A floating type self-adaptive pier scour depth monitoring device and method

Abstract

The application discloses a kind of floating self-adapting pier scouring depth monitoring device and method, including floating platform and its moving parts, image positioning system, water level calibration system and data processing terminal;Floating platform and its moving parts include: annular floating ring, monitoring floating platform, ultrasonic depth measuring module and power supply module.The application realizes the all-weather, full-domain dynamic monitoring of scouring depth around pier by floating self-adaptive monitoring and multi-sensor fusion positioning technology, breaks through the limitation of single-point monitoring.Using tidal reciprocating flow, promote the position change of measuring device, the device naturally floats and moves and automatically completes the three-dimensional coordinate test of riverbed surface adaptively, so as to realize the coverage of the area around pier, synchronously monitor scouring and silting process, and obtain complete riverbed topography dynamic data.

Description

Technical Field

[0001] This invention relates to a bridge pier scour monitoring device, and more particularly to a floating adaptive bridge pier scour depth monitoring device and method. Background Technology

[0002] Bridge pier scour is a major cause of bridge damage caused by floods, and accurate monitoring of the scour depth around the piers is crucial to ensuring bridge safety. Currently, the primary method for monitoring bridge pier scour is ultrasonic testing. Ultrasonic topographic surveying is divided into single-point measurement and side-scan. Single-point measurement, using a fixed installation method, can only measure the scour depth at a single location. Ultrasonic side-scan can measure the scour topography along a straight line, but the location is fixed and lacks variation, and it is also costly. Summary of the Invention

[0003] Purpose of the invention: To address the above problems, this invention proposes a floating adaptive bridge pier scour depth monitoring device and method, which can realize topographic measurement of different locations of scour pits and fuse them to obtain three-dimensional monitoring results of the scour depth around the bridge pier.

[0004] Technical solution: The technical solution adopted in this invention is a floating adaptive bridge pier scour depth monitoring device, including a floating platform and its moving parts, an image positioning system, a water level calibration system and a data processing terminal;

[0005] The floating platform and its moving parts include: an annular floating ring, which is fitted onto the outer wall of the pier and slides vertically along the pier; a monitoring platform, which floats on the water surface and is movably connected to the annular floating ring through a chain connection assembly; an ultrasonic depth sounding module is provided at the bottom of the monitoring platform to obtain the height data of the monitoring platform from the bed surface; and a power supply module to supply power to the electrical devices on the floating platform and its moving parts.

[0006] The image positioning system is used to acquire images of the water surface around the bridge pier in real time, and calculate the planar coordinate data of the monitoring platform on the horizontal plane in the real world coordinate system based on the images of the water surface around the bridge pier.

[0007] The water level calibration system is used to acquire water level elevation data in real time;

[0008] The data processing terminal is used to calculate the three-dimensional coordinates of the scour bed corresponding to the current location of the monitoring floating platform based on the height data, planar coordinate data and water level elevation data of the monitoring floating platform, and to fuse the three-dimensional monitoring results of the scour depth around the bridge pier based on the three-dimensional coordinate data of different locations.

[0009] The ultrasonic depth sounding module includes multiple ultrasonic sensors. These sensors take measurements at rapid intervals to ensure that the signals do not interfere with each other. Based on the data from multiple probes, the data with large errors are removed using the "three-times-standard-deviation" (3σ) rule, and then the average is calculated to obtain more accurate elevation data.

[0010] The image positioning system includes a camera installed on the side wall of the bridge pier for capturing real-time images of the water surface around the bridge pier.

[0011] The water level calibration system includes a water level gauge, which is installed on the bridge or on the bank, with the installation position above the highest water level coverage area.

[0012] The annular floating ring is made of polyurethane, which has good corrosion resistance and buoyancy. The chain connection assembly includes multiple retractable chain links, which can adapt to different water level changes.

[0013] The bottom of the monitoring platform is equipped with a counterweight, so that the center of gravity of the monitoring platform is located underwater.

[0014] The power supply module is powered by solar energy.

[0015] This invention proposes a method for monitoring the scour depth of bridge piers, comprising:

[0016] Based on the monitoring data of the height from the floating platform to the bed surface and water level elevation data The bed elevation data of the monitoring platform's location was obtained. ;

[0017] Based on the water surface images around the bridge piers, the planar coordinate data of the monitoring floating platform on the horizontal plane in the real-world coordinate system are obtained using an image target localization algorithm. , );

[0018] Based on the bed surface elevation data z and the plane coordinate data ( , The three-dimensional coordinates of the scouring bed surface corresponding to the current location of the monitoring floating platform are obtained by fusion. , ,z);

[0019] The monitoring platform moves around the bridge pier as the water flow and tides change, and can obtain the three-dimensional coordinates of different points on the scour bed around the bridge pier, thus obtaining the three-dimensional monitoring results of the scour depth around the bridge pier.

[0020] The image target localization algorithm includes: obtaining the mapping relationship between the pixel coordinates of the image captured by the camera and the real-world physical coordinates through a camera calibration method; performing real-time analysis of the acquired image using a machine learning algorithm to identify the monitoring platform in the image and obtain the pixel coordinates of the monitoring platform in the image; and transforming the image coordinates of the monitoring platform to the real-world coordinate system based on the pixel coordinates of the monitoring platform in the image and the mapping relationship between the obtained image pixel coordinates and the real-world physical coordinates to obtain the planar coordinate data of the monitoring platform on the horizontal plane in the real-world coordinate system.

[0021] The machine learning algorithm uses the YOLO object detection model.

[0022] Beneficial effects: Compared with the prior art, the present invention has the following advantages: (1) The present invention realizes all-weather, full-area dynamic monitoring of the scour depth around the bridge pier through floating adaptive monitoring and multi-sensor fusion positioning technology, breaking through the limitations of single-point monitoring. By utilizing the tidal reciprocating water flow, the position of the measuring device changes, the device floats and moves naturally and automatically completes the three-dimensional coordinate test of the riverbed surface, thereby realizing the coverage of the area around the bridge pier, synchronously monitoring the scour and siltation process, and obtaining complete dynamic data of the riverbed topography. (2) The device of the present invention is easy to install and has strong stability, solving the problems of difficult installation and easy damage in the prior art. (3) The present invention eliminates underwater construction operations, realizes rapid installation and deployment, avoids the sensor being directly buried in the riverbed or in contact with the riverbed, and prevents damage from impact of the bedload. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the structure of the floating platform and its moving parts described in this invention;

[0024] Figure 2 This is an overall schematic diagram of the floating adaptive pier scour depth monitoring device described in this invention.

[0025] Figure 3 This is a flowchart illustrating the image target localization algorithm;

[0026] Figure 4 This is the training process for the target detection and positioning model of the monitoring floating platform described in this invention. Detailed Implementation

[0027] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0028] Appendix Figures 1 to 2 The accompanying figure labels are as follows:

[0029] 1. Circular floating ring; 2. Monitoring platform; 3. Chain connection assembly; 4. Ultrasonic depth sounding module; 5. Counterweight; 6. Power supply module; 7. Camera; 8. Water level gauge; 9. Solar panel; 10. Pier.

[0030] Example 1:

[0031] The floating adaptive bridge pier scour depth monitoring device of the present invention includes a floating platform and its moving parts, an image positioning system, a water level calibration system, and a data processing terminal. Figure 1 The diagram shows a bottom view of the floating platform and its moving parts, including: an annular floating ring 1, a monitoring floating platform 2, a chain connection assembly 3, an ultrasonic depth sounding module 4, and a counterweight 5. The annular floating ring 1 is fitted onto the outer wall of the pier 10 and moves vertically along the pier 10 with the water level. The monitoring floating platform 2 floats on the water surface and is movably connected to the annular floating ring 1 via the chain connection assembly 3. The ultrasonic depth sounding module 4 is installed at the bottom of the monitoring floating platform, with its transducer pointing vertically downwards to acquire the height data of the monitoring floating platform from the water surface. A counterweight 5 is provided at the bottom of the monitoring floating platform to keep its center of gravity underwater. Figure 2 The figure shown is an overall schematic diagram of the floating adaptive pier scour depth monitoring device of the present invention.

[0032] The annular floating ring 1 is made of polyurethane material, which has good corrosion resistance and buoyancy. The chain connection assembly 3 consists of multiple retractable chain links, which can adapt to different water level changes. Figure 1 As shown, in practical applications, the annular floating ring 1 is fitted onto the outer wall of the pier 10, and the monitoring platform 2 is connected to the annular floating ring 1 through the chain connection component 3, and floats on the water surface.

[0033] The ultrasonic depth sounding module 4 includes four ultrasonic sensors. When the position of the monitoring platform 2 changes with the water flow, the four ultrasonic sensors quickly take measurements at intervals to ensure that the signals do not interfere with each other. Then, based on the data from multiple probes, the data with large errors are deleted using the "three-times standard deviation" (3σ) rule, and then the average is calculated to obtain more accurate elevation data.

[0034] The counterweight 5 keeps the center of gravity of the monitoring platform underwater, ensuring that the transducer is always facing the riverbed.

[0035] In this embodiment, the power supply module 6 is powered by solar energy. The power supply module 6 is charged by the solar panel 9 and supplies power to the electrical structures on the floating platform and its moving parts, such as the ultrasonic depth sounding module 4.

[0036] The image positioning system is used to acquire real-time images of the water surface around the bridge pier and calculate the planar coordinates of the monitoring floating platform 2 on the horizontal plane in the real-world coordinate system based on these images. The image positioning system includes a high-resolution camera 7 mounted on the side wall of the bridge pier to capture real-time images of the water surface around the pier. The system calculates the real-time position coordinates of the monitoring floating platform based on these images. The system periodically captures reference images of known locations and automatically corrects camera parameter deviations to ensure stable positioning accuracy.

[0037] The water level calibration system is used to acquire water level elevation data in real time. The water level calibration system includes a water level gauge 8, which is installed on a bridge or on the bank. The installation position of the water level gauge is above the highest water level coverage area, so that the water level elevation data can be measured.

[0038] The data processing terminal includes a memory and a processor. It is used to calculate the three-dimensional coordinates of various measuring points around the bridge pier based on the height data from the monitoring platform to the bed surface, the planar coordinates of the monitoring platform, and the water level elevation data, thus obtaining three-dimensional monitoring data on the scour depth around the bridge pier. The data processing terminal can also create a three-dimensional terrain model of the scour depth around the bridge pier based on the three-dimensional monitoring data. The data processing terminal is typically located in the bridge's electrical box, but can also be a computer device.

[0039] The monitoring platform floats and moves around the area around the bridge pier as the water flow and tides change, which can obtain the three-dimensional coordinates of different points on the scour bed around the bridge pier, thus obtaining the three-dimensional monitoring results of the scour depth around the bridge pier.

[0040] Example 2:

[0041] The bridge pier scour depth monitoring method of the present invention uses the floating adaptive bridge pier scour depth monitoring device described in Example 1 to monitor the bridge pier scour depth. The specific method includes:

[0042] Based on the monitoring data of the height from the floating platform to the bed surface and water level elevation data The bed elevation of the monitoring platform location is obtained. ;

[0043] Based on the water surface images around the bridge piers, the planar coordinates of the monitoring floating platform in the real scene are obtained using an image target localization algorithm. , );

[0044] Based on the bed elevation z of the monitoring platform's location and the planar coordinates of the monitoring platform in the real scene ( , The three-dimensional coordinates of the scouring bed surface corresponding to the current location of the monitoring floating platform are obtained by fusion. , The monitoring platform moves with the tides and water flow, allowing for the acquisition of scour data at different locations around the bridge pier. Based on the three-dimensional coordinate data from these different locations, a three-dimensional monitoring result of the scour depth around the bridge pier is obtained.

[0045] The flowchart of the image target localization algorithm is as follows: Figure 3 As shown, the core idea is as follows: First, through camera calibration, the mapping relationship between the pixel coordinates of the image captured by camera 7 and the real-world physical coordinates is obtained. Through a pre-calibration process, the camera's intrinsic parameters, extrinsic parameters, and distortion parameters are obtained; these parameters together constitute a mathematical model describing the camera's imaging geometric characteristics. Second, machine learning algorithms are used to analyze the images or video streams captured by camera 7 in real time, identifying and obtaining the pixel coordinates of the monitoring platform in the image. This invention uses the YOLO target detection model, and the pixel coordinates of the monitoring platform in the image are marked with center point coordinates. Finally, based on the pixel coordinates of the monitoring platform in the image, and the mapping relationship between the obtained image pixel coordinates and the real-world physical coordinates, the image coordinates of the monitoring platform are back-projected into the real-world coordinate system to complete planar positioning. The following section details the process of obtaining the pixel coordinates of the monitoring platform in the image, using a general image target localization algorithm:

[0046] 1. Camera calibration method.

[0047] Camera calibration is the first step in establishing an accurate mapping between image pixel coordinates and real-world physical coordinates. This invention employs the "Zhang Zhengyou calibration method" for camera calibration. First, a calibration board (e.g., a checkerboard) is held handheld, and multiple images (usually 10-20) are taken from different angles and positions within the camera's field of view. The images need to cover the entire field of view and include sufficient pose variations to ensure the stability of the solution. For each image, the pixel coordinates of all feature points (e.g., checkerboard points) on the calibration board are automatically detected. For each image, a 3 × 3 homography matrix H is calculated based on the known physical coordinates of the calibration board and its corresponding image pixel coordinates. This homography matrix describes the mapping relationship between points on the planar calibration board and their projection points on the image. Using the homography matrix calculated from multiple images, a system of linear equations is constructed to solve for the camera's intrinsic parameter matrix K. The intrinsic parameter matrix K describes the camera's internal geometric and optical characteristics. After obtaining the intrinsic parameter matrix, the extrinsic parameter matrix [R|t] for each image can be calculated based on the homography matrix and intrinsic parameter matrix of each image: H = K*[R|t], where H is the homography matrix (3×3 matrix) used to calculate the mapping between image coordinates and world coordinates; R is the rotation matrix (3×3 matrix) representing the rotation of the camera coordinate system relative to the world coordinate system; t is the translation vector (3×1 matrix) representing the translation of the camera coordinate system relative to the world coordinate system; and K is the intrinsic parameter matrix. The extrinsic parameter matrix [R|t] consists of a rotation matrix R and a translation vector t.

[0048] Nonlinear optimization is used to simultaneously optimize all parameters (intrinsic, extrinsic, and distortion coefficients) to minimize the reprojection error.

[0049] The camera intrinsic parameter K is obtained through calibration. , camera extrinsic matrix [R|t], and distortion coefficients (including radial and tangential distortion). and These represent the camera's focal length in the x and y directions (in pixels); s represents the slope of the x and y axes in the image coordinate system. and This represents the principal point in the image coordinate system. Camera calibration can be used to locate three-dimensional points in the world coordinate system. Points transformed to the camera coordinate system , . The world coordinate system is a predefined coordinate system used to describe the target's position in the real world. The camera coordinate system, centered on the camera, describes the target's position within the camera's viewpoint. Obtaining extrinsic parameters aims to correlate the scene observed by the camera with a predefined coordinate system of the floating platform on the water surface.

[0050] 2. Target detection and localization.

[0051] This invention uses YOLO and a trained model as the target detection and localization model for monitoring floating platforms. The model training process is described below. Figure 4 In real-time images or video streams captured by the camera, the pixel position of the monitoring platform in the image is identified. The output of the target detection algorithm is usually one or more bounding boxes. The bounding box is the output of the target detection algorithm, representing the range of the target's position in the image. In this embodiment, the bounding box is a rectangle, and the coordinates of the center point of the rectangle are extracted ( , ).

[0052] The coordinates of the center point can be calculated using the arithmetic mean: and ,this( , The coordinates, which are the pixel coordinates of the monitoring platform in the image, will be used as input for the subsequent coordinate mapping stage. , These are the left and right boundaries of the bounding box. , These are the upper and lower boundaries of the bounding box.

[0053] 3. Coordinate mapping.

[0054] The pixel coordinates of the target on the two-dimensional image plane ( , Convert it to its position in the three-dimensional real-world coordinate system. , , ), this is mainly ( , Elevation A water level gauge is needed to obtain the data.

[0055] Before using image coordinates for calculations, distortion correction is usually required. This invention uses a "single-point correction" approach, employing the OpenCV library's `cv2.undistortPoints()` function, taking the original coordinates, intrinsic parameter matrix, and distortion coefficients as input to obtain the corrected ideal pixel coordinates. Next, the ray direction is calculated, converting the pixel coordinates into y-order coordinate form (…). Homogeneous coordinates convert two-dimensional image coordinates into homogeneous coordinate form, facilitating matrix operations. Then, the vector is obtained by left multiplying by the inverse of the intrinsic parameter matrix. This is the direction vector of the ray in the camera coordinate system. The ray direction vector represents the direction vector from the camera center to the target point in the camera coordinate system. The final step is to solve for the intersection point of this ray and the plane containing the target. We will use the ray equation in the camera coordinate system... Transform to world coordinate system The scalar parameters representing the intersection of the ray and the plane are obtained by solving linear equations. This is achieved using the inverse transformation of the extrinsic parameters. , Represents the coordinates of a point in the world coordinate system. Representing the coordinates of a point in the camera coordinate system, we can obtain... Substitute this expression into the plane equation , Representing the water surface, we obtain a value about... Solving this linear equation yields a unique solution. Value. Substituting these coordinates into the ray equation, we can obtain the coordinates of the intersection point in the camera coordinate system. Finally, by using extrinsic parameter transformation again, we can transform these intersection point coordinates to the world coordinate system, thus obtaining the final position of the target in the real world. , , ).

[0056] 4. Output the results.

[0057] After coordinate mapping calculation, the system obtains the target's position coordinates in the real-world coordinate system. This invention only needs to monitor the two-dimensional coordinates of the floating platform and directly outputs (…). , That's it. At this point, the data terminal has obtained the planar coordinates of the monitoring floating platform, i.e., the ultrasonic sensor measurement point, on the water surface.

Claims

1. A floating adaptive pier scour depth monitoring device, characterized in that: It consists of a floating platform and its moving parts, an image positioning system, a water level calibration system, and a data processing terminal; The floating platform and its moving parts are composed of the following components: an annular floating ring (1), which is fitted on the outer wall of the pier and slides vertically along the pier; a non-powered monitoring floating platform (2), which floats on the water surface and is movably connected to the annular floating ring (1) through a chain connection component (3); an ultrasonic depth sounding module (4) is provided at the bottom of the non-powered monitoring floating platform (2) to obtain the height data of the non-powered monitoring floating platform to the bed surface; and a power supply module (6) to supply power to the electrical devices on the floating platform and its moving parts. The image positioning system is used to acquire images of the water surface around the bridge pier in real time, and to calculate the planar coordinate data of the unpowered monitoring platform (2) on the horizontal plane in the real world coordinate system based on the images of the water surface around the bridge pier. The water level calibration system is used to acquire water level elevation data in real time; The data processing terminal is used to calculate the three-dimensional coordinates of the scour bed corresponding to the current location of the unpowered monitoring floating platform based on the height data, plane coordinate data and water level elevation data of the unpowered monitoring floating platform to the bed surface, and to fuse the three-dimensional monitoring results of the scour depth around the bridge pier based on the three-dimensional coordinate data of different locations. The monitoring process for the three-dimensional monitoring results of the scour depth around the bridge pier includes: Based on the height data from the non-powered monitoring pontoon to the bed surface and the water level elevation data, the bed surface elevation data at the location of the non-powered monitoring pontoon is obtained. The bed surface elevation data is obtained by subtracting the height data from the non-powered monitoring pontoon to the bed surface from the water level elevation data. Based on images of the water surface around the bridge piers, an image target localization algorithm is used to obtain the planar coordinate data of the unpowered monitoring pontoon on the horizontal plane in the real-world coordinate system. The image target localization algorithm includes: obtaining the mapping relationship between the pixel coordinates of the images captured by the camera and the real-world physical coordinates through camera calibration; using a machine learning algorithm to analyze the acquired images in real time, identifying the unpowered monitoring pontoon in the images, and obtaining the pixel coordinates of the unpowered monitoring pontoon in the images; based on the pixel coordinates of the unpowered monitoring pontoon in the images, and the mapping relationship between the obtained image pixel coordinates and the real-world physical coordinates, transforming the image coordinates of the unpowered monitoring pontoon to the real-world coordinate system to obtain the planar coordinate data of the unpowered monitoring pontoon on the horizontal plane in the real-world coordinate system; the machine learning algorithm uses the YOLO target detection model. Based on the bed surface elevation data and the plane coordinate data, the three-dimensional coordinates of the scouring bed surface corresponding to the current location of the non-powered monitoring floating platform are obtained by fusing them. The non-powered monitoring platform moves under the action of water flow and tide to obtain the three-dimensional coordinates of different positions of the scour bed surface around the bridge pier, and obtain the three-dimensional monitoring results of the scour depth around the bridge pier. The process of transforming the image coordinates of the unpowered monitoring platform to the real-world coordinate system includes: using the formula: ， In the formula, , , The coordinates of the unpowered monitoring platform are shown in the world coordinate system, and x, y represent the coordinates of the unpowered monitoring platform in the image coordinate system. Scalar parameters representing the intersection of a ray and a plane; Substituting the above formula into the plane equation , get about The linear equation, solving the equation yields a unique solution. value; Will Substitute into the ray equation ,in This represents the coordinates of a point in the camera coordinate system, and we obtain the coordinates of the intersection point in the camera coordinate system. Finally, using extrinsic parameter transformation, we transform the coordinates of the intersection point in the camera coordinate system to the world coordinate system, thus obtaining the final position of the target in the real world. , , ).

2. The floating adaptive pier scour depth monitoring device according to claim 1, characterized in that: The ultrasonic depth sounding module (4) includes multiple ultrasonic sensors, which measure the height data of the monitoring platform from the bed surface at intervals. Based on the measurement data of the multiple ultrasonic sensors, the data with large errors are deleted using the "three standard deviations" rule, and then the average value is calculated as the final height data of the monitoring platform from the bed surface.

3. The floating adaptive pier scour depth monitoring device according to claim 1, characterized in that: The image positioning system includes a camera (7) installed on the side wall of the pier for capturing real-time images of the water surface around the pier.

4. The floating adaptive pier scour depth monitoring device according to claim 1, characterized in that: The water level calibration system includes a water level gauge (8), which is installed on the bridge or on the bank, with the installation position above the highest water level coverage area.

5. The floating adaptive pier scour depth monitoring device according to claim 1, characterized in that: The annular floating ring (1) is made of polyurethane material; the chain connection assembly (3) includes multiple retractable chain links.

6. The floating adaptive pier scour depth monitoring device according to claim 1, characterized in that: The bottom of the monitoring platform (2) is provided with a counterweight (5), which makes the center of gravity of the monitoring platform (2) located underwater.

7. The floating adaptive pier scour depth monitoring device according to claim 1, characterized in that: The power supply module (6) is powered by solar energy.

Images