification measures, shutdown measures, reporting measures, and inspection measures in power grid operation records and historical violation logs are located, identified, and relationship extracted, and then converted into response rules for violation behavior events with a standard format, and a response rule ID is assigned to each response rule for violation behavior events.

[0012] The standard format of the response rules for violation behavior events is specifically: response rule ID + violation behavior type + response measures.

[0013] As a preferred embodiment of the present invention, all extracted determination rules and response rules for violation behavior events are converted into machine language based on logical judgment and output to the knowledge base generation and management module.

[0014] The knowledge base generation and management module is used to integrate all determination rules and response rules for violation behavior events, and construct and manage a knowledge base related to the power grid.

[0015] Three sub-databases are created: the first sub-database, the second sub-database, and the third sub-database. Among them, the first sub-database is used to store the determination rules for violation behavior events and their violation behavior IDs; the second sub-database is used to store the response rules for violation behavior events and their response rule IDs; the third sub-database is used to store the mapping relationship between violation behavior IDs and response rule IDs;

[0016] In the first sub-database and the second sub-database, the data integration module is re-accessed every preset update period to discover newly extracted determination rules and response rules for violation behavior events, as well as corrections to existing determination rules and response rules for violation behavior events, so as to ensure the timely update of the knowledge base.

[0017] The version control technology is used for rule version tracking, recording each modified and newly added determination rules and response rules for violation behavior events, saving the update time when they enter the database, and providing a historical version backtracking window.

[0018] The historical version backtracking window includes the content and update time of each historical version of the determination rules and response rules for violation behavior events, and highlights the modified content of each historical version compared with the previous version.

[0019] As a preferred embodiment of the present invention, in the third sub-database, the mapping relationship between violation behavior IDs and response rule IDs is saved in a graph structure.

[0020] Construct a graph structure containing a five-dimensional space in the third sub-database, with each dimension corresponding to a keyword, including: the first dimension: device name; the second dimension: device parameters; the third dimension: environmental conditions; the fourth dimension: numerical range; the fifth dimension: safety conditions;

[0021] Extract the keywords related to the first to fifth dimensions in the violation event determination rules corresponding to each violation behavior ID through a text matching algorithm based on keywords and context, and map them to each specific dimension according to the semantic dictionary to obtain the specific coordinates of the violation behavior ID in each dimension.

[0022] Extract the keywords related to the first to fifth dimensions in the violation event response rules corresponding to each response rule ID through a text matching algorithm based on keywords and context, and map them to each specific dimension according to the semantic dictionary to obtain the specific coordinates of the response rule ID in each dimension.

[0023] Convert the violation behavior ID and the response rule ID into nodes in the graph structure according to the mapped coordinates, and perform automated association of violation behavior - response rules based on the Euclidean distance between the nodes. Calculate the Euclidean distance between each violation behavior ID and all other response rule IDs in the graph structure. Denote the obtained Euclidean distance as the matching similarity between each violation behavior ID and all other response rule IDs, and perform index association between each violation behavior ID and the response rule IDs whose matching similarity is greater than the preset threshold to obtain the mapping relationship between the violation behavior ID and the response rule ID.

[0024] The violation behavior recognition module monitors the power grid operation site in real time through sensors and image recognition technologies, identifies and judges potential violation behaviors, and ensures that equipment operations and work processes comply with industry standard specifications.

[0025] Use a high-resolution camera to monitor the power grid operation site in real time, especially the equipment operation area. Capture the behaviors of the operators at the power grid operation site, the states of the equipment, and the changes in the surrounding environment. Identify potential violation behaviors through image recognition technologies including object detection, action recognition, and scene analysis.

[0026] Obtain the working parameters of the equipment in real time through temperature and humidity sensors, voltage, current, equipment temperature, and equipment position sensors installed on the power grid equipment, and use them to detect whether the equipment is in a normal working state and complies with industry standard specifications.

[0027] Convert the collected potential violation behaviors and the working parameters of the equipment into electronic operation records, and input them into the first sub-database in the knowledge base generation and management module for violation behavior matching and determination to obtain the violation behavior ID and determination result corresponding to the electronic operation record.

[0028] As a preferred embodiment of the present invention, the mapping relationship between the violation behavior IDs and the response rule IDs in the third sub-database of the knowledge base generation and management module is optimized through a graph neural network, and the positions of each node in the third sub-database are iteratively updated, so that each node not only depends on its own features, but also aggregates information from neighboring nodes through latent connection capture.

[0029] The intelligent Q&A module is responsible for implementing intelligent query, identification, and guidance of power grid operation violation behaviors.

[0030] Collect the violation behavior questions related to power grid operations input by the user, understand and identify the content through natural language processing, match the response rule IDs in the knowledge base generation and management module, and output them to the user, providing detailed violation behavior types and their corresponding countermeasures.

[0031] As a preferred embodiment of the present invention, windows for accessing and retrieving the first sub-database and the second sub-database are provided.

[0032] The progress tracking module tracks the progress of each violation behavior ID and determination result in the electronic operation record, continuously records the processing progress of the violation behavior, automatically records each step in the processing process, and provides corresponding solutions through the knowledge base to ensure that each violation behavior is effectively responded to.

[0033] Generate a unique tracking ID for each identified violation behavior. The tracking ID is associated with the violation behavior event determination rule and the response rule of this violation behavior, and the rectification status of this violation behavior is updated in real time. Access the electronic operation record at preset time intervals, record the progress and results of each processing of the violation behavior corresponding to each tracking ID, and update the status flag of this violation behavior in real time. For newly generated violation behaviors, assign them a to-be-processed flag; if it is identified in the electronic operation record that countermeasures are taken for a violation behavior with a to-be-processed flag, then adjust its corresponding status flag to in-processing; if it is identified in the electronic operation record that a certain violation behavior is resolved, then adjust its corresponding status flag to resolved.

[0034] As a preferred embodiment of the present invention, the rectification progress of violation behaviors is statistically obtained through task status, the number of violation behaviors with to-be-processed flags, in-processing flags, and resolved flags is acquired, and a rectification progress statistical chart is generated.

[0035] If it is identified that there is a violation behavior that has not obtained a resolved flag within the specified time, an alert notice is automatically sent to require the relevant responsible person to process it as soon as possible.

[0036] Compared with the prior art, the beneficial effects of the present invention are:

[0037] 1. The present invention extracts violation behavior determination rules and response rules from data sources such as power grid operation records, industry standard specifications, and historical violation logs. Through natural language processing technology, these rules are standardized and transformed, further improving the intelligent recognition efficiency of violation behaviors. The system can monitor the power grid operation site in real time and accurately identify violation behaviors, ensuring that equipment operations and operation processes always comply with industry standards, thereby greatly reducing the workload of manual recognition and improving the automation and accuracy of operations;

[0038] 2. The present invention constructs multiple sub-databases and combines version control technology to track rule versions, ensuring the timely update of violation behavior determination rules and response rules related to the power grid. Every preset period, the system will automatically re-access the data integration module to extract new violation behavior determination rules and response rules, or correct and supplement existing rules. This dynamic update mechanism not only ensures the timeliness of the knowledge base content but also provides the ability to track rule changes through the historical version backtracking function. This feature enables the system to continuously learn and adapt to changes in power grid industry standards, further improving the accuracy and timeliness of power grid operation violation behavior recognition and handling;

[0039] 3. The present invention uses a graph neural network to optimize the mapping relationship between violation behavior IDs and response rule IDs. By iteratively updating the node positions, capturing potential connections, and aggregating neighbor node information, it can more accurately automate the association of violation behaviors with corresponding response rules. This optimization method based on the graph structure improves the system's judgment ability in complex scenarios and can automatically recommend the most appropriate response measures according to different types of violation behaviors. In addition, the application of the graph neural network enables the system to make intelligent decisions in a wider range of scenarios, further improving the safety and efficiency of power grid operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] For the convenience of those skilled in the art to understand, the present invention will be further described below in conjunction with the accompanying drawings:

[0041] Figure 1 is the system block diagram of the present invention;

[0042] Figure 2 is the statistical chart of the improvement progress proposed in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0044] Please refer to Figure 1 shown in the figure. An intelligent recognition Q&A system for illegal behaviors in power grid information operations includes a data integration module, a knowledge base generation and management module, an illegal behavior recognition module, an intelligent Q&A module, and a progress tracking module.

[0045] The data integration module is responsible for accessing industry standard specification documents, operation manuals, power grid operation records, and historical violation logs, and extracting the machine language regarding the determination rules and response rules for illegal behaviors.

[0046] Through natural language processing technology, keywords including equipment names, equipment parameters, operation requirements, environmental requirements, numerical ranges, and personnel configurations in industry standard specification documents and operation manuals are located, identified, and relationship-extracted, and are transformed into determination rules for illegal behavior events with a standard format, and an illegal behavior ID is assigned to each determination rule for illegal behavior events.

[0047] The standard format of the determination rule for illegal behavior events is specifically: illegal behavior ID + judgment principle + judgment result.

[0048] For example, the following information is extracted from the industry standard specification document:

[0049] "When operating the equipment, the voltage value of the switchgear must be between 400 kV and 800 kV, and the current should not exceed 1000 A; the equipment temperature must be maintained between -10°C and 60°C, and any abnormal equipment situation should immediately stop the machine for inspection;

[0050] A safety warning line of 3 meters should be set in the operation area of all high-voltage equipment, and non-staff are prohibited from entering;

[0051] At least two temperature and humidity monitoring devices must be equipped, and it is ensured that the sensor of the monitoring device is no more than 5 meters from the nearest point of the equipment operation area."

[0052] After performing keyword location, identification, and relationship extraction, the following determination rules for illegal behavior events are obtained:

[0053] "Determination rule 1, judgment principle: The voltage of the switchgear is greater than 800 kV or less than 400 kV → judgment result: illegal;

[0054] Determination rule 2, judgment principle: The current of the switchgear is greater than 1000 A → judgment result: illegal;

[0055] Determination rule 3, judgment principle: The temperature of the switchgear is less than -10°C or greater than 60°C → judgment result: illegal;

[0056] Determination rule 4, judgment principle: If there is no 3-meter safety warning line set at the operation site → judgment result: illegal;

[0057] Determination Rule 5, Judgment Principle: The distance of the on-site monitoring equipment for operations is greater than 5 meters → Judgment Result: Violation.

[0058] Furthermore, through natural language processing technology, keywords including rectification measures, shutdown measures, reporting measures, and inspection measures in power grid operation records and historical violation logs are located, identified, and relationship-extracted, and then converted into response rules for violation event with a standard format, and a response rule ID is assigned to each response rule for violation event.

[0059] The standard format of the response rule for violation event is specifically: Response Rule ID + Violation Behavior Type + Response Measure.

[0060] For example, the following information is extracted from power grid operation records:

[0061] "When the voltage of the switchgear is lower than 400 kV or higher than 800 kV, the equipment should be immediately shut down, the voltage of the equipment should be adjusted, and it can continue to operate only after ensuring that it is restored to the normal range;

[0062] When the current of the equipment is greater than 1000 A, the power supply should be immediately disconnected, and by checking the equipment line and load, ensure that the current returns to the safe range;

[0063] If the temperature of the equipment is lower than -10 °C or higher than 60 °C, the operation of the equipment should be immediately stopped, and temperature control measures should be started to ensure that the equipment can resume normal operation only within the safe temperature range;

[0064] If there is no 3-meter safety warning line set at the operation site, non-staff should be immediately evacuated to a safe area, and a safety warning line should be set up as required as soon as possible to ensure the safety of the operating personnel;

[0065] If the installation distance of the monitoring equipment sensor exceeds 5 meters, it should be immediately reinstalled and ensure that its distance complies with the operation specifications to ensure the accuracy of real-time monitoring."

[0066] By performing keyword location, identification, and relationship extraction, the following response rules for violation event are obtained:

[0067] "Response Rule 1, Violation Behavior Type: The voltage of the switchgear exceeds the specified range (lower than 400 kV or higher than 800 kV) → Response Measure: Immediately shut down the equipment and adjust the voltage of the equipment to restore it between 400 kV and 800 kV, and continue the operation only after confirmation of normality;

[0068] Response Rule 2, Violation Behavior Type: The current of the switchgear is greater than 1000 A → Response Measure: Disconnect the power supply, check the load and line conditions, and restore the power supply only after confirming that the current has returned to the safe range.

[0069] Response Rule 3, Type of Violation: Equipment temperature is lower than -10°C or higher than 60°C → Response Measure: Immediately stop the equipment operation, take temperature control measures, and ensure that the equipment temperature returns to the range of -10°C to 60°C before resuming operation.

[0070] Response Rule 4, Type of Violation: No 3-meter safety warning line is set at the operation site → Response Measure: Immediately evacuate non-staff, set a 3-meter safety warning line, and ensure that operators do not enter the dangerous area.

[0071] Response Rule 5, Type of Violation: The distance of the monitoring equipment sensor exceeds 5 meters → Response Measure: Redeploy the monitoring equipment to ensure that the distance between the sensor and the operation area complies with the specification and ensure real-time and accurate monitoring.

[0072] Furthermore, convert all the extracted violation event determination rules and violation event response rules into machine language based on logical judgment and output them to the knowledge base generation and management module.

[0073] The knowledge base generation and management module is used to integrate all the violation event determination rules and violation event response rules to build and manage the knowledge base related to the power grid.

[0074] Create three sub-databases: the first sub-database, the second sub-database, and the third sub-database. Among them, the first sub-database is used to store the violation event determination rules and their violation IDs; the second sub-database is used to store the violation event response rules and their response IDs; the third sub-database is used to store the mapping relationship between the violation IDs and the response IDs;

[0075] In the first sub-database and the second sub-database, re-access the data integration module every preset update period to discover the newly extracted violation event determination rules and violation event response rules, as well as the amendments to the existing violation event determination rules and violation event response rules, so as to ensure the timely update of the knowledge base.

[0076] Use version control technology to track the rule versions, record each modified and newly added violation event judgment rule and violation event response rule, save the update time when they enter the database, and provide a historical version backtracking window.

[0077] The historical version backtracking window contains the content and update time of each historical version of the violation event determination rule and the violation event response rule, and highlights the modified content of each historical version compared with the previous version.

[0078] In the third sub-database, save the mapping relationship between the violation ID and the response ID through a graph structure.

[0079] Construct a graph structure containing a five - dimensional space in the third sub - database, with each dimension corresponding to a keyword, including: the first dimension: device name; the second dimension: device parameters; the third dimension: environmental conditions; the fourth dimension: numerical range; the fifth dimension: safety conditions;

[0080] Extract the keywords related to the first to fifth dimensions in the violation event determination rules corresponding to each violation behavior ID through a text matching algorithm based on keywords and context, and map them to each specific dimension according to the semantic dictionary to obtain the specific coordinates of the violation behavior ID in each dimension.

[0081] Extract the keywords related to the first to fifth dimensions in the violation event response rules corresponding to each response rule ID through a text matching algorithm based on keywords and context, and map them to each specific dimension according to the semantic dictionary to obtain the specific coordinates of the response rule ID in each dimension.

[0082] Convert the violation behavior ID and the response rule ID into nodes in the graph structure according to the coordinates obtained by mapping, and perform automated association of violation behavior - response rules based on the Euclidean distance between nodes. Calculate the Euclidean distance between each violation behavior ID and all other response rule IDs in the graph structure. Denote the obtained Euclidean distance as the matching similarity between each violation behavior ID and all other response rule IDs, and perform index association between each violation behavior ID and the response rule IDs whose matching similarity is greater than the preset threshold to obtain the mapping relationship between the violation behavior ID and the response rule ID.

[0083] The violation behavior recognition module monitors the power grid operation site in real - time through sensors and image recognition technologies, identifies and judges potential violation behaviors, and ensures that equipment operations and operation processes comply with industry standard specifications.

[0084] Use a high - resolution camera to monitor the power grid operation site in real - time, especially the equipment operation area. Capture the behaviors of the operators at the power grid operation site, the states of the equipment, and the changes in the surrounding environment. Identify potential violation behaviors through image recognition technologies including object detection, action recognition, and scene analysis.

[0085] Through temperature - humidity sensors, voltage, current, device temperature, and device location sensors installed on power grid equipment, obtain the working parameters of the equipment in real - time, which are used to detect whether the equipment is in a normal working state and complies with industry standard specifications.

[0086] Convert the collected potential violation behaviors and the working parameters of the equipment into electronic operation records, and input them into the first sub - database in the knowledge base generation and management module for violation behavior matching and determination to obtain the violation behavior ID and determination result corresponding to the electronic operation record.

[0087] Furthermore, the mapping relationship between the violation behavior ID and the response rule ID in the third sub-database of the knowledge base generation and management module is optimized through a graph neural network, and the positions of each node in the third sub-database are iteratively updated, so that each node not only depends on its own features, but also aggregates information from neighboring nodes through potential connection capture.

[0088] Obtain the information set in the third sub-database: G = (V, E), where V is the set of nodes and E is the set of edges, that is, the set generated by the index association connection relationship between all nodes.

[0089] Set the node update formula:

[0090] where k is the number of iterations, is the coordinate obtained by node v after the (k + 1)-th operation, where is the coordinate obtained by node v after the k-th operation;

[0091] where v is the node number corresponding to the violation behavior, specifically the violation behavior ID; where u is the node number corresponding to the response rule, specifically the response rule ID; where σ is the ReLU non-linear activation function, and the function expression is σ(x) = max(0, x);

[0092] where A vu is an element of the adjacency matrix, representing the index association connection relationship between nodes u and v. Specifically, when there is an index association between nodes u and v, let A vu have a value of 1; otherwise, let A vu have a value of 0;

[0093] where Ν(v) is the set of neighboring nodes of node v, that is, the set formed by all nodes having an index association connection relationship with node v;

[0094] where W (k) is the trainable matrix in the k-th operation, and its specific value is obtained through training and learning, and is used to adjust the weight of each node update; where b (k) is the trainable bias term in the k-th operation, which is used to increase the flexibility of each node update.

[0095] The trainable matrix W (k) and the trainable bias term b (k) The specific training process is as follows:

[0096] In the first operation, the trainable matrix W (0) and the trainable bias term b (0)Initialize the specific value to a random value; in each operation, substitute the trainable matrix and trainable bias term from the previous operation into the set node update formula to update the nodes, and calculate the error between the predicted result and the true label through the cross-entropy loss function. Calculate the gradient of each parameter through backpropagation and update the values of the trainable matrix and trainable bias term.

[0097] The intelligent Q&A module is responsible for implementing intelligent query, identification, and guidance of grid operation violations.

[0098] Collect the user input of violation behavior questions related to grid operations, understand and identify the content through natural language processing, match the corresponding rule ID in the knowledge base generation and management module, and output it to the user, providing detailed violation behavior types and their corresponding countermeasures.

[0099] Furthermore, provide windows for accessing and retrieving the first sub-database and the second sub-database.

[0100] The progress tracking module tracks the progress of each violation behavior ID and determination result in the electronic operation record, continuously records the processing progress of the violation behavior, automatically records each step in the processing process, and provides corresponding solutions through the knowledge base to ensure that each violation behavior is effectively responded to.

[0101] Generate a unique tracking ID for each identified violation behavior. The tracking ID is associated with the violation behavior event determination rule and countermeasure rule of this violation behavior, and the rectification status of this violation behavior is updated in real time. Access the electronic operation record at preset time intervals, record the progress and results of each processing of the violation behavior corresponding to each tracking ID, and update the status flag of this violation behavior in real time. For newly generated violation behaviors, assign them a to-be-processed flag; if it is identified in the electronic operation record that countermeasures are taken for a violation behavior with a to-be-processed flag, adjust its corresponding status flag to being processed; if it is identified in the electronic operation record that a certain violation behavior is resolved, adjust its corresponding status flag to resolved.

[0102] Please refer to Figure 2 As shown, count the rectification progress of violation behaviors through the task status, obtain the number of violation behaviors with to-be-processed flags, being-processed flags, and resolved flags, and generate a rectification progress statistical chart.

[0103] If it is identified that there are violation behaviors that have not obtained a resolved flag within the specified time, automatically send a reminder notice to require the relevant responsible person to handle it as soon as possible.

[0104] It should be understood that the terms “comprising” and “including” as used in the specification and claims of this disclosure indicate the presence of the described features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or their groups.

[0105] It should also be understood that the terminology used herein in the specification of this disclosure is for the purpose of describing particular embodiments only and is not intended to limit the disclosure. As used in the specification and claims of this disclosure, unless the context clearly dictates otherwise, the singular forms “a,” “an,” and “the” are intended to include the plural forms. It should be further understood that the term “and / or” as used in the specification and claims of this disclosure refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations;

[0106] The preferred embodiments of the present invention disclosed above are only used to assist in the explanation of the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only the specific embodiments. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the present invention, so that those skilled in the art can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. An intelligent recognition Q&A system for grid information operation violation behaviors, including a data integration module, a knowledge base generation and management module, and a violation behavior recognition module, characterized in that ； The data integration module is responsible for accessing industry standard specification documents, operation manuals, power grid operation records, and historical violation logs, and performing keyword extraction through natural language processing technology to generate violation event determination rules and violation event response rules with standard formats; The knowledge base generation and management module is responsible for integrating all violation event determination rules and violation event response rules, constructing and managing a knowledge base related to the power grid; rule management is carried out through three sub-databases: the first sub-database stores violation event determination rules; the second sub-database stores response rules; the third sub-database stores the mapping relationship between violation behaviors and response rules; The violation behavior identification module monitors the power grid operation site in real time through sensors and image recognition technology, identifies and judges potential violation behaviors, and ensures that equipment operations and operation processes comply with industry standard specifications; through a graph neural network, it optimizes the mapping relationship between violation behaviors and response rules in the third sub-database of the knowledge base generation and management module, and iteratively updates the positions of each node in the third sub-database, so that each node not only depends on its own features, but also aggregates information from neighboring nodes through potential connection capture.

2. The intelligent recognition Q&A system for illegal acts in power grid information operations according to claim 1, wherein, It also includes an intelligent question answering module and a progress tracking module; The intelligent question answering module identifies and processes questions about power grid operation violations input by users through natural language processing technology; this module matches the user's questions to the response rules in the knowledge base and outputs the types of violation behaviors and corresponding response measures to help users understand how to handle violation behaviors; The progress tracking module generates a unique tracking ID for each violation behavior and its determination result, and records the processing progress of this violation behavior in real time; this module continuously tracks the rectification process of violation behaviors to ensure that each violation behavior is responded to and resolved in a timely manner.

3. An intelligent recognition Q&A system for illegal acts in power grid information operations according to claim 1, characterized in that, The specific process of performing keyword extraction through natural language processing technology is as follows: Through natural language processing technology, keywords including equipment names, equipment parameters, operation requirements, environmental requirements, numerical ranges, and personnel configurations in industry standard specification documents and operation manuals are located, identified, and relationship extracted, and they are converted into violation event determination rules with standard formats, and a violation behavior ID is assigned to each violation event determination rule; Through natural language processing technology, keywords including rectification measures, shutdown measures, reporting measures, and inspection measures in power grid operation records and historical violation logs are located, identified, and relationship extracted, and they are converted into violation event response rules with standard formats, and a response rule ID is assigned to each violation event response rule.

4. The intelligent recognition Q&A system for illegal acts in power grid information operations according to claim 3, characterized in that, The standard formats of violation event determination rules and violation event response rules are specifically as follows: The standard format of violation event determination rules is specifically: violation behavior ID + judgment principle + judgment result; The standard format of violation event response rules is specifically ID: response rule + violation behavior type + response measure.

5. An intelligent recognition Q&A system for illegal behaviors in power grid information operations according to claim 1, characterized in that, The specific process of performing rule management through three sub-databases is as follows: Create three sub - databases: the first sub - database, the second sub - database, and the third sub - database; among them, the first sub - database is used to store the determination rules for violation behavior events and their violation behavior IDs; the second sub - database is used to store the response rules for violation behavior events and their response rule IDs; the third sub - database is used to store the mapping relationship between violation behavior IDs and response rule IDs; In the first sub - database and the second sub - database, access the data integration module again every preset update period to discover the newly extracted determination rules for violation behavior events and response rules for violation behavior events, as well as the corrections to the existing determination rules for violation behavior events and response rules for violation behavior events, so as to ensure the timely update of the knowledge base; Use version control technology to track rule versions, record each modified and newly added determination rule for violation behavior events and response rule for violation behavior events, save the update time when they enter the database, and provide a historical version backtracking window; The historical version backtracking window contains the content and update time of each historical version of the determination rule for violation behavior events and response rule for violation behavior events, and highlights the modified content of each historical version compared with the previous version; In the third sub - database, save the mapping relationship between violation behavior IDs and response rule IDs through a graph structure.

6. An intelligent identification Q&A system for illegal behaviors in power grid information operations according to claim 1, characterized in that, The specific process of saving the mapping relationship between violation behavior IDs and response rule IDs through a graph structure is as follows: Construct a graph structure with a five - dimensional space in the third sub - database, and each dimension corresponds to a keyword, including: the first dimension: device name; the second dimension: device parameters; the third dimension: environmental conditions; the fourth dimension: numerical range; the fifth dimension: safety conditions; Extract the keywords related to the first to fifth dimensions in the determination rule for the violation behavior event corresponding to each violation behavior ID through a text matching algorithm based on keywords and context, and map them to each specific dimension according to the semantic dictionary to obtain the specific coordinates of the violation behavior ID in each dimension; Extract the keywords related to the first to fifth dimensions in the response rule for the violation behavior event corresponding to each response rule ID through a text matching algorithm based on keywords and context, and map them to each specific dimension according to the semantic dictionary to obtain the specific coordinates of the response rule ID in each dimension; Convert the violation behavior ID and response rule ID into nodes in the graph structure according to the mapped coordinates, and perform automated association of violation behavior - response rules based on the Euclidean distance between the nodes, and calculate the Euclidean distance between each violation behavior ID and all other response rule IDs in the graph structure; Record the obtained Euclidean distance as the matching similarity between each violation behavior ID and all other response rule IDs, and perform index association between each violation behavior ID and the response rule ID whose matching similarity is greater than the preset threshold to obtain the mapping relationship between violation behavior IDs and response rule IDs.

7. An intelligent recognition and Q&A system for illegal behaviors in power grid information operations according to claim 1, characterized in that, The specific process of identifying and judging potential violation behaviors is as follows: Use a high-resolution camera to monitor the power grid operation site in real time, especially the equipment operation area; capture the behaviors of the operators, the states of the equipment, and the changes in the surrounding environment at the power grid operation site; identify potential violations through image recognition technologies including object detection, action recognition, and scene analysis; Through temperature and humidity sensors, voltage, current, equipment temperature, and equipment position sensors installed on power grid equipment, obtain the working parameters of the equipment in real time, which are used to detect whether the equipment is in a normal working state and complies with industry standard specifications; Convert the collected potential violations and the working parameters of the equipment into electronic operation records, and input them into the first sub-database in the knowledge base generation and management module to perform violation matching and determination, and obtain the violation ID and determination result corresponding to the electronic operation record.

8. An intelligent recognition Q&A system for illegal behaviors in power grid information operations according to claim 1, characterized in that, The specific process of continuously tracking the rectification process of violations is as follows: Generate a unique tracking ID for each identified violation; the tracking ID is associated with the violation event determination rules and response rules of this violation, and the rectification status of this violation is updated in real time; Access the electronic operation record at preset time intervals, record the progress and results of each handling of the violations corresponding to each tracking ID, and update the status flag of this violation in real time; for newly generated violations, assign them a to-be-handled flag; if it is identified in the electronic operation record that a response measure is taken for a violation with a to-be-handled flag, adjust its corresponding status flag to being handled; if it is identified in the electronic operation record that a certain violation is resolved, adjust its corresponding status flag to resolved; Perform visual display and automatic reminder according to the status flag.

9. An intelligent recognition Q&A system for grid information operation violation behaviors according to claim 1, characterized in that, The specific process of performing visual display and automatic reminder according to the status flag is as follows: Statistically analyze the rectification progress of violations through the task status, obtain the number of violations with to-be-handled flags, being-handled flags, and resolved flags, and generate a rectification progress statistical chart; If it is identified that there are violations that have not obtained a resolved flag within the specified time, automatically send a reminder notice to require the relevant responsible persons to handle them as soon as possible.

Images