A beidou-based intelligent planning of sand mining area terrain and unmanned aerial vehicle operation navigation method

Abstract

The embodiment of the application discloses a sand mining area terrain intelligent planning and unmanned aerial vehicle operation navigation method based on Beidou. The technical scheme provided by the embodiment of the application generates a mining event point by real-time acquisition of the ship position and mining data reported by a sand mining ship, performs settlement prediction based on a riverbed terrain settlement influence function and generates a hotspot area, and then dynamically re-plans an unmanned aerial vehicle cruising route for execution of review, and finally compares and identifies an abnormal mining area by comparing real-time terrain with an initial digital twin model and outputs a report, realizing deep linkage of real-time operation data of the sand mining ship and unmanned aerial vehicle operation navigation, improving real-time response of terrain change in the sand mining area and accuracy of unmanned aerial vehicle operation, enabling the unmanned aerial vehicle to actively predict and preferentially review a high-risk area, significantly improving the intelligent level of sand mining supervision and the acquisition efficiency of supervision basis, and meeting the needs of fine and dynamic management and control of the sand mining area.

Description

Technical Field

[0001] This application relates to the field of sand mining supervision technology, and in particular to a method for intelligent terrain planning and UAV operation navigation in sand mining areas based on Beidou. Background Technology

[0002] Currently, terrain management and drone operations in sand mining areas primarily rely on BeiDou high-precision positioning, drone aerial surveying, and shipborne depth sounding technologies. Specifically, drones equipped with lidar or oblique photography cameras, combined with BeiDou RTK, conduct aerial surveys to acquire topographic data of the water surface and shallow water areas within the sand mining area. Simultaneously, shipborne single-beam or multi-beam depth sounders collect underwater topographic point clouds, which are then converted and registered to construct a 3D digital model of the sand mining area. Based on this, a fixed flight path planning method is employed, with parameters such as drone altitude, speed, and overlap preset at the ground station, enabling the drone to fly autonomously along a predetermined trajectory, completing periodic mapping or inspections of the sand mining area. Furthermore, sand dredging vessels are equipped with BeiDou terminals for real-time reporting of vessel location and dredging data, enabling remote monitoring of sand mining activities.

[0003] However, in existing drone operation navigation solutions, the drone's flight path planning and the real-time operational data of the sand dredger are independent of each other, lacking an effective linkage mechanism. Because sand dredgers constantly change the riverbed topography during operation, while drones usually execute tasks according to pre-set static flight paths, drones often can only collect evidence after violations have occurred, thus affecting the real-time nature and effectiveness of sand mining area supervision, resulting in relatively poor drone operation results. Summary of the Invention

[0004] This application provides a BeiDou-based method for intelligent terrain planning and UAV operation navigation in sand mining areas. It enables deep integration of real-time operation data of sand dredgers and UAV operation navigation, solving the technical problem that UAV route planning and real-time data of sand dredgers are independent of each other and lack a linkage mechanism, resulting in only post-event evidence collection after violations occur and poor real-time supervision.

[0005] In a first aspect, embodiments of this application provide a BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas, including: Real-time acquisition of vessel location data and mining data reported by target sand dredgers within the sand mining area; and generation of mining event points with timestamps and spatial coordinates based on vessel location data and mining data. Based on the mining event points and the preset riverbed topography settlement influence function, the settlement of the surrounding riverbed topography area is predicted to obtain the predicted settlement distribution field of the sand mining area. Based on the predicted settlement distribution field, the preset sand mining permit boundary and settlement gradient threshold, one or more hot spots are generated. Hot spots include ultra-deep areas to be reviewed, boundary risk areas and / or abnormal settlement areas. The drone's cruise route is dynamically replanned based on the location of the hotspot area. The drone is then controlled to fly to the hotspot area in sequence to perform verification work, thereby obtaining real-time terrain data of the hotspot area. The real-time terrain data is compared with the initial digital twin model of the sand mining area. Based on the comparison results, abnormal mining areas are identified from the hot spots, and an abnormal mining report is output based on the abnormal mining areas.

[0006] Furthermore, based on the mining event points and a pre-defined riverbed topographic settlement influence function, settlement prediction is performed for the surrounding riverbed topographic area to obtain the predicted settlement distribution field of the sand mining area, including: The sand mining area is divided into regular grids, and each grid independently stores the expected settlement. For a single mining event point, the instantaneous settlement contribution of the mining event point to the surrounding grid is calculated using the riverbed topographic settlement influence function; Within a preset time window, the instantaneous settlement contribution values ​​of all mining event points to the same grid are linearly superimposed to obtain the cumulative expected settlement of the corresponding grid. By iterating through the cumulative expected settlement of all grids, the predicted settlement distribution field of the sand mining area is obtained.

[0007] Furthermore, the influence function of riverbed topographic settlement is expressed as:

[0008] in, This represents the instantaneous settlement contribution value. Represents the normalization coefficient. This indicates the excavation volume corresponding to the excavation event point. This represents the set Gaussian kernel standard deviation. This represents the Euclidean distance from the center of the corresponding grid to the mining event point.

[0009] Furthermore, based on the predicted settlement distribution field, the preset sand mining permit boundary, and the settlement gradient threshold, one or more hotspot areas are generated, including: Query the cumulative expected settlement for each grid in the predicted settlement distribution field, as well as the settlement gradient between the corresponding grid and its adjacent grids; Grids with accumulated expected settlement exceeding a preset threshold are marked as abnormal settlement grids; continuous grid areas with settlement gradients exceeding the settlement gradient threshold are marked as potentially ultra-deep grids; and grids with accumulated expected settlement reaching the set index and intersecting with the sand mining permit boundary are marked as boundary crossing risk grids. Adjacent grids with the same labeling type in abnormal settlement grids, ultra-deep grids, and out-of-bounds risk grids are merged into polygonal regions, which are then used as hotspot regions of the corresponding labeling type.

[0010] Furthermore, the process of building the initial digital twin model includes: Point cloud and image data of the sand mining area and the designated shallow water area were acquired by drones. The point cloud and image data were then used to perform three-dimensional reconstruction to generate an initial digital elevation model.

[0011] Furthermore, the drone's cruise route is dynamically replanned based on the location of hotspot areas, including: During the process of the drone performing sand mining monitoring and inspection according to the initial cruise route, the drone's cruise route is dynamically replanned using the drone's current position as the starting point and the hot spot area as the necessary target point. The shortest path planning algorithm is set to dynamically replan the drone's cruise route.

[0012] Furthermore, the real-time terrain data is compared with the initial digital twin model of the sand mining area, including: Generate a local digital elevation model based on real-time terrain data; The local digital elevation model is compared with the digital elevation of the corresponding area in the initial digital twin model by grid-by-grid difference, and the corresponding elevation change is output as the comparison result.

[0013] In a second aspect, embodiments of this application provide a BeiDou-based intelligent terrain planning and UAV operation navigation system for sand mining areas, comprising: The data acquisition module is used to acquire in real time the vessel location data and mining data reported by the target sand dredging vessel in the sand mining area, and generate mining event points with timestamps and spatial coordinates based on the vessel location data and mining data. The settlement prediction module is used to predict the settlement of the surrounding riverbed topography area based on the mining event point and the preset riverbed topography settlement influence function, to obtain the predicted settlement distribution field of the sand mining area. Based on the predicted settlement distribution field, the preset sand mining permit boundary and settlement gradient threshold, one or more hot spots are generated. Hot spots include ultra-deep areas to be reviewed, boundary risk areas and / or abnormal settlement areas. The verification module is used to dynamically replan the drone's cruise route based on the location of the hotspot area. Based on the replanned cruise route, the drone is controlled to fly to the hotspot area in sequence to perform verification operations and obtain real-time terrain data of the hotspot area. The report output module is used to compare real-time terrain data with the initial digital twin model of the sand mining area, identify abnormal mining areas from the hot spots based on the comparison results, and output an abnormal mining report based on the abnormal mining areas.

[0014] In a third aspect, embodiments of this application provide an electronic device, including: Memory and one or more processors; The memory is used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas as described in the first aspect.

[0015] In a fourth aspect, embodiments of this application provide a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas as described in the first aspect.

[0016] This application embodiment acquires real-time vessel location data and mining data reported by target sand dredgers within the sand mining area. Based on the vessel location data and mining data, mining event points with timestamps and spatial coordinates are generated. Based on the mining event points and a preset riverbed topographic settlement influence function, settlement prediction is performed on the surrounding riverbed topographic area to obtain the predicted settlement distribution field of the sand mining area. According to the predicted settlement distribution field, the preset sand mining permit boundary, and the settlement gradient threshold, one or more hotspot areas are generated. Hotspot areas include ultra-deep areas to be reviewed, boundary risk areas, and / or abnormal settlement areas. The drone's cruise route is dynamically replanned according to the location of the hotspot areas. Based on the replanned cruise route, the drone is controlled to fly to the hotspot areas sequentially to perform review operations, obtaining real-time topographic data of the hotspot areas. The real-time topographic data is compared with the initial digital twin model of the sand mining area. Based on the comparison results, abnormal mining areas are identified from the hotspot areas, and an abnormal mining report is output based on the abnormal mining areas. By employing the aforementioned technical means, mining event points are generated by acquiring real-time vessel positions and mining data reported by sand dredgers. Settlement prediction is performed based on the riverbed topography settlement influence function, and hotspot areas are generated. Then, the drone patrol route is dynamically replanned for verification. Finally, the real-time terrain is compared with the initial digital twin model to identify abnormal mining areas and output a report. This achieves deep linkage between real-time sand dredger operation data and drone operation navigation, improving the real-time response to terrain changes in the sand mining area and the accuracy of drone operations. This enables drones to proactively predict and prioritize the verification of high-risk areas, significantly improving the intelligence level of sand mining supervision and the efficiency of obtaining regulatory evidence, and meeting the needs of refined and dynamic management of sand mining areas. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou, provided in an embodiment of this application. Figure 2 This is a flowchart illustrating the determination of the settlement distribution field in an embodiment of this application; Figure 3 This is a flowchart illustrating the determination of hotspot areas in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of a BeiDou-based intelligent terrain planning and UAV operation navigation system for sand mining areas, provided in an embodiment of this application. Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, specific embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely for explaining this application and not for limiting it. It should also be noted that, for ease of description, only the parts relevant to this application are shown in the drawings, not all of them. Before discussing exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe operations (or steps) as sequential processes, many of these operations can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but may also have additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0020] Example: Figure 1A flowchart of a BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas is provided in this embodiment. This BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas can be executed by a BeiDou-based intelligent terrain planning and UAV operation navigation device. This BeiDou-based intelligent terrain planning and UAV operation navigation device can be implemented through software and / or hardware. This device can consist of two or more physical entities, or it can consist of a single physical entity. Generally, this BeiDou-based intelligent terrain planning and UAV operation navigation device can be a processing device such as a sand mining area management server.

[0021] The following description uses a sand mining area management server as the main entity for implementing a BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas. (Refer to...) Figure 1 The BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas specifically includes: S110. Real-time acquisition of vessel location data and mining data reported by target sand dredgers within the sand mining area, and generation of mining event points with timestamps and spatial coordinates based on vessel location data and mining data.

[0022] In the process of terrain management in sand mining areas, this application predicts changes in the sand mining terrain and controls drones to conduct real-time patrol operations based on these changes, thereby achieving refined and dynamic sand mining terrain management.

[0023] Specifically, during the sand dredging terrain management process, an integrated intelligent terminal is installed on the sand dredging vessel to collect vessel position data and dredging data. The integrated intelligent terminal integrates a BeiDou RTK positioning module, attitude sensor, water depth sensor, and robotic arm encoder. The BeiDou RTK module receives satellite differential signals, calculates the real-time latitude, longitude, and ellipsoidal height of the vessel's antenna point, and obtains the precise three-dimensional position of the vessel's center of gravity through coordinate transformation; this position is defined as the vessel's position data. On the other hand, the robotic arm encoder records the moment the sand dredging equipment contacts the riverbed, thus counting the number of sand dredging operations. The dredging volume can be determined based on the number of dredging operations. The calculation of dredging volume is divided into two types according to the sand dredging operation mode: for grab bucket sand dredging, the dredging volume equals the caliber bucket capacity multiplied by the number of dredging operations; for cutter suction sand dredging, the dredging volume equals the dredging thickness multiplied by the area covered by a single dredging operation (this area can be determined by the lateral swing amplitude and longitudinal forward distance of the dredging head). The dredging volume can be used as dredging data for subsequent riverbed topographic settlement calculations. The integrated intelligent terminal reports the collected vessel location data and mining data to the sand mining area management server equipment at the back end of the system, thereby realizing the acquisition of the vessel location data and mining data of the target sand dredger in the sand mining area.

[0024] Optionally, the integrated terminal equipment of the target sand dredging vessel can also combine the known vertical distance from the ship's antenna point to the water surface and instantaneous tide data to calculate the real-time water surface elevation at the moment of operation. Shipborne depth sensors (such as multibeam echo sounders) emit sound waves towards the riverbed, calculating the vertical water depth from the ship to the riverbed based on the round-trip time of the sound waves and the speed of sound in water. Simultaneously, attitude sensors record the ship's roll, pitch, and heading angles to correct for depth measurement deviations caused by ship rolling. Subtracting the water depth value corrected for attitude and draft from the water surface elevation yields the riverbed elevation at the sand dredging point at that moment. Further, the initial riverbed elevation of the sand dredging point is recorded before the sand dredging operation. Then, during the sand dredging operation, the moment the sand dredging equipment contacts the riverbed is recorded by the robotic arm encoder, and the terminal simultaneously measures the real-time riverbed elevation at that moment. The difference between the two elevations represents the excavation thickness at that sand dredging point. The instantaneous water depth, riverbed elevation after attitude and tide level correction, and the thickness of this excavation obtained from the above calculations can also be reported to the server as excavation data for the generation of subsequent excavation reports.

[0025] Upon receiving the vessel location data and mining data, the server performs spatiotemporal alignment locally to ensure that each mining data point corresponds to a unique vessel location time. Then, based on the vessel location data and mining data, it generates mining event points with timestamps and spatial coordinates. A mining event point treats each effective mining action of a sand dredger as a discrete spatial point, simultaneously recording the time of the mining, precise geographic coordinates (i.e., vessel location data or real-time location data of the sand dredger calculated from its relative position), and the mining volume at that point (obtained from the mining data). Based on this data, using BeiDou time as a unified time reference, each mining action is transformed into a mining event point with a timestamp and three-dimensional spatial coordinates (X, Y, Z), and stored in a rolling time window (e.g., the past hour) of event points for subsequent settlement prediction.

[0026] S120. Based on the mining event points and the preset riverbed topographic settlement influence function, the settlement of the surrounding riverbed topographic area is predicted to obtain the predicted settlement distribution field of the sand mining area. Based on the predicted settlement distribution field, the preset sand mining permit boundary and settlement gradient threshold, one or more hot spots are generated. The hot spots include ultra-deep areas to be reviewed, boundary risk areas and / or abnormal settlement areas.

[0027] Furthermore, based on the aforementioned mining event points, this application first predicts riverbed topographic settlement. The riverbed topographic settlement influence function refers to a mathematical model describing the contribution of a single mining event point to the settlement of the riverbed at different distances around it, such as a single-point disturbance Gaussian diffusion model. The predicted settlement distribution field refers to a two-dimensional continuous field formed by dividing the sand mining area into regular grids and independently calculating the expected settlement accumulated under the combined effect of all mining event points within the current time window for each grid. The sand mining permit boundary refers to the pre-defined polygonal boundary of the area where sand mining activities are permitted. The settlement gradient threshold is the maximum allowable value of the rate of change of expected settlement between adjacent grids; exceeding this value indicates a risk of slope instability.

[0028] Optionally, refer to Figure 2 Based on the mining event points and a preset riverbed topographic settlement influence function, settlement prediction is performed on the surrounding riverbed topographic area to obtain the predicted settlement distribution field of the sand mining area, including: S1201. Divide the sand mining area into regular grids, with each grid independently storing the expected settlement amount; S1202. For a single mining event point, use the riverbed topographic settlement influence function to calculate the instantaneous settlement contribution of the mining event point to the surrounding grids. S1203. Within a preset time window, the instantaneous settlement contribution values ​​of all mining event points to the same grid are linearly superimposed to obtain the cumulative expected settlement of the corresponding grid. S1204. Traverse all grids to obtain the cumulative expected settlement, and obtain the predicted settlement distribution field of the sand mining area.

[0029] In settlement prediction calculations, the sand mining area is discretized into a regular grid on a horizontal plane. Each grid cell has a uniform size (e.g., 1 meter × 1 meter), and each cell independently stores a variable representing the cumulative expected settlement amount for that cell, with an initial value of zero. This grid partitioning allows subsequent calculations to be performed in discrete space, facilitating parallel processing and rapid updates. Upon receiving a new mining event point, the server invokes a preset riverbed topographic settlement influence function. This function employs a single-point perturbation Gaussian diffusion model, and the riverbed topographic settlement influence function is expressed as:

[0030] in, This represents the instantaneous settlement contribution value. Represents the normalization coefficient. This indicates the excavation volume corresponding to the excavation event point. This represents the set Gaussian kernel standard deviation, which is related to the excavation volume and substrate type. This represents the Euclidean distance from the center of the corresponding grid to the mining event point.

[0031] For a single mining event point, the server calculates the settlement contribution of the grid within a radius of 3σ centered on that point (set according to actual calculation requirements). Outside this range, the settlement contribution has decayed to a negligible level. For each grid within this range, the instantaneous settlement contribution of the event point to that grid is calculated by substituting the Euclidean distance r from the grid center to the mining event point into the above function, and this value is added to the original cumulative expected settlement of the grid.

[0032] Since sand mining operations are continuous, this application maintains a preset time window (e.g., the past hour), which includes all mining event points from the current moment back to the start of the window. Whenever a new event point arrives and its contribution value is calculated, the system simultaneously checks if there are any expired event points (i.e., old event points that have exceeded the time window). If so, the instantaneous contribution value of that old event point to each grid is subtracted from the cumulative amount, thus ensuring that the cumulative expected settlement only reflects the mining activities within the window. This sliding window mechanism allows the predicted distribution field to respond in real time to the latest sand mining activities while eliminating the outdated effects of historical events.

[0033] After completing the above calculations for all mining event points, the entire sand mining area is traversed through all grids, and the cumulative expected settlement of each grid is read, thus forming a two-dimensional predicted settlement distribution field. The value of each grid in this distribution field represents the expected riverbed settlement depth at that location under the combined effect of all sand mining activities within the current window, thereby providing a quantitative basis for the generation of subsequent hotspot areas.

[0034] By using regular grid division and linear superposition of Gaussian diffusion functions, combined with a sliding time window mechanism, the cumulative settlement impact of sand mining activities on the surrounding riverbed can be calculated efficiently and in real time. This allows the prediction results to reflect both spatial diffusion patterns and temporal cumulative effects, significantly improving the accuracy and dynamic response capability of settlement prediction.

[0035] Optionally, this application may also employ a riverbed sediment transport equation based on the finite difference method, where each mining event point is considered as a source term. The settlement distribution field is obtained by solving a predefined two-dimensional diffusion empirical equation and then iteratively updated, thereby achieving the transformation from mining event points to the predicted settlement distribution field. This application does not impose fixed restrictions on the specific calculation of the predicted settlement distribution field, and will not elaborate further here.

[0036] In one embodiment, after obtaining the predicted settlement distribution field based on the real-time reported data of dredging events from the sand dredger, a Long Short-Term Memory (LSTM) network model is further introduced to perform deep learning prediction of the settlement time series. The server collects historical data of dredging events from the sand mining area over several consecutive days, using the historical cumulative settlement sequence of each grid as the input feature, and simultaneously introducing auxiliary features such as water flow velocity, bottom sediment particle size, and sand mining intensity to construct a multi-layer LSTM network. The network structure includes an input layer (with a time window length of 24 hours), two hidden layers (128 neurons each), and an output layer (predicting the settlement amount for the next 2 hours). In the offline stage, the LSTM network is trained in a supervised manner using historical sand mining data, with the root mean square error (RMSE) loss function. In actual operation, the cumulative settlement sequence of each grid within the current time window is input into the trained LSTM model in real time, outputting the settlement change trend over a period of time, and then weighted and fused with the result of superposition of the riverbed topography settlement influence function (e.g., assigning 0.6 weight to the LSTM prediction result and 0.4 weight to the riverbed topography settlement influence function) to obtain a more accurate predicted settlement distribution field, thereby improving the settlement prediction accuracy under complex sand mining conditions.

[0037] Furthermore, referring to Figure 3 Based on the predicted settlement distribution field, the preset sand mining permit boundary, and the settlement gradient threshold, one or more hotspot areas are generated, including: S1205, Query the cumulative expected settlement of each grid in the predicted settlement distribution field, as well as the settlement gradient between the corresponding grid and its adjacent grids; S1206. Mark the grids whose cumulative expected settlement exceeds the preset threshold as abnormal settlement grids, mark the continuous grid areas whose settlement gradient exceeds the settlement gradient threshold as potential ultra-deep grids, and mark the grids whose cumulative expected settlement reaches the set index and intersects with the sand mining permit boundary as boundary crossing risk grids. S1207. Merge adjacent grids with the same labeling type in abnormal settlement grids, ultra-deep grids, and out-of-bounds risk grids into polygonal regions, which are then used as hotspot regions of the corresponding labeling type.

[0038] Furthermore, by traversing each grid in the predicted settlement distribution field, the cumulative expected settlement of each grid is read, and the settlement gradient between that grid and its neighboring grids is calculated, i.e., the rate of change of settlement per unit distance. This gradient reflects the local steepness of the riverbed topography. At the same time, a preset sand mining permit boundary is loaded. This boundary is a polygonal area delineated for permitted sand mining activities and is stored in vector form.

[0039] For each grid, three types of judgments are performed: First, if the cumulative expected settlement of the grid exceeds a preset abnormal settlement threshold (e.g., 0.2 meters), the grid is marked as an abnormal settlement grid. Such grids indicate that significant settlement has occurred or is about to occur at this location, and there may be a risk of over-excavation. Second, if the settlement gradient between the grid and any of its adjacent grids exceeds a preset settlement gradient threshold (e.g., 0.3 meters / meter), and these consecutive adjacent grids together form a connected region, all grids within this connected region are marked as potentially ultra-deep grids. This marking reflects drastic local topographic changes and may indicate slope instability of the mining pit or steep slopes formed by over-excavation. Third, if the cumulative expected settlement of the grid is greater than zero (i.e., there is a measurable settlement contribution) and the geographical coordinates of the grid are outside the permitted sand mining boundary, or the cumulative expected settlement range extends outward by a buffer distance and intersects with the permitted boundary, the grid is marked as a boundary violation risk grid. This marking is used to identify that sand mining activities may have exceeded the boundary of the legal area.

[0040] After completing the full mesh labeling, a connected component analysis method is used to aggregate adjacent meshes with the same labeling type into connected polygonal regions for three categories of labeled meshes: anomalous settlement meshes, potentially ultra-deep meshes, and boundary risk meshes. For each generated polygonal region, its geometric characteristics are further calculated, such as the bounding rectangle, center coordinates, area, and the average or maximum value of the cumulative expected settlement of the internal meshes. Finally, these three types of polygonal regions are output as anomalous settlement regions, ultra-deep regions, and boundary risk regions, respectively, collectively referred to as hotspot regions, and sorted according to the severity of risk (such as maximum cumulative settlement or average settlement gradient) for subsequent UAV dynamic replanning.

[0041] By comprehensively utilizing the triple constraints of settlement threshold, settlement gradient threshold, and sand mining permit boundary, different types of terrain anomaly risks can be accurately identified. Discrete grid markers are transformed into polygonal hotspot areas with clear spatial ranges, providing UAVs with clear and actionable target locations, thereby improving the accuracy and targeting of anomaly area identification.

[0042] Optionally, this application can also employ an unsupervised classification method based on clustering algorithms for hotspot region classification. The cumulative settlement, settlement gradient, and distance to the permissible boundary of each grid in the predicted settlement distribution field are directly used as three-dimensional feature vectors and input into a pre-defined clustering model. Grids with similar features are clustered into one class, and then, based on the relationship between the cluster center and the judgment threshold, each cluster is classified into the corresponding hotspot region type. This application does not impose fixed restrictions on the specific method for determining hotspot regions, and will not elaborate further here.

[0043] S130. Based on the location of the hotspot area, dynamically replan the drone's cruise route, and control the drone to fly to the hotspot area in sequence to perform verification work, thereby obtaining real-time terrain data of the hotspot area.

[0044] Based on the aforementioned hotspot areas, the server can replan the training flight paths of the UAVs, thereby enabling verification operations in these hotspot areas. Dynamic replanning refers to the UAV, during mission execution, regenerating its flight path online from its current location, sequentially covering all hotspot areas, based on a real-time pushed list of hotspot areas. Real-time terrain data refers to high-density point cloud and imagery data collected by the UAV over the hotspot areas. This data carries BeiDou RTK real-time differential positioning information, reflecting the current surface morphology of the riverbed.

[0045] The server sorts the generated list of hotspot areas according to comprehensive risk scores (e.g., areas with the fastest settlement rate or the highest risk of crossing boundaries have the highest priority), then reads the current precise location provided by the UAV's onboard navigation system as the starting point of the path, takes the center point of each polygonal area in the list of hotspot areas as the necessary node, and takes minimizing the total flight mileage as the optimization objective to perform path search in three-dimensional space, thereby realizing the dynamic replanning of the UAV's cruise route.

[0046] Optionally, the drone's cruise route can be dynamically replanned based on the location of the hotspot area, including: During the process of the drone performing sand mining monitoring and inspection according to the initial cruise route, the drone's cruise route is dynamically replanned using the drone's current position as the starting point and the hot spot area as the necessary target point. The shortest path planning algorithm is set to dynamically replan the drone's cruise route.

[0047] During routine monitoring of the sand mining area, the UAV initially performs periodic inspection tasks according to the static cruise route preset by the server. When one or more hotspot areas are generated, the server uses the sequence of hotspot area locations as the necessary target points to be traversed sequentially, and calls a pre-defined shortest path planning algorithm (such as the improved A algorithm) to search for a path in three-dimensional space. During the search, it considers prohibited mining area polygons, fixed obstacles (such as shoreline structures), and dynamic obstacles perceived in real-time by airborne millimeter-wave radar (such as moving sand dredgers). With the shortest flight distance as the optimization objective, it generates a smooth three-dimensional trajectory connecting the starting point to all necessary target points and visiting them sequentially, while avoiding all obstacles. The generated new trajectory replaces the original initial cruise route and is synchronized to the UAV. The UAV then flies to each hotspot area sequentially according to the replanned route to perform key verification operations. Through route replanning, the UAV can seamlessly switch from routine inspection to target-oriented verification mode, responding in real-time to dynamic changes in sand mining and shortening the response time from anomaly prediction to on-site verification.

[0048] Upon reaching the airspace above each hotspot area, the drone reduces its flight speed and adjusts to a preset scanning altitude (e.g., 15 to 20 meters above the riverbed or water surface). It then activates the lidar to collect high-density point clouds and simultaneously engages the oblique photography camera for multi-angle imaging. During the data collection process, the onboard edge computing unit performs preliminary filtering and noise reduction on the raw data and reports it to the server in real time, forming real-time terrain data for that hotspot area.

[0049] S140. Compare the real-time terrain data with the initial digital twin model of the sand mining area, identify abnormal mining areas from the hot spots based on the comparison results, and output an abnormal mining report based on the abnormal mining areas.

[0050] Based on real-time topographic data collected in practice, and compared with the initial digital twin model of the sand mining area, abnormal mining areas can be accurately identified. The initial digital twin model refers to a three-dimensional digital elevation model of the sand mining area constructed before the start of sand mining operations using first-ever full-coverage aerial surveys conducted by UAVs and integrated with shipborne depth sounding data. It represents the original riverbed topographic benchmark before sand mining disturbance. Abnormal mining areas refer to continuous spatial areas that exceed the legal mining scope or permitting excavation depth, identified by comparing real-time topographic data after sand mining with the initial digital twin model. These include ultra-deep areas (exceeding the permitted depth), boundary-crossing areas (horizontal coordinates exceeding the permitted boundary), and abnormal settlement areas (settlement gradient exceeding the safety value).

[0051] Specifically, the process of building the initial digital twin model includes: Point cloud and image data of the sand mining area and the designated shallow water area were acquired by drones. The point cloud and image data were then used to perform three-dimensional reconstruction to generate an initial digital elevation model.

[0052] Before sand mining operations commence, drones equipped with lidar and oblique photography cameras, combined with BeiDou RTK positioning, conduct full-coverage aerial surveys of the sand mining area, acquiring high-density point cloud and high-resolution image data of the water surface and designated shallow water areas. The lidar directly collects the three-dimensional coordinates of the ground surface by emitting laser pulses and receiving echoes, forming a discrete point cloud. The oblique photography camera simultaneously captures images from multiple angles, obtaining lateral texture information of ground features. After receiving the data, the server registers and fuses the laser point cloud and the dense image point cloud based on BeiDou coordinates. The fused point cloud is then reconstructed using Poisson surface mapping to generate an initial digital elevation model with a regular grid. This model records the original topographic morphology of the riverbed and shoreline before sand mining, serving as a benchmark for subsequent subsidence prediction and anomaly comparison.

[0053] Furthermore, the real-time terrain data is compared with the initial digital twin model of the sand mining area, including: Generate a local digital elevation model based on real-time terrain data; The local digital elevation model is compared with the digital elevation of the corresponding area in the initial digital twin model by grid-by-grid difference, and the corresponding elevation change is output as the comparison result.

[0054] After the drone performs verification operations in the hotspot area, the point cloud data is unified to an independent coordinate system using precise exterior orientation elements provided by BeiDou RTK. This regularizes the discrete point cloud into a local digital elevation model with consistent grid spacing, representing the riverbed surface morphology of the hotspot area after sand mining. Subsequently, digital elevation data of the corresponding area that completely overlaps with the spatial range of this local digital elevation model is extracted from the initial digital twin model. A grid-by-grid difference operation is then performed between the two, i.e., the elevations at corresponding locations are subtracted to obtain the elevation change for each grid. A negative change indicates riverbed subsidence (due to mining or erosion), while a positive change indicates riverbed uplift (due to siltation or measurement error). This set of elevation changes serves as the comparison result, used for subsequent identification and determination of abnormal mining areas.

[0055] Based on the comparison results, if the elevation change of a grid is negative and its absolute value exceeds the permissible excavation depth threshold, and the grid is located within the boundary of the sand mining area, it is marked as an over-depth anomaly; if the elevation change of a grid is negative and the grid is located outside the boundary of the sand mining area, it is marked as an out-of-bounds anomaly; if the elevation change gradient of multiple consecutive grids exceeds the settlement gradient threshold, it is marked as abnormal settlement. All marked grids are merged into an abnormal mining region through connected components. Furthermore, the geometric center coordinates, outer envelope area, and excavation volume obtained by integrating the elevation change can be calculated for each abnormal mining region. Finally, the server generates an abnormal mining report based on the abnormal mining region. The abnormal mining report is a control basis document that includes quantitative information such as the spatial location, area, volume, over-excavation depth, or out-of-bounds length of the aforementioned abnormal region, and is overlaid on a 3D model in a visual form. This report can serve as a technical basis for on-site verification by the control department. By introducing a linkage mechanism between real-time dredging event points of sand dredgers for settlement prediction and dynamic replanning cruise of drones, proactive prediction and accurate verification of topographic changes in sand mining areas are achieved, improving the response speed of sand mining area management and the accuracy of anomaly identification.

[0056] On the other hand, this application can also introduce a time dimension to construct a dynamic monitoring model in three dimensions of space and time based on the existing digital twin model, thereby realizing the whole process of topographic evolution in sand mining areas and future early warning. Specifically, the server continuously receives real-time mining event points from sand dredgers and real-time topographic data after each review by drones, generating a series of time-series digital elevation models in one-hour time slices. A spatiotemporal database stores the elevation values ​​of each grid at different times, forming a topographic evolution curve for each spatial point. The server calculates two key indicators in the three-dimensional space and time dimensions. One is the cumulative impact index, which is the difference between the current elevation and the initial elevation divided by time, reflecting the intensity of sand mining; the other is the recovery potential index, which predicts the time required for future natural recovery based on historical siltation rates. When the cumulative impact index of a certain area exceeds the ecological threshold or the recovery potential index falls below the preset lower limit, the system automatically generates an early warning message; if both excessive depth, boundary crossing, and extremely low recovery potential occur simultaneously, a higher-level early warning is generated. The early warning information is overlaid on the three-dimensional digital twin model and displayed in the form of heat map and time slider. Users can drag the timeline to view the terrain status at any historical moment, and can also play evolution animation to intuitively show the damage of the sand mining process to the riverbed, thus providing visualized data support for ecological damage assessment, evidence collection and restoration decision-making in sand mining areas.

[0057] The above-mentioned process involves acquiring real-time vessel location data and mining data reported by target sand dredgers within the sand mining area, generating mining event points with timestamps and spatial coordinates based on the vessel location data and mining data, predicting the settlement of surrounding riverbed topography based on the mining event points and a preset riverbed topography settlement influence function, obtaining the predicted settlement distribution field of the sand mining area, and generating one or more hotspot areas based on the predicted settlement distribution field, preset sand mining permit boundaries, and settlement gradient thresholds. Hotspot areas include ultra-deep areas to be reviewed, boundary risk areas, and / or abnormal settlement areas. The drone's cruise route is dynamically replanned based on the location of the hotspot areas, and the drone is controlled to fly to the hotspot areas sequentially to perform review operations based on the replanned cruise route, obtaining real-time topographic data of the hotspot areas. The real-time topographic data is compared with the initial digital twin model of the sand mining area, and abnormal mining areas are identified from the hotspot areas based on the comparison results, and an abnormal mining report is output based on the abnormal mining areas. By employing the aforementioned technical means, the real-time vessel positions and mining data reported by sand dredgers are acquired to generate mining event points. Based on the riverbed topography subsidence influence function, subsidence prediction is performed and hotspot areas are generated. Subsequently, the drone patrol routes are dynamically replanned for verification. Finally, the real-time terrain is compared with the initial digital twin model to identify abnormal mining areas and output a report. This achieves deep linkage between real-time sand dredger operation data and drone operation navigation, improving the real-time response to terrain changes in the sand mining area and the accuracy of drone operations. This enables drones to proactively predict and prioritize the verification of high-risk areas, significantly improving the intelligence level of sand mining supervision and the efficiency of obtaining regulatory evidence, and meeting the needs of refined and dynamic management of sand mining areas.

[0058] Based on the above embodiments, Figure 4 This is a schematic diagram of a BeiDou-based intelligent terrain planning and UAV operation navigation system for sand mining areas, provided as an embodiment of this application. (Reference) Figure 4 The BeiDou-based intelligent terrain planning and UAV operation navigation system for sand mining areas provided in this embodiment specifically includes: Data acquisition module 21 is used to acquire in real time the ship position data and mining data reported by the target sand mining vessel in the sand mining area, and generate mining event points with timestamps and spatial coordinates based on the ship position data and mining data. Settlement prediction module 22 is used to predict the settlement of the surrounding riverbed topography area based on the mining event point and the preset riverbed topography settlement influence function, to obtain the predicted settlement distribution field of the sand mining area, and to generate one or more hot spots based on the predicted settlement distribution field, the preset sand mining permit boundary and settlement gradient threshold. The hot spots include ultra-deep areas to be reviewed, boundary risk areas and / or abnormal settlement areas. The verification module 23 is used to dynamically replan the UAV's cruise route based on the location of the hotspot area, and control the UAV to fly to the hotspot area in sequence to perform verification operations based on the replanned cruise route, so as to obtain the real-time terrain data of the hotspot area. The report output module 24 is used to compare real-time terrain data with the initial digital twin model of the sand mining area, identify abnormal mining areas from the hot spots based on the comparison results, and output an abnormal mining report based on the abnormal mining areas.

[0059] Specifically, based on the mining event points and a pre-defined riverbed topographic settlement influence function, settlement prediction is performed on the surrounding riverbed topographic area to obtain the predicted settlement distribution field of the sand mining area, including: The sand mining area is divided into regular grids, and each grid independently stores the expected settlement. For a single mining event point, the instantaneous settlement contribution of the mining event point to the surrounding grid is calculated using the riverbed topographic settlement influence function; Within a preset time window, the instantaneous settlement contribution values ​​of all mining event points to the same grid are linearly superimposed to obtain the cumulative expected settlement of the corresponding grid. By iterating through the cumulative expected settlement of all grids, the predicted settlement distribution field of the sand mining area is obtained.

[0060] Specifically, the riverbed topographic settlement influence function is expressed as:

[0061] in, This represents the instantaneous settlement contribution value. Represents the normalization coefficient. This indicates the excavation volume corresponding to the excavation event point. This represents the set Gaussian kernel standard deviation. This represents the Euclidean distance from the center of the corresponding grid to the mining event point.

[0062] Specifically, based on the predicted settlement distribution field, the preset sand mining permit boundary, and the settlement gradient threshold, one or more hotspot areas are generated, including: Query the cumulative expected settlement for each grid in the predicted settlement distribution field, as well as the settlement gradient between the corresponding grid and its adjacent grids; Grids with accumulated expected settlement exceeding a preset threshold are marked as abnormal settlement grids; continuous grid areas with settlement gradients exceeding the settlement gradient threshold are marked as potentially ultra-deep grids; and grids with accumulated expected settlement reaching the set index and intersecting with the sand mining permit boundary are marked as boundary crossing risk grids. Adjacent grids with the same labeling type in abnormal settlement grids, ultra-deep grids, and out-of-bounds risk grids are merged into polygonal regions, which are then used as hotspot regions of the corresponding labeling type.

[0063] Specifically, the process of building the initial digital twin model includes: Point cloud and image data of the sand mining area and the designated shallow water area were acquired by drones. The point cloud and image data were then used to perform three-dimensional reconstruction to generate an initial digital elevation model.

[0064] Specifically, the drone's cruise route is dynamically replanned based on the location of hotspot areas, including: During the process of the drone performing sand mining monitoring and inspection according to the initial cruise route, the drone's cruise route is dynamically replanned using the drone's current position as the starting point and the hot spot area as the necessary target point. The shortest path planning algorithm is set to dynamically replan the drone's cruise route.

[0065] Specifically, real-time terrain data is compared with the initial digital twin model of the sand mining area, including: Generate a local digital elevation model based on real-time terrain data; The local digital elevation model is compared with the digital elevation of the corresponding area in the initial digital twin model by grid-by-grid difference, and the corresponding elevation change is output as the comparison result.

[0066] The above-mentioned process involves acquiring real-time vessel location data and mining data reported by target sand dredgers within the sand mining area, generating mining event points with timestamps and spatial coordinates based on the vessel location data and mining data, predicting the settlement of surrounding riverbed topography based on the mining event points and a preset riverbed topography settlement influence function, obtaining the predicted settlement distribution field of the sand mining area, and generating one or more hotspot areas based on the predicted settlement distribution field, preset sand mining permit boundaries, and settlement gradient thresholds. Hotspot areas include ultra-deep areas to be reviewed, boundary risk areas, and / or abnormal settlement areas. The drone's cruise route is dynamically replanned based on the location of the hotspot areas, and the drone is controlled to fly to the hotspot areas sequentially to perform review operations based on the replanned cruise route, obtaining real-time topographic data of the hotspot areas. The real-time topographic data is compared with the initial digital twin model of the sand mining area, and abnormal mining areas are identified from the hotspot areas based on the comparison results, and an abnormal mining report is output based on the abnormal mining areas. By employing the aforementioned technical means, the real-time vessel positions and mining data reported by sand dredgers are acquired to generate mining event points. Based on the riverbed topography subsidence influence function, subsidence prediction is performed and hotspot areas are generated. Subsequently, the drone patrol routes are dynamically replanned for verification. Finally, the real-time terrain is compared with the initial digital twin model to identify abnormal mining areas and output a report. This achieves deep linkage between real-time sand dredger operation data and drone operation navigation, improving the real-time response to terrain changes in the sand mining area and the accuracy of drone operations. This enables drones to proactively predict and prioritize the verification of high-risk areas, significantly improving the intelligence level of sand mining supervision and the efficiency of obtaining regulatory evidence, and meeting the needs of refined and dynamic management of sand mining areas.

[0067] The BeiDou-based intelligent terrain planning and UAV operation navigation system for sand mining areas provided in this application embodiment can be used to execute the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas provided in the above embodiment, and has corresponding functions and beneficial effects.

[0068] This application provides an electronic device, which is described in reference to... Figure 5 The electronic device includes a processor 31, a memory 32, a communication module 33, an input device 34, and an output device 35. The electronic device may have one or more processors and one or more memories. The processor, memory, communication module, input device, and output device of the electronic device can be connected via a bus or other means.

[0069] Memory, as a computer-readable storage medium, can be used to store software programs, computer-executable programs, and modules, such as the program instructions / modules corresponding to the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas (e.g., the data acquisition module, settlement prediction module, verification module, and report output module in the BeiDou-based intelligent terrain planning and UAV operation navigation system for sand mining areas). Memory may primarily include a program storage area and a data storage area. The program storage area may store the operating system and at least one application program required for a function; the data storage area may store data created based on device usage. Furthermore, memory may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some instances, memory may further include memory remotely located relative to the processor, which can be connected to the device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0070] The communication module is used for data transmission.

[0071] The processor executes various functional applications and data processing of the device by running software programs, instructions, and modules stored in memory, thereby realizing the above-mentioned Beidou-based intelligent terrain planning and UAV operation navigation method for sand mining areas.

[0072] Input devices can be used to receive input numerical or character information, and to generate key signal inputs related to user settings and function control of the device. Output devices may include display devices such as displays.

[0073] The electronic device provided above can be used to execute the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas provided in the above embodiments, and has corresponding functions and beneficial effects.

[0074] This application embodiment also provides a storage medium containing computer-executable instructions. When executed by a computer processor, these computer-executable instructions are used to execute a BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas. This BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas includes: real-time acquisition of vessel position data and mining data reported by target sand mining vessels within the sand mining area; generating mining event points with timestamps and spatial coordinates based on the vessel position data and mining data; and predicting the settlement of the surrounding riverbed topography area based on the mining event points and a preset riverbed topography settlement influence function, thereby obtaining the settlement data for the sand mining area. The system predicts the settlement distribution field, and generates one or more hotspot areas based on the predicted settlement distribution field, preset sand mining permit boundaries, and settlement gradient thresholds. Hotspot areas include ultra-deep areas to be verified, boundary risk areas, and / or abnormal settlement areas. The system dynamically replans the drone's cruise route based on the location of the hotspot areas, and controls the drone to fly to the hotspot areas sequentially to perform verification operations, obtaining real-time terrain data of the hotspot areas. The real-time terrain data is compared with the initial digital twin model of the sand mining area, and abnormal mining areas are identified from the hotspot areas based on the comparison results. An abnormal mining report is then output based on the abnormal mining areas.

[0075] Storage medium – any type of memory device or storage device. The term “storage medium” is intended to include: mounting media, such as CD-ROM, floppy disk, or magnetic tape devices; computer system memory or random access memory, such as DRAM, DDR RAM, SRAM, EDO RAM, Rambus RAM, etc.; non-volatile memory, such as flash memory, magnetic media (e.g., hard disk or optical storage); registers or other similar types of memory elements, etc. Storage medium may also include other types of memory or combinations thereof. Furthermore, storage medium may reside in a first computer system in which the program is executed, or it may reside in a different second computer system connected to the first computer system via a network (such as the Internet). The second computer system can provide program instructions to the first computer for execution. The term “storage medium” can include two or more storage media residing in different locations (e.g., in different computer systems connected via a network). Storage medium may store program instructions (e.g., specifically implemented as a computer program) executable by one or more processors.

[0076] Of course, the computer-executable instructions provided in the embodiments of this application are not limited to the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas as described above, but can also execute related operations in the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas provided in any embodiment of this application.

[0077] The BeiDou-based intelligent terrain planning and UAV operation navigation system, storage medium, and electronic equipment provided in the above embodiments can execute the BeiDou-based intelligent terrain planning and UAV operation navigation method provided in any embodiment of this application. For technical details not described in detail in the above embodiments, please refer to the BeiDou-based intelligent terrain planning and UAV operation navigation method provided in any embodiment of this application.

[0078] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of this application. Therefore, although this application has been described in detail through the above embodiments, this application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of this application, the scope of which is determined by the scope of the claims.

Claims

1. A method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou, characterized in that, include: Real-time acquisition of vessel location data and mining data reported by target sand dredgers within the sand mining area; and generation of mining event points with timestamps and spatial coordinates based on the vessel location data and the mining data. Based on the mining event points and the preset riverbed topography subsidence influence function, the subsidence of the surrounding riverbed topography area is predicted to obtain the predicted subsidence distribution field of the sand mining area. According to the predicted subsidence distribution field, the preset sand mining permit boundary and the subsidence gradient threshold, one or more hot spots are generated. The hot spots include ultra-deep areas to be reviewed, boundary risk areas and / or abnormal subsidence areas. The drone's cruise route is dynamically replanned based on the location of the hotspot area, and the drone is controlled to fly to the hotspot area in sequence to perform verification work according to the replanned cruise route, so as to obtain the real-time terrain data of the hotspot area. The real-time terrain data is compared with the initial digital twin model of the sand mining area. Based on the comparison results, abnormal mining areas are identified from the hotspot areas, and an abnormal mining report is output based on the abnormal mining areas.

2. The method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou as described in claim 1, characterized in that, The settlement prediction of the surrounding riverbed topography area based on the mining event point and a preset riverbed topography settlement influence function is obtained, resulting in the predicted settlement distribution field of the sand mining area, including: The sand mining area is divided into regular grids, and each grid independently stores the expected settlement. For a single mining event point, the instantaneous settlement contribution of the mining event point to each surrounding grid is calculated using the riverbed topographic settlement influence function; Within a preset time window, the instantaneous settlement contribution values ​​of all the mining event points to the same grid are linearly superimposed to obtain the cumulative expected settlement of the corresponding grid. By iterating through all the grids and calculating the cumulative expected settlement, the predicted settlement distribution field of the sand mining area is obtained.

3. The method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou as described in claim 2, characterized in that, The riverbed topographic settlement influence function is expressed as follows: in, This represents the instantaneous settlement contribution value. Represents the normalization coefficient. This indicates the excavation volume corresponding to the excavation event point. This represents the set Gaussian kernel standard deviation. This represents the Euclidean distance from the center of the corresponding grid to the mining event point.

4. The method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou as described in claim 2, characterized in that, Based on the predicted settlement distribution field, the preset sand mining permit boundary, and the settlement gradient threshold, one or more hotspot areas are generated, including: Query the cumulative expected settlement for each grid in the predicted settlement distribution field, and the settlement gradient between the corresponding grid and its adjacent grids; The grids whose cumulative expected settlement exceeds a preset threshold are marked as abnormal settlement grids; the continuous grid regions whose settlement gradient exceeds the settlement gradient threshold are marked as potentially ultra-deep grids; and the grids whose cumulative expected settlement reaches a set index and intersects with the sand mining permit boundary are marked as boundary crossing risk grids. Adjacent grids with the same labeling type among the abnormal settlement grid, the ultra-deep grid, and the boundary risk grid are merged into polygonal regions, which are then designated as hotspot regions of the corresponding labeling type.

5. The method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou as described in claim 1, characterized in that, The construction process of the initial digital twin model includes: The initial digital elevation model is generated by using point cloud and image data of the sand mining area and the designated shallow water area acquired by UAVs to perform three-dimensional reconstruction of the point cloud and image data.

6. The method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou as described in claim 1, characterized in that, The dynamic replanning of the drone's cruise route based on the location of the hotspot area includes: During the sand mining monitoring and inspection process of the UAV following the initial cruise route, the cruise route of the UAV is dynamically replanned using the current position of the UAV as the starting point and the position of the hot spot area as the necessary target point.

7. The method for intelligent terrain planning and UAV operation navigation in sand mining areas based on BeiDou as described in claim 1, characterized in that, The step of comparing the real-time terrain data with the initial digital twin model of the sand mining area includes: A local digital elevation model is generated based on the real-time terrain data; The local digital elevation model is compared with the digital elevation of the corresponding region in the initial digital twin model by grid-by-grid difference, and the corresponding elevation change is output as the comparison result.

8. A BeiDou-based intelligent terrain planning and UAV operation navigation system for sand mining areas, characterized in that, include: The data acquisition module is used to acquire in real time the vessel location data and mining data reported by the target sand dredging vessel in the sand mining area, and generate mining event points with timestamps and spatial coordinates based on the vessel location data and the mining data. The settlement prediction module is used to predict the settlement of the surrounding riverbed topography area based on the mining event point and the preset riverbed topography settlement influence function, to obtain the predicted settlement distribution field of the sand mining area, and to generate one or more hot spots based on the predicted settlement distribution field, the preset sand mining permit boundary and settlement gradient threshold. The hot spots include ultra-deep areas to be reviewed, boundary risk areas and / or abnormal settlement areas. The verification module is used to dynamically replan the drone's cruise route based on the location of the hotspot area, and control the drone to fly to the hotspot area in sequence to perform verification operations according to the replanned cruise route, so as to obtain the real-time terrain data of the hotspot area. The report output module is used to compare the real-time terrain data with the initial digital twin model of the sand mining area, identify abnormal mining areas from the hotspot areas based on the comparison results, and output an abnormal mining report based on the abnormal mining areas.

9. An electronic device, characterized in that, include: Memory and one or more processors; The memory is used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas as described in any one of claims 1-7.

10. A storage medium containing computer-executable instructions, characterized in that, The computer-executable instructions, when executed by a computer processor, are used to perform the BeiDou-based intelligent terrain planning and UAV operation navigation method for sand mining areas as described in any one of claims 1-7.

Images