Method and system for intelligently measuring underwater topography of power transmission and transformation project pond and accurately calculating backfill amount

Abstract

The invention discloses a power transmission and transformation project pond underwater terrain intelligent measurement and backfill amount accurate calculation method and system. According to the method, high-precision three-dimensional data acquisition is realized through a multi-beam sounding system carried by an unmanned ship; an improved ant colony algorithm is combined with B-spline interpolation to optimize path planning, a safe steering angle and a minimum turning path are set, and dynamic obstacle avoidance is achieved; processing data based on a layered sound ray tracking algorithm, CUBE filtering and an attitude correction technology, and eliminating sound velocity errors and environmental interference; performing triangulation network topological structure optimization on the point cloud data by adopting a Delaunay triangulation algorithm, and generating a DEM (Digital Elevation Model) through a contour extraction technology based on region growth; the calculation units are divided through a square grid method, and the backfill amount is accurately calculated through a triangular pyramid volume formula. According to the method, the problems of low precision and poor efficiency of a traditional measurement method are solved, the measurement error is reduced, and the automation degree of underwater topographic measurement and the engineering quantity calculation accuracy are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of water conservancy engineering measurement technology, and specifically to a method and system for intelligent measurement of underwater terrain and precise calculation of backfill volume in rivers and ponds of power transmission and transformation projects, which is particularly suitable for underwater terrain survey and earthwork volume calculation in power infrastructure construction. Background Art

[0002] In the construction of power transmission and transformation projects, accurately measuring the backfill volume of rivers and ponds has always been a technical challenge. Traditional underwater topographic survey methods mainly rely on manual pumping and cross-sectional measurement or the use of single-beam bathymetry systems. These methods have the following drawbacks: Inadequate measurement accuracy: Traditional cross-sectional surveying methods rely on manual point selection and calculation, resulting in rough estimates and large errors. Single-beam bathymetry systems can only obtain water depth data at a single point at a time, failing to achieve full coverage. This results in incomplete underwater topography data, impacting the accuracy of engineering quantity calculations. Inefficiency: Traditional methods require draining rivers and ponds before measurement, which is not only time-consuming and labor-intensive but also subject to weather and hydrological conditions, severely impacting project progress. Furthermore, manual operation and data processing are inefficient, making them inadequate for large-scale projects. Safety hazards: Manual or manned vessel surveying in complex waters (such as rapids and deep waters) presents significant safety risks, especially in inclement weather, where the safety of surveyors cannot be guaranteed. Poor data reliability: Traditional bathymetry methods are prone to data errors due to interference from environmental factors such as water turbidity, aquatic plants, and fish schools. Furthermore, they lack effective noise filtering and error correction methods, resulting in significant deviations from the actual topography.

[0003] In recent years, unmanned vessels equipped with multi-beam bathymetry technology have gradually been applied to underwater terrain surveying, but its practical application in power transmission and transformation projects still faces the following problems: the route planning algorithm lacks obstacle avoidance capability and path smoothness in complex waters; the multi-sensor data fusion and error correction technology is imperfect, especially the impact of sound velocity profile and attitude deviation on bathymetry accuracy has not been effectively solved; the efficiency and accuracy of point cloud data processing methods (such as contour extraction and three-dimensional modeling) need to be improved, and it is difficult to meet the centimeter-level DEM modeling requirements; the backfill volume calculation model lacks dynamic adaptation to high-precision terrain data, resulting in significant differences between the calculation results and the actual construction volume.

[0004] Therefore, there is an urgent need to develop an underwater topographic measurement method for power transmission and transformation projects that integrates high-precision data acquisition, intelligent path planning, multi-source data fusion processing, and precise engineering quantity calculation, so as to solve the problems of low precision, poor efficiency, and insufficient safety in existing technologies and provide reliable technical support for engineering construction. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a safe, efficient, accurate and reliable method and system for intelligent measurement of underwater topography and accurate calculation of backfill volume in rivers and ponds of power transmission and transformation projects.

[0006] In a first aspect, an embodiment of the present application provides a method for intelligently measuring underwater terrain and accurately calculating backfill volume in a river or pond in a power transmission and transformation project, the method comprising: S1. Use an unmanned vessel equipped with a multi-beam bathymetric system, integrated with a surface sound velocity meter, attitude indicator, and positioning system, to collect high-precision three-dimensional data of the underwater terrain of rivers and ponds, and obtain information on water depth, terrain, and obstacles; S2: Smoothing the planned path based on the improved ant colony algorithm and B-spline interpolation method, setting the safe steering angle and minimum turning path, and achieving dynamic obstacle avoidance and trajectory optimization; S3, intelligently process the collected underwater data through layered ray tracking algorithm, CUBE filtering algorithm, and attitude deviation correction algorithm to eliminate the influence of sound velocity error, environmental noise interference and sounding deviation; S4. Use the Delaunay triangulation algorithm to optimize the triangulated network topology of the point cloud data, and generate a high-precision underwater digital elevation model (DEM) through contour extraction technology based on region growing; S5. Based on the DEM model, the grid method is used to divide the calculation units and calculate the backfill volume.

[0007] Optionally, in an implementation of the first aspect of the present invention, S1, using an unmanned vessel equipped with a multi-beam bathymetry system, integrating a surface sound velocity meter, an attitude meter, and a positioning system, to perform high-precision three-dimensional data acquisition of underwater terrain of rivers and ponds to obtain water depth, terrain, and obstacle information, includes: S1.1. Use an unmanned vessel equipped with a multi-beam bathymetry system, integrating a surface sound velocity meter, an attitude meter, and a positioning system. This system uses multi-beam bathymetry technology to acquire underwater topographic data within a water depth range of 0.15-300 meters. It also simultaneously collects sound velocity profiles, vessel attitude, and high-precision positioning information to construct an original underwater point cloud dataset. S1.2. Preprocess the original underwater point cloud dataset, including: S1.2.1. Data validity check: Eliminate invalid sounding points with signal strength below the threshold, and filter out abnormal data caused by water surface reflection or instrument noise; S1.2.2, Time and Space Synchronization Calibration: Time align multi-sensor data based on GNSS timestamps, and unify each measurement point to the engineering coordinate system through coordinate transformation; S1.2.3, Attitude compensation correction: Based on the real-time collected roll and pitch data, the quaternion rotation matrix is ​​used to dynamically compensate the sounding point position; S1.2.4, Sound velocity profile correction: Based on the surface sound velocity meter and CTD profiler data, a layered sound velocity model is established to perform sound line bending correction on the original water depth value; S1.2.5. Data standardization: Convert the processed point cloud data into a standardized format containing three-dimensional coordinates, intensity values, and confidence levels to build a preprocessed point cloud database.

[0008] Optionally, in an implementation of the first aspect of the present invention, in S2, smoothing the planned path based on an improved ant colony algorithm and a B-spline interpolation method includes: S2.1. Constructing an improved ant colony algorithm path planning model: S2.1.1. Design Heuristic Function : , in, , , , , , in, is the distance function, 、 Is the current ant The next step is to select the maximum and minimum distances from all grid centers to the target grid center; Indicates the grid to be walked The distance heuristic factor from the center of to the starting grid and the target grid center; 、 is the distance coefficient, is the distance heuristic information coefficient; and Represents the next feasible grid Euclidean distance to the center of the starting grid and the target grid, Represents a feasible grid Euclidean distance to the center of the target grid, Represents ants A set of optional next feasible grids, express Time Path distance on; S2.1.2、Optimize path transfer probability: , in, Characterize the importance of pheromones, Indicates the importance of the heuristic function, express Time Path Pheromone concentration on; S2.2. Introducing path smoothness constraints: , in, is the distance heuristic information coefficient; is the coefficient representing the importance of straight travel; Indicates the direction the ant turned in the previous step. Indicates the direction to be turned next.

[0009] Optionally, in an implementation of the first aspect of the present invention, in S2, setting a safe steering angle and a minimum turning path to achieve dynamic obstacle avoidance and track optimization includes: S2.3. Set mechanical motion constraints: , , in, is the heading angle of the unmanned ship, is the offset angle between the ship's center axis and the steering direction, and the minimum turning path of the unmanned ship is , is the difference in center axis distance of the ship; S2.4. Implement path optimization: Use B-spline curve to smoothly interpolate the planned path, establish trajectory planning guide lines through the Freynes coordinate system, and construct the optimization objective function : , in, , , for the coordinate system Each discrete point in the axis direction The lateral offset , select the lateral offset of the point 、 and As the optimization variable, and For horizontal Offset relative to The first and second derivatives of ; 、 and is the corresponding weight coefficient; S2.5, Dynamic Obstacle Avoidance Processing: Create a route-time obstacle map, use the A* algorithm to perform global path search, and implement speed planning optimization: , in, , , Timestamp The corresponding longitudinal distance, To optimize the decision variables.

[0010] Optionally, in an implementation of the first aspect of the present invention, S3, performing intelligent processing on the collected underwater data through a layered ray tracking algorithm, a CUBE filtering algorithm, and an attitude deviation correction algorithm to eliminate the influence of sound velocity error, environmental noise interference, and bathymetric deviation, includes: S3.1. Using the layered ray tracing algorithm, a sound velocity profile model is established based on Snell's law. S3.2. Use the CUBE algorithm for automatic filtering, calculate the influence radius of the sounding point, calculate the grid point capture radius, and implement Bayesian filtering; S3.3. Based on the roll and pitch data collected by the POS system, dynamic compensation is performed using the quaternion rotation matrix to eliminate the "crying face" or "smiling face" deformation caused by residual errors; S3.4. Eliminate abnormal points with signal strength below the threshold, and perform manual interactive filtering to supplement the processing of complex terrain areas, outputting centimeter-level accuracy.

[0011] Optionally, in an implementation of the first aspect of the present invention, the step S3.1, using a layered ray tracing algorithm to establish a sound velocity profile model based on Snell's law, includes: a) Constant sound velocity layered model: The sound velocity profile is divided into N layers. The sound velocity in each layer is constant, and the sound line propagates in a straight line. The propagation time of each layer is calculated. and horizontal distance : , , in, is the initial incident angle of the beam, is the initial sound velocity, according to Snell's law and the sound velocity of each layer , the incident angle of the sound ray in each layer can be obtained, is the distance on the Z axis; b) Constant gradient layered model: The speed of sound changes linearly within each layer, and the sound line propagates in a circular arc. Calculate the curvature of the sound line: , in, The sound ray element The incident angle, is its grazing angle, for The speed of sound at the infinitesimal point is determined by Snell's law when the sound line is at Initial grazing angle within the layer and the initial sound speed , given time is a constant, is the distance element.

[0012] Optionally, in an implementation of the first aspect of the present invention, in S4, the Delaunay triangulation algorithm is used to optimize the triangulated network topology structure of the point cloud data, and a high-precision underwater digital elevation model DEM is generated by a contour extraction technology based on region growing, including: S4.1. Delaunay triangulation optimization: Perform Delaunay triangulation on the multi-beam Ping point cloud data, construct triangular facets by maximizing the minimum internal angle criterion, and ensure the geometric uniformity and topological stability of the triangulation network. Verify the legitimacy of the triangles using the empty circumscribed circle criterion, remove triangles that do not meet the Delaunay conditions, and optimize the triangulation network structure. S4.2. Contour extraction based on region growing: The Ping point cloud data collected by multi-beam sonar is spatially gridded, and distance and density thresholds are set. Using the seed point as the core, adjacent points that meet the threshold conditions are grouped into the same contour line using Euclidean distance clustering and region growing algorithms to avoid contour line breakage or distortion. For contour lines with "branching" problems, a breakpoint detection method based on curvature analysis is used to split complex contours into multiple simple sub-contours to ensure a "one-to-one" correspondence. S4.3. Contour patch filling: For the contour lines of adjacent pings, a corresponding point matching algorithm based on the minimum spanning tree (MST) is used to establish topological connections between the contour lines. Dynamic programming is used to optimize the generation path of triangular patches to ensure smooth transitions between adjacent contour lines. The optimized triangulated network model is then integrated with the measured microtopography data to generate a centimeter-level underwater digital elevation model (DEM). S4.4, DEM post-processing: Use the Laplacian smoothing algorithm to eliminate local noise in the triangulation network and improve the smoothness of the DEM surface; use Kriging interpolation to fill in data gaps and ensure the integrity and continuity of the DEM model.

[0013] Optionally, in an implementation of the first aspect of the present invention, in S5, the backfill amount is calculated by combining the grid method and the DEM model through the following steps: S5.1. Divide the DEM model into uniform grid cells at a preset resolution, with each cell corresponding to a planar area of ​​actual water; S5.2. Extract the elevation data of each grid cell and calculate the height difference between the cell and the cut and fill height based on the designed backfill elevation. S5.3, by bidirectionally splitting the triangular pyramid average value algorithm, according to the formula Calculate the backfill volume of each triangular pyramid element, where is the side length of the square, Fill and cut height difference for cell vertices; S5.4. Accumulate the backfill volume of all units, output the total backfill volume, and compare and verify it with the actual construction value.

[0014] In a second aspect, an embodiment of the present application provides a system for intelligently measuring underwater terrain and accurately calculating backfill volume in rivers and ponds for power transmission and transformation projects, which is applied to the method for intelligently measuring underwater terrain and accurately calculating backfill volume in rivers and ponds for power transmission and transformation projects as described in the first aspect. The system includes: The data acquisition module is used to use an unmanned vessel equipped with a multi-beam bathymetric system, integrating a surface sound velocity meter, attitude indicator, and positioning system to collect high-precision three-dimensional data of underwater terrain in rivers and ponds, and obtain information on water depth, terrain, and obstacles; The path planning module is used to smooth the planned path based on the improved ant colony algorithm and B-spline interpolation method, set the safe steering angle and minimum turning path, and achieve dynamic obstacle avoidance and trajectory optimization; The data processing module is used to intelligently process the collected underwater data through the layered ray tracking algorithm, CUBE filtering algorithm, and attitude deviation correction algorithm to eliminate the influence of sound velocity error, environmental noise interference, and bathymetric deviation; The 3D modeling module uses the Delaunay triangulation algorithm to optimize the triangulated network topology of point cloud data and generates a high-precision underwater digital elevation model (DEM) through contour extraction technology based on region growing; The engineering quantity calculation module is used to divide the calculation units using the grid method based on the DEM model and calculate the backfill quantity.

[0015] In a third aspect, an embodiment of the present application provides an electronic device, including: processor; a memory for storing processor-executable instructions; Among them, the processor is configured to implement the method for intelligent measurement of underwater terrain and accurate calculation of backfill volume of rivers and ponds in power transmission and transformation projects as described in the first aspect when executing the instructions.

[0016] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing a program, wherein the program instructs a device to execute the method for intelligent measurement of underwater terrain and precise calculation of backfill volume of rivers and ponds in power transmission and transformation projects as described in the first aspect.

[0017] The present invention discloses a method and system for intelligent measurement of underwater terrain and precise calculation of backfill volume in rivers and ponds of power transmission and transformation projects. The method uses an unmanned boat equipped with a multi-beam sounding system, integrated with a surface sound velocity meter, attitude meter and positioning system, to achieve high-precision three-dimensional data acquisition in a water depth range of 0.15-300 meters; an improved ant colony algorithm combined with B-spline interpolation is used to optimize path planning, set a safe turning angle and a minimum turning path, and achieve dynamic obstacle avoidance; data is processed based on a layered sound ray tracking algorithm, CUBE filtering and attitude correction technology to eliminate sound velocity errors and environmental interference; a Delaunay triangulation algorithm is used to optimize the triangulated network topology structure of point cloud data, and a high-precision underwater digital elevation model DEM is generated through a contour extraction technology based on regional growth; finally, a grid method is used to divide the calculation unit, and the backfill volume is accurately calculated using a triangular pyramid volume formula. The present invention solves the problems of low precision and poor efficiency of traditional measurement methods, reduces measurement errors, significantly improves the degree of automation of underwater terrain measurement and the accuracy of engineering quantity calculation, and provides reliable technical support for the construction of power transmission and transformation projects.

[0018] Beneficial effects:

[0019] 1. Significantly improved measurement accuracy. The multi-beam bathymetry system and high-precision sensors work together, combined with a layered ray tracking algorithm and attitude compensation technology to improve underwater terrain measurement accuracy, reduce backfill calculation errors, and improve calculation accuracy.

[0020] 2. Significantly improved operational efficiency: The improved ant colony algorithm is used to achieve autonomous path planning, combined with B-spline curve smoothing to improve measurement track optimization efficiency.

[0021] 3. Enhanced safety and reliability. A dynamic obstacle avoidance algorithm (A* search + speed planning optimization) enables the unmanned vessel to automatically avoid underwater obstacles, reducing operational risks. The CUBE filtering algorithm, combined with manual interactive verification, effectively eliminates interference from aquatic plants, fish schools, and other sources, improving data reliability. Mechanical motion constraints ensure navigation stability and prevent rollover accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A flow chart of a method for intelligent measurement of underwater terrain and precise calculation of backfill volume for power transmission and transformation projects provided in one embodiment of the present application.

[0023] Figure 2 This is a simplified model of a simple unmanned boat provided in one embodiment of the present application.

[0024] Figure 3 A comparison diagram of paths before and after smoothing provided in an embodiment of the present application.

[0025] Figure 4 A speed planning optimization diagram is provided for an embodiment of the present application.

[0026] Figure 5 This is a sound velocity layer diagram of a model with constant sound velocity within a layer provided by an embodiment of the present application.

[0027] Figure 6 This is a sound velocity layer diagram of a constant gradient assumption model within a layer provided by an embodiment of the present application.

[0028] Figure 7 This is an architecture diagram of a system for intelligent measurement of underwater terrain and precise calculation of backfill volume for power transmission and transformation projects, provided in one embodiment of the present application.

[0029] Figure 8 A schematic diagram of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments.

[0031] It should be noted that, in the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art in the art to which this application relates. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0032] It should be noted that, in the embodiments of the present application, words such as "first" and "second" are only used for the purpose of distinguishing descriptions, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying an order. Features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a concrete way.

[0033] Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0034] Example 1

[0035] Figure 1This is a flow chart of a method for intelligent measurement of underwater terrain and accurate calculation of backfill volume for power transmission and transformation projects provided in one embodiment of the present application. Figure 1 As shown, a method for intelligent measurement of underwater terrain and accurate calculation of backfill volume of rivers and ponds in power transmission and transformation projects includes: S1. Use an unmanned boat equipped with a multi-beam bathymetric system, integrated with a surface sound velocity meter, attitude meter, and positioning system, to collect high-precision three-dimensional data of the underwater topography of rivers and ponds, and obtain information on water depth, terrain, and obstacles.

[0036] For underwater topography surveying, the use of unmanned vessels equipped with multibeam bathymetry systems is an efficient and highly accurate method. By emitting fan-shaped beams and receiving echo signals reflected from the bottom, multibeam bathymetry systems can acquire high-density 3D point cloud data, thereby reconstructing complex underwater topography. Furthermore, the integration of surface velocity meters, attitude indicators, and positioning systems (such as RTK) further enhances measurement accuracy and ensures the accuracy of depth data. Furthermore, the automated operation capabilities of unmanned vessels make the survey process more efficient and reduce the need for human intervention.

[0037] Specifically, based on the DEM model, a grid method can be used to divide the underwater area into uniform calculation units. Each unit corresponds to a planar area of ​​the actual water area, and its elevation data is extracted. Combined with the designed backfill elevation, the fill-cut height difference of each unit is calculated. Using the bidirectional triangular pyramid average algorithm, the backfill volume of each triangular pyramid unit is calculated, and the backfill volumes of all units are accumulated to obtain the total backfill volume. Finally, the calculated results are compared and verified with the actual construction values ​​to ensure that the error is controlled within the effective range.

[0038] In this embodiment, S1 uses an unmanned vessel equipped with a multi-beam bathymetry system, integrated with a surface sound velocity meter, attitude meter, and positioning system, to collect high-precision three-dimensional data of the underwater terrain of rivers and ponds, and obtain information on water depth, terrain, and obstacles, including: S1.1. Use an unmanned vessel equipped with a multi-beam bathymetry system, which integrates a surface sound velocity meter, an attitude meter, and a positioning system. Multi-beam bathymetry technology is used to obtain underwater topographic data within a water depth range of 0.15-300 meters. The sound velocity profile, ship attitude, and high-precision positioning information are simultaneously collected to construct an original underwater point cloud dataset.

[0039] During data preprocessing, the raw underwater point cloud data must first be validated. Invalid sounding points with signal strength below a threshold are removed, and abnormal data caused by water surface reflections or instrument noise is filtered out. This process can be accomplished using specialized post-processing software (such as CARIS) to ensure data accuracy and reliability.

[0040] Specifically, in this embodiment, S1.2, pre-processing of the original underwater point cloud data set, including: S1.2.1, data validity check: eliminating invalid sounding points with signal strength below the threshold, and filtering abnormal data caused by water surface reflection or instrument noise. S1.2.2, time and space synchronization calibration: time alignment of multi-sensor data based on GNSS timestamp, and unification of each measuring point to the engineering coordinate system through coordinate transformation. In order to ensure the consistency of multi-sensor data, it is necessary to time align the multi-sensor data based on the GNSS timestamp, and unify each measuring point to the engineering coordinate system through coordinate transformation. This step can be achieved by the GNSS-RTK system carried by the unmanned vessel to ensure the time and space consistency of the data.

[0041] S1.2.3. Attitude Compensation: Based on real-time roll and pitch data, a quaternion rotation matrix is ​​used to dynamically compensate for the sounding point positions. During the measurement process, changes in the ship's attitude can affect the accuracy of the sounding data. Therefore, dynamic compensation of the sounding point positions is performed using a quaternion rotation matrix based on real-time roll and pitch data. This method effectively reduces the impact of attitude errors on measurement results and improves data accuracy.

[0042] S1.2.4 Sound Velocity Profile Correction: Based on the data from the surface sound velocity meter and CTD profiler, a layered sound velocity model is established to correct the raw water depth values ​​for ray curvature. Sound velocity is a key factor affecting the accuracy of water depth measurements. To eliminate the impact of sound velocity errors on measurement results, a layered sound velocity model is established based on the data from the surface sound velocity meter and CTD profiler, and ray curvature correction is applied to the raw water depth values. This process can be performed using multibeam post-processing software (such as CMS) to ensure the accuracy of water depth data.

[0043] S1.2.5 Data Standardization: Convert the processed point cloud data into a standardized format containing 3D coordinates, intensity values, and confidence levels to construct a preprocessed point cloud database. Finally, convert the processed point cloud data into a standardized format containing 3D coordinates, intensity values, and confidence levels to construct a preprocessed point cloud database. This process can be performed using specialized post-processing software (such as CARIS) to ensure data standardization and processability.

[0044] S2. Based on the improved ant colony algorithm and B-spline interpolation method, the planned path is smoothed, the safe steering angle and minimum turning path are set, and dynamic obstacle avoidance and trajectory optimization are achieved.

[0045] The unmanned vessel turning radius algorithm determines the minimum radius required for a vessel to turn during navigation. This algorithm plays a crucial role in ensuring both safe and efficient navigation. The calculation of the unmanned vessel turning radius algorithm requires consideration of various vessel parameters, such as length, width, displacement, and speed. These parameters all affect the turning radius, necessitating precise calculation. The ship turning radius algorithm also considers the vessel's navigation environment, such as water depth, current, and wind direction. These environmental factors also affect the turning radius, necessitating comprehensive consideration. Traditional ant colony algorithms fail to account for turning radius during path changes. However, in actual navigation, it is important to consider turning energy consumption and avoid capsizing during turns. Therefore, this study, based on an improved ant colony algorithm for navigation path planning, smoothes the path using B-spline interpolation to ensure vessel stability.

[0046] Path planning is a critical step in the unmanned vessel survey process. To ensure smooth measurement, the planned path must be smoothed, and safe turning angles and minimum turning paths must be set to achieve dynamic obstacle avoidance and track optimization. An improved ant colony algorithm and B-spline interpolation method can effectively optimize path planning, enabling unmanned vessels to efficiently and safely complete survey missions in complex waters. These algorithms not only improve path smoothness but also reduce unnecessary turns, thereby improving measurement efficiency.

[0047] In traditional ant colony algorithms, the heuristic function is only related to path length. However, in actual path planning, the evaluation of path quality cannot only consider path length. Path smoothness, adaptive distance heuristic factor, and algorithm runtime are also factors that need to be considered. Therefore, this paper improves the heuristic function by considering the total path length, path smoothness, and adaptive distance heuristic factor.

[0048] Specifically, in S2, the planned path is smoothed based on the improved ant colony algorithm and B-spline interpolation method, including: S2.1. Build an improved ant colony algorithm path planning model: Improve the heuristic function based on the total length of the path, the smoothness of the path, and the adaptive distance heuristic factor. S2.1.1. Design the heuristic function : .

[0049] Among them, the path length factor is related to the distance factor between the grid to be walked to the target grid and the grid to be walked to the starting grid. The distance function is introduced , defined as follows: , , , , , in, is the distance function, 、 Is the current ant The next step is to select the maximum and minimum distances from all grid centers to the target grid center; Indicates the grid to be walked The distance heuristic factor from the center of to the starting grid and the target grid center; 、 is the distance coefficient, is the distance heuristic information coefficient; and Represents the next feasible grid Euclidean distance to the center of the starting grid and the target grid, Represents a feasible grid Euclidean distance to the center of the target grid, Represents ants A set of optional next feasible grids, express Time Path The distance on.

[0050] Ants search for paths based on the guidance of the heuristic function, the pheromone concentration along the path, and the transition probability of the current node state.

[0051] S2.1.2, Being an Ant Located in the current grid When choosing the next direction of travel, the pheromone concentration on each path will be used. and heuristic function To determine the next path and optimize the path transfer probability: , in, Characterize the importance of pheromones, Indicates the importance of the heuristic function, express Time Path Pheromone concentration on; The value of is related to the global search capability of the algorithm. The value affects how the ant chooses the next adjacent grid; Represents ants A set of optional next feasible grids, Inversely proportional to the Euclidean distance between the two grid centers, the grid 、 distance The smaller it is, the larger the heuristic function for selecting the path is. The formula is as follows: .

[0052] S2.2. Introducing path smoothness constraints: , in, is the distance heuristic information coefficient; is the coefficient representing the importance of straight travel; Indicates the direction the ant turned in the previous step. Indicates the direction to be turned next.

[0053] S2.3, path smoothing step. Figure 2 , a simplified model of a simple unmanned ship. Assume that the unmanned ship is at a certain moment in the navigation process, and the heading of the unmanned ship is , the offset angle between the ship's center axis and the steering direction is , the steering angle of the unmanned ship has a mechanical characteristic constraint, that is, the mechanical characteristic constraint is set on the steering angle of the unmanned ship: , , in, is the heading angle of the unmanned ship, is the offset angle between the ship's center axis and the steering direction, and the minimum turning path of the unmanned ship is , is the center axis distance difference of the ship.

[0054] The smoothing result is as follows Figure 3 As shown, without smoothing, the path has obvious turning points, and some steering angles reach , which does not meet the actual navigation needs; the smoothed path can effectively meet the requirements of safe steering angle and minimum turning path during turning.

[0055] Extract key nodes from the output optimal path and perform B-spline interpolation on the inflection points. Select and set the optimization objective function. First, for each discrete point along the axis of the coordinate system, the lateral offset should be as close to the guide line as possible without considering the influence of surrounding environmental factors. and the second-order derivative , as small as possible, which can reduce the acceleration of lateral motion, avoid frequent steering, and save energy. According to the above analysis, the objective function is defined as: S2.4. Implement path optimization: Use B-spline curve to smoothly interpolate the planned path, establish trajectory planning guide lines through the Freynes coordinate system, and construct the optimization objective function : , In order to ensure the continuity of the path, the horizontal distance The relevant derivatives of are approximated by the difference at each discrete point, where , , for the coordinate system Each discrete point in the axis direction The lateral offset , select the lateral offset of the point 、 and As the optimization variable, and For horizontal Offset relative to The first and second derivatives of ; 、 and is the corresponding weight coefficient.

[0056] In the coordinate system of the route-time obstacle map, the cumulative distance along the time axis and the route is discretized to form a grid map of equal time and distance. Then, according to the planning requirements, various environmental factors are considered and different weights are used. A universal and unified cost function is established. Adjacent grid nodes in the grid map are assigned a cost. After the grid map is established, the entire solution process is equivalent to a typical graph search process for path planning. The search result is a globally optimal path. The search algorithm samples the typical A* algorithm. From the results of A*, it can be seen that a set of cumulative route distances S and time t can be obtained. Each point in this sequence represents the time (t) and the expected location (S) of the unmanned boat.

[0057] The search results are composed of multiple straight line segments connected in sequence, which does not meet the smoothness requirements of the autonomous navigation of the unmanned ship and does not meet the kinematic constraints. The results need to be optimized. First, calculate the timestamp Corresponding longitudinal distance , by fixing the timestamp Bundle As the decision variable for optimization, the objective function of the optimization problem is defined as follows according to the above analysis: S2.5, Dynamic Obstacle Avoidance Processing: Create a route-time obstacle map, use the A* algorithm to perform global path search, and implement speed planning optimization: , In order to ensure the continuity of speed distribution, The relevant derivatives are approximated by differences at each discrete point, i.e. , , Timestamp The corresponding longitudinal distance, To optimize the decision variables.

[0058] The above constraints are all linear constraints. The feasible domain of the decision variables has become a convex set through the obstacle avoidance constraint relationship. The objective function of the optimization problem is a quadratic convex function. Speed ​​optimization becomes a quadratic programming problem. The quadratic programming solver can be used to quickly solve it. The optimization results are as follows: Figure 4 Show.

[0059] S3. Intelligently process the collected underwater data through layered ray tracking algorithm, CUBE filtering algorithm, and attitude deviation correction algorithm to eliminate the influence of sound velocity error, environmental noise interference, and depth deviation.

[0060] During underwater data processing, sound velocity errors, environmental noise interference, and bathymetric deviations are the main factors affecting measurement accuracy. To mitigate these effects, layered ray tracking algorithms, CUBE filtering algorithms, and attitude deviation correction algorithms can be employed. These algorithms effectively correct for sound velocity errors, filter for environmental noise, and correct for bathymetric deviations, thereby improving the accuracy and reliability of underwater data. Furthermore, the combined use of these algorithms enables intelligent processing of underwater data, enhancing overall measurement quality.

[0061] Specifically, the S3, intelligently processing the collected underwater data through a layered ray tracking algorithm, a CUBE filtering algorithm, and an attitude deviation correction algorithm to eliminate the effects of sound velocity error, environmental noise interference, and sounding deviation, includes: S3.1. A layered ray-tracking algorithm is used to establish a sound velocity profile model based on Snell's law. S3.2. The CUBE algorithm is used for automatic filtering to calculate the influence radius of the sounding point, the grid point capture radius, and Bayesian filtering. S3.3. Based on the roll and pitch data collected by the POS system, a quaternion rotation matrix is ​​used for dynamic compensation to eliminate "crying face" or "smiling face" deformation caused by residual errors. S3.4. Outliers with signal strength below the threshold are eliminated, and manual interactive filtering is used to supplement the processing of complex terrain areas to output centimeter-level accuracy.

[0062] The impact of the sound velocity profile on bathymetry results is reflected in changes in the trajectory of sound rays, thereby affecting the spatial alignment of the beam footprint. Tracking the propagation path of sound rays through the water during multi-beam depth calculation is called ray tracking. The theoretical basis of ray tracking is the velocity stratification hypothesis, which states that any complex velocity profile structure can be approximated as consisting of multiple layers of simpler structures. This assumption replaces the continuous variation of the overall velocity profile with a broken-line distribution of the velocity within each individual layer. In practical applications, two common types of velocity stratification are constant velocity stratification and constant velocity gradient stratification. The former assumes that the velocity within each layer remains constant, and sound rays propagate along a straight line. The latter assumes that the velocity within a layer varies linearly, and sound rays propagate along a curved line.

[0063] Wherein, the step S3.1, using a layered ray tracing algorithm to establish a sound velocity profile model based on Snell's law, includes: a) Constant sound velocity layered model. The constant sound velocity model within the layer assumes that the sound velocity layered model is as follows Figure 5 As shown. When the sound velocity is constant in the layer, the grazing angle of the sound line in the layer remains unchanged and the trajectory is a straight line. Assume that the initial incident angle of the beam is , the initial sound speed is According to Snell's law and the sound speed of each layer , the incident angle of the sound ray in each layer can be obtained.

[0064] b) Divide the sound velocity profile into N layers. The sound velocity in each layer is constant and the sound line propagates in a straight line. Calculate the propagation time of each layer. and horizontal distance : , , in, is the initial incident angle of the beam, is the initial sound velocity, according to Snell's law and the sound velocity of each layer , the incident angle of the sound ray in each layer can be obtained, is the distance on the Z axis.

[0065] The cumulative propagation time and horizontal distance after the sound line propagates through the complete N-1 layer are as follows: , .

[0066] If the sound line still has time left after N-1 layers of refraction, it is recorded as , but it is not enough to pass through the complete Nth layer. The depth and horizontal distance propagated in the Nth layer are: , .

[0067] b) Constant gradient layered model. The constant gradient model assumes that the sound velocity layering within the layer is as follows: Figure 6 As shown. According to the basic theory of ray acoustics, The curvature of the sound ray trajectory of each layer, that is, the sound speed in each layer changes linearly, and the sound ray propagates in a circular arc. The curvature of the sound ray is calculated as: , in, The sound ray element The incident angle, is its grazing angle, for The speed of sound at the infinitesimal point, is the distance element, and according to Snell's law, when the sound line is Initial grazing angle within the layer and the initial sound speed , given time is a constant. For a constant velocity gradient layer is a constant. Therefore, is a constant, that is, In the layer, the curvature of the sound line is equal everywhere, which is The trajectory of the sound ray within the layer is an arc.

[0068] S4. Use the Delaunay triangulation algorithm to optimize the triangulated network topology structure of the point cloud data, and generate a high-precision underwater digital elevation model (DEM) through the contour extraction technology based on region growing.

[0069] In the above S4, the Delaunay triangulation algorithm is used to optimize the triangulated network topology structure of the point cloud data, and a high-precision underwater digital elevation model (DEM) is generated by a contour extraction technique based on region growing. Specifically, the following steps are included: S4.1. Delaunay triangulation optimization: Perform Delaunay triangulation on the multi-beam Ping point cloud data, construct triangular facets by maximizing the minimum internal angle criterion, and ensure the geometric uniformity and topological stability of the triangulation network. Verify the legitimacy of the triangles using the empty circumscribed circle criterion, eliminate triangles that do not meet the Delaunay conditions, and optimize the triangulation network structure.

[0070] Delaunay triangulation is a widely used algorithm in computational geometry. Its core concept is to construct triangular patches by maximizing the minimum interior angle criterion, ensuring the geometric uniformity and topological stability of the triangulated mesh. The empty circumcircle property of the Delaunay triangulation (i.e., the circumcircle of any triangle contains no other points) is a key basis for its validity. By removing triangles that do not meet the Delaunay condition, the triangulated mesh structure can be optimized, improving its applicability in underwater terrain modeling. In practical applications, Delaunay triangulation algorithms can be categorized into several types, including divide-and-conquer algorithms, triangulation growing methods, and random growing methods. Random growing methods are widely used for triangulating point cloud data due to their ease of implementation and low memory usage. Furthermore, when generating a triangulated mesh, the Delaunay triangulation ensures that the angles of each triangle are as close to equilateral as possible, thus avoiding the appearance of "slender" triangles and improving model stability and accuracy.

[0071] S4.2. Contour extraction based on region growing: The Ping point cloud data collected by multi-beam sonar is spatially gridded, and distance thresholds and density thresholds are set. With the seed point as the core, adjacent points that meet the threshold conditions are divided into the same contour line through Euclidean distance clustering and region growing algorithms to avoid contour line breakage or distortion. For contour lines with "branching" problems, a breakpoint detection method based on curvature analysis is used to split the complex contour into multiple simple sub-contours to ensure a "one-to-one" correspondence.

[0072] Region growing-based contour extraction is a commonly used method for point cloud data segmentation. Its core concept is to group nearby points that meet certain criteria into the same contour line by setting distance and density thresholds, thereby avoiding contour breakage or distortion. This method typically uses a seed point as its core and, through Euclidean distance clustering and region growing algorithms, gradually expands the contour range to ensure its continuity and integrity.

[0073] For contour lines with branching problems, a breakpoint detection method based on curvature analysis can be used to split complex contours into multiple simple sub-contours, ensuring a "one-to-one" correspondence. This method has shown good results in processing point clouds for complex building rooftop segmentation, effectively solving segmentation problems caused by point cloud fragmentation and errors.

[0074] S4.3. Contour line patch filling: For the contour lines of adjacent pings, a corresponding point matching algorithm based on the minimum spanning tree (MST) is used to establish topological connections between the contour lines. The generation path of triangular patches is optimized through dynamic programming to ensure a smooth transition between adjacent contour lines. The optimized triangulated network model is integrated with the measured microtopography data to generate a centimeter-level underwater digital elevation model (DEM).

[0075] During the contour patch filling phase, the contour lines of adjacent pings are topologically connected using a corresponding point matching algorithm based on the minimum spanning tree (MST). Dynamic programming optimizes the generation path of triangular patches to ensure smooth transitions between adjacent contour lines. Finally, the optimized triangulated mesh model is fused with measured microtopography data to generate an underwater digital elevation model (DEM) with centimeter-level accuracy.

[0076] S4.4, DEM post-processing: Use the Laplacian smoothing algorithm to eliminate local noise in the triangulation network and improve the smoothness of the DEM surface; use Kriging interpolation to fill in data gaps and ensure the integrity and continuity of the DEM model.

[0077] After DEM generation, the Laplacian smoothing algorithm is often used to eliminate local noise in the triangulation network to improve surface smoothness and data integrity. Furthermore, Kriging interpolation can be used to fill in data gaps to ensure the continuity and consistency of the DEM model.

[0078] The optimized triangulated network model is integrated with the measured micro-topography data to generate a centimeter-level underwater digital elevation model (DEM), which is used to accurately calculate the backfill volume of rivers and ponds and analyze underwater terrain.

[0079] S5. Based on the DEM model, the grid method is used to divide the calculation units and calculate the backfill volume.

[0080] Specifically, in S5, the grid method and the DEM model are combined to calculate the backfill amount through the following steps: S5.1, dividing the DEM model into uniform grid cells according to a preset resolution, each cell corresponding to a plane area of ​​the actual water area; S5.2, extract the elevation data of each grid unit, and calculate the height difference of the unit cut and fill combined with the designed backfill elevation; S5.3, use the bidirectional segmentation triangular pyramid average value algorithm, according to the formula Calculate the backfill volume of each triangular pyramid element, where is the side length of the square, The height difference between the fill and cut of the unit vertex; S5.4, accumulate the backfill volume of all units, output the total backfill volume, and compare and verify it with the actual construction value.

[0081] Example 2

[0082] like Figure 7 As shown, the present application provides an architecture diagram of a system for intelligent measurement of underwater terrain and precise calculation of backfill volume for power transmission and transformation projects, which is applied to the system for intelligent measurement of underwater terrain and precise calculation of backfill volume for power transmission and transformation projects as described in Example 1, including a data acquisition module 11, a path planning module 12, a data processing module 13, a three-dimensional modeling module 14, and an engineering quantity calculation module 15.

[0083] The data acquisition module 11 is used to use an unmanned vessel equipped with a multi-beam bathymetric system, which integrates a surface sound velocity meter, an attitude meter, and a positioning system to collect high-precision three-dimensional data of the underwater terrain of rivers and ponds, and obtain information on water depth, terrain, and obstacles; The path planning module 12 is used to smooth the planned path based on the improved ant colony algorithm and B-spline interpolation method, set the safe steering angle and the minimum turning path, and realize dynamic obstacle avoidance and track optimization; The data processing module 13 is used to intelligently process the collected underwater data through a layered ray tracking algorithm, a CUBE filtering algorithm, and an attitude deviation correction algorithm to eliminate the effects of sound velocity error, environmental noise interference, and sounding deviation; 3D modeling module 14 uses the Delaunay triangulation algorithm to optimize the triangulated network topology of point cloud data and generates a high-precision underwater digital elevation model (DEM) through contour extraction technology based on region growing; The engineering quantity calculation module 15 is used to divide the calculation units into grids based on the DEM model and calculate the backfill quantity.

[0084] Figure 8 This is an electronic device provided by an embodiment of the present application. Figure 8 As shown, the electronic device includes at least the following parts: a processor 101 and a memory 100 , a communication interface 103 , and a bus 102 .

[0085] In an embodiment of the present application, the memory 100 is used to store instructions executable by the processor 101, and the processor 101 is configured to implement the method of the first aspect when executing the instructions.

[0086] In an embodiment of the present application, a computer-readable storage medium includes instructions, and the instructions instruct a device to execute the method of the first aspect. For example, the instructions instruct the device to execute Figure 1 The method is shown in the process steps.

[0087] The program running in the electronic device involved in one embodiment of the present application may be a program that controls a central processing unit (CPU) and the like to implement the functions of the above-mentioned embodiment involved in one embodiment of the present invention (a program that causes a computer to function). The information processed by these devices is temporarily stored in random access memory (RAM) while being processed, and then stored in various ROMs such as read-only memory (Flash ROM) and hard disk drives (HDDs), where it is read, modified, and written as needed by the CPU.

[0088] It should be noted that a portion of the electronic device of the above embodiment may also be implemented by a computer. In this case, a program for implementing the control function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read into a computer and executed.

[0089] It should be noted that the "computer" mentioned here refers to a computer built into an electronic device, employing hardware including an operating system (OS) and peripheral devices. Furthermore, "computer-readable recording medium" refers to removable media such as floppy disks, magneto-optical disks, ROMs, and CD-ROMs, as well as storage devices such as hard disks built into computers.

[0090] Furthermore, "computer-readable recording media" may include: media that dynamically store programs for a short period of time, such as communication lines when transmitting programs via networks such as the Internet or communication lines such as telephone lines; and media that store programs for a fixed period of time, such as volatile memory within computers acting as servers or clients in this context. Furthermore, the aforementioned program may be a program for implementing a portion of the aforementioned functions, or a program that can achieve the aforementioned functions by combining with a program already stored in a computer.

[0091] Furthermore, the electronic device in the above-described embodiments can also be implemented as a collection of multiple devices (a device group). Each device comprising the device group may include some or all of the functions or functional blocks of the electronic device in the above-described embodiments. A device group only needs to include all of the functions or functional blocks of the electronic device.

[0092] Those skilled in the art should recognize that the above embodiments are merely intended to illustrate the present application and are not intended to limit the present application. As long as they are within the spirit of the present application, appropriate changes and modifications to the above embodiments are within the scope of protection claimed in the present application.

Claims

1. A method for intelligent measurement of underwater terrain and accurate calculation of backfill volume in rivers and ponds for power transmission and transformation projects, characterized by: The method comprises: S1. Use an unmanned vessel equipped with a multi-beam bathymetric system, integrated with a surface sound velocity meter, attitude indicator, and positioning system, to collect high-precision three-dimensional data of the underwater terrain of rivers and ponds, and obtain information on water depth, terrain, and obstacles; S2: Smoothing the planned path based on the improved ant colony algorithm and B-spline interpolation method, setting the safe steering angle and minimum turning path, and achieving dynamic obstacle avoidance and trajectory optimization; S3, intelligently process the collected underwater data through layered ray tracking algorithm, CUBE filtering algorithm, and attitude deviation correction algorithm to eliminate the influence of sound velocity error, environmental noise interference and sounding deviation; S4. Use the Delaunay triangulation algorithm to optimize the triangulated network topology of the point cloud data, and generate a high-precision underwater digital elevation model (DEM) through contour extraction technology based on region growing; S5. Based on the DEM model, the grid method is used to divide the calculation units and calculate the backfill volume.

2. The method for intelligent measurement of underwater terrain and accurate calculation of backfill volume of rivers and ponds in power transmission and transformation projects according to claim 1 is characterized in that: S1 uses an unmanned vessel equipped with a multi-beam bathymetry system, integrated with a surface sound velocity meter, attitude meter, and positioning system, to collect high-precision three-dimensional data of the underwater terrain of rivers and ponds, and obtain information on water depth, terrain, and obstacles, including: S1.

1. Use an unmanned vessel equipped with a multi-beam bathymetry system, integrating a surface sound velocity meter, an attitude meter, and a positioning system. This system uses multi-beam bathymetry technology to acquire underwater topographic data within a water depth range of 0.15-300 meters. It also simultaneously collects sound velocity profiles, vessel attitude, and high-precision positioning information to construct an original underwater point cloud dataset. S1.

2. Preprocess the original underwater point cloud dataset, including: S1.2.

1. Data validity check: Eliminate invalid sounding points with signal strength below the threshold, and filter out abnormal data caused by water surface reflection or instrument noise; S1.2.2, Time and Space Synchronization Calibration: Time align multi-sensor data based on GNSS timestamps, and unify each measurement point to the engineering coordinate system through coordinate transformation; S1.2.3, Attitude compensation correction: Based on the real-time collected roll and pitch data, the quaternion rotation matrix is ​​used to dynamically compensate the sounding point position; S1.2.4, Sound velocity profile correction: Based on the surface sound velocity meter and CTD profiler data, a layered sound velocity model is established to perform sound line bending correction on the original water depth value; S1.2.

5. Data standardization: Convert the processed point cloud data into a standardized format containing three-dimensional coordinates, intensity values, and confidence levels to build a preprocessed point cloud database.

3. The method for intelligent measurement of underwater terrain and accurate calculation of backfill volume of rivers and ponds in power transmission and transformation projects according to claim 1 is characterized in that: In S2, the planned path is smoothed based on the improved ant colony algorithm and B-spline interpolation method, including: S2.

1. Constructing an improved ant colony algorithm path planning model: S2.1.

1. Design Heuristic Function : ， in, ， ， ， ， ， in, is the distance function, 、 Is the current ant The next step is to select the maximum and minimum distances from all grid centers to the target grid center; Indicates the grid to be walked The distance heuristic factor from the center of to the starting grid and the target grid center; 、 is the distance coefficient, is the distance heuristic information coefficient; and Represents the next feasible grid Euclidean distance to the center of the starting grid and the target grid, Represents a feasible grid Euclidean distance to the center of the target grid, Represents ants A set of optional next feasible grids, express Time Path distance on; S2.1.2、Optimize path transfer probability: ， in, Characterize the importance of pheromones, Indicates the importance of the heuristic function, express Time Path Pheromone concentration on; S2.

2. Introducing path smoothness constraints: ， in, is the distance heuristic information coefficient; is a coefficient representing the importance of straight travel; Indicates the direction the ant turned in the previous step. Indicates the direction to be turned next.

4. The method for intelligent measurement of underwater terrain and accurate calculation of backfill volume of rivers and ponds in power transmission and transformation projects according to claim 2 is characterized in that: In S2, a safe steering angle and a minimum turning path are set to achieve dynamic obstacle avoidance and track optimization, including: S2.

3. Set mechanical motion constraints: ， ， in, is the heading angle of the unmanned ship, is the offset angle between the ship's center axis and the steering direction, and the minimum turning path of the unmanned ship is , is the difference in center axis distance of the ship; S2.

4. Implement path optimization: Use B-spline curve to smoothly interpolate the planned path, establish trajectory planning guide lines through the Freynes coordinate system, and construct the optimization objective function : ， in, , , for the coordinate system Each discrete point in the axis direction The lateral offset , select the lateral offset of the point 、 and As the optimization variable, and For horizontal Offset relative to The first and second derivatives of ; 、 and is the corresponding weight coefficient; S2.5, Dynamic Obstacle Avoidance Processing: Create a route-time obstacle map, use the A* algorithm to perform global path search, and implement speed planning optimization: ， in, , , Timestamp The corresponding longitudinal distance, To optimize the decision variables.

5. The method for intelligent measurement of underwater terrain and accurate calculation of backfill volume of rivers and ponds in power transmission and transformation projects according to claim 2 is characterized in that: S3, intelligently processing the collected underwater data through a layered ray tracking algorithm, a CUBE filtering algorithm, and an attitude deviation correction algorithm to eliminate the effects of sound velocity error, environmental noise interference, and sounding deviation, including: S3.

1. Using the layered ray tracing algorithm, a sound velocity profile model is established based on Snell's law. S3.

2. Use the CUBE algorithm for automatic filtering, calculate the influence radius of the sounding point, calculate the grid point capture radius, and implement Bayesian filtering; S3.

3. Based on the roll and pitch data collected by the POS system, dynamic compensation is performed using the quaternion rotation matrix to eliminate the "crying face" or "smiling face" deformation caused by residual errors; S3.

4. Eliminate abnormal points with signal strength below the threshold, and perform manual interactive filtering to supplement the processing of complex terrain areas, outputting centimeter-level accuracy.

6. The method for intelligent measurement of underwater terrain and accurate calculation of backfill volume for power transmission and transformation projects according to claim 5 is characterized in that: S3.1, using a layered ray tracing algorithm to establish a sound velocity profile model based on Snell's law, includes: a) Constant sound velocity layered model: The sound velocity profile is divided into N layers. The sound velocity in each layer is constant, and the sound line propagates in a straight line. The propagation time of each layer is calculated. and horizontal distance : ， ， in, is the initial incident angle of the beam, is the initial sound velocity, according to Snell's law and the sound velocity of each layer , the incident angle of the sound ray in each layer can be obtained, is the distance on the Z axis; b) Constant gradient layered model: The speed of sound changes linearly within each layer, and the sound line propagates in a circular arc. Calculate the curvature of the sound line: ， in, The sound ray element The incident angle, is its grazing angle, for The speed of sound at the infinitesimal point is determined by Snell's law when the sound line is at Initial grazing angle within the layer and the initial sound speed , given time is a constant, is the distance element.

7. The method for intelligent measurement of underwater terrain and accurate calculation of backfill volume for power transmission and transformation projects according to claim 1 is characterized in that: In S4, the Delaunay triangulation algorithm is used to optimize the triangulated network topology of the point cloud data, and a high-precision underwater digital elevation model (DEM) is generated by a contour extraction technique based on region growing, including: S4.

1. Delaunay triangulation optimization: Perform Delaunay triangulation on the multi-beam Ping point cloud data, construct triangular facets by maximizing the minimum internal angle criterion, and ensure the geometric uniformity and topological stability of the triangulation network. Verify the legitimacy of the triangles using the empty circumscribed circle criterion, remove triangles that do not meet the Delaunay conditions, and optimize the triangulation network structure. S4.

2. Contour extraction based on region growing: The Ping point cloud data collected by multibeam sonar is spatially gridded, and distance and density thresholds are set. Using the seed point as the core, adjacent points that meet the threshold conditions are grouped into the same contour line using Euclidean distance clustering and region growing algorithms to avoid contour line breakage or distortion. For contour lines with "branching" problems, a breakpoint detection method based on curvature analysis is used to split complex contours into multiple simple sub-contours to ensure a "one-to-one" correspondence. S4.

3. Contour patch filling: For the contour lines of adjacent pings, a corresponding point matching algorithm based on the minimum spanning tree (MST) is used to establish topological connections between the contour lines. Dynamic programming is used to optimize the generation path of triangular patches to ensure smooth transitions between adjacent contour lines. The optimized triangulated network model is then integrated with the measured microtopography data to generate a centimeter-level underwater digital elevation model (DEM). S4.4, DEM post-processing: Use the Laplacian smoothing algorithm to eliminate local noise in the triangulation network and improve the smoothness of the DEM surface; use Kriging interpolation to fill in data gaps and ensure the integrity and continuity of the DEM model.

8. The method for intelligent measurement of underwater terrain and accurate calculation of backfill volume for power transmission and transformation projects according to claim 5 is characterized in that: In S5, the backfill volume is calculated by combining the grid method and the DEM model through the following steps: S5.

1. Divide the DEM model into uniform grid cells at a preset resolution, with each cell corresponding to a planar area of ​​actual water; S5.

2. Extract the elevation data of each grid cell and calculate the height difference between the cell and the cut and fill height based on the designed backfill elevation. S5.3, by bidirectionally splitting the triangular pyramid average value algorithm, according to the formula Calculate the backfill volume of each triangular pyramid element, where is the side length of the square, Fill and cut height difference for cell vertices; S5.

4. Accumulate the backfill volume of all units, output the total backfill volume, and compare and verify it with the actual construction value.

9. A system for intelligent measurement of underwater terrain and precise calculation of backfill volume for power transmission and transformation projects, applied to the method for intelligent measurement of underwater terrain and precise calculation of backfill volume for power transmission and transformation projects as claimed in any one of claims 1 to 8, characterized in that: The system comprises: The data acquisition module is used to use an unmanned vessel equipped with a multi-beam bathymetric system, integrating a surface sound velocity meter, attitude indicator, and positioning system to collect high-precision three-dimensional data of underwater terrain in rivers and ponds, and obtain information on water depth, terrain, and obstacles; The path planning module is used to smooth the planned path based on the improved ant colony algorithm and B-spline interpolation method, set the safe steering angle and minimum turning path, and achieve dynamic obstacle avoidance and trajectory optimization; The data processing module is used to intelligently process the collected underwater data through the layered ray tracking algorithm, CUBE filtering algorithm, and attitude deviation correction algorithm to eliminate the influence of sound velocity error, environmental noise interference, and bathymetric deviation; The 3D modeling module uses the Delaunay triangulation algorithm to optimize the triangulated network topology of point cloud data and generates a high-precision underwater digital elevation model (DEM) through contour extraction technology based on region growing; The engineering quantity calculation module is used to divide the calculation units using the grid method based on the DEM model and calculate the backfill quantity.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a program, and the program instructs the device to execute the method for intelligent measurement of underwater terrain and precise calculation of backfill volume of rivers and ponds in power transmission and transformation projects as described in any one of claims 1 to 8.

Images