A method and system for intelligent operation and maintenance monitoring and management of water conservancy facilities

Abstract

The application is suitable for the technical field of water conservancy projects, and provides a water conservancy project facility intelligent operation and maintenance monitoring management method and system, obtains a patrol inspection plan of a water conservancy project facility, generates a patrol inspection task according to the patrol inspection plan, generates a patrol inspection two-dimensional code according to the patrol inspection task, scans the patrol inspection two-dimensional code through a mobile terminal, compares on-site information with the patrol inspection two-dimensional code, pre-processes a hidden danger on-site map, identifies a hidden danger category in an image, judges hidden danger severity according to the hidden danger category, matches a corresponding approval process template according to the hidden danger severity, sends the process template to a corresponding terminal, uploads a rectification result through the mobile terminal, accepts the rectification result, uploads points according to a closed-loop hidden danger, and has the advantages of a complete technical chain from intelligent patrol inspection, automatic grading, dynamic routing, closed-loop acceptance to point incentives.

Description

Technical Field

[0001] This invention belongs to the field of water conservancy engineering technology, and in particular relates to a method and system for intelligent operation and maintenance monitoring and management of water conservancy engineering facilities. Background Technology

[0002] With the continuous advancement of the national water network construction and the ongoing expansion of water conservancy infrastructure, the supervision of water conservancy project safety production faces severe challenges, including a wide scope, long operational lines, and numerous risk points. The traditional supervision model, relying on manual inspections, paper-based records, and on-site verification, is no longer sufficient to meet the precise and efficient requirements of modern water conservancy project management, exhibiting pain points such as poor data flow, difficulty in tracing hazard records, and inadequate rectification closure. In recent years, some water conservancy projects have begun to introduce information management methods, such as using WeChat mini-programs for online hazard entry and process management, utilizing mobile apps to complete the closed-loop process of "reporting—review—rectification—acceptance—closure," or installing QR codes at key project locations for maintenance personnel to scan and report inspection findings.

[0003] However, the aforementioned information technology practices still have the following shortcomings at the technical level. Regarding hazard identification, while existing systems support uploading photos, the determination of hazard categories and severity levels still largely rely on the personal experience and subjective judgment of reporting and reviewing personnel. This results in inconsistent grading standards and low processing efficiency. Even in systems that have introduced image recognition technology, most only detect single-category defects, lacking the multi-task processing capability to simultaneously perform category identification, severity grading, and area localization of hazard images. In terms of process management, existing systems' approval processes are mostly based on preset fixed paths. Once a hazard is reported, it enters a unified review chain, unable to automatically match differentiated approval templates and responsible persons based on the severity of the hazard. This leads to a mismatch between the handling authority and approval nodes for major and general hazards, resulting in rigid processes and low efficiency. Regarding the application of QR codes in inspections, most QR codes used in existing water conservancy projects are static and fixed, serving only as information display entry points or fixed check-in points. Their encoded information does not include dynamic elements such as task ID, validity period, and responsible person ID. The system cannot perform real-time comparison and verification of scanning location, scanning time, and the identity of the scanning personnel, making it difficult to effectively prevent violations such as reporting from different locations, overdue reporting, and proxy signing. As for incentive points, although some projects have attempted to introduce safety point mechanisms, point calculation still relies mainly on manual registration and entry, lacking automated multi-dimensional calculation capabilities. Point rules typically only consider the single dimension of reporting behavior, failing to comprehensively reflect factors such as hazard level, completion timeliness, and contribution quality, resulting in limited incentive effects.

[0004] In summary, existing water conservancy project operation and maintenance monitoring and management technologies have significant shortcomings in areas such as intelligent identification of hidden dangers, dynamic adaptation of processes, spatiotemporal binding of QR codes, and automated management of points. There is an urgent need for a smart operation and maintenance monitoring and management method that can achieve the coordinated operation of the above functions. Summary of the Invention

[0005] The purpose of this invention is to provide a smart operation and maintenance monitoring and management method for water conservancy engineering facilities, which aims to solve the problems raised in the third part of the background technology.

[0006] The present invention is implemented as follows: a method for intelligent operation and maintenance monitoring and management of water conservancy engineering facilities, the method comprising: Obtain the inspection plan for water conservancy engineering facilities, and generate inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information. An inspection QR code is generated based on the inspection task. The inspection QR code is scanned by a mobile terminal, and the on-site information is compared with the inspection QR code. If a match is found, the on-site image of the potential hazard is taken by the mobile terminal, and a text description is filled in before the data is submitted for reporting. Preprocessing is performed on the on-site images of potential hazards to identify the types of hazards in the images. The severity of the hazards is determined based on the types of hazards, and the corresponding approval process template is matched based on the severity of the hazards. The process template is then sent to the corresponding terminal. The rectification results are uploaded via mobile terminal, and the rectification results are accepted. The acceptance involves comparing the hidden danger areas to determine whether the hidden dangers have been eliminated, obtaining preset point rules for different types of contribution behaviors, and uploading points based on the closed-loop hidden dangers.

[0007] Preferably, the steps of generating an inspection QR code based on the inspection task, scanning the inspection QR code with a mobile terminal, comparing the on-site information with the inspection QR code, and if a match is found, taking an image of the potential hazard site with the mobile terminal, filling in a text description, and submitting the reported data specifically include: An inspection QR code is generated based on the inspection task. The inspection QR code includes a timestamp, validity period and responsible person ID. The inspection QR code is scanned by a mobile terminal. Compare the on-site information with the inspection QR code; if a mismatch is found, access is denied and an anomaly log is recorded. If a match is found, the user takes an image of the potential hazard site using a mobile terminal, fills in a text description, and submits the reported data. The reported data includes the site image, text description, reporter ID, reporting time, GPS coordinates, and the corresponding inspection task ID.

[0008] Preferably, the steps of preprocessing the site hazard image, identifying the hazard category in the image, determining the severity of the hazard based on the hazard category, matching the corresponding approval process template based on the severity of the hazard, and sending the process template to the corresponding terminal specifically include: Preprocessing is performed on the site map of the potential hazard. The preprocessing includes size normalization, illumination equalization and noise filtering to obtain the preprocessed image. Identify the category of potential hazards in the image, and determine the severity of the hazard based on the category. The severity includes major hazards, general hazards, minor hazards, and no hazard. Match the corresponding approval process template according to the severity of the potential hazard, and send the process template to the corresponding terminal.

[0009] Preferably, the steps of uploading rectification results via mobile terminal, accepting the rectification results, comparing the hazard areas to determine whether the hazards have been eliminated, obtaining preset point rules for different types of contribution behaviors, and uploading points based on the closed-loop hazards specifically include: The rectification results are uploaded via mobile terminal. The rectification results include on-site images and rectification descriptions after rectification. The rectification results are then accepted, and the acceptance involves comparing the areas with potential hazards. Determine whether the hidden danger has been eliminated. If the acceptance is deemed unsuccessful, the process is reverted to the rectification stage, and the original person responsible for rectification is required to rectify it again. If the acceptance is deemed successful, the process ends, and the hidden danger status is updated to closed loop. The system acquires preset point rules for different types of contribution behaviors, including hazard reporting, hazard rectification, knowledge base upload, and knowledge base review, and awards points based on hazard uploads that have been closed-loop.

[0010] Preferably, the categories of potential hazards include body cracks, seepage, landslides, piping, equipment failures, structural deformation, and abnormal water quality.

[0011] Another objective of this invention is to provide an intelligent operation and maintenance monitoring and management system for water conservancy engineering facilities, the system comprising: The inspection task module obtains the inspection plan of water conservancy engineering facilities and generates inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information. The on-site comparison module generates an inspection QR code based on the inspection task. The on-site information is compared with the inspection QR code by scanning the inspection QR code with a mobile terminal. If a match is found, the on-site image of the potential hazard is taken by the mobile terminal, and a text description is filled in before the data is submitted for reporting. The hazard assessment module preprocesses the hazard site image, identifies the hazard category in the image, determines the severity of the hazard based on the hazard category, matches the corresponding approval process template based on the severity of the hazard, and sends the process template to the corresponding terminal. The points module allows users to upload rectification results via mobile terminals and verify the results. The verification process involves comparing potential hazard areas to determine whether the hazards have been eliminated, obtaining preset points rules for different types of contribution behaviors, and uploading points based on closed-loop hazards.

[0012] Preferably, the on-site comparison module includes: The QR code scanning unit generates an inspection QR code based on the inspection task. The inspection QR code includes a timestamp, validity period, and responsible person ID. The inspection QR code is scanned by a mobile terminal. The comparison unit compares the on-site information with the inspection QR code. If a mismatch is determined, access is denied and an anomaly log is recorded. If a match is found, the data reporting unit will take a picture of the potential hazard site using a mobile terminal, fill in a text description, and then submit the reported data. The reported data includes the site image, text description, reporter ID, reporting time, GPS coordinates, and the corresponding inspection task ID.

[0013] Preferably, the hazard assessment module includes: The image preprocessing unit performs preprocessing based on the hazard site image. The preprocessing includes size normalization, illumination equalization, and noise filtering to obtain the preprocessed image. The hazard category unit identifies the hazard category in the image and determines the severity of the hazard based on the hazard category. The severity includes major hazard, general hazard, minor hazard, and no hazard. The template unit matches the corresponding approval process template based on the severity of the potential hazard and sends the process template to the corresponding terminal.

[0014] Preferably, the integration module includes: The acceptance unit uploads the rectification results via a mobile terminal. The rectification results include on-site images and rectification descriptions after rectification. The acceptance process involves comparing the areas with potential hazards. The hazard elimination unit determines whether the hazard has been eliminated. If the acceptance is deemed unsuccessful, the process is rolled back to the rectification node, and the original person responsible for rectification must rectify it again. If the acceptance is deemed successful, the process ends, and the hazard status is updated to closed loop. The points unit obtains preset points rules for different types of contribution behaviors. The points rules include hazard reporting, hazard rectification, knowledge base upload, and knowledge base review. Points are uploaded based on closed-loop hazard uploads.

[0015] Preferably, the categories of potential hazards include body cracks, seepage, landslides, piping, equipment failures, structural deformation, and abnormal water quality.

[0016] This invention provides a smart operation and maintenance monitoring and management method for water conservancy engineering facilities. It uses a multi-task convolutional neural network to simultaneously perform category recognition, level classification, and region segmentation on hazard images, combined with semantic fusion verification, to eliminate subjective differences in manual judgment and effectively solve the problems of hazard classification relying on experience, inconsistent standards, and low efficiency. Simultaneously, it constructs a dynamic routing engine based on a directed acyclic graph, automatically matching differentiated approval templates according to the four hazard levels, enabling dynamic adaptation of process paths according to risk levels. Combined with timeout warnings and rollback operations, it solves the technical problems of rigid processes and the inability to dynamically adjust.

[0017] This invention generates a dynamic QR code for each inspection task, containing the task ID, coordinates, timestamp, validity period, and responsible person ID. During scanning, the system simultaneously collects on-site GPS coordinates, current time, and logged-in user ID, completing triple verification of coordinate deviation, time validity, and identity consistency. If a match is not found, the scan is rejected and logged, making the QR code a dynamic credential that is "one code per time, one code per person, one code per location," solving the problems of static QR code generation and lack of spatiotemporal traceability. Simultaneously, a multi-dimensional automatic points calculation method is established, comprehensively assigning points based on multiple dimensions such as grade weight, timeliness, quality, and frequency, achieving a strong and automated correlation between points and contributions, solving the problems of points relying on manual calculation and loose incentives. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a smart operation and maintenance monitoring and management method for water conservancy engineering facilities, provided as an embodiment of the present invention; Figure 2 A flowchart of the steps for generating an inspection QR code based on an inspection task and comparing on-site information with the inspection QR code, provided in an embodiment of the present invention; Figure 3 This is a flowchart illustrating the steps of determining the severity of a hazard based on its category, matching a corresponding approval process template based on the severity of the hazard, and sending the process template to the corresponding terminal, as provided in an embodiment of the present invention. Figure 4 The flowchart illustrates the steps of uploading rectification results via a mobile terminal, accepting the rectification results, and uploading points based on closed-loop hidden dangers, as provided in this embodiment of the invention. Figure 5 An architecture diagram of a smart operation and maintenance monitoring and management system for water conservancy engineering facilities provided in an embodiment of the present invention; Figure 6 This is an architecture diagram of the on-site comparison module provided in an embodiment of the present invention; Figure 7 This is an architecture diagram of the hazard assessment module provided in an embodiment of the present invention; Figure 8 This is an architecture diagram of the integration module provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0020] It is understood that the terms "first," "second," etc., used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are used only to distinguish one element from another. For example, without departing from the scope of this application, a first script may be referred to as a second script, and similarly, a second script may be referred to as a first script.

[0021] like Figure 1 As shown in the figure, a smart operation and maintenance monitoring and management method for water conservancy engineering facilities is provided by an embodiment of the present invention. The method is characterized by comprising: S100: Obtain the inspection plan for water conservancy engineering facilities, and generate inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information.

[0022] In this step, the inspection plan for water conservancy facilities is obtained. First, the configured inspection plan for water conservancy facilities is obtained. Then, the plan is automatically broken down into executable task instances by start date and cycle granularity through a timed task trigger. Each instance is assigned a status identifier, and the inspection cycle is converted into an effective time window.

[0023] For each generated task instance, a unique facility identifier is bound to ensure that the facility can be accurately identified; the inspection cycle is refined into a combination of frequency and duration logic, based on which the generation time of the next task is automatically calculated.

[0024] When executing task instantiation, the jurisdiction area and operation and maintenance team are retrieved according to the facility identification code, the corresponding responsible department is matched, and the primary responsible person and their backup personnel (B role) are selected from the on-duty personnel of the department. The on-duty status and task load are comprehensively considered to ensure balanced assignment. Finally, the responsible person information is solidified into the task instance, forming a four-dimensional task assignment closed loop of "facility-cycle-department-personnel".

[0025] S200: Generate an inspection QR code based on the inspection task. Scan the inspection QR code with a mobile terminal and compare the on-site information with the inspection QR code. If a match is found, take an image of the potential hazard site with the mobile terminal, fill in a text description, and submit the data for reporting.

[0026] In this step, an inspection QR code is generated based on the inspection task. A unique QR code is dynamically generated for each task based on the generated inspection task instances. The QR code's encoding information includes the task ID, facility identification code, geographic coordinates, generation timestamp, validity period, and responsible person ID. Upon arrival at the site, inspection personnel scan the QR code. The terminal automatically reads the preset information and uses the device's GPS to obtain the current coordinates, system time, and logged-in user ID, creating a comparison data pair between the "preset QR code information" and the "real-time on-site information."

[0027] The comparison process is completed locally and in the cloud: the local machine calculates the coordinate deviation and time validity. If the distance difference between the scanned coordinates and the preset coordinates exceeds the preset threshold (e.g., 50 meters), or the current time exceeds the validity period, or the scanning personnel ID does not match the responsible person ID, the terminal will prompt an error and refuse to enter the reporting page.

[0028] The S300 preprocesses the site image of the potential hazard, identifies the hazard category in the image, determines the severity of the hazard based on the hazard category, matches the corresponding approval process template based on the severity of the hazard, and sends the process template to the corresponding terminal.

[0029] In this step, preprocessing is performed based on the on-site images of potential hazards. After receiving the reported data packets, the backend server first extracts the on-site images of potential hazards for preprocessing. Taking a dam inspection scenario as an example: the identification category is "dam body cracks" with a confidence level of 0.92, the severity is "general hazard (assignable)", and the hazard area accounts for 3.5%. Subsequently, the identification results are semantically fused and verified with the text description. If the similarity is lower than a preset threshold, manual review is triggered; after verification, the process enters the routing stage.

[0030] The system automatically matches an approval process template based on the severity level, and sends the template to the current reporting user's mobile device to-do list. Clicking the template initiates the assignment process. For major hazards, the template is sent to the project manager's device; if assignment is not possible, it is simultaneously pushed to the to-do lists of three users, with the first user to handle it assigning the task. The entire process is automated, requiring no manual intervention.

[0031] S400: Upload rectification results via mobile terminal, and accept the rectification results. The acceptance involves comparing the hidden danger areas to determine whether the hidden dangers have been eliminated, obtaining preset point rules for different types of contribution behaviors, and uploading points based on the closed-loop hidden dangers.

[0032] In this step, the rectification results are uploaded via mobile terminal. The person responsible for rectification uploads the rectified images and descriptions via mobile terminal. After receiving the data, the server performs three acceptance verifications: image quality verification, location consistency verification, and image content change verification. Taking the aforementioned dam crack as an example, the location deviation is 8 meters, and the pixel change rate of the potential hazard area is 95%, indicating that the potential hazard has been eliminated.

[0033] The rectification results were sequentially pushed to the to-do lists on the mobile terminals of the head of the safety department and the project manager. After the two reviewed and confirmed them, the process status was updated to "closed loop". The complete handling log recorded the time, personnel and coordinate information of each link of reporting, review, rectification and acceptance, forming a traceable closed loop file.

[0034] like Figure 2 As shown in the preferred embodiment of the present invention, the steps of generating an inspection QR code based on the inspection task, scanning the inspection QR code with a mobile terminal, comparing the on-site information with the inspection QR code, and if a match is found, taking an image of the potential hazard site with the mobile terminal, filling in a text description, and submitting the reported data specifically include: S201, Generate an inspection QR code based on the inspection task. The inspection QR code includes a timestamp, validity period, and responsible person ID. Scan the inspection QR code using a mobile terminal.

[0035] In this step, an inspection QR code is generated based on the inspection task. A unique QR code is dynamically generated for each task based on the generated inspection task instances. The QR code encoding information includes a generation timestamp, validity period, and responsible person ID. The QR code automatically expires after the validity period and must be regenerated to ensure timeliness.

[0036] Upon arrival at the site, inspection personnel scan a QR code using a WeChat mini-program. The terminal reads the pre-set information and immediately retrieves the current system time and logged-in user ID for verification. If the current time has expired, a message will be displayed stating "QR code has expired"; if the logged-in user ID does not match the responsible person's ID, a message will be displayed stating "Not the person responsible for this task". Only after both verifications are successful can the personnel access the hazard reporting page.

[0037] S202. Compare the on-site information with the inspection QR code. If a mismatch is found, access is denied and an anomaly log is recorded.

[0038] In this step, the on-site information is compared with the inspection QR code. After the inspection personnel scan the inspection QR code, the system automatically performs a three-dimensional comparison and verification between the on-site information and the pre-set information in the QR code: the distance deviation between the scan coordinates and the pre-set facility coordinates, the matching between the current time and the validity period, and the consistency between the currently logged-in user ID and the responsible person ID. If any dimension is not met, the matching is deemed to have failed, the terminal displays the corresponding error message, and access to the hazard reporting page is refused.

[0039] At the same time, the backend automatically records anomaly logs, including the anomaly type, QR code ID, scanner ID, scanning time, scanning coordinates, and terminal device ID. These logs are stored on the server for administrators to audit and trace, effectively preventing violations such as proxy signing, reporting from different locations, and reporting after the deadline.

[0040] S203, if a match is determined, the user takes a picture of the potential hazard site using a mobile terminal, fills in a text description, and submits the reported data. The reported data includes the site image, text description, reporter ID, reporting time, GPS coordinates, and the corresponding inspection task ID.

[0041] In this step, if a match is found, the inspection personnel will verify the location, time, and identity through 3D comparison, and the terminal will automatically unlock the hazard reporting function. The inspection personnel will then use the camera to capture images of the hazard site. The system will automatically embed the capture timestamp, GPS coordinates, and device ID into the image metadata. At the same time, the personnel will fill in a text description. After the report is completed, the system will automatically package the site image, text description, reporter ID, reporting time, GPS coordinates, and corresponding inspection task ID into a reporting data package and submit it to the backend server.

[0042] like Figure 3 As shown, in a preferred embodiment of the present invention, the steps of preprocessing the site hazard image, identifying the hazard category in the image, determining the severity of the hazard based on the hazard category, matching the corresponding approval process template based on the severity of the hazard, and sending the process template to the corresponding terminal specifically include: S301, Preprocessing is performed based on the site map of the potential hazard. The preprocessing includes size normalization, illumination equalization, and noise filtering to obtain the preprocessed image.

[0043] In this step, preprocessing is performed based on the site image of the potential hazard. After the server extracts the site image, it performs three preprocessing steps: size normalization, which scales the original image to 224×224 pixels to meet the model input requirements; illumination equalization, which uses a histogram equalization algorithm to adjust the brightness distribution and eliminate local over-darkness caused by backlighting or shadows; and noise filtering, which uses Gaussian filtering to remove sensor noise and environmental interference introduced during acquisition and transmission, while preserving edge details of the potential hazard area.

[0044] Taking the image of a dam crack as an example, the original image was taken under cloudy conditions and contained dark areas and raindrop reflection noise. After preprocessing, the contrast was significantly enhanced, the noise was eliminated, and the crack edges were clearly distinguishable. The preprocessing process was completed automatically, with an average time of no more than 200 milliseconds. The processed image was then input into a multi-task convolutional neural network model in tensor form for subsequent inference.

[0045] S302, identify the category of hidden danger in the image, and determine the severity of the hidden danger based on the category of hidden danger. The severity includes major hidden danger, general hidden danger, general hidden danger and no hidden danger.

[0046] In this step, the system identifies the categories of potential hazards in the image. The preprocessed image is input into a multi-task convolutional neural network model, which first outputs the hazard category identification results. Supported categories include dam cracks, seepage, landslides, piping, equipment failures, structural deformation, and abnormal water quality, with each category accompanied by a confidence score. Based on the identified category and confidence score, combined with preset grading rules (e.g., "through cracks" are mapped as major hazards, "surface non-structural cracks" as general hazards, and "minor coating peeling" as not constituting a hazard), the system comprehensively determines the severity of the hazard and classifies it into four levels: major hazard, general hazard, and no hazard.

[0047] After the judgment is completed, the system performs semantic fusion verification between the recognition result and the text description filled in by the inspector, extracts keyword vectors to calculate similarity, and triggers manual review if the similarity is lower than a preset threshold to prevent grade deviation. The final judgment result is written into the hazard record as the basis for subsequent process routing.

[0048] S303: Match the corresponding approval process template according to the severity of the hidden danger, and send the process template to the corresponding terminal.

[0049] In this step, the corresponding approval process template is matched according to the severity of the hazard. Based on the severity classification results, the corresponding approval template is automatically matched from the process template library. The template library is configured with four types of templates: Major hazards match "Project Manager Assigns Order → Rectification → Safety Department Head Acceptance → Project Manager Final Review"; General hazards (which can be assigned) match "Reporter Assigns Rectification Person → Rectification → Safety Department Head Acceptance → Project Manager / Safety Director / Deputy Manager Review"; General hazards (which cannot be assigned) match "Project Manager / Safety Director / Deputy Manager Assigns Order → Rectification → Safety Department Head Acceptance → Project Manager / Safety Director / Deputy Manager Review"; Hazards that do not constitute a single "Process End" node do not proceed to approval.

[0050] like Figure 4 As shown, in a preferred embodiment of the present invention, the steps of uploading rectification results via a mobile terminal, accepting the rectification results, comparing the hazard areas to determine whether the hazards have been eliminated, obtaining preset point rules for different types of contribution behaviors, and uploading points based on the closed-loop hazards specifically include: S401, upload the rectification results via mobile terminal. The rectification results include on-site images and rectification descriptions after rectification. The rectification results are then inspected and verified, and the inspection involves comparing the areas with potential hazards.

[0051] In this step, the rectification results are uploaded via mobile terminal. After the person in charge of rectification completes the rectification, they take a picture of the site after rectification and fill in the rectification description via mobile terminal. The terminal automatically embeds the shooting timestamp, GPS coordinates and device ID into the image metadata and submits it to the backend server in a package.

[0052] After receiving the data, the backend performs a comparison and acceptance test of the hazard area: It extracts a pixel-level mask of the original hazard area, performs image registration (based on SIFT feature point matching) between the rectified image and the original hazard image, maps the mask to the corresponding position in the rectified image, and calculates the pixel change rate within that area. Taking the aforementioned dam crack as an example, the hazard area is the crack pixel area (accounting for 3.5%). After rectification, the pixel change rate of this area reaches 95%, which is greater than the preset threshold of 90%, indicating that the hazard has been eliminated.

[0053] If the change rate is less than 90%, the system determines that the rectification is incomplete and prompts "Rectification not up to standard, please rectify again" on the acceptor's terminal and rolls back; if the change rate reaches the threshold and the GPS coordinate deviation is within 50 meters, the acceptance is deemed passed and pushed to the pending list of the next approval node terminal.

[0054] S402, determine whether the hidden danger has been eliminated. If the acceptance is deemed unsuccessful, the process is rolled back to the rectification node, and the original person responsible for rectification is required to rectify it again. If the acceptance is deemed successful, the process ends and the hidden danger status is updated to closed loop.

[0055] In this step, it is determined whether the hidden danger has been eliminated by comparing the pixel change rate of the hidden danger area with a preset threshold (90%). Taking the aforementioned dam crack as an example, the change rate is 95%, which is greater than the threshold. The acceptance is deemed passed, the process ends, the hidden danger status is updated from "under rectification" to "closed loop", and the complete handling traceability chain is generated and archived.

[0056] If the rate of change is lower than the threshold (e.g., only 60%), the acceptance is deemed unsuccessful. The system will display "Rectification not up to standard" on the acceptor's terminal, automatically roll back the process to the rectification node, and send a re-rectification notice to the original person responsible for rectification. After the person responsible completes the rectification, they will upload the rectification results again and re-enter the acceptance comparison process.

[0057] If multiple rectification attempts fail, the system will trigger an alert on the third rollback, sending a "Hazard Rectification Abnormality" notification to the superior terminal and recommending manual on-site verification. During the rectification period, the process status remains "Rectifying" until acceptance is passed and updated to "Closed Loop," ensuring sufficient and effective rectification verification before closure.

[0058] S403, obtain preset point rules for different types of contribution behaviors, the point rules include hidden danger reporting, hidden danger rectification, knowledge base upload and knowledge base review, and upload points according to the closed-loop hidden danger upload.

[0059] In this step, preset point rules for different types of contribution behaviors are obtained, and a point rule library is pre-configured: 10 points for reporting hidden dangers, 8 points for rectifying hidden dangers, 5 points for uploading to the knowledge base, and 3 points for passing the knowledge base review. Each rule is set by the administrator through point management - point rules according to the process type, and only one valid rule is set for each process.

[0060] Taking the closed-loop dam crack hazard as an example, two contribution behaviors are triggered. When the process is closed, the following is automatically calculated for the reporter, Zhang: a base score of 10 points multiplied by a level weight of 1.5, a contribution type weight of 1.0, a timeliness correction of 1.1, a quality correction of 1.05, and a frequency decay of 1.0, resulting in 17.325 points, which are written into the points details; and for the person responsible for rectification, Li: a base score of 8 points multiplied by a level weight of 1.5, a contribution type weight of 0.8, a timeliness correction of 1.1, a quality correction of 1.0, and a frequency decay of 1.0, resulting in 10.56 points, which are written into the points details.

[0061] After points are entered, the system automatically updates the accumulated points and leaderboard. Users can view details through a mini-program, and administrators can query all data by user, time, behavior type, and other criteria through the web interface, achieving fully automated points management.

[0062] like Figure 5 As shown, an intelligent operation and maintenance monitoring and management system for water conservancy engineering facilities is provided in an embodiment of the present invention. The system includes: The inspection task module 100 is used to obtain the inspection plan of water conservancy engineering facilities and generate inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information.

[0063] In this system, the inspection task module 100 obtains the inspection plan of water conservancy facilities. First, it obtains the configured water conservancy facility inspection plan, and then automatically breaks it down into executable task instances by start date and cycle granularity through a timed task trigger. Each instance is assigned a status identifier, and the inspection cycle is converted into an effective time window.

[0064] For each generated task instance, a unique facility identifier is bound to ensure that the facility can be accurately identified; the inspection cycle is refined into a combination of frequency and duration logic, based on which the generation time of the next task is automatically calculated.

[0065] When executing task instantiation, the jurisdiction area and operation and maintenance team are retrieved according to the facility identification code, the corresponding responsible department is matched, and the primary responsible person and their backup personnel (B role) are selected from the on-duty personnel of the department. The on-duty status and task load are comprehensively considered to ensure balanced assignment. Finally, the responsible person information is solidified into the task instance, forming a four-dimensional task assignment closed loop of "facility-cycle-department-personnel".

[0066] The on-site comparison module 200 is used to generate an inspection QR code based on the inspection task. The inspection QR code is scanned by a mobile terminal, and the on-site information is compared with the inspection QR code. If a match is found, the on-site image of the potential hazard is taken by the mobile terminal, and a text description is filled in before the data is submitted for reporting.

[0067] In this system, the on-site comparison module 200 generates inspection QR codes based on inspection tasks. Based on the generated inspection task instances, it dynamically generates a unique inspection QR code for each task. The QR code's encoding information includes the task ID, facility identification code, geographic coordinates, generation timestamp, validity period, and responsible person ID. Upon arrival at the site, inspection personnel scan the QR code. The terminal automatically reads the preset information and uses the device's GPS to obtain the current coordinates, system time, and logged-in user ID, forming a comparison data pair between the "preset QR code information" and the "real-time on-site information."

[0068] The comparison process is completed locally and in the cloud: the local machine calculates the coordinate deviation and time validity. If the distance difference between the scanned coordinates and the preset coordinates exceeds the preset threshold (e.g., 50 meters), or the current time exceeds the validity period, or the scanning personnel ID does not match the responsible person ID, the terminal will prompt an error and refuse to enter the reporting page.

[0069] The hazard assessment module 300 is used to preprocess the hazard site map, identify the hazard category in the image, determine the severity of the hazard based on the hazard category, match the corresponding approval process template based on the severity of the hazard, and send the process template to the corresponding terminal.

[0070] In this system, the hazard identification module 300 preprocesses the hazard site image. After receiving the reported data packet, the backend server first extracts the hazard site image for preprocessing. Taking a dam inspection scenario as an example: the identification category is "dam body cracks" with a confidence level of 0.92, the severity is "general hazard (assignable)", and the hazard area accounts for 3.5%. Subsequently, the identification results are semantically fused and verified with the text description. If the similarity is lower than a preset threshold, manual review is triggered; after verification, the process routing stage begins.

[0071] The system automatically matches an approval process template based on the severity level, and sends the template to the current reporting user's mobile device to-do list. Clicking the template initiates the assignment process. For major hazards, the template is sent to the project manager's device; if assignment is not possible, it is simultaneously pushed to the to-do lists of three users, with the first user to handle it assigning the task. The entire process is automated, requiring no manual intervention.

[0072] The points module 400 is used to upload rectification results via mobile terminal, and to accept the rectification results. The acceptance involves comparing the hidden danger areas to determine whether the hidden dangers have been eliminated, obtaining preset points rules for different types of contribution behaviors, and uploading points based on the closed-loop hidden dangers.

[0073] In this system, the integration module 400 uploads the rectification results via mobile terminal, and the person responsible for rectification uploads the rectified images and descriptions via mobile terminal. Upon receiving these, the server performs three acceptance verifications: image quality verification, location consistency verification, and image content change verification. Taking the aforementioned dam crack as an example, with a location deviation of 8 meters and a pixel change rate of 95% in the hazard area, the hazard is determined to have been eliminated.

[0074] The rectification results were sequentially pushed to the to-do lists on the mobile terminals of the head of the safety department and the project manager. After the two reviewed and confirmed them, the process status was updated to "closed loop". The complete handling log recorded the time, personnel and coordinate information of each link of reporting, review, rectification and acceptance, forming a traceable closed loop file.

[0075] like Figure 6 As shown, in a preferred embodiment of the present invention, the on-site comparison module 200 includes: The QR code scanning unit 201 is used to generate an inspection QR code according to the inspection task. The inspection QR code includes a timestamp, validity period and responsible person ID, and is scanned by a mobile terminal.

[0076] In this module, the QR code scanning unit 201 generates inspection QR codes based on the inspection tasks. It dynamically generates a unique inspection QR code for each task based on the generated inspection task instances. The QR code encoding information includes a generation timestamp, validity period, and responsible person ID. The QR code automatically expires after the validity period and must be regenerated to ensure timeliness.

[0077] Upon arrival at the site, inspection personnel scan a QR code using a WeChat mini-program. The terminal reads the pre-set information and immediately retrieves the current system time and logged-in user ID for verification. If the current time has expired, a message will be displayed stating "QR code has expired"; if the logged-in user ID does not match the responsible person's ID, a message will be displayed stating "Not the person responsible for this task". Only after both verifications are successful can the personnel access the hazard reporting page.

[0078] The comparison unit 202 is used to compare the on-site information with the inspection QR code. If a mismatch is found, access is denied and an anomaly log is recorded.

[0079] In this module, comparison unit 202 compares the on-site information with the inspection QR code. After the inspection personnel scan the inspection QR code, the system automatically performs a three-dimensional comparison and verification between the on-site information and the preset information in the QR code: the distance deviation between the scan coordinates and the preset facility coordinates, the matching between the current time and the validity period, and the consistency between the currently logged-in user ID and the responsible person ID. If any dimension is not satisfied, the matching is deemed to have failed, the terminal displays the corresponding error message, and access to the hazard reporting page is refused.

[0080] At the same time, the backend automatically records anomaly logs, including the anomaly type, QR code ID, scanner ID, scanning time, scanning coordinates, and terminal device ID. These logs are stored on the server for administrators to audit and trace, effectively preventing violations such as proxy signing, reporting from different locations, and reporting after the deadline.

[0081] The data reporting unit 203 is used to, if a match is determined, take a picture of the potential hazard site through a mobile terminal, fill in a text description, and then submit the reported data. The reported data includes the site image, text description, reporter ID, reporting time, GPS coordinates, and the corresponding inspection task ID.

[0082] In this module, if the data reporting unit 203 determines a match, the inspection personnel will automatically unlock the hazard reporting function after verifying the location, time, and identity through 3D comparison. The inspection personnel will then use the camera to capture images of the hazard site. The system will automatically embed the capture timestamp, GPS coordinates, and device ID into the image metadata. Simultaneously, a text description will be filled in. After the data is completed, the system will automatically package the site image, text description, reporter ID, reporting time, GPS coordinates, and corresponding inspection task ID into a single report data package and submit it to the backend server.

[0083] like Figure 7 As shown, in a preferred embodiment of the present invention, the hazard judgment module 300 includes: Image preprocessing unit 301 is used to preprocess the image of the potential hazard site. The preprocessing includes size normalization, illumination equalization and noise filtering to obtain the preprocessed image.

[0084] In this module, the image preprocessing unit 301 performs preprocessing based on the hazard site image. After the server extracts the hazard site image, it performs three preprocessing steps: size normalization, which scales the original image to 224×224 pixels to meet the model input requirements; illumination equalization, which uses a histogram equalization algorithm to adjust the brightness distribution and eliminate local over-darkness caused by backlighting or shadows; and noise filtering, which uses Gaussian filtering to remove sensor noise and environmental interference introduced during acquisition and transmission, while preserving edge details of the hazard area.

[0085] Taking the image of a dam crack as an example, the original image was taken under cloudy conditions and contained dark areas and raindrop reflection noise. After preprocessing, the contrast was significantly enhanced, the noise was eliminated, and the crack edges were clearly distinguishable. The preprocessing process was completed automatically, with an average time of no more than 200 milliseconds. The processed image was then input into a multi-task convolutional neural network model in tensor form for subsequent inference.

[0086] The hazard classification unit 302 is used to identify the hazard category in the image and determine the severity of the hazard based on the hazard category. The severity includes major hazard, general hazard, minor hazard, and no hazard.

[0087] In this module, the hazard classification unit 302 identifies the hazard categories in the image. The preprocessed image is input into a multi-task convolutional neural network model, which first outputs the hazard category identification result. Supported categories include dam cracks, seepage, landslides, piping, equipment failures, structural deformation, and abnormal water quality, etc. Each category is accompanied by a confidence score. Based on the identified category and confidence score, combined with preset classification rules, such as mapping "through cracks" to major hazards, "surface non-structural cracks" to general hazards, and "minor coating peeling" to not constitute a hazard, the system comprehensively judges the severity of the hazard and classifies it into four levels: major hazard, general hazard, and not constitute a hazard.

[0088] After the judgment is completed, the system performs semantic fusion verification between the recognition result and the text description filled in by the inspector, extracts keyword vectors to calculate similarity, and triggers manual review if the similarity is lower than a preset threshold to prevent grade deviation. The final judgment result is written into the hazard record as the basis for subsequent process routing.

[0089] Template unit 303 is used to match the corresponding approval process template according to the severity of the hidden danger and send the process template to the corresponding terminal.

[0090] In this module, template unit 303 matches the corresponding approval process template based on the severity of the hazard. Based on the severity classification results, it automatically matches the corresponding approval template from the process template library. The template library is configured with four types of templates: Major hazards match "Project Manager Assigns Order → Rectification → Safety Department Head Acceptance → Project Manager Final Review"; General hazards (assignable) match "Reporter Assigns Rectification Person → Rectification → Safety Department Head Acceptance → Project Manager / Safety Director / Deputy Manager Review"; General hazards (not assignable) match "Project Manager / Safety Director / Deputy Manager Assigns Order → Rectification → Safety Department Head Acceptance → Project Manager / Safety Director / Deputy Manager Review"; Hazards that do not constitute a single "Process End" node do not proceed to approval.

[0091] like Figure 8 As shown, in a preferred embodiment of the present invention, the integration module 400 includes: The acceptance unit 401 is used to upload the rectification results via a mobile terminal. The rectification results include on-site images and rectification descriptions after rectification. The acceptance is to compare the rectification results with the potential hazard areas.

[0092] In this module, the acceptance unit 401 uploads the rectification results via a mobile terminal. After the person in charge of rectification completes the rectification, they take a picture of the site after rectification and fill in the rectification description via the mobile terminal. The terminal automatically embeds the shooting timestamp, GPS coordinates and device ID into the image metadata and submits it to the backend server in a package.

[0093] After receiving the data, the backend performs a comparison and acceptance test of the hazard area: It extracts a pixel-level mask of the original hazard area, performs image registration (based on SIFT feature point matching) between the rectified image and the original hazard image, maps the mask to the corresponding position in the rectified image, and calculates the pixel change rate within that area. Taking the aforementioned dam crack as an example, the hazard area is the crack pixel area (accounting for 3.5%). After rectification, the pixel change rate of this area reaches 95%, which is greater than the preset threshold of 90%, indicating that the hazard has been eliminated.

[0094] If the change rate is less than 90%, the system determines that the rectification is incomplete and prompts "Rectification not up to standard, please rectify again" on the acceptor's terminal and rolls back; if the change rate reaches the threshold and the GPS coordinate deviation is within 50 meters, the acceptance is deemed passed and pushed to the pending list of the next approval node terminal.

[0095] The hazard elimination unit 402 is used to determine whether the hazard has been eliminated. If the acceptance is deemed unsuccessful, the process is rolled back to the rectification node, and the original person responsible for rectification is required to rectify it again. If the acceptance is deemed successful, the process ends and the hazard status is updated to closed loop.

[0096] In this module, the hazard elimination unit 402 determines whether the hazard has been eliminated by comparing the pixel change rate of the hazard area with a preset threshold (90%). Taking the aforementioned dam crack as an example, the change rate is 95%, which is greater than the threshold. Therefore, the acceptance is deemed successful, the process ends, the hazard status is updated from "under rectification" to "closed loop", and a complete handling traceability chain is generated and archived.

[0097] If the rate of change is lower than the threshold (e.g., only 60%), the acceptance is deemed unsuccessful. The system will display "Rectification not up to standard" on the acceptor's terminal, automatically roll back the process to the rectification node, and send a re-rectification notice to the original person responsible for rectification. After the person responsible completes the rectification, they will upload the rectification results again and re-enter the acceptance comparison process.

[0098] If multiple rectification attempts fail, the system will trigger an alert on the third rollback, sending a "Hazard Rectification Abnormality" notification to the superior terminal and recommending manual on-site verification. During the rectification period, the process status remains "Rectifying" until acceptance is passed and updated to "Closed Loop," ensuring sufficient and effective rectification verification before closure.

[0099] The scoring unit 403 is used to obtain preset scoring rules for different types of contribution behaviors. The scoring rules include hazard reporting, hazard rectification, knowledge base uploading, and knowledge base review. Scores are uploaded based on closed-loop hazard reporting.

[0100] In this module, the points unit 403 obtains preset points rules for different types of contribution behaviors and pre-configures the points rule library: 10 points for reporting hidden dangers, 8 points for rectifying hidden dangers, 5 points for uploading to the knowledge base, and 3 points for passing the knowledge base review. Each rule is set by the administrator through points management - points rules according to the process type, and there is only one valid rule for each process.

[0101] Taking the closed-loop dam crack hazard as an example, two contribution behaviors are triggered. When the process is closed, the following is automatically calculated for the reporter, Zhang: a base score of 10 points multiplied by a level weight of 1.5, a contribution type weight of 1.0, a timeliness correction of 1.1, a quality correction of 1.05, and a frequency decay of 1.0, resulting in 17.325 points, which are written into the points details; and for the person responsible for rectification, Li: a base score of 8 points multiplied by a level weight of 1.5, a contribution type weight of 0.8, a timeliness correction of 1.1, a quality correction of 1.0, and a frequency decay of 1.0, resulting in 10.56 points, which are written into the points details.

[0102] After points are entered, the system automatically updates the accumulated points and leaderboard. Users can view details through a mini-program, and administrators can query all data by user, time, behavior type, and other criteria through the web interface, achieving fully automated points management.

[0103] In one embodiment, a computer device is provided, the computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, performs the following steps: Obtain the inspection plan for water conservancy engineering facilities, and generate inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information. An inspection QR code is generated based on the inspection task. The inspection QR code is scanned by a mobile terminal, and the on-site information is compared with the inspection QR code. If a match is found, the on-site image of the potential hazard is taken by the mobile terminal, and a text description is filled in before the data is submitted for reporting. Preprocessing is performed on the on-site images of potential hazards to identify the types of hazards in the images. The severity of the hazards is determined based on the types of hazards, and the corresponding approval process template is matched based on the severity of the hazards. The process template is then sent to the corresponding terminal. The rectification results are uploaded via mobile terminal, and the rectification results are accepted. The acceptance involves comparing the hidden danger areas to determine whether the hidden dangers have been eliminated, obtaining preset point rules for different types of contribution behaviors, and uploading points based on the closed-loop hidden dangers.

[0104] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, causes the processor to perform the following steps: Obtain the inspection plan for water conservancy engineering facilities, and generate inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information. An inspection QR code is generated based on the inspection task. The inspection QR code is scanned by a mobile terminal, and the on-site information is compared with the inspection QR code. If a match is found, the on-site image of the potential hazard is taken by the mobile terminal, and a text description is filled in before the data is submitted for reporting. Preprocessing is performed on the on-site images of potential hazards to identify the types of hazards in the images. The severity of the hazards is determined based on the types of hazards, and the corresponding approval process template is matched based on the severity of the hazards. The process template is then sent to the corresponding terminal. The rectification results are uploaded via mobile terminal, and the rectification results are accepted. The acceptance involves comparing the hidden danger areas to determine whether the hidden dangers have been eliminated, obtaining preset point rules for different types of contribution behaviors, and uploading points based on the closed-loop hidden dangers.

[0105] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps. Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0106] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0107] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

[0108] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent operation, maintenance, monitoring, and management of water conservancy engineering facilities, characterized in that, The method includes: Obtain the inspection plan for water conservancy engineering facilities, and generate inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information. An inspection QR code is generated based on the inspection task. The inspection QR code is scanned by a mobile terminal, and the on-site information is compared with the inspection QR code. If a match is found, the on-site image of the potential hazard is taken by the mobile terminal, and a text description is filled in before the data is submitted for reporting. Preprocessing is performed on the on-site images of potential hazards to identify the types of hazards in the images. The severity of the hazards is determined based on the types of hazards, and the corresponding approval process template is matched based on the severity of the hazards. The process template is then sent to the corresponding terminal. The rectification results are uploaded via mobile terminal, and the rectification results are accepted. The acceptance involves comparing the hidden danger areas to determine whether the hidden dangers have been eliminated, obtaining preset point rules for different types of contribution behaviors, and uploading points based on the closed-loop hidden dangers.

2. The intelligent operation and maintenance monitoring and management method for water conservancy engineering facilities according to claim 1, characterized in that, The steps of generating an inspection QR code based on the inspection task, scanning the inspection QR code with a mobile terminal, comparing the on-site information with the inspection QR code, and if a match is found, taking an image of the potential hazard site with the mobile terminal, filling in a text description, and submitting the reported data specifically include: An inspection QR code is generated based on the inspection task. The inspection QR code includes a timestamp, validity period and responsible person ID. The inspection QR code is scanned by a mobile terminal. Compare the on-site information with the inspection QR code; if a mismatch is found, access is denied and an anomaly log is recorded. If a match is found, the user takes an image of the potential hazard site using a mobile terminal, fills in a text description, and submits the reported data. The reported data includes the site image, text description, reporter ID, reporting time, GPS coordinates, and the corresponding inspection task ID.

3. The intelligent operation and maintenance monitoring and management method for water conservancy engineering facilities according to claim 1, characterized in that, The steps of preprocessing the site hazard image, identifying the hazard category in the image, determining the severity of the hazard based on the hazard category, matching the corresponding approval process template based on the severity of the hazard, and sending the process template to the corresponding terminal specifically include: Preprocessing is performed on the site map of the potential hazard. The preprocessing includes size normalization, illumination equalization and noise filtering to obtain the preprocessed image. Identify the category of potential hazards in the image, and determine the severity of the hazard based on the category. The severity includes major hazards, general hazards, minor hazards, and no hazard. Match the corresponding approval process template according to the severity of the potential hazard, and send the process template to the corresponding terminal.

4. The intelligent operation and maintenance monitoring and management method for water conservancy engineering facilities according to claim 1, characterized in that, The steps of uploading rectification results via mobile terminal, accepting the rectification results (which involves comparing the hazard areas to determine whether the hazards have been eliminated), obtaining preset point rules for different types of contribution behaviors, and uploading points based on closed-loop hazards specifically include: The rectification results are uploaded via mobile terminal. The rectification results include on-site images and rectification descriptions after rectification. The rectification results are then accepted, and the acceptance involves comparing the areas with potential hazards. Determine whether the hidden danger has been eliminated. If the acceptance is deemed unsuccessful, the process is reverted to the rectification stage, and the original person responsible for rectification is required to rectify it again. If the acceptance is deemed successful, the process ends, and the hidden danger status is updated to closed loop. The system acquires preset point rules for different types of contribution behaviors, including hazard reporting, hazard rectification, knowledge base upload, and knowledge base review, and awards points based on hazard uploads that have been closed-loop.

5. The intelligent operation and maintenance monitoring and management method for water conservancy engineering facilities according to claim 3, characterized in that, The categories of potential hazards include body cracks, seepage, landslides, piping, equipment failures, structural deformation, and abnormal water quality.

6. A smart operation and maintenance monitoring and management system for water conservancy engineering facilities, characterized in that, The system includes: The inspection task module obtains the inspection plan of water conservancy engineering facilities and generates inspection tasks based on the inspection plan. The inspection tasks include facility identification, inspection cycle, responsible department and responsible person information. The on-site comparison module generates an inspection QR code based on the inspection task. The on-site information is compared with the inspection QR code by scanning the inspection QR code with a mobile terminal. If a match is found, the on-site image of the potential hazard is taken by the mobile terminal, and a text description is filled in before the data is submitted for reporting. The hazard assessment module preprocesses the hazard site image, identifies the hazard category in the image, determines the severity of the hazard based on the hazard category, matches the corresponding approval process template based on the severity of the hazard, and sends the process template to the corresponding terminal. The points module allows users to upload rectification results via mobile terminals and verify the results. The verification process involves comparing potential hazard areas to determine whether the hazards have been eliminated, obtaining preset points rules for different types of contribution behaviors, and uploading points based on closed-loop hazards.

7. The intelligent operation and maintenance monitoring and management system for water conservancy engineering facilities according to claim 6, characterized in that, The on-site comparison module includes: The QR code scanning unit generates an inspection QR code based on the inspection task. The inspection QR code includes a timestamp, validity period, and responsible person ID. The inspection QR code is scanned by a mobile terminal. The comparison unit compares the on-site information with the inspection QR code. If a mismatch is found, access is denied and an anomaly log is recorded. If a match is found, the data reporting unit will take a picture of the potential hazard site using a mobile terminal, fill in a text description, and then submit the reported data. The reported data includes the site image, text description, reporter ID, reporting time, GPS coordinates, and the corresponding inspection task ID.

8. The intelligent operation and maintenance monitoring and management system for water conservancy engineering facilities according to claim 7, characterized in that, The hazard assessment module includes: The image preprocessing unit performs preprocessing based on the hazard site image. The preprocessing includes size normalization, illumination equalization, and noise filtering to obtain the preprocessed image. The hazard category unit identifies the hazard category in the image and determines the severity of the hazard based on the hazard category. The severity includes major hazard, general hazard, minor hazard, and no hazard. The template unit matches the corresponding approval process template based on the severity of the potential hazard and sends the process template to the corresponding terminal.

9. A smart operation and maintenance monitoring and management system for water conservancy engineering facilities according to claim 8, characterized in that, The integration module includes: The acceptance unit uploads the rectification results via a mobile terminal. The rectification results include on-site images and rectification descriptions after rectification. The acceptance process involves comparing the areas with potential hazards. The hazard elimination unit determines whether the hazard has been eliminated. If the acceptance is deemed unsuccessful, the process is rolled back to the rectification node, and the original person responsible for rectification must rectify it again. If the acceptance is deemed successful, the process ends, and the hazard status is updated to closed loop. The points unit obtains preset points rules for different types of contribution behaviors. The points rules include hazard reporting, hazard rectification, knowledge base upload, and knowledge base review. Points are uploaded based on closed-loop hazard uploads.

10. A smart operation and maintenance monitoring and management system for water conservancy engineering facilities according to claim 9, characterized in that, The categories of potential hazards include body cracks, seepage, landslides, piping, equipment failures, structural deformation, and abnormal water quality.

Images