A relay protection operation state monitoring system and method

Abstract

The application discloses a kind of relay protection operation state monitoring system and method, the system includes: data processing module, for obtaining the pre-processing equipment image before or after relay protection operation, generates equipment state code;Intelligent analysis engine, it includes: safety measure card execution sequence modeling module, for parsing safety measure ticket text, to construct operation sequence rule base;Guidance operation module, for generating operation step according to operation sequence rule base, to guide user operation;Operation record module, for structured generation user's safety measure operation behavior record;Intelligent state comparison module, for comparing the equipment state difference before and after relay protection operation according to equipment state code, operation sequence rule base and safety measure operation behavior record;Miss recovery early warning module, for triggering miss recovery early warning according to comparison result.Adopt the application can improve the safety and efficiency of relay protection operation.

Description

Technical Field

[0001] This invention relates to the field of power grid safety technology, and in particular to a relay protection operation status monitoring system and method. Background Technology

[0002] In power system maintenance, the safety and reliability of relay protection operations are directly related to the stable operation of the power grid. The correct execution and complete restoration of secondary safety measures are crucial. Secondary safety measures prevent secondary accidents (such as equipment damage or power outages) caused by human error through operations such as isolating equipment and disconnecting circuits. Accurate restoration of these measures after the operation is completed is a key guarantee against leaving any safety hazards. With the expansion and increasing complexity of power systems, the number and types of secondary equipment in relay protection cabinets (such as pressure plates, circuit breakers, and terminal blocks) are increasing, making the execution and restoration process of secondary safety measures increasingly cumbersome.

[0003] Traditional secondary safety measures operations mainly rely on manual labor, involving a process of "drawing review - on-site confirmation - manual recording - operation execution - post-operation verification." In this process, confirming the equipment status before and after the operation is a core step to ensure that safety measures are implemented without omissions or errors. However, this step has long been limited by the inherent defects of manual operation, failing to meet the needs of intelligent power grid operation and maintenance in terms of data accuracy, process efficiency, and risk control. Summary of the Invention

[0004] The purpose of this invention is to provide a relay protection operation status monitoring system and method to improve the safety, efficiency and controllability of relay protection operations.

[0005] To achieve the above objectives, the present invention provides a relay protection operation status monitoring system, including a data processing module, an intelligent analysis engine, and a leakage recovery early warning module;

[0006] The data processing module is used to acquire pre-processed equipment images before or after relay protection operations and generate equipment status codes.

[0007] The intelligent analysis engine includes:

[0008] The Ancuoka execution sequence modeling module is used to parse the safety measure ticket text to build an operation sequence rule base;

[0009] A guided operation module is used to generate operation steps based on the operation sequence rule base to guide the user's operation;

[0010] The operation log module is used to generate structured records of users' security measures and operational behaviors.

[0011] The intelligent status comparison module is used to compare the differences in equipment status before and after relay protection operation based on the equipment status code, the operation sequence rule base and the safety measure operation behavior record;

[0012] The leak recovery early warning module is used to trigger a leak recovery early warning based on the comparison results.

[0013] Optionally, the intelligent state comparison module is used for:

[0014] Obtain the equipment status code before relay protection operation to generate a reference status matrix;

[0015] Obtain the equipment status code after relay protection operation to generate a recovery status matrix;

[0016] A matrix-by-bit comparison algorithm is used to perform difference analysis between the recovery status matrix and the baseline status matrix. When an inconsistency is detected in the status code at the corresponding position, the corresponding device is marked as a suspected device with missing recovery.

[0017] Based on the pre-set execution sequence rule base and the operation records of safety measure cards in the relay protection operation process, the marked suspected leakage recovery equipment is verified to obtain information on the leakage recovery equipment and the status to be restored.

[0018] Optionally, the anthoka execution sequence modeling module is used for:

[0019] The preprocessed security ticket text is segmented into words to obtain a segmented text sequence;

[0020] The device operation object entity is extracted from the segmented text sequence using a trained named entity recognition model.

[0021] Map the operation terms in the security measure ticket text to standard operation types and corresponding operation codes;

[0022] The sequential relationship of operation steps in the security measure ticket text is analyzed using dependency parsing and semantic role labeling techniques.

[0023] Based on the device operation object entity, the standard operation type and corresponding operation code, and the sequence of operation steps, a machine-recognizable instruction set is generated.

[0024] Based on the machine-recognizable instruction set, an operation sequence rule base is constructed.

[0025] Optionally, the data processing module is used for:

[0026] An attention-based object detection model is used to detect targets in equipment images before or after relay protection operations to obtain the target area of ​​the equipment.

[0027] A convolutional neural network is used to classify the device status of the target area of ​​the device to obtain the device status code.

[0028] Optionally, the guiding operation module is used for:

[0029] An operation step list is generated based on the machine-recognizable instruction set in the operation sequence rule base to instruct the user to perform operation steps according to the operation step list; wherein, each operation step includes the device's positioning information, reference image, operation instruction and confirmation mechanism.

[0030] Optionally, the leak recovery early warning module is used for:

[0031] When a difference in equipment status is detected before and after relay protection operation, an audible and visual alarm and a warning message are triggered.

[0032] Based on the differences in device status, a handling suggestion is generated using a pre-set rule engine that governs common operational errors and omissions.

[0033] Optionally, it also includes:

[0034] The image acquisition module is used to acquire equipment images before or after relay protection operations.

[0035] The data processing module is used to preprocess equipment images before or after relay protection operations.

[0036] The communication module is used to transmit preprocessed equipment images before or after relay protection operations to the data processing module.

[0037] Optionally, it also includes a closed-loop management module for performing execution verification after the user performs an operation based on the leak recovery warning.

[0038] Optionally, it also includes a visual monitoring module for displaying the real-time status of the device, a list of operation steps, information on devices that have been missed for recovery and the status that should be recovered.

[0039] To achieve the above objectives, the present invention also provides a relay protection operation status monitoring method, applied to the relay protection operation status monitoring system as described in any of the preceding claims, comprising:

[0040] Collect equipment images before relay protection operations and generate a baseline state matrix;

[0041] The safety measure ticket text is parsed to generate a machine-readable instruction set to guide operators in performing operations and record the operation log.

[0042] Collect equipment images after relay protection operation and generate a recovery state matrix;

[0043] A matrix-by-matrix comparison algorithm is used to perform a difference analysis between the recovered state matrix and the baseline state matrix to obtain the state difference results.

[0044] Based on the state difference results, and using a pre-set execution order rule base and the operation records, a leak recovery warning and recovery status information are triggered.

[0045] Compared with existing technologies, the relay protection operation status monitoring system and method provided by this invention have the following advantages: The data processing module generates standardized equipment status codes through image preprocessing, replacing traditional manual recording. This improves the accuracy and consistency of status data and reduces errors caused by ambiguity or omissions in manual descriptions. In the intelligent analysis engine, the Ancuoka execution sequence modeling module transforms unstructured text into a standardized rule base, and combined with the guided operation module, provides standardized step guidance for operations. This effectively avoids human errors such as reversed operation sequences and reduces the risk of equipment malfunction. The structured recording function of the operation record module enables full traceability of operational behavior. The intelligent status comparison module accurately identifies differences in equipment status before and after operations through multi-dimensional data linkage analysis, improving the efficiency and accuracy of leak recovery detection. The leak recovery early warning module triggers timely warnings based on the comparison results, preventing power grid safety accidents caused by equipment leak recovery, thereby improving the safety, efficiency, and controllability of relay protection operations. Attached Figure Description

[0046] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a structural block diagram of a relay protection operation status monitoring system provided in an embodiment of the present invention;

[0048] Figure 2 This is a flowchart of an anti-slip card execution sequence modeling module provided in an embodiment of the present invention;

[0049] Figure 3 This is a flowchart of the operation of an intelligent state comparison module provided in an embodiment of the present invention;

[0050] Figure 4 This is a flowchart of a relay protection operation status monitoring method provided in an embodiment of the present invention. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] See Figure 1 , Figure 1 This is a structural block diagram of a relay protection operation status monitoring system provided in an embodiment of the present invention. Figure 1 As shown, the relay protection operation status monitoring system includes an image acquisition module 101, a data acquisition module 102, a communication module 103, a data processing module 104, an intelligent analysis engine 105, a leakage recovery early warning module 106, a closed-loop management module 107, and a visualization monitoring module 108.

[0053] It should be noted that the relay protection operation status monitoring system includes a hardware layer, a software layer, and an application layer. The hardware layer includes an image acquisition module 101, a data acquisition module 102, and a communication module 103; the software layer includes a data processing module 104, an intelligent analysis engine 105, and a leakage recovery early warning module 106; and the application layer includes a closed-loop management module 107 and a visual monitoring module 108.

[0054] The image acquisition module 101 is used to acquire equipment images before or after relay protection operations.

[0055] Preferably, the image acquisition module 101 is a sensor array. The image acquisition device for the sensor array can be a 20-megapixel industrial-grade high-definition camera, equipped with a large F / 1.8 aperture and infrared night vision function, enabling clear imaging in low-light environments of ≤5 lux. The camera lens supports 5-50mm electric focal length adjustment, ensuring that minute features such as terminal block numbers and pressure plate markings can be completely captured within a distance of 1-3 meters.

[0056] It should be noted that, in the embodiments of the present invention, the secondary equipment mainly refers to the pressure plates (such as trip output pressure plates, functional pressure plates), circuit breakers (DC / AC power circuit breakers), and terminal blocks (used for wiring connections between devices) within the relay protection cabinet. For example, the "#1 busbar protection cabinet" is a cabinet specifically used to install the relay protection devices and supporting components related to the #1 busbar.

[0057] For example, before relay protection work, the relay protection work status monitoring system triggers a "baseline status acquisition" prompt. The operator (user) can then use a mobile terminal to control a camera to scan the cabinet along a "Z" shaped path, with the scanning step distance set to 50cm and the acquisition frame rate at 30fps. After the relay protection work is completed, a second acquisition is performed using the same parameters to ensure 100% coverage of the equipment status image.

[0058] The data acquisition module 102 is used to preprocess equipment images before or after relay protection operations.

[0059] In one alternative embodiment, the preprocessing includes:

[0060] Remove image noise from equipment images before or after relay protection operations;

[0061] After removing image noise, the device image is enhanced with global contrast and then a multi-level edge detection algorithm is used to outline the device contour.

[0062] By stitching together the partial images of the outlines of each device, a preprocessed image of the device is obtained before or after the relay protection operation.

[0063] For example, a preprocessing pipeline can be built based on the OpenCV vision library (a powerful computer vision library): First, image noise is removed from the acquired device images using 5×5 Gaussian kernel convolution. After caching, histogram equalization (equalizeHist) is used to enhance global contrast. Finally, the Canny multi-level edge detection algorithm (threshold 100-200) is used to extract the device contours, thus effectively solving the imaging blurring problem under complex lighting conditions. The image stitching stage can employ the SIFT (Scale Invariant Feature Transform) feature point matching algorithm. Through feature point detection, descriptor calculation, and matching optimization, multiple frames of local images are stitched into a complete panoramic image. Actual measurements show that the stitching error is controlled within 1mm, achieving seamless coverage of secondary equipment such as display cabinets.

[0064] The communication module 103 is used to transmit the preprocessed equipment image before or after relay protection operation to the data processing module 104.

[0065] For example, the communication module 103 can be a 5G / WiFi 6 dual-mode wireless transmission module built into the image acquisition device and the mobile operating terminal. It is equipped with a wireless transmission component to realize the efficient transmission of structured data output by the data acquisition module to the software layer, while supporting real-time data interaction between the mobile terminal and the system, and completing the communication connection between the hardware layer and the software layer.

[0066] The data processing module 104 is used to acquire pre-processed equipment images before or after relay protection operations and generate equipment status codes.

[0067] The intelligent analysis engine 105 includes:

[0068] The Ancuoka execution sequence modeling module 1051 is used to parse the safety measure ticket text to build an operation sequence rule base;

[0069] The guided operation module 1052 is used to generate operation steps according to the operation sequence rule base to guide the user operation;

[0070] Operation logging module 1053 is used to generate structured records of users' security measures and operational behaviors.

[0071] The intelligent status comparison module 1054 is used to compare the differences in equipment status before and after relay protection operation based on the equipment status code, the operation sequence rule base and the safety measure operation behavior record.

[0072] The leak recovery early warning module 106 is used to trigger a leak recovery early warning based on the comparison results.

[0073] The closed-loop management module 107 is used to perform execution verification after the user executes the operation based on the leak recovery warning;

[0074] The visualization monitoring module 108 is used to display the real-time status of the device, the list of operation steps, the devices that have been missed for recovery, and the status information that should be recovered.

[0075] In one alternative embodiment, the data processing module 104 is configured to:

[0076] An attention-based object detection model is used to detect targets in equipment images before or after relay protection operations to obtain the target area of ​​the equipment.

[0077] A convolutional neural network is used to classify the device status of the target area of ​​the device to obtain the device status code.

[0078] For example, a high-precision device recognition model based on YOLOv12 was first constructed. Over 100,000 on-site images containing various device models were finely annotated. Utilizing the powerful feature extraction capabilities of the YOLOv12 model, millimeter-level precise positioning of device type and spatial location was achieved, yielding the target device area. After multiple rounds of field testing, the recognition accuracy remained consistently above 99%.

[0079] Then, a convolutional neural network (such as CNN) is used to classify the device status of the target region, mainly including:

[0080] Pressure plate status: By capturing the "protrusion / concavity" shape characteristics of the pressure plate, the protrusion status is encoded as 1 (entry) and the concavity status is encoded as 0 (exit).

[0081] Circuit breaker status: Based on the handle angle recognition algorithm, 0° corresponds to the open state (code 0), and 90° corresponds to the closed state (code 1).

[0082] Terminal block status: Based on the wire connection characteristics detection, the code for a wire connection is 1 (connected), and the code for no wire connection is 0 (disconnected).

[0083] Finally, backup status codes are generated and integrated into a standardized status code matrix to provide structured data support for subsequent intelligent comparison.

[0084] In one optional embodiment, the Anzoka execution sequence modeling module 1051 is specifically used for:

[0085] A1. Segment the preprocessed security ticket text into words to obtain a segmented text sequence;

[0086] A2. Extract the device operation object entity from the segmented text sequence using the trained named entity recognition model;

[0087] A3. Map the operation terms in the security measure ticket text to standard operation types and corresponding operation codes;

[0088] A4. The sequential relationship of operation steps in the security measure ticket text is analyzed using dependency parsing and semantic role labeling techniques;

[0089] A5. Generate a machine-recognizable instruction set based on the device operation object entity, the standard operation type and corresponding operation code, and the sequence of operation steps;

[0090] A6. Based on the machine-recognizable instruction set, construct an operation sequence rule base;

[0091] See Figure 2 , Figure 2 This is a flowchart of an anti-slip card execution sequence modeling module provided in an embodiment of the present invention.

[0092] For example, the safety measure ticket text (hereinafter referred to as the safety measure ticket text) is as follows:

[0093] "Work task: Regular inspection of the #1 busbar protection panel."

[0094] Safety measures implementation steps:

[0095] 1. First, engage the #1 busbar protection panel function switch 1 and confirm that the device power indicator light is on.

[0096] 2. Next, exit the trip output pressure plate 3 of the #1 bus protection screen and record the current status as "in operation".

[0097] 3. Then, disconnect the DC circuit breaker of the #1 bus protection panel by 2 steps to verify that the device has no voltage output.

[0098] 4. Finally, short-circuit terminal block 5 of the #1 busbar protection panel to complete the current loop isolation.

[0099] First, the original text of the safety check ticket is cleaned to remove redundant punctuation, special characters, and invalid spaces, and the text encoding format is standardized. At the same time, a word segmentation tool is used to break long sentences into word sequences and perform part-of-speech tagging. For example, “#1 busbar protection screen trip outlet pressure plate 3” is tagged as “equipment name” and “put into operation” is tagged as “operation verb” to avoid ambiguity.

[0100] For example, see Table 1, which shows the word segmentation and word tagging results of the security measure ticket text sample mentioned above.

[0101] Table 1. Word segmentation results and part-of-speech tagging results for a sample security measure ticket text.

[0102]

[0103] Then, using the trained named entity recognition model, relevant entities of the device operation objects are extracted from the segmented text, such as the protection panel name, pressure plate number, and switch name, and assigned a unique device ID. See Table 2; for example, recognizing "#1 busbar protection panel trip outlet pressure plate 3" generates the corresponding device ID "BP-001-P3". It should be noted that the device ID uses the format "protection panel identifier-device type-number" to ensure uniqueness.

[0104] Table 2 Entity Extraction Results of Safety Measures Ticket Text

[0105]

[0106] Then, an operation type classifier is constructed to map operation terms in the text to standard operation types, such as "engage," "exit," "open," and "close," and assign corresponding operation codes. See Table 3; for example, "engage" corresponds to operation code "OP-01," and "exit" corresponds to operation code "OP-02." Operation codes can also be extended to include operation types (such as "short-circuit"), which can be supplemented by custom codes.

[0107] Table 3. Text manipulation vocabulary mapping results for security measure tickets

[0108]

[0109] Then, dependency parsing and semantic role labeling techniques are used to analyze the sequential relationship of operation steps in the text. If logical connectors such as "first...then..." or "firstly...then..." appear, the operation order is clarified; if there are no explicit connectors, the operation order is inferred through event timestamps, action logical relationships, etc., and finally a machine-recognizable instruction set containing device ID, operation code and sequential relationship is formed, as shown in Table 4.

[0110] Table 4 shows the sequence of steps in the security measure ticket text operation and the set of machine-readable instructions.

[0111]

[0112] It's important to note that in a machine-readable instruction set, "null" indicates that the step has no preceding step. Taking the step with "Device ID" "BP-001-P1" as an example, it is the starting step of the entire operation process. There are no other operation steps that need to be completed before it; therefore, its "preceding step" field is marked as "null" as an identifier of the process's initiation. Subsequent steps are then linked numerically to clearly present the sequential dependencies between steps. It's worth noting that the operation sequence is logically defined through explicit connectors, generating an instruction set that includes preceding relationships, providing a machine-readable process benchmark for subsequent intelligent execution. Therefore, a machine-readable instruction set is a structured data set containing unique device identifiers, operation codes, and timing constraints.

[0113] Furthermore, an operation sequence rule base is constructed based on the machine-recognizable instruction set.

[0114] It should be noted that the operation sequence rule base aims to integrate standardized operation logic under different work scenarios, transforming the operation specifications of various equipment into a set of machine-recognizable instructions that includes unique equipment identifiers, operation codes, and timing constraints.

[0115] For example, in the field of pressure plate operation: a special instruction subset is constructed based on the core specification of "interlocking control of pressure plates in the same group". It is clear that the predecessor operation of "exiting 'trip output pressure plate' (device ID: BP-002)" must be "exiting 'function pressure plate' (device ID: BP-001, operation code: OP-02)". The corresponding instruction set rule is: `[{"device ID":"BP-002","operation code":"OP-02","predecessor constraint":"BP-001 completes OP-02"}]`.

[0116] In the field of circuit breaker status: a status quantification benchmark is established through a subset of binary encoded instructions, forming an instruction set rule: `[{"Equipment Type":"Circuit Breaker","Opening Status Code":0,"Closing Status Code":1,"Status Transition Instruction":{"Opening→Closing":"OP-03","Closing→Opening":"OP-04"}}]`, to achieve full-domain standardization of equipment status definition.

[0117] In the field of terminal block operation: Based on the connection status mark, a subset of safe operation instructions is constructed. It is clear that the "terminal block disconnection" operation must meet the mandatory logic of "power test before operation". The instruction set rule is: `[{"Device ID":"TP-005","Operation Code":"OP-05 (disconnection)","Predecessor Condition":"Power test operation record completed (REC-001)"}]`.

[0118] The operation sequence rule base dynamically integrates machine-recognizable instruction sets from multiple device types and operation scenarios to achieve full coverage of operation sequence for various devices, providing standardized instruction benchmarks for intelligent guidance and status comparison analysis of subsequent work processes.

[0119] In addition, non-standard equipment operation rules (such as old cabinet equipment) can be manually entered, and the rule base can be iteratively optimized through historical operation data (the rule weight is updated once every 100 operation data are accumulated) to ensure that the rule coverage of the on-site scenario is ≥98%.

[0120] In one optional embodiment, the guiding operation module 1052 is specifically used for:

[0121] An operation step list is generated based on the machine-recognizable instruction set in the operation sequence rule base to instruct the user to perform operation steps according to the operation step list; wherein, each operation step includes the device's positioning information, reference image, operation instruction and confirmation mechanism.

[0122] Specifically, the guided operation module 1052 adopts a lightweight interactive design paradigm, dynamically generating a standardized list of operation steps based on a machine-recognizable instruction set. Each operation step node contains four core pieces of information about the operated equipment: ① Positioning information: precise coordinate positioning is used (e.g., row 3, column 2 of cabinet #1) to achieve accurate mapping of the equipment's spatial position; ② Reference image: pre-stored real-time images of the equipment's reference state before operation, providing a visual reference for operation comparison; ③ Operation instructions: clearly define the operation execution standards (e.g., "switch the 'overcurrent protection' pressure plate to the recessed state"), standardizing the boundaries of operation actions; ④ Confirmation mechanism: a "Confirm Operation Completion" interactive button is configured. After being clicked, the system automatically records the operation timestamp accurate to the second and the operator's unique identifier, constructing a closed loop for the entire operation log.

[0123] In one optional embodiment, before the user executes the operation steps according to the operation step list, it can be determined whether the content of the safety measure card has been dynamically adjusted (such as adding steps or changing equipment). If so, a dual reminder is triggered: ① Visual warning: the changed content is highlighted in red; ② Pop-up prompt: a notification is immediately pushed that "the safety measure card has been updated, please check it again before execution" to ensure that the operator obtains the latest information in a timely manner.

[0124] In one optional embodiment, when the user performs operation steps according to the operation step list, a built-in intelligent detection engine can also be used to monitor the operation process in real time. Specifically, this includes:

[0125] If a skipped operation is detected (such as skipping step 2 and directly executing step 3), a warning window containing a device location diagram and operation diagram will immediately pop up, forcibly interrupting the violation until the process returns to normal.

[0126] If the target object is misjudged (such as incorrectly operating pressure plate B instead of the target pressure plate A), the blocking mechanism will be triggered and visual guidance will be provided to help operators quickly locate and correct the problem.

[0127] Furthermore, the user's security action behavior records are generated in a structured manner through the operation record module 1053. For example, each structured operation record contains seven core fields:

[0128] Operation ID: Uniquely identifies a single operation;

[0129] Unique equipment code: follows the format of cabinet ID-equipment type-serial number (e.g., P1-pressure plate-005);

[0130] Pre-operation status code / Post-operation status code: Digital identifier of equipment status;

[0131] Operation time: accurate to the second with timestamp;

[0132] Operator ID: Associated with the identity of the operator;

[0133] Terminal ID: Records information about the operating device.

[0134] Preferably, a hierarchical storage strategy is adopted to record structured data of operation behavior, such as storing it in the local database of the mobile terminal, thereby supporting rapid retrieval and analysis. The original images of the device can be named using the job ID + timestamp + device code (e.g., J20250801-1023-P1-Blow-002.jpg) and stored in the system's local storage directory, achieving precise association between job images and operation records.

[0135] In addition, it is recommended to set up data security protections during implementation. For example, automatically attach a digital signature and timestamp based on the SHA-256 algorithm to each record when it is generated to implement an anti-tampering mechanism; at the same time, modification operations should be authorized electronically by both the operator and the guardian; and all modification actions should be recorded in a complete log, including key information such as the content before and after the modification, the operator, and the operation time, to ensure that the data is traceable throughout its entire lifecycle.

[0136] In one optional embodiment, the intelligent state comparison module 1054 is specifically used for:

[0137] B1. Obtain the equipment status code before relay protection operation to generate a reference status matrix;

[0138] See Figure 3 , Figure 3 This is a flowchart illustrating the workflow of an intelligent state comparison module provided in an embodiment of the present invention. Figure 3 As shown, before the relay protection operation is initiated, the data processing module 104 performs preprocessing operations such as noise reduction and enhancement on the acquired equipment images. It then identifies the equipment status code using a feature extraction algorithm to obtain the equipment status code before the relay protection operation. A baseline status matrix (denoted as ) is then constructed based on this equipment status code. ).

[0139] In this embodiment of the invention, the baseline status matrix is ​​stored in the structure format of [Device 1: Status Code, Device 2: Status Code, ..., Device n: Status Code], where 1 represents that the device is in normal operation and 0 represents that there is an abnormal condition. The generated baseline status matrix is ​​synchronously stored in the local database of the mobile terminal as a reference benchmark for subsequent comparisons.

[0140] B2. Obtain the equipment status code after relay protection operation to generate a recovery status matrix;

[0141] like Figure 3 As shown, after the relay protection operation is completed, the system triggers a secondary image acquisition and processing flow to obtain the equipment status code before the relay protection operation. Using the same data processing link as when the baseline state was constructed, a recovery state matrix (denoted as...) is generated. Maintain data format and Consistency is ensured to guarantee the compatibility and accuracy of subsequent comparison processes.

[0142] B3. A matrix-by-bit comparison algorithm is used to perform difference analysis between the recovery status matrix and the baseline status matrix. When an inconsistency is detected in the status code at the corresponding position, the corresponding device is marked as a suspected device with missing recovery.

[0143] like Figure 3 As shown, a matrix-by-matrix comparison algorithm is used to compare the matrix bit by bit. and Perform automated difference analysis. When an inconsistency in the status code at the corresponding location is detected (i.e., ... If the device is found to be in a suspected recovery failure, it should be immediately marked as such and included in the scope of key investigations.

[0144] B4. Based on the pre-set execution sequence rule base and the operation records of safety measure cards in the relay protection operation process, the marked suspected missing recovery equipment is verified to obtain information on the missing recovery equipment and the status to be restored.

[0145] It is worth noting that the intelligent status comparison module 1054 is the core functional unit of the system. Relying on the built-in execution order rule library and combined with the operation records of the safety measure cards in the work process, it performs secondary verification on the initially marked suspected equipment. It can eliminate interference from status changes of equipment not involved in the operation and effectively avoid the risk of misjudgment.

[0146] like Figure 3 As shown, the visualization monitoring module 108 can also accurately mark the devices that failed to recover with a highlighted red border in the panoramic equipment image interface, and display detailed information on the "status to be recovered" of the devices in real time (e.g., "circuit breaker needs to be closed to the 90° position"). Simultaneously, a structured difference report is automatically generated, comprehensively covering core data such as the physical location of the equipment, its current operating status, and standard recovery status. The report will be automatically archived to the mobile terminal, providing detailed evidence for subsequent review and traceability.

[0147] In one optional embodiment, the leakage recovery early warning module 106 includes:

[0148] When a difference in equipment status is detected before and after relay protection operation, an audible and visual alarm and a warning message are triggered.

[0149] Based on the differences in device status, a handling suggestion is generated using a pre-set rule engine that governs common operational errors and omissions.

[0150] For example, when the intelligent status comparison module 1054 detects one or more missing recovery devices, the missing recovery early warning module 106 immediately activates a three-level linkage early warning, triggering level by level in the order of local early warning to monitoring early warning. If the mobile terminal held by the operator activates the high-frequency alarm program, the 85dB buzzer will sound continuously for 30 seconds, and the screen will simultaneously activate red light flashing and vibration prompts; the alarm interface will automatically be placed on top, displaying the list of missing recovery devices in real time, and presenting a spatial distribution map of the devices through sub-meter positioning technology (accuracy ±3cm).

[0151] At the same time, a warning text message containing information about the work scenario, details of the missing recovery equipment, and real-time images of the site is sent to the monitoring personnel, along with a 15-minute emergency response time limit to ensure timely risk response.

[0152] In addition, a built-in rule engine for common operational errors and omissions can be implemented to accurately identify leakage recovery issues and automatically generate targeted handling suggestions. This rule engine integrates common error-prone scenarios in power operations, equipment operating characteristics, and standardized handling procedures. Through preset "problem phenomenon-cause correlation-solution" mapping rules, it improves the accuracy and compliance of handling suggestions. For example, when leakage recovery pressure plate X is detected, based on historical error-prone data matching, the following judgment is made: [Leakage recovery pressure plate X: Rule matching determines that the operating force did not reach the standard threshold, resulting in a half-engaged state. It is recommended to apply a uniform pushing force along the preset "engagement" direction of the equipment until mechanical locking feedback is triggered (presenting a stable protruding state)]. Simultaneously, a visual guidance diagram containing operational step breakdowns and force application point annotations is pushed, intuitively presenting the standardized operating trajectory. This can effectively avoid secondary problems caused by non-standard operation and strengthen the standardization and safety of leakage recovery handling.

[0153] After the operators complete the rectification, a secondary image acquisition and re-verification are triggered through the closed-loop management module 107. Using the YOLOv12 target detection algorithm, and through multi-scale feature fusion and an improved attention mechanism, sub-pixel-level localization of key equipment components is achieved. Combined with temporal comparison analysis technology, changes in equipment status parameters are accurately identified. When the status parameters of each component are detected to be completely matched with the state before the rectification, the equipment is deemed to have returned to normal, and the warning is immediately lifted. A closed-loop management report containing before-and-after image comparisons and records of the handling process is generated in real time and stored in the local database of the mobile terminal using the AES-256 encryption algorithm.

[0154] Based on the above system items, the present invention provides corresponding embodiments of the method items.

[0155] See Figure 3 , Figure 3 This is a flowchart of a relay protection operation status monitoring method provided by an embodiment of the present invention. The relay protection operation status monitoring method includes steps S1 to S2:

[0156] S1. Collect equipment images before relay protection operation and generate a baseline state matrix;

[0157] S2. Parse the safety measure ticket text to generate a machine-readable instruction set to guide operators in performing operations and record operation logs;

[0158] S3. Collect equipment images after relay protection operation and generate a recovery state matrix;

[0159] S4. Using a matrix-by-matrix comparison algorithm, perform a difference analysis between the recovered state matrix and the baseline state matrix to obtain the state difference results.

[0160] S5. Based on the state difference results, and using the preset execution order rule base and the operation records, trigger a missed recovery warning and the state information to be recovered.

[0161] It should be noted that the relay protection operation status monitoring system provided in this embodiment of the invention is used to execute all the process steps of the relay protection operation status monitoring method in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0162] In summary, the specific implementation process of the relay protection operation status monitoring system provided by the embodiments of the present invention is as follows:

[0163] (1) Preparation before the operation:

[0164] After the operators import the safety measure card, the system uses the safety measure card execution sequence modeling module to parse the safety measure ticket text and generate standardized operating rules. The operators use a mobile terminal camera to collect equipment images before the operation. After image preprocessing, feature recognition and other technical processing, a baseline state matrix is ​​generated and stored in the mobile terminal's local database.

[0165] (2) Execution during the operation:

[0166] The guided operation module displays the operation process step by step on the guided interface according to the operation rules. After completing each step, the operator clicks "confirm". The operation recording module captures the operation behavior in real time and records key information such as operation time, operation content, and equipment status codes before and after the operation.

[0167] (3) Post-work verification:

[0168] The recovered image data, acquired using camera acquisition devices, is processed to generate a recovery state matrix. The intelligent state comparison module performs pixel-level and logical-level dual comparisons between the recovered state matrix and the baseline state matrix, marking devices that were missed in the recovery process and generating a visual difference analysis report.

[0169] (4) Early warning and rectification:

[0170] The leak recovery early warning module triggers local audible and visual alarms and remote push notifications, simultaneously issuing a rectification work order containing the risk level and handling plan. After the operators complete the rectification, the system performs a secondary verification. Upon successful verification, it automatically generates a closed-loop report and encrypts and stores it on the mobile terminal, forming a traceable work file.

[0171] In summary, the relay protection operation status monitoring system and method provided by this invention have the following advantages: The data processing module generates standardized equipment status codes through image preprocessing, replacing traditional manual recording. This improves the accuracy and consistency of status data and reduces errors caused by ambiguity or omissions in manual descriptions. In the intelligent analysis engine, the Ancuoka execution sequence modeling module transforms unstructured text into a standardized rule base, and combined with the guided operation module, provides standardized step guidance for operations. This effectively avoids human errors such as reversed operation sequences and reduces the risk of equipment malfunction. The structured recording function of the operation record module enables full traceability of operational behavior. The intelligent status comparison module accurately identifies differences in equipment status before and after operations through multi-dimensional data linkage analysis, improving the efficiency and accuracy of leak recovery detection. The leak recovery early warning module triggers timely warnings based on the comparison results, preventing power grid safety accidents caused by equipment leak recovery, thereby improving the safety, efficiency, and controllability of relay protection operations.

[0172] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A relay protection operation state monitoring system characterized by comprising: The device comprises a data processing module, an intelligent analysis engine and a missing recovery early warning module. The data processing module is configured to acquire a pre-treatment device image before or after relay protection operation, and generate a device state code. The intelligent analysis engine comprises: A safety measure card execution sequence modeling module configured to parse a safety measure card text to construct an operation sequence rule base; A guiding operation module configured to generate operation steps according to the operation sequence rule base to guide user operation; An operation record module configured to generate a structured safety measure operation behavior record of a user; An intelligent state comparison module configured to compare device state differences before and after relay protection operation according to the device state code, the operation sequence rule base and the safety measure operation behavior record; The missing recovery early warning module is configured to trigger a missing recovery early warning according to a comparison result. The intelligent state comparison module is specifically configured to: Acquire a device state code before relay protection operation to generate a baseline state matrix; Acquire a device state code after relay protection operation to generate a recovery state matrix; Use a matrix bit-by-bit comparison algorithm to analyze differences between the recovery state matrix and the baseline state matrix, and when detecting that state codes at corresponding positions are inconsistent, mark the corresponding device as a suspected missing recovery device; Verify the suspected missing recovery device based on a pre-set execution sequence rule base and safety measure card operation records in a relay protection operation process to obtain missing recovery device and recovery state information.

2. The protective relay operating state monitoring system of claim 1, wherein, The safety measure card execution sequence modeling module is configured to: Carry out word segmentation on a pre-processed safety measure card text to obtain a segmented text sequence; Extract device operation object entities from the segmented text sequence through a trained named entity recognition model; Map operation words in the safety measure card text into standard operation types and corresponding operation codes; Analyze operation step sequence relationships in the safety measure card text through dependency syntax analysis and semantic role labeling technology; Generate a machine recognizable instruction set according to the device operation object entities, the standard operation types and corresponding operation codes, and the operation step sequence relationships; Construct an operation sequence rule base based on the machine recognizable instruction set.

3. The protective relay operating state monitoring system of claim 1, wherein, The data processing module is configured to: Use an object detection model based on an attention mechanism to detect targets in a device image before or after relay protection operation to obtain a device target region; Use a convolutional neural network to classify device states in the device target region to obtain a device state code.

4. The protective relay operating state monitoring system of claim 1, wherein, The guiding operation module is configured to: Generate an operation step list based on the machine recognizable instruction set in the operation sequence rule base to instruct a user to perform operation steps according to the operation step list; wherein each operation step comprises positioning information of a device, a baseline image, an operation instruction and a confirmation mechanism.

5. The protective relay operating state monitoring system of claim 1, wherein, The missing recovery early warning module is configured to: Trigger an audible and visual alarm and push pre-warning information when detecting that there are differences in device states before and after relay protection operation; Generate a treatment suggestion based on a pre-set operation error-prone and missing-prone rule engine according to the device state differences.

6. The protective relay operating state monitoring system of claim 1, wherein, The device further comprises: An image acquisition module is configured to acquire an image of the device before or after the relay protection operation; A data processing module is configured to preprocess the image of the device before or after the relay protection operation; A communication module is configured to transmit the preprocessed image of the device before or after the relay protection operation to the data processing module.

7. The protective relay operating state monitoring system of claim 1, wherein, A closed-loop management module is further included to perform verification after the user performs the operation according to the missing recovery early warning.

8. The protective relay operating state monitoring system of claim 1, wherein, A visual monitoring module is further included to display the real-time state of the device, the operation step list, the missing recovery device, and the should-recovery state information.

9. A method of monitoring the state of a protective relay operation, characterized by, The application is applied to the relay protection operation state monitoring system as claimed in any one of claims 1 to 8, and includes: An image of the device before the relay protection operation is acquired to generate a reference state matrix; A safety measure ticket text is parsed to generate a machine-recognizable instruction set to guide the operator to perform the operation and record the operation record; An image of the device after the relay protection operation is acquired to generate a recovery state matrix; A matrix bit-by-bit comparison algorithm is adopted to perform difference analysis on the recovery state matrix and the reference state matrix to obtain a state difference result; According to the state difference result, a pre-set execution sequence rule library and the operation record are used to trigger a missing recovery early warning and should-recovery state information.

Images