A safety inspection system for large and medium-sized reservoirs based on virtual reality technology

Abstract

This invention proposes a safety inspection system for large and medium-sized reservoirs based on virtual reality technology, comprising: a sensing system: multiple sensors, including at least a water and rainfall monitor, a humidity sensor, and a liquid level sensor, are installed at the physical dam site to collect status information, performance information, and surrounding environmental information of the physical dam. Data is obtained through various preset sensors and transmitted in real time; a BIM model digital twin dam: a virtual BIM model is constructed based on the physical dam, forming a one-to-one mapping with the physical dam, reflecting the dam's structural characteristics, material characteristics, construction process characteristics, operation and maintenance process characteristics, and performance change characteristics. This model is used for calculation and analysis and is updated in real time according to changes in the physical dam; a data processing system; and an Internet of Things (IoT) system to connect the sensing system, the BIM model digital twin dam, and the data processing system. The aim is to achieve information interaction and virtual control between the digital twin and the physical entity.

Description

Technical Field

[0001] This invention belongs to the field of smart reservoir and safety monitoring technology for large and medium-sized reservoir dams, and particularly relates to a safety inspection system for large and medium-sized reservoirs based on virtual reality technology. Background Technology

[0002] Reservoirs and dams can intercept flood peaks during the flood season and conserve water resources for water supply, power generation, and ecological protection during the non-flood season, serving as essential infrastructure for water conservancy and flood hazard mitigation. Dam safety monitoring, as a crucial aspect of dam safety management, is a vital measure for predicting dam risks.

[0003] Dam safety monitoring refers to the monitoring and observation of the main structure, foundation, slopes on both banks, related facilities, and surrounding environment of a dam project through instrument monitoring and inspection. It serves to verify design, improve construction, and evaluate the dam's safety status. The main monitoring includes dam deformation, seepage, stress and strain, and environmental parameters. As the most direct and important means of monitoring dam operation, dam safety monitoring plays an irreplaceable role in dam construction and operation management. Through safety monitoring, people can understand the dam's past and current operational status, providing a reference for dam safety management.

[0004] Currently, due to the dispersed nature of the monitoring sensors on dams, and the fact that most operate independently, it is impossible to intuitively and comprehensively monitor and manage the overall safety status of the dam. In actual engineering projects, another common situation is that the safety inspection of large and medium-sized reservoir dams largely relies on routine inspection personnel visually observing the dams and then adding, deleting, modifying, and querying inspection reports. Given the limited professional knowledge of these personnel, they cannot quickly assess the dam's safety status and require the reports to be returned for analysis by professionals, i.e., structural behavior prediction based on measured data and on-site inspection photos.

[0005] Current traditional solutions for dam safety monitoring still have the following prominent problems:

[0006] 1. Insufficient management and control capabilities: Dam construction has a long construction period, many influencing factors, and overlapping construction, such as dam concrete pouring, temperature control measures for poured concrete, foundation grouting, transverse joint grouting, and orifice metal structure construction. During the process, the structure, materials, and process measures are frequently adjusted and changed, and the existing information integration and management and control capabilities are insufficient.

[0007] 2. Poor Data Integrity: During dam operation, monitoring relies heavily on numerous instruments installed during the construction phase. However, the level of data fusion and analysis is low, resulting in poor overall integrity and real-time performance, making it difficult to effectively support real-time assessment and decision-making. Therefore, there is an urgent need to utilize new technologies to further improve dam safety management.

[0008] 3. Low timeliness: Dam projects involve numerous fields and have complex operational characteristics, making safety management and decision-making difficult. On-site inspection personnel often have limited professional knowledge and need to relay information from the site to experienced experts for decision-making, leading to delays and making it difficult to address safety hazards in a timely manner.

[0009] 4. Low level of visualization: Dam safety monitoring involves various factors that affect the daily operation of the dam. The current structured reports are poorly presented and have a high cost of understanding. Summary of the Invention

[0010] To address the shortcomings and gaps in existing technologies, the present invention aims to provide a safety inspection system for large and medium-sized reservoirs based on virtual reality technology. The system aims to construct a BIM digital twin of the physical dam in virtual space, enabling information interaction and virtual control between the digital twin and the physical entity. This will improve the dam's information integration and management capabilities, assist in rapid decision-making, and enhance the industry's level of intelligence.

[0011] The present invention specifically adopts the following technical solution:

[0012] A safety inspection system for large and medium-sized reservoirs based on virtual reality technology, characterized in that it includes:

[0013] Sensing system: Install at the physical dam site a variety of sensors, including at least water and rainfall monitors, humidity sensors, and liquid level sensors, to collect the physical dam's status information, performance information, and surrounding environmental information. The system obtains data through various preset sensors and transmits it in real time.

[0014] BIM Model Digital Twin Dam: A virtual BIM model is built based on the physical dam, forming a one-to-one mapping with the physical dam. It reflects the dam's structural characteristics, material characteristics, construction process characteristics, operation and maintenance process characteristics, and behavior change characteristics. It is used for calculation and analysis and is updated in real time according to changes in the physical dam.

[0015] Data Processing System: Designed to address the operational characteristics of water resources knowledge graphs and hydrological data, this system performs data processing algorithms and rules to clean, filter, and validate the raw data. It converts real-time monitoring data collected from field sensors into various data formats that meet operational requirements.

[0016] And IoT systems, used to connect sensing systems, BIM model digital twin dams, and data processing systems.

[0017] Furthermore, the real-time monitoring data collected by the sensing system is preprocessed and then transmitted to the Internet of Things (IoT) system; the data processing system creates a database according to water conservancy industry standards to store topographic data, meteorological data, hydrological data, socio-economic data, real-time data and historical data, and stores the preprocessed real-time monitoring data provided by the IoT system into the standard database after data cleaning;

[0018] Furthermore, the method for constructing the BIM model digital twin dam is as follows:

[0019] Drones were used for aerial photography and laser scanning to obtain point cloud data and video images of the area to be modeled.

[0020] Enhancement, segmentation, and classification operations are performed on point cloud data to clean and enrich the point cloud model, and noise reduction and simplification are performed on the point cloud data.

[0021] The ground dense point cloud and the air dense point cloud are registered by the iterative nearest point method, and the rigid body transformation during registration is derived. The ground dense point cloud and the air dense point cloud are fused according to the rigid body transformation to obtain the air and ground three-dimensional point cloud.

[0022] By fusing panoramic photogrammetry images with 3D point clouds, a precise geometric model and all-around texture information of the ground and objects in the target area are obtained, realistically reconstructing the 3D appearance of the scene.

[0023] By using close-up panoramic data, the shortcomings of the obtained 3D model in covering the side and bottom scenes can be compensated for, thereby further improving the modeling effect;

[0024] The color binarization method is used to perform color balance processing on the modeled image. The steps are as follows:

[0025] Step 1: Let f(n,k) be the gray value of the image at pixel (n,k). Consider a (2i+1)×(2i+1) window centered at pixel (n,k).

[0026] Step 2: Calculate the threshold T(n,k) for each pixel (n,k) in the image;

[0027] Step 3: Binarize each pixel (n, k) in the image using the value of b(n, k).

[0028] Furthermore, within the BIM model digital twin dam:

[0029] The 3D real-scene model data of water conservancy projects such as reservoirs, dikes, and sluices, as well as field sensors such as video monitors, water and rainfall monitoring devices, humidity sensors, and liquid level sensors are overlaid on the GIS map for display.

[0030] After on-site surveying, shooting points were selected, and photos were taken using tripods and panoramic gimbals. The resulting on-site dam photos were then edited and stitched together to export 720VR panoramic images, which were then overlaid onto a GIS map for display.

[0031] The real-time monitoring data from the obtained standard database, along with the on-site monitoring video transmitted from the video monitor, are overlaid onto the GIS map for display.

[0032] Furthermore, in the implementation of 3D modeling, the point cloud data is processed in a lightweight manner. The specific processing method is as follows: using point cloud processing software, feature extraction algorithms are used to obtain the feature point cloud data of the building outline. While retaining the feature point cloud data with an accuracy of centimeters, some outlier point clouds and some point cloud data inside the building are removed.

[0033] Furthermore, the data cleaning specifically includes the following steps:

[0034] Based on the operational characteristics of various hydrological data, including water level and rainfall, different abnormal data filtering rules, time period data transformation rules, and missing value compensation algorithms are designed.

[0035] Filter various real-time monitoring data using abnormal data filtering rules;

[0036] According to the time period data conversion rules, the real-time monitoring data collected from the field sensors will be converted into various data formats that meet business needs;

[0037] Verify all real-time monitoring data after data cleaning and remove monitoring data that fails verification.

[0038] Furthermore, the data processing system processes the raw data collected from the standard database to establish a data processing model; in response to potential dam safety hazards, it combines various data analysis methods, including multivariate nonlinear regression and machine learning, to determine the dam's safety status, providing support for dam safety early warning judgment.

[0039] Furthermore, the data processing system identifies the warning level, generates warning decisions, and issues warning information based on the water conservancy industry's early warning grading standards; and automatically exports various business reports periodically based on historical monitoring data, historical early warning records, historical inspection records, and historical maintenance records.

[0040] Furthermore, the data processing system includes a real-time monitoring subsystem, a monitoring and analysis subsystem, a security inspection subsystem, an early warning release subsystem, a video surveillance subsystem, and a monitoring report subsystem;

[0041] The real-time monitoring subsystem is used to store various monitoring data and, based on a geographic information system, overlays them onto two-dimensional maps and three-dimensional BIM models for intuitive display.

[0042] The monitoring and analysis subsystem is used to classify, calculate, and statistically analyze the various monitoring data in the database;

[0043] The security inspection subsystem is based on a business flow engine and manages inspection points, inspection plans, and inspection personnel, registers inspection personnel information, and assigns inspection plans to them.

[0044] The early warning release subsystem is used to issue alarms for various dam monitoring data that exceed the threshold based on the construction and operation parameters that exceed the threshold.

[0045] The video surveillance subsystem is connected to the dam safety monitoring system, and through the video surveillance module, it can quickly retrieve, locate, compare and back up the surveillance video files, realize unified identity authentication and access management, and ensure the security of information and data.

[0046] The monitoring and reporting subsystem is used to display various reports generated by the dam's daily operation, including: inspection reports, maintenance reports, and upkeep reports.

[0047] Compared with the prior art, the beneficial effects of the present invention and its preferred embodiments include:

[0048] 1. BIM Digital Twin Technology: Utilizing BIM technology, a digital twin of the physical dam is constructed, virtually simulating the dam and surrounding sensors and other equipment, and recreating them on a software platform. By establishing a VR panoramic environment covering the entire reservoir, the limitation of blind spots in dam observation is overcome. This allows project operation and management personnel to view the site conditions on the software platform.

[0049] 2. VR Virtual Reality Technology: Currently, dam safety monitoring departments primarily use structured reports for data presentation, which lacks intuitiveness. This invention, based on VR virtual reality technology, overlays water conservancy-related data onto a virtual dam, providing operators with an immersive spatial experience during remote monitoring. It also supports interactive displays in various aspects, including augmented reality, interactive experiences, and interactive sensor control, thus improving the visualization effect of data presentation.

[0050] 3. Water Conservancy Standard Database Design: This invention designs a comprehensive and standardized water conservancy database, improving the current situation in the water conservancy industry where data is diverse, large in volume, and difficult to manage. The database references various national standards in the water conservancy industry and covers all hydrological and water affairs data related to the daily operation of reservoirs and dams, including topographic data, meteorological data, hydrological data, and socio-economic data.

[0051] 4. Unmanned Inspection Function: Through the safety inspection subsystem, this invention can transmit the inspection indicators recorded by on-site inspectors, on-site photos of inspection points, and inspection routes in real time, which can help water conservancy experts make quick and accurate judgments and take measures. It can also directly contact the person in charge of flood control during flood control scheduling through mobile communication terminal devices, thereby improving the efficiency of command and dispatch.

[0052] 5. Lightweight Model Technology: Traditional modeling methods involve massive amounts of data, extremely long processing times, and put significant pressure on the system. This invention employs lightweight model technology to reduce the weight of point cloud data during 3D modeling. Specifically, point cloud processing software is used to extract feature point cloud data of building outlines using a feature extraction algorithm. While maintaining centimeter-level accuracy, some outliers and internal building point cloud data are removed.

[0053] 6. Internet of Things (IoT) Platform: Construct an IoT platform comprising device, support, platform, and business layers. Link the digital twin dam to the IoT platform to achieve unified access management of water conservancy sensor equipment terminals, including management and maintenance, data uploading to the cloud, and data analysis.

[0054] 7. Dedicated Evaluation System for Dams: Based on dam safety standards, a dam safety analysis method integrating model-driven and data-driven approaches is used to construct a dam safety inversion model and a smart monitoring and early warning system for each dam. For dams of different sizes and types, evaluation standards for monitoring indicators are formulated, and a dedicated safety evaluation standard system for each dam is constructed, enabling accurate real-time evaluation of the safety status of dam monitoring items.

[0055] 8. Historical Data Restoration and Analysis: This solution supports the replay of historical data from different monitoring items and time periods within the data pool. It can provide staff with historical event restoration and analysis, summarize historical flood prevention and flood control work, and design key work arrangements for future flood prevention and flood control.

[0056] 9. Automatic Generation of Special Reports: The system supports user-defined report templates and can automatically export special reports on dam safety monitoring from the platform's data pool at regular intervals based on the set business templates. The report export frequency can be set according to business needs, including daily, monthly, quarterly, and annual reports. This reduces the workload of reservoir management personnel and makes data reporting more timely. Attached Figure Description

[0057] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0058] Figure 1 This is a schematic diagram of the overall construction process of the embodiment of the present invention;

[0059] Figure 2 This is a flowchart illustrating the data cleaning process according to an embodiment of the present invention.

[0060] Figure 3 This is a flowchart illustrating the three-dimensional modeling process of an embodiment of the present invention;

[0061] Figure 4 This is a system architecture diagram of an embodiment of the present invention. Detailed Implementation

[0062] To make the features and advantages of this patent more apparent and understandable, specific embodiments are provided below for detailed explanation:

[0063] like Figures 1-4 As shown in the figure, this embodiment provides a safety inspection system for large and medium-sized reservoirs based on virtual reality technology, involving: physical dam, BIM model digital twin dam, dam inspection VR technology, equipment maintenance system, Internet of Things (IoT) technology, etc. It utilizes virtual reality technology combined with IoT technology in the water conservancy industry, providing a system to support safety inspection operations for safety monitoring and management departments of large and medium-sized reservoirs and dams. Among them:

[0064] The physical dam is the actual application scenario of the system, including the dam's physical structure, various connected sensors, other auxiliary facilities and equipment, and the surrounding physical environment; the various sensors include sensors that monitor the status and performance information of the dam's physical structure, and sensors that detect the status of the surrounding environment of the dam;

[0065] A virtual dam is a virtual BIM model built on top of a physical dam, forming a one-to-one mapping with the physical dam. It reflects the dam's structural characteristics, material characteristics, construction process characteristics, operation and maintenance process characteristics, and behavior change characteristics. It is used for calculation and analysis and is updated in real time according to changes in the physical dam.

[0066] The VR technology for dam inspection uses VR virtual reality technology to virtually simulate the parts of a reservoir dam that cannot be seen in person, and then combines this with the physical world to digitally recreate and present the scene.

[0067] The equipment maintenance subsystem is connected to the dam safety monitoring system, and through the integration of equipment management modules, equipment type management modules, spare parts management modules, etc., it can uniformly manage business data related to equipment maintenance, such as equipment inherent information and equipment maintenance information. This technology can improve the efficiency of maintenance personnel in performing maintenance and repair.

[0068] The Internet of Things (IoT) technology includes hardware systems and software platforms. It serves as a data storage center and is directly connected to various sensor acquisition systems, virtual dams, computing and analysis systems, as well as key technology knowledge graph information and decision control information. All data obtained from these connected systems are stored on the IoT central platform.

[0069] Its construction process includes the following steps:

[0070] Step S1: Install various sensors such as water and rainfall monitoring devices, humidity sensors, and liquid level sensors at the physical dam site to collect the status information, performance information, and surrounding environmental information of the physical dam. Data is obtained through the preset sensors and transmitted in real time.

[0071] Step S2: Preprocess the collected real-time monitoring data and transmit the processed real-time monitoring data to the IoT central platform;

[0072] Step S3: Create a database according to water conservancy industry standards to store topographic data, meteorological data, hydrological data, socio-economic data, real-time data, and historical data. The standard documents are as follows:

[0073] DL_T 1321-2014 Standard for Database Table Structure and Identifiers for Dam Safety Monitoring

[0074] SL 323-2011 Real-time Rainfall and Water Level Database Table Structure and Identifiers

[0075] SL 715-2015 Water Resources Information System Operation and Maintenance Specification 20201111-001

[0076] SLT 809-2021 Basic Database Table Structure and Identifiers for Water Conservancy Objects

[0077] The table structure and data reporting technical requirements for the special database on flash flood disasters (revised version)

[0078] National Hydrological Database Table Structure and Identifiers;

[0079] Step S4: After data cleaning, store the preprocessed real-time monitoring data into a standard database;

[0080] like Figure 2 As shown, the data cleaning method in step S4 specifically includes:

[0081] Step S41: Based on the operational characteristics of various hydrological data such as water level and rainfall, design different abnormal data filtering rules, time period data conversion rules, and missing value compensation algorithms;

[0082] Step S42: Filter the various real-time monitoring data using abnormal data filtering rules;

[0083] Step S43: According to the time period data conversion rules, convert the various real-time monitoring data collected from the field sensors into various data formats that meet business requirements;

[0084] Step S44: Verify all real-time monitoring data after data cleaning and remove monitoring data that fail to be verified.

[0085] Step S5: Based on the actual physical condition of the dam, use BIM technology to create a three-dimensional model to reflect the dam's structural characteristics, material characteristics, construction process characteristics, operation and maintenance process characteristics, behavior change characteristics, and sensor status.

[0086] like Figure 3 As shown, the 3D modeling method in step S5 specifically includes:

[0087] Step S51: Use drone equipment to perform aerial photography and laser scanning to obtain point cloud data and video images of the area to be modeled;

[0088] Step S52: Perform enhancement, segmentation, and classification operations on the point cloud data to clean and enrich the point cloud model, and denoise and simplify the point cloud data.

[0089] Step S53: Register the ground dense point cloud and the air dense point cloud using the iterative nearest point method, derive the rigid body transformation during registration, and fuse the ground dense point cloud and the air dense point cloud according to the rigid body transformation to obtain the air-ground three-dimensional point cloud.

[0090] Step S54: Effectively fuse panoramic photogrammetry images with 3D point clouds to obtain accurate geometric models and all-round texture information of the ground and objects in the target area, and realistically reconstruct the 3D appearance of the scene.

[0091] Step S55: Use close-up panoramic data to compensate for the deficiencies in the coverage of the side and bottom scenes of the obtained 3D model, and further improve the modeling effect;

[0092] Step S56: Perform color balance processing on the modeling image using a color binarization method. The steps are as follows:

[0093] Step 1: Let f(n,k) be the gray value of the image at pixel (n,k). Consider a (2i+1)×(2i+1) window centered at pixel (n,k).

[0094] Step 2: Calculate the threshold T(n,k) for each pixel (n,k) in the image;

[0095] Step 3: Binarize each pixel (n, k) in the image using the value of b(n, k).

[0096] Step S6: Overlay the 3D real-scene model data of water conservancy projects such as reservoirs, dikes, and sluices, as well as on-site sensors such as video monitors, water and rainfall monitoring devices, humidity sensors, and liquid level sensors onto the GIS map for display;

[0097] Step S7: After on-site survey, select shooting nodes, use a tripod and panoramic gimbal to take pictures, and trim and stitch the obtained dam site pictures to export 720VR panoramic images and overlay them onto a GIS map for display.

[0098] Step S8: In the real-time monitoring module and video surveillance, the system overlays the real-time monitoring data in the obtained standard database and the on-site monitoring video returned by the video monitor onto the GIS map for a vivid, comprehensive and intuitive display.

[0099] Step S9: In the monitoring and analysis module, the system processes the raw data collected from the standard database and establishes a data processing model. For potential dam safety hazards, it combines multiple nonlinear regression, machine learning, and other data analysis methods to accurately determine the dam's safety status, providing support for dam safety early warning judgments.

[0100] Step S10: In the early warning release module, based on the safety early warning judgment in step S8, the system intelligently identifies the early warning level, generates an early warning decision, and issues an early warning message according to the water conservancy industry early warning classification standard.

[0101] Step S11: Based on historical monitoring data, historical early warning records, historical inspection records, and historical maintenance records, the system automatically exports various business reports on a regular basis to enable staff to reconstruct and analyze historical events, summarize historical flood prevention and flood control work, and provide theoretical support for future flood prevention and flood control work arrangements.

[0102] Based on the above steps, this invention designs a complete security inspection software system, which includes a homepage overview, a real-time monitoring subsystem, a monitoring and analysis subsystem, a security inspection subsystem, an early warning release subsystem, a video surveillance subsystem, a monitoring report subsystem, and a system management module.

[0103] The real-time monitoring subsystem stores various monitoring data and, based on a geographic information system, overlays it onto two-dimensional maps and three-dimensional BIM models for intuitive display.

[0104] The monitoring and analysis subsystem is used to classify, calculate, and statistically analyze the various monitoring data from the database subsystem.

[0105] The safety inspection subsystem is based on a business flow engine to intelligently manage inspection points, inspection plans, and inspection personnel, registering inspection personnel information and assigning them inspection plans.

[0106] The early warning release subsystem is used to issue alarms for various precise monitoring data of the dam that exceed the threshold based on the construction and operation parameters.

[0107] The video surveillance subsystem is connected to the safety monitoring system of large and medium-sized reservoir dams, and through the video surveillance module, it can quickly retrieve, locate, compare and back up the surveillance video files, realize unified identity authentication and access management, and ensure the security of information and data.

[0108] The monitoring and reporting subsystem is used to display various reports generated by the dam's daily operation, including inspection reports, maintenance reports, and upkeep reports.

[0109] In summary, the main design features of this embodiment also include:

[0110] 1. Apply VR virtual reality technology to the water conservancy industry. Through 3D modeling, the parts of the reservoir dam that cannot be seen can be virtually simulated using digital twin technology. Combined with the physical world site, the virtual reality environment of the reservoir is digitally restored and presented, creating a VR panoramic environment covering the entire reservoir area. This digital and realistic approach supports the command and decision-making work of the command staff.

[0111] 2. The system supports interactive operation via VR terminals. Staff can use VR glasses or VR terminals for guidance and prompts to reach the corresponding potential hazard areas. By combining digital twin system integration with virtual display technology, the system can display the location of problems and related phenomena. By combining the current situation in reality with the information from the digital system, safety issues can be addressed in a timely manner, thereby improving and optimizing inspection efficiency.

[0112] 3. Employ BIM digital twin technology to digitally represent the reservoir dam. Construct digital twins of key structures such as the dam, reservoir, intake, spillway, gate control rooms, and monitoring stations. Combined with IoT sensor technology, the algorithm model can display the specific conditions of various points within the dam in real time, as well as predictive data. Risk levels in early warning areas are marked.

[0113] 4. Design a comprehensive water resources database based on national standards for the water conservancy industry. Data sources include, but are not limited to, hydrological and rainfall data, engineering data, meteorological data, water conservancy census data, flash flood disaster data, basic data, measurement data, and external system results data; data types include structured data as well as unstructured data such as documents, images, and videos. A data pool characterized by "wide sources, diverse types, large data volume, and high real-time performance" will be used to organize and display hydrological and water affairs data related to the daily operation of reservoirs and dams.

[0114] 5. In the safety inspection subsystem, the system can transmit the inspection indicators recorded by on-site inspectors, on-site photos of inspection points, and inspection routes in real time, which can help water conservancy experts make quick and accurate judgments and take measures. It can also directly contact the person in charge of flood control during flood control dispatch through mobile communication terminal devices, thereby improving the efficiency of command and dispatch.

[0115] 6. In oblique photogrammetry modeling, data preprocessing operations such as distortion correction, enhancement, format conversion, and image rotation are used to quickly solve the problems of image orientation and ground point densification. Simultaneously, the SIFT algorithm is used for image matching, which has a significant advantage in the number of feature points extracted from the point cloud model.

[0116] 7. Link the digital twin dam with the Internet of Things platform to achieve unified access management of water conservancy sensor equipment terminals, as well as management and maintenance, data uploading to the cloud, and data analysis.

[0117] 8. Based on dam safety standards, a dam safety analysis method integrating model-driven and data-driven approaches is used to construct a dam safety inversion model and a smart monitoring and early warning system for each dam. For dams of different sizes and types, evaluation standards for monitoring indicators are formulated, and a dedicated safety evaluation standard system for each dam is constructed, enabling accurate real-time evaluation of the safety status of dam monitoring items.

[0118] 9. Enable integrated display of information from key dam monitoring points. During the flood season, the real-time situation is complex and constantly changing. The system can access real-time video feeds from engineering safety monitoring and water and rainfall monitoring at key locations, combined with real-time dam safety monitoring data, to provide command personnel with comprehensive and real-time data information. This digital and real-scene approach supports command personnel in their decision-making.

[0119] This patent is not limited to the above-described preferred embodiments. Anyone can derive other forms of safety inspection systems for large and medium-sized reservoirs based on virtual reality technology under the guidance of this patent. All equivalent changes and modifications made within the scope of this patent application shall fall within the scope of this patent.

Claims

1. A large and medium-sized reservoir safety inspection system based on virtual reality technology, characterized in that, include: Sensing system: Install at the physical dam site a variety of sensors, including at least water and rainfall monitors, humidity sensors, and liquid level sensors, to collect the physical dam's status information, performance information, and surrounding environmental information. The system obtains data through various preset sensors and transmits it in real time. BIM Model Digital Twin Dam: A virtual BIM model is built based on the physical dam, forming a one-to-one mapping with the physical dam. It reflects the dam's structural characteristics, material characteristics, construction process characteristics, operation and maintenance process characteristics, and behavior change characteristics. It is used for calculation and analysis and is updated in real time according to changes in the physical dam. Data processing system; And IoT systems, used to connect sensing systems, BIM model digital twin dams, and data processing systems; The real-time monitoring data collected by the sensing system is preprocessed and then transmitted to the Internet of Things (IoT) system. The data processing system creates a database according to water conservancy industry standards, storing topographic data, meteorological data, hydrological data, socio-economic data, real-time data, and historical data. The preprocessed real-time monitoring data provided by the IoT system is cleaned and then stored in the standard database. Within the digital twin dam in the BIM model: The 3D reality model data of field sensors, including water conservancy projects such as reservoirs, dikes, and sluices, video monitors, water and rainfall monitoring devices, humidity sensors, and liquid level sensors, are overlaid onto a GIS map for display. After on-site surveying, shooting points were selected, and photos were taken using tripods and panoramic gimbals. The resulting on-site dam photos were then edited and stitched together to export 720VR panoramic images, which were then overlaid onto a GIS map for display. The real-time monitoring data in the obtained standard database, as well as the on-site monitoring video transmitted back by the video monitor, are overlaid on the GIS map for display. In the implementation of 3D modeling, point cloud data is processed in a lightweight manner. The specific processing method is as follows: using point cloud processing software, feature extraction algorithms are used to obtain the feature point cloud data of building outlines. While retaining the feature point cloud data with an accuracy of centimeters, some outlier point clouds and some point cloud data inside the building are removed. The data processing system utilizes a dam safety analysis method that integrates model-driven and data-driven approaches to construct a dam safety inversion model and a smart monitoring and early warning system for each dam. It also formulates monitoring indicator evaluation standards for dam body monitoring items based on the different sizes and types of dams, and constructs a dedicated safety evaluation standard system for each dam.

2. The large and medium-sized reservoir safety inspection system based on virtual reality technology according to claim 1, characterized in that: The method for constructing the BIM model digital twin dam is as follows: Drones were used for aerial photography and laser scanning to obtain point cloud data and video images of the area to be modeled. Enhancement, segmentation, and classification operations are performed on point cloud data to clean and enrich the point cloud model, and noise reduction and simplification are performed on the point cloud data. The ground dense point cloud and the air dense point cloud are registered by the iterative nearest point method, and the rigid body transformation during registration is derived. The ground dense point cloud and the air dense point cloud are fused according to the rigid body transformation to obtain the air and ground three-dimensional point cloud. By fusing panoramic photogrammetry images with 3D point clouds, a precise geometric model and all-around texture information of the ground and objects in the target area are obtained, realistically reconstructing the 3D appearance of the scene. By using close-up panoramic data, the shortcomings of the obtained 3D model in covering the side and bottom scenes can be compensated for, thereby further improving the modeling effect; The color binarization method is used to perform color balance processing on the modeled image. The steps are as follows: Step 1: Let f(n,k) be the gray value of the image at pixel (n,k). Consider a (2i+1)×(2i+1) window centered at pixel (n,k). Step 2: Calculate the threshold T(n,k) for each pixel (n,k) in the image; Step 3: Binarize each pixel (n, k) in the image using the value of b(n, k). 3.The large and medium-sized reservoir safety inspection system based on virtual reality technology according to claim 1, characterized in that: The data cleaning process specifically includes the following steps: Based on the operational characteristics of various hydrological data, including water level and rainfall, different abnormal data filtering rules, time period data transformation rules, and missing value compensation algorithms are designed. Filter various real-time monitoring data using abnormal data filtering rules; According to the time period data conversion rules, the real-time monitoring data collected from the field sensors will be converted into various data formats that meet business needs; Verify all real-time monitoring data after data cleaning and remove monitoring data that fails verification.

4. The large and medium-sized reservoir safety inspection system based on virtual reality technology according to claim 1, characterized in that: The data processing system processes the raw data collected from the standard database and establishes a data processing model. In response to potential safety hazards of dams, it combines various data analysis methods, including multivariate nonlinear regression and machine learning, to determine the safety status of dams and provide support for dam safety early warning judgment.

5. The large and medium-sized reservoir safety inspection system based on virtual reality technology according to claim 1, characterized in that: The data processing system identifies early warning levels, generates early warning decisions, and issues early warning information based on the water conservancy industry's early warning grading standards; it also automatically exports various business reports periodically based on historical monitoring data, historical early warning records, historical inspection records, and historical maintenance records.

6. The safety inspection system for large and medium-sized reservoirs based on virtual reality technology according to claim 1, characterized in that: The data processing system includes a real-time monitoring subsystem, a monitoring and analysis subsystem, a security inspection subsystem, an early warning release subsystem, a video surveillance subsystem, and a monitoring report subsystem. The real-time monitoring subsystem is used to store various monitoring data and, based on a geographic information system, overlays them onto two-dimensional maps and three-dimensional BIM models for intuitive display. The monitoring and analysis subsystem is used to classify, calculate, and statistically analyze the various monitoring data in the database; The security inspection subsystem is based on a business flow engine and manages inspection points, inspection plans, and inspection personnel, registers inspection personnel information, and assigns inspection plans to them. The early warning release subsystem is used to issue alarms for various dam monitoring data that exceed the threshold based on the construction and operation parameters that exceed the threshold. The video surveillance subsystem is connected to the dam safety monitoring system, and through the video surveillance module, it can quickly retrieve, locate, compare and back up the surveillance video files, realize unified identity authentication and access management, and ensure the security of information and data. The monitoring and reporting subsystem is used to display various reports generated by the dam's daily operation, including: inspection reports, maintenance reports, and upkeep reports.

Images