A method and device for synchronous acquisition of multi-source data suitable for rockfill dam surface
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CCCC YANGTZE CONSTRUCTION & DEVELOPMENT GROUP CO LTD
- Filing Date
- 2026-02-04
- Publication Date
- 2026-06-02
Smart Images
Figure CN122131349A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of intelligent detection technology for water conservancy projects, and in particular relates to a method and equipment for synchronous acquisition of multi-source data on the surface of rockfill dams. Background Technology
[0002] Rockfill dams are one of the mainstream dam types in water conservancy and hydropower projects. Their filling quality is directly related to the overall stability and long-term safety of the dam. During the construction of rockfill dams, the particle size distribution and compaction of the rockfill material are two important quality control indicators. If the particle size distribution is discontinuous or the internal compaction is not compacted, it is very easy to cause uneven settlement or seepage damage to the dam body in the later stage.
[0003] Traditional quality inspection of rockfill dam construction mainly relies on manual methods. For compaction, the "digging pits and filling with water" method is commonly used. This method not only damages the dam structure but also has low work efficiency, sparse sample points, and difficulty in reflecting the overall quality of the work surface. For particle size distribution, on-site sieving is often used, which is time-consuming, labor-intensive, and has a strong time lag. In recent years, ground penetrating radar (GPR) and machine vision technology have been gradually introduced into the field of water conservancy engineering inspection. GPR can invert the internal porosity and compaction layer structure by analyzing the difference in dielectric constant, while machine vision can extract the surface particle size distribution features through image segmentation algorithms.
[0004] However, existing multi-source data acquisition technologies face significant technical bottlenecks when directly applied to rugged rockfill dam construction sites. The rockfill dam fill surface is composed of large-diameter boulders, and the road conditions are extremely bumpy. Existing general-purpose mobile measurement systems, when acquiring images during continuous movement, are prone to motion blur due to high-frequency vibrations, resulting in the loss of image edge details. This severely affects the accuracy of subsequent particle size distribution analysis based on edge recognition. Therefore, how to obtain high-quality, clear texture images while ensuring operational efficiency is a major challenge. In addition, the "apparent characteristics" of rockfill materials are often strongly correlated with their "internal structure," but existing equipment usually operates independently. Radar data and image data lack a unified time reference and coordinate reference for rockfill dam projects. This makes it impossible to accurately align the surface gradation quality and internal density later, making it difficult to establish an accurate "apparent-internal" multimodal correlation model and provide reliable data support for the digital control and intelligent evaluation of dam construction quality. Summary of the Invention
[0005] The purpose of this invention is to address the problems existing in the prior art by providing a method and device for synchronous acquisition of multi-source data on the surface of rockfill dams.
[0006] To achieve the above objectives, the invention employs the following technical solution: a method for synchronous acquisition of multi-source data on the surface of a rockfill dam, the method comprising the following steps:
[0007] The GNSS positioning system continuously acquires the three-dimensional real-time position coordinates, velocity vector and azimuth of the acquisition equipment on the working face of the rockfill dam, and marks the three-dimensional real-time position coordinates with GNSS timestamps. The three-dimensional real-time position coordinates are then converted into rockfill dam engineering coordinates according to the engineering coordinate transformation model.
[0008] Based on the acquired velocity vector, the operating status of the acquisition device at the working face of the rockfill dam is determined, i.e., the operating status of the intelligent sensing device.
[0009] If it is determined that the acquisition device is stationary at the working face of the rockfill dam, the image acquisition module on the acquisition device is triggered to acquire image data of the rockfill material surface of the rockfill dam and extract the particle size distribution characteristics of the rockfill dam.
[0010] If it is determined that the acquisition device is in a moving state on the working face of the rockfill dam, the ground penetrating radar system on the acquisition device continuously scans the interior of the rockfill dam compaction layer to obtain the rockfill dam compaction layer data, extract the density characteristics of the rockfill dam, and combine the rockfill dam engineering coordinates to perform dynamic measurement line unit division to obtain measurement line trajectory data.
[0011] Using the GNSS timestamp of the GNSS positioning system as the time reference, the image data of the rockfill surface and the data of the rockfill dam compaction layer are processed for multi-source data time synchronization alignment to obtain time synchronization data.
[0012] Based on all the above data, a structured multimodal dataset is generated.
[0013] As a further technical solution, the method for converting the three-dimensional real-time position coordinates into rockfill dam engineering coordinates is as follows:
[0014] An engineering coordinate system is established based on the coordinates of the control points of the rockfill dam axis. By translating and rotating the coordinates in the Gaussian projection of the engineering coordinate transformation model, the three-dimensional real-time position coordinates obtained by the GNSS positioning system are converted into the station number, offset, and elevation relative to the dam axis and the reference plane. The position of the acquisition point on the working face of the rockfill dam is located in this way, and the engineering coordinates of the rockfill dam are obtained.
[0015] As a further technical solution, the method for determining the operating status of the data acquisition device at the working face of the rockfill dam is as follows:
[0016] When the moving speed vector of the acquisition device is less than the stationary determination threshold V for n consecutive seconds static At that time, it is determined that the data acquisition device is in a stationary state at the working face of the rockfill dam;
[0017] When the moving speed vector of the acquisition device is greater than the movement determination threshold V for n consecutive seconds moveAt that time, it is determined that the data acquisition device is in a moving state at the working face of the rockfill dam.
[0018] By adopting the above technical solution, the operating status of the acquisition equipment is determined, thereby achieving high-precision acquisition of surface image data of rockfill and compacted layer data of rockfill dam.
[0019] As a further technical solution, the V static The value range is 0.01 m / s - 0.2 m / s, and the V... move The condition for taking the value is δv is the hysteresis threshold, with a value of 0.05 m / s, and n ranges from 1 to 5 s.
[0020] As a further technical solution, the method for obtaining survey line trajectory data by dividing the dynamic survey line unit is as follows:
[0021] Each radar sampling point in the data of the rockfill dam compaction layer is associated with the GNSS timestamp and azimuth of the GNSS positioning system and the engineering coordinates of the rockfill dam. Based on the engineering coordinates of the rockfill dam and the strip-shaped characteristics of the rockfill dam filling and compaction process, the position coordinate sequence of the coordinate path is obtained. Based on the position coordinate sequence, the least squares method is used to fit the survey line, and the vertical distance d between the current position of the acquisition device and the current fitted survey line is calculated.
[0022] If d < preset survey line attribution threshold D threshold If the acquisition device is still within the range of the current fitted survey line, the current position point of the acquisition device is added to the position coordinate sequence of the current fitted survey line, the current fitted survey line is updated, and the survey line trajectory data is obtained.
[0023] If d > preset survey line attribution threshold D threshold If the acquisition device deviates from the range of the currently fitted survey line, the currently fitted survey line is marked as "completed". Then, taking the current position of the acquisition device as the starting point, a new position coordinate sequence of the coordinate path is re-acquired. Based on the new position coordinate sequence, the least squares method is used to fit a new survey line to obtain another survey line trajectory data.
[0024] By adopting the above technical solution, the ground penetrating radar operating trajectory is dynamically divided into multiple survey lines, thereby adapting to the irregular coordinate path formed by the acquisition equipment for rockfill dam engineering coordinates under complex terrain, and providing high-precision data support for intelligent detection of surface compaction quality of rockfill dams.
[0025] As a further technical solution, the method for acquiring time synchronization data is as follows:
[0026] The image data of the rockfill surface is associated with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project to generate a time series of rockfill images;
[0027] The radar sampling point in the data of the rockfill dam compacted layer is associated with the GNSS timestamp, rockfill dam project coordinates and azimuth angle obtained by the GNSS positioning system to generate a time series of the rockfill dam compacted layer.
[0028] The time series of the rockfill image is time-aligned with the time series of the rockfill dam compacted layer using the GNSS timestamp obtained by the GNSS positioning system, thereby obtaining time synchronization data between the time series of the rockfill image and the time series of the rockfill dam compacted layer.
[0029] By adopting the above technical solution, GNSS timestamps are used to align the time series of rockfill material images with the time series of the rockfill dam compaction layer.
[0030] As a further technical solution, the method for generating structured multimodal datasets is as follows:
[0031] Create a root directory named after the fill layer number and elevation of the rockfill dam;
[0032] Create a GNSS trajectory log folder in the root directory, and integrate the GNSS timestamps, three-dimensional real-time position coordinates and azimuth angles obtained by the GNSS positioning system with the coordinate records of the rockfill dam project according to the GNSS timestamp order;
[0033] Create a folder named "Rockfill Image Data" in the root directory to store the surface image data of the rockfill and the particle size distribution characteristics of the rockfill dam corresponding to each frame of the image. Associate each frame of the rockfill surface image data with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project, and annotate the file name of each frame of the rockfill surface image data.
[0034] Create a ground-penetrating radar survey line data folder in the root directory to store the survey line trajectory data obtained by the dynamic survey line unit division and the density characteristics of the rockfill dam corresponding to each survey line trajectory data. Associate each survey line trajectory data obtained by the dynamic survey line unit division with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project. Add metadata tags to the folder name of each survey line file.
[0035] Based on equipment status perception and adaptive update algorithm of measuring line, this invention realizes synchronous high-precision fusion of multi-source data in spatial and temporal dimensions, effectively solves the problem of asynchronous acquisition, outputs data with high spatial correlation, significantly improves the reliability and efficiency of intelligent detection of compaction quality of rockfill dams, and provides a solid data foundation for engineering safety assessment.
[0036] A device for synchronous acquisition of multi-source data on the surface of a rockfill dam is provided to implement a method for synchronous acquisition of multi-source data on the surface of a rockfill dam. The acquisition device includes a GNSS positioning system, an image acquisition module, a ground-penetrating radar system, and a data processing unit, all integrated and mounted on a mobile carrier platform.
[0037] The mobile platform is also equipped with a power module to provide stable power.
[0038] As a further technical solution, the image acquisition module uses an industrial camera equipped with a wide-angle lens, the GNSS positioning system uses a dual-frequency GNSS receiver, the ground penetrating radar system uses an antenna array, and the data processing unit is either an embedded processor or an industrial control computer.
[0039] By adopting the above technical solution, the industrial camera trigger signal is issued by the data processing unit based on the obtained velocity vector to determine the operating status of the acquisition device, thereby ensuring that high-definition images of the dam surface are acquired when the acquisition device is stationary.
[0040] As a further technical solution, the image acquisition module is connected to the data processing unit via a gigabit network port, and both the GNSS positioning system and the ground penetrating radar system are connected to the data processing unit via a wireless network.
[0041] By adopting the above technical solutions, gigabit Ethernet ports can provide sufficiently high transmission rates, avoiding frame drops or stuttering caused by insufficient bandwidth; wireless networks offer high flexibility and facilitate the use and adjustment of GNSS positioning systems and ground-penetrating radar systems.
[0042] Compared with the prior art, the beneficial effects of the present invention are:
[0043] 1. This invention is based on the GNSS positioning system, which performs precise positioning, intelligent determination of operating status, targeted collection and extraction of features, dynamic division of survey line units, and time synchronization and alignment of multi-source data, thereby generating a structured multimodal dataset, improving the registration accuracy of heterogeneous data for rockfill dam quality inspection, and providing reliable data support for the digital management and intelligent evaluation of rockfill dam construction quality.
[0044] 2. This invention automatically determines the operating status of the equipment based on velocity vectors, collecting image data of the rockfill surface when stationary and data of the compacted layer of the rockfill dam when in motion, reducing the amount of invalid data and improving the capture rate of key information;
[0045] 3. This invention uses GNSS timestamps as a unified time reference to ensure that the surface image data of the rockfill material and the compacted layer data of the rockfill dam are aligned in time scale, thereby achieving high-precision spatiotemporal synchronization and avoiding registration errors in the later stages.
[0046] 4. This invention innovatively proposes a dynamic survey line division method, which combines path fitting and distance criteria to achieve adaptive division of survey line units, adapting to irregular walking paths in complex terrain;
[0047] 5. This invention integrates a GNSS positioning system, an image acquisition module, a ground-penetrating radar system, a data processing unit, and a power supply module onto a mobile carrier platform, enabling flexible and regular inspection of various data on the rockfill dam, reducing repetitive work, minimizing manual labor, and improving inspection efficiency. Attached Figure Description
[0048] Figure 1 This is a schematic diagram of the process for synchronous acquisition of multi-source data on the surface of a rockfill dam, as described in this invention.
[0049] Figure 2 This is a schematic diagram of the acquisition device structure in the multi-source data synchronous acquisition device for the surface of a rockfill dam, applicable to the present invention;
[0050] Figure 3 This is a schematic diagram of the operating status of the acquisition equipment in the multi-source data synchronous acquisition method for the surface of a rockfill dam, applicable to the present invention.
[0051] Figure 4 This is a schematic diagram illustrating the dynamic measurement line unit division of the acquisition equipment in the multi-source data synchronous acquisition method for the surface of rockfill dams, applicable to the present invention.
[0052] Figure 5 (a) is a schematic diagram of the labeled image data of the rockfill surface in the multi-source data synchronous acquisition method for the surface of rockfill dams according to the present invention;
[0053] Figure 5 (b) is a schematic diagram of the data of the compacted layer of the rockfill dam in the multi-source data synchronous acquisition method of the rockfill dam surface of the present invention;
[0054] Figure 5 (c) is a schematic diagram of a single survey line in the multi-source data synchronous acquisition method for the surface of rockfill dams according to the present invention.
[0055] In the diagram: 101, GNSS positioning system; 102, image acquisition module; 103, ground penetrating radar system; 104, data processing unit; 105, power supply module; 106, mobile support platform; 107, casters. Specific implementation methods
[0056] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0057] In the description of this invention, it should be noted that the terms "middle", "upper", "lower", "left", "right", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.
[0058] like Figure 1-5 As shown, the specific solution of the embodiment is as follows: A method for synchronous acquisition of multi-source data on the surface of a rockfill dam, the method includes the following steps:
[0059] The GNSS positioning system continuously acquires the three-dimensional real-time position coordinates, velocity vector and azimuth of the acquisition equipment on the working face of the rockfill dam, and marks the three-dimensional real-time position coordinates with GNSS timestamps. The three-dimensional real-time position coordinates are then converted into rockfill dam engineering coordinates according to the engineering coordinate transformation model.
[0060] Based on the acquired velocity vector, the operating status of the data acquisition equipment at the rockfill dam working face is determined (i.e., the operating status of the intelligent sensing equipment).
[0061] If it is determined that the acquisition device is stationary at the working face of the rockfill dam, the image acquisition module on the acquisition device is triggered to obtain image data of the rockfill material surface of the rockfill dam and extract the particle size distribution characteristics of the rockfill dam.
[0062] If it is determined that the data acquisition equipment is in a moving state on the working face of the rockfill dam, the ground penetrating radar system on the data acquisition equipment continuously scans the interior of the rockfill dam compaction layer to obtain the data of the rockfill dam compaction layer, extract the density characteristics of the rockfill dam, and combine the rockfill dam engineering coordinates to divide the dynamic survey line unit and obtain the survey line trajectory data.
[0063] Using the GNSS timestamp of the GNSS positioning system as the time reference, the surface image data of the rockfill and the compacted layer data of the rockfill dam are processed for multi-source data time synchronization and alignment to obtain time synchronization data;
[0064] Based on all the above data, a structured multimodal dataset is generated.
[0065] It should be noted that all the above data include the three-dimensional real-time position coordinates, velocity vectors and azimuth angles of the GNSS positioning system, coordinates of the rockfill dam project, surface image data of the rockfill material and particle size distribution characteristics of the rockfill dam, data of the compacted layer of the rockfill dam and density characteristics of the rockfill dam, survey line trajectory data, and time synchronization data.
[0066] Among them, the engineering coordinate transformation model is the coordinate translation and rotation method in Gaussian projection. Based on the coordinates of the control points of the dam axis of the rockfill dam, an engineering coordinate system is established. Through coordinate translation and rotation in Gaussian projection, the three-dimensional real-time position coordinates obtained by the GNSS positioning system are converted into the station number, offset and elevation relative to the dam axis and the reference plane. In this way, the position of the acquisition point on the working face of the rockfill dam is located and the engineering coordinates of the rockfill dam are obtained.
[0067] Specifically, the GNSS positioning system sends NMEA0183 protocol data containing position, velocity, azimuth, and GNSS time to the data processing unit on the acquisition equipment at a frequency of 1Hz. The data processing unit then uses the coordinate translation and rotation method in the Gaussian projection method of the engineering coordinate transformation model to convert the three-dimensional real-time position coordinates into the engineering coordinates of the rockfill dam. The calculation formula is as follows:
[0068] ,
[0069] In the formula, x and y are the ordinate and abscissa of the rockfill dam project coordinates (Gauss projection plane coordinates), respectively; X0 is the meridian length between the equator and the projection point p0 on the reference ellipsoid from the ground point p; N is the radius of curvature of the primordial circle at point p0; l is the difference between the longitude of point p0 and the longitude of the central meridian of the projection zone; Lat is the latitude of point p0; and e' is the second eccentricity of the ellipsoid.
[0070] In this embodiment, the method for determining the operating status of the data acquisition device at the working face of the rockfill dam is as follows:
[0071] When the moving velocity vector of the data acquisition device is less than the stationary determination threshold V for n consecutive seconds static At that time, it was determined that the data acquisition equipment was stationary at the working face of the rockfill dam;
[0072] When the moving speed vector of the acquisition device is greater than the movement detection threshold V for n consecutive seconds move At that time, it was determined that the data acquisition equipment was in a moving state at the working face of the rockfill dam.
[0073] Among them, the static determination threshold V static and the movement determination threshold V move All of these are pre-set known values.
[0074] Specifically, let P be the data stream of velocity vectors continuously received by the data processing unit from the GNSS positioning system (e.g., ...). Figure 3 Let the velocities at corresponding times be P1, P2, P3, ..., P12 (marked in the middle). Let the current time be i. Then the velocity vector transmitted by the GNSS positioning system at time i is v. k To eliminate instantaneous fluctuations caused by vibration of the acquisition equipment or signal noise, the raw velocity data is smoothed and averaged to obtain the average moving speed v. f At this point, the window size is 5 seconds. The specific formula is as follows:
[0075] ,
[0076] Where k is the index variable for summation, and its value ranges from i-4 to i.
[0077] In this embodiment, V static The value range is 0.01 m / s - 0.2 m / s, V move The condition for taking the value is δv is the hysteresis threshold, with a value of 0.05 m / s, and n ranges from 1 to 5 s.
[0078] Since the window size for the smoothing average filter is 5 seconds, n is set to 5 seconds, and the optimal value is V. static It is 0.1 m / s, according to V move Given the conditions, calculate V. move The average moving speed after filtering is 0.15 m / s. f When the speed is greater than 0.15 m / s for 5 consecutive seconds, the data processing unit determines that the acquisition device is stationary, does not trigger the image acquisition module on the acquisition device, and directly starts the ground penetrating radar system to continuously acquire data of the rockfill dam compaction layer.
[0079] In this embodiment, the method for obtaining survey line trajectory data by dynamically dividing survey line units is as follows:
[0080] Each radar sampling point in the data of the rockfill dam compaction layer is associated with the GNSS timestamp and azimuth of the GNSS positioning system and the rockfill dam engineering coordinates. Based on the rockfill dam engineering coordinates and the strip-shaped characteristics of the rockfill dam filling and compaction process, the position coordinate sequence of the coordinate path is obtained. The least squares method is used to fit the survey line based on the position coordinate sequence, and the vertical distance d between the current position of the acquisition device and the current fitted survey line is calculated.
[0081] If d < preset survey line attribution threshold D thresholdIf the acquisition device is still within the range of the currently fitted measurement line, add the current position point of the acquisition device to the position coordinate sequence of the currently fitted measurement line, update the currently fitted measurement line, and obtain the measurement line trajectory data;
[0082] If d > preset survey line attribution threshold D threshold If the acquisition device deviates from the range of the currently fitted survey line, the currently fitted survey line is marked as "completed". Then, taking the current position of the acquisition device as the starting point, a new position coordinate sequence of the coordinate path is reacquired. Based on the new position coordinate sequence, the least squares method is used to fit a new survey line to obtain the trajectory data of another survey line.
[0083] Among them, the preset survey line attribution threshold D threshold The pre-set known value.
[0084] Specifically, the equation of the current survey line is obtained by selecting the 10 current valid positioning points and performing least squares fitting:
[0085] ,
[0086] Where A is the parameter in the x-direction of the measurement line, B is the parameter in the y-direction of the measurement line, and C is the position intercept of the measurement line, the vertical distance d between the current position of the acquisition device and the current fitted measurement line can be calculated.
[0087] In this embodiment, the method for obtaining time synchronization data is as follows:
[0088] By associating each frame of the rockfill surface image data with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project, a time series of rockfill images is generated.
[0089] By associating each radar sampling point in the data of the rockfill dam compacted layer with the GNSS timestamp obtained by the GNSS positioning system, the coordinates of the rockfill dam project, and the azimuth, a time series of the rockfill dam compacted layer is generated.
[0090] The time series of rockfill images and the time series of rockfill dam compacted layer are time-aligned using GNSS timestamps obtained from the GNSS positioning system to obtain time synchronization data between the time series of rockfill images and the time series of rockfill dam compacted layer.
[0091] It should be noted that for the surface image data of the riprap, each successfully acquired image has been associated with a GNSS timestamp T. image For the data of the compacted layer of the rockfill dam, each radar sampling point is also associated with its GNSS timestamp T at the time of acquisition. radar ;
[0092] The data processing unit extracts the timestamps of all images to form a time series T of the riprap images. SimageExtract the timestamps of all radar sampling points to form the time series T of the rockfill dam compaction layer. Sradar ;
[0093] By aligning with timestamps, we can accurately know the location and orientation of the acquisition device at any given time T, as well as the data on the compacted layer of the rockfill dam and the surface image data of the rockfill material that may have been acquired before and after that time.
[0094] For image frames for which a precise timestamp could not be directly obtained, the most recent GNSS timestamp was correlated with the internal clock of the image acquisition module and the GNSS clock through the calibration relationship between them.
[0095] In this embodiment, the method for generating a structured multimodal dataset is as follows:
[0096] Create a root directory named after the fill layer number and elevation of the rockfill dam;
[0097] Create a GNSS trajectory log folder in the root directory, and integrate the GNSS timestamps, three-dimensional real-time position coordinates and azimuth angles obtained by the GNSS positioning system with the coordinate records of the rockfill dam project in the order of GNSS timestamps;
[0098] Create a folder named "Rockfill Image Data" in the root directory to store the surface image data of the rockfill and the particle size distribution characteristics of the rockfill dam corresponding to each frame of the image. Associate each frame of the rockfill surface image data with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project, and mark it in the file name of each frame of the rockfill surface image data.
[0099] Create a ground-penetrating radar survey line data folder in the root directory to store the survey line trajectory data obtained from the dynamic survey line unit division and the corresponding rockfill dam compaction characteristics of each survey line trajectory data. Associate each survey line trajectory data obtained from the dynamic survey line unit division with the GNSS timestamp and azimuth of the GNSS positioning system and the rockfill dam project coordinates. Add metadata tags to the folder name of each survey line file.
[0100] Specifically, after data acquisition and synchronization are completed, the data processing unit integrates all the information to generate a structured multimodal dataset, which includes GNSS trajectory log files (GNSS trajectory log folder), geographic label image sequences (rockfill image data folder), and organized radar profile data (ground penetrating radar survey line data folder).
[0101] The GNSS trajectory log file contains a CSV or JSON file that records all GNSS data during the entire operation in timestamp order, including GNSS timestamps, 3D real-time position coordinates, rockfill dam project coordinates, velocity, azimuth, etc.
[0102] The geotagged image sequence contains a folder storing all the acquired high-resolution images. Each image filename, or its associated EXIF / JSON metadata, includes the coordinates, azimuth, and timestamp of the rockfill dam project at the time of acquisition. Figure 5 As shown, location information can be overlaid on the image;
[0103] In organized radar profile data, radar data is classified and stored according to dynamic survey line units. Each survey line corresponds to an independent radar data file (such as .DZT format) or folder. Each survey line file has detailed metadata tags, including: survey line unit number, start and end timestamps, start and end spatial coordinates, spatial range (boundary box), average velocity, and quality markers for rockfill dam compaction characteristics, etc.
[0104] Among them, by using equipment status perception and adaptive update algorithm for measuring lines, synchronous high-precision fusion of multi-source data in spatial and temporal dimensions was achieved, effectively solving the problem of asynchronous acquisition, outputting data with high spatial correlation, significantly improving the reliability and efficiency of intelligent detection of compaction quality of rockfill dams, and providing a solid data foundation for engineering safety assessment.
[0105] like Figure 2 As shown, a multi-source synchronous data acquisition device suitable for the surface of a rockfill dam is disclosed. The acquisition device includes a GNSS positioning system 101, an image acquisition module 102, a ground penetrating radar system 103, and a data processing unit 104 integrated and installed on a mobile support platform 106. A power module 105 is also provided on the mobile support platform 106 to provide stable power.
[0106] It should be noted that the mobile carrier platform 106 integrates a GNSS positioning system 101, an image acquisition module 102, a ground penetrating radar system 103, a data processing unit 104, and a power module 105. It has four casters and can be used as an AGV (Automated Guided Vehicle). It can move according to the instructions sent by the data processing unit 104 to achieve flexible data acquisition.
[0107] The GNSS positioning system 101 can be a dual-antenna GNSS receiver from Sinan Navigation, used to acquire the three-dimensional real-time position coordinates, velocity vector and azimuth of the acquisition equipment on the working face of the rockfill dam, and to mark the three-dimensional real-time position coordinates with GNSS timestamps. After transmitting the data to the data processing unit 104, the three-dimensional real-time position coordinates are converted into rockfill dam engineering coordinates using the rockfill dam engineering coordinate transformation model.
[0108] The image acquisition module 102 uses a high-resolution industrial camera equipped with a wide-angle lens and is connected to the data processing unit 104 via a gigabit network port. The data processing unit 104 receives the output signal from the GNSS positioning system 101 and processes it to determine the operating status of the acquisition device. When stationary, it acquires high-definition image data of the surface of the rockfill material of the rockfill dam.
[0109] The ground-penetrating radar system 103 uses a 400MHz antenna array and is connected to the data processing unit (104) via a wireless network. The data processing unit 104 receives the output signal of the GNSS positioning system 101 and processes and determines the operating status of the acquisition equipment. When in motion, it continuously scans the interior of the rockfill dam compaction layer to acquire data of the rockfill dam compaction layer.
[0110] The power module 105 integrates a high-capacity lithium battery to provide stable power support for all units;
[0111] The data processing unit 104 is either an embedded processor or an industrial control computer. It receives data from various units, determines the operating status of the acquisition equipment, extracts the particle size distribution characteristics of the rockfill dam, extracts the particle size distribution characteristics and survey trajectory data of the rockfill dam, performs time synchronization and alignment processing on various data, and generates a structured multimodal dataset.
[0112] Specifically, the mobile support platform 106 is equipped with a support frame, on which a power module 105 and a data processing unit 104 are mounted. A symmetrical GNSS positioning system 101 is also mounted on the support frame via a carrier frame. An image acquisition module 102 is mounted on the carrier frame via an extension frame, so that the image acquisition module 102 is not obstructed by the mobile support platform 106. A ground penetrating radar system 103 is mounted on the mobile support platform 106.
[0113] In this embodiment, the image acquisition module 102 is connected to the data processing unit 104 via a gigabit Ethernet port, and the GNSS positioning system 101 and the ground penetrating radar system 103 are both connected to the data processing unit 104 via a wireless network.
[0114] In some implementations, the image acquisition module 102 can also be connected to the data processing unit 104 via a wireless network, while the GNSS positioning system 101 and the ground penetrating radar system 103 can both be connected to the data processing unit 104 via a wired method.
[0115] The working principle of the above embodiments is as follows:
[0116] Using GNSS positioning and timestamps as a global benchmark, the operating status of the acquisition equipment is determined by velocity vectors. When stationary, the industrial camera is triggered to acquire image data of the rockfill surface, and the GNSS timestamp and location information of each frame are recorded synchronously. When moving, the data of the rockfill dam compaction layer are continuously acquired and the survey line units are automatically divided. The GNSS timestamp and location information of each radar sampling point are recorded synchronously. Furthermore, the image data of the rockfill surface and the data of the rockfill dam compaction layer are synchronized in time based on the GNSS timestamp. Finally, the trajectory coordinates, images and data of the rockfill dam compaction layer are fused to generate a spatiotemporally synchronized structured multimodal dataset, providing high-precision data support for intelligent detection of the surface compaction quality of rockfill dams.
[0117] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A method for synchronous acquisition of multi-source data on the surface of a rockfill dam, characterized in that, The method includes the following steps: The GNSS positioning system continuously acquires the three-dimensional real-time position coordinates, velocity vector and azimuth of the acquisition equipment on the working face of the rockfill dam, and marks the three-dimensional real-time position coordinates with GNSS timestamps. The three-dimensional real-time position coordinates are then converted into rockfill dam engineering coordinates according to the engineering coordinate transformation model. Based on the acquired velocity vector, the operating status of the acquisition device at the working face of the rockfill dam is determined; If it is determined that the acquisition device is stationary at the working face of the rockfill dam, the image acquisition module on the acquisition device is triggered to acquire image data of the rockfill material surface of the rockfill dam and extract the particle size distribution characteristics of the rockfill dam. If it is determined that the acquisition device is in a moving state on the working face of the rockfill dam, the ground penetrating radar system on the acquisition device continuously scans the interior of the rockfill dam compaction layer to obtain the rockfill dam compaction layer data, extract the density characteristics of the rockfill dam, and combine the rockfill dam engineering coordinates to perform dynamic measurement line unit division to obtain measurement line trajectory data. Using the GNSS timestamp of the GNSS positioning system as the time reference, the image data of the rockfill surface and the data of the rockfill dam compaction layer are processed for multi-source data time synchronization alignment to obtain time synchronization data. Based on all the above data, a structured multimodal dataset is generated.
2. The method for synchronous acquisition of multi-source data on the surface of a rockfill dam according to claim 1, characterized in that: The method for converting the three-dimensional real-time location coordinates into rockfill dam engineering coordinates is as follows: An engineering coordinate system is established based on the coordinates of the control points of the rockfill dam axis. By translating and rotating the coordinates in the Gaussian projection of the engineering coordinate transformation model, the three-dimensional real-time position coordinates obtained by the GNSS positioning system are converted into the station number, offset, and elevation relative to the dam axis and the reference plane. The position of the acquisition point on the working face of the rockfill dam is located in this way, and the engineering coordinates of the rockfill dam are obtained.
3. The method for synchronous acquisition of multi-source data on the surface of a rockfill dam according to claim 1, characterized in that: The method for determining the operating status of the data acquisition device at the working face of the rockfill dam is as follows: When the moving speed vector of the acquisition device is less than the stationary determination threshold V for n consecutive seconds static At that time, it is determined that the data acquisition device is in a stationary state at the working face of the rockfill dam; When the moving speed vector of the acquisition device is greater than the movement determination threshold V for n consecutive seconds move At that time, it is determined that the data acquisition device is in a moving state at the working face of the rockfill dam.
4. The method for synchronous acquisition of multi-source data on the surface of a rockfill dam according to claim 3, characterized in that: The V static The value range is 0.01 m / s - 0.2 m / s, and the V... move The condition for taking the value is δv is the hysteresis threshold, with a value of 0.05 m / s, and n ranges from 1 to 5 s.
5. A method for synchronous acquisition of multi-source data on the surface of a rockfill dam according to claim 1, characterized in that: The method for obtaining survey line trajectory data by dividing the dynamic survey line unit is as follows: Each radar sampling point in the data of the rockfill dam compaction layer is associated with the GNSS timestamp and azimuth of the GNSS positioning system and the engineering coordinates of the rockfill dam. Based on the engineering coordinates of the rockfill dam and the strip-shaped characteristics of the rockfill dam filling and compaction process, the position coordinate sequence of the coordinate path is obtained. Based on the position coordinate sequence, the least squares method is used to fit the survey line, and the vertical distance d between the current position of the acquisition device and the current fitted survey line is calculated. If d < preset survey line attribution threshold D threshold If the acquisition device is still within the range of the current fitted survey line, the current position point of the acquisition device is added to the position coordinate sequence of the current fitted survey line, the current fitted survey line is updated, and the survey line trajectory data is obtained. If d > preset survey line attribution threshold D threshold If the acquisition device deviates from the range of the currently fitted survey line, the currently fitted survey line is marked as "completed". Then, taking the current position of the acquisition device as the starting point, a new position coordinate sequence of the coordinate path is reacquired. Based on the new position coordinate sequence, the least squares method is used to fit a new survey line to obtain another survey line trajectory data.
6. A method for synchronous acquisition of multi-source data on the surface of a rockfill dam according to claim 5, characterized in that: The method for obtaining time synchronization data is as follows: The image data of the rockfill surface is associated with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project to generate a time series of rockfill images; The radar sampling point in the data of the rockfill dam compacted layer is associated with the GNSS timestamp, rockfill dam project coordinates and azimuth angle obtained by the GNSS positioning system to generate a time series of the rockfill dam compacted layer. The time series of the rockfill image is time-aligned with the time series of the rockfill dam compacted layer using the GNSS timestamp obtained by the GNSS positioning system, thereby obtaining time synchronization data between the time series of the rockfill image and the time series of the rockfill dam compacted layer.
7. A method for synchronous acquisition of multi-source data on the surface of a rockfill dam according to claim 6, characterized in that: The method for generating structured multimodal datasets is as follows: Create a root directory named after the fill layer number and elevation of the rockfill dam; Create a GNSS trajectory log folder in the root directory, and integrate the GNSS timestamps, three-dimensional real-time position coordinates and azimuth angles obtained by the GNSS positioning system with the coordinate records of the rockfill dam project according to the GNSS timestamp order; Create a folder named "Rockfill Image Data" in the root directory to store the surface image data of the rockfill and the particle size distribution characteristics of the rockfill dam corresponding to each frame of the image. Associate each frame of the rockfill surface image data with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project, and annotate the file name of each frame of the image data. Create a ground-penetrating radar survey line data folder in the root directory to store the survey line trajectory data obtained by the dynamic survey line unit division and the density characteristics of the rockfill dam corresponding to each survey line trajectory data. Associate each survey line trajectory data obtained by the dynamic survey line unit division with the GNSS timestamp and azimuth of the GNSS positioning system and the coordinates of the rockfill dam project. Add metadata tags to the folder name of each survey line file.
8. A device for synchronous acquisition of multi-source data on the surface of a rockfill dam, used to implement the method for synchronous acquisition of multi-source data on the surface of a rockfill dam as described in claims 1-7, characterized in that, The acquisition equipment includes the GNSS positioning system, the image acquisition module, the ground penetrating radar system, and the data processing unit, all integrated and installed on a mobile carrier platform. The mobile platform is also equipped with a power module to provide stable power.
9. A multi-source data synchronous acquisition device for the surface of a rockfill dam according to claim 8, characterized in that: The image acquisition module uses an industrial camera equipped with a wide-angle lens; the GNSS positioning system uses a dual-frequency GNSS receiver; the ground penetrating radar system uses an antenna array; and the data processing unit is either an embedded processor or an industrial control computer.
10. A multi-source data synchronous acquisition device for the surface of a rockfill dam according to claim 8, characterized in that: The image acquisition module is connected to the data processing unit via a gigabit Ethernet port, and both the GNSS positioning system and the ground-penetrating radar system are connected to the data processing unit via a wireless network.