A multi-factor UAV flow measurement method and system

Abstract

The present invention relates to the technical field of remote sensing hydrological measurement, and in particular to a multi-element unmanned aerial vehicle flow measurement method and system. The method comprises the following steps: obtaining river data and unmanned aerial vehicle positioning data, and constructing a three-dimensional model of a river water body; respectively calculating the water body flow velocity data detected by radar and image through reflection radar signal and water body image analysis; fusing the two flow velocity data to obtain a more reliable final flow velocity result; inverting standard radar and image detection process data to optimize data acquisition; analyzing the standard height coordinates of flow measurement and the safe flight altitude, determining the standard flow measurement path and the safe flight path, constructing a closed-loop control model, and realizing the standard and safe flow measurement of unmanned aerial vehicles; transmitting the water body section model and flow velocity data to the ground in real time, and completing the unmanned aerial vehicle flow measurement task; the present invention can realize the comprehensive collection, high-precision fusion and intelligent control of multi-element flow measurement data, and greatly improve the flow measurement efficiency and data quality.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing hydrological measurement, and in particular to a multi-factor unmanned aerial vehicle flow measurement method and system. Background Art

[0002] In the early stage, multi-factor UAV flow measurement technology mainly used a single sensor (such as electromagnetic flowmeter) mounted on the UAV platform to achieve simple water flow velocity measurement. However, this method has limitations, such as a small measurement range, the need for manual control of the UAV, and low data collection efficiency. For example, the use of radar wave flow measurement technology is easily affected by weather conditions, such as rain, snow, and clouds, which will reduce the measurement accuracy; the use of image flow measurement data is extremely dependent on clear images of the water surface, which are greatly affected by sunlight and shadow factors. In the transition stage (mid-2000s), researchers began to try to integrate multiple sensors (such as acoustic Doppler flowmeters and water depth meters) on UAVs, which can simultaneously measure multiple hydrological parameters such as flow velocity and water depth. However, this method has complex sensor integration, difficult data processing, and measurement accuracy needs to be further improved. Entering the intelligent stage (since the 2010s), researchers use autonomous navigation UAV platforms to achieve automatic cruise measurement. At the same time, combined with remote sensing image analysis and hydrological models, the measurement accuracy and data analysis capabilities are improved. Although this method has made progress, it still has some defects. For example, the autonomy and adaptability of drones need to be further enhanced, the comprehensiveness and real-time measurement of hydrological parameters need to be improved, and the data processing and analysis capabilities need to be improved. Summary of the invention

[0003] Based on this, it is necessary to provide a multi-factor UAV flow measurement method and system to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a multi-factor UAV flow measurement method includes the following steps:

[0005] Step S1: obtaining a designated flow measurement river channel, a vertical water depth data set, an underwater terrain data set, and a real-time positioning data of a UAV; performing a water depth remote sensing stereo reconstruction on the vertical water depth data set and the underwater terrain data set based on the real-time positioning data of the UAV to obtain a river water body cross-section model;

[0006] Step S2: Acquire the reflected radar signal; calculate the water flow velocity of the reflected radar signal to obtain radar detected water flow velocity data;

[0007] Step S3: Acquire a river water body image set; annotate the river water body image set with key points to generate water body image key points; map the water body flow velocity of the river water body image set based on the water body image key points to generate image detection water body flow velocity data;

[0008] Step S4: performing water velocity measurement calculation by fusing the reflected radar signal and the river water image set according to the radar detected water velocity data and the image detected water velocity data, and generating fused measured water velocity data; performing standard measurement process inversion on the radar detected water velocity data and the image detected water velocity data according to the fused measured water velocity data, and obtaining standard radar detection process data and standard image detection process data;

[0009] Step S5: extracting the intersection height of the standard radar detection process data and the standard image detection process data to generate standard flow measurement height data; extracting the intersection coordinates of the standard radar detection process data and the standard image detection process data to obtain standard flow measurement coordinate data;

[0010] Step S6: Define the safe height for the fused measured water flow velocity data to obtain the safe flight height data; construct a closed-loop control model for the flow measurement standard height data, flow measurement standard coordinate data and safe flight height data, so as to achieve a safe UAV control task; transmit the river water cross-section diagram and the fused measured water flow velocity data to the ground station in real time, so as to achieve the UAV flow measurement task.

[0011] The present invention realizes the three-dimensional perception of river water bodies by combining multi-source information such as drone positioning data, water depth data, and underwater terrain data, and fully obtains relevant data on water flow velocity by using multiple measurement methods such as reflected radar signals and river water body images. Through the fusion calculation of radar detection and image detection water flow velocity data, more accurate and reliable fusion measurement water flow velocity data is generated, and the standard detection process data is inverted based on the fusion data, which improves the accuracy and reliability of the measurement results. The safe flight altitude is defined according to the fusion measurement data, and a closed-loop control model is constructed to ensure that the drone completes the flow measurement task within a safe range. The river water body cross-section diagram and flow velocity data are transmitted to the ground station in real time to provide timely and reliable hydrological information for decision support. The full process automation of drone flow measurement operations is realized by integrating hydrological monitoring, remote sensing measurement, and intelligent control technologies. The advantages of various technical means are fully utilized to realize the comprehensive collection, high-precision fusion, and intelligent control of flow measurement data, which greatly improves the flow measurement efficiency and data quality.

[0012] In this specification, a multi-factor UAV flow measurement system is also provided, which is used to execute the multi-factor UAV flow measurement method as described above, and the multi-factor UAV flow measurement system includes:

[0013] The cross-section construction module is used to obtain the designated flow measurement river channel, vertical water depth data set, underwater terrain data set and drone real-time positioning data; based on the drone real-time positioning data, the vertical water depth data set and underwater terrain data set are reconstructed by water depth remote sensing to obtain the river water body cross-section model;

[0014] The radar flow measurement module is used to obtain the reflected radar signal; the water flow velocity is calculated based on the reflected radar signal to obtain the radar detected water flow velocity data;

[0015] The image flow measurement module is used to obtain a river water body image set; mark key points of the river water body image set to generate water body image key points; map the water body flow velocity of the river water body image set based on the water body image key points to generate image detection water body flow velocity data;

[0016] The fusion flow measurement module is used to perform water velocity measurement calculation on the reflected radar signal and the river water image set according to the radar detected water velocity data and the image detected water velocity data, and generate fused measured water velocity data; perform standard measurement process inversion on the radar detected water velocity data and the image detected water velocity data according to the fused measured water velocity data, and obtain standard radar detection process data and standard image detection process data;

[0017] The flow measurement height calculation module is used to extract the intersection height of the standard radar detection process data and the standard image detection process data to generate the flow measurement standard height data; extract the intersection coordinates of the standard radar detection process data and the standard image detection process data to obtain the flow measurement standard coordinate data;

[0018] The UAV control module is used to define the safe height for the fused measured water flow velocity data to obtain the safe flight height data; to construct a closed-loop control model for the flow measurement standard height data, flow measurement standard coordinate data and safe flight height data, thereby realizing the safe UAV control task; and to transmit the river water cross-section diagram and the fused measured water flow velocity data to the ground station in real time, thereby realizing the UAV flow measurement task.

[0019] The invention is beneficial in that it integrates multiple data sources such as drone positioning data, water depth data, reflection radar data, and water body image data, which makes up for the limitations of a single data source and improves the accuracy and reliability of the final water body flow velocity measurement. Based on the drone positioning data, the vertical water depth and underwater terrain data are three-dimensionally reconstructed, and a detailed river water body cross-section model is established, which provides a more accurate geospatial reference for subsequent flow velocity calculation. By analyzing the reflection radar signal and the water body image set, water body flow velocity data of two different dimensions are obtained, and then a more reliable final flow velocity result is obtained through fusion calculation. Inverting the fused measurement results back to the standard radar detection process and image detection process helps to further optimize and standardize the data acquisition link and improve data quality. A closed-loop control model is constructed according to the standard flow measurement height, standard flow measurement coordinate data, and safe flight height, which can dynamically adjust the flight height and path of the drone to ensure the safety of the flow measurement operation. The measurement results are transmitted to the ground in a timely manner, which can realize real-time monitoring and visual display of the river hydrological conditions, and provide support for related applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A schematic diagram of the steps of a multi-factor UAV flow measurement method;

[0021] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;

[0022] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0023] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0024] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0025] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0026] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0027] To achieve this, please refer to Figures 1 to 3 , a multi-factor UAV flow measurement method, comprising the following steps:

[0028] Step S1: obtaining a designated flow measurement river channel, a vertical water depth data set, an underwater terrain data set, and a real-time positioning data of a UAV; performing a water depth remote sensing stereo reconstruction on the vertical water depth data set and the underwater terrain data set based on the real-time positioning data of the UAV to obtain a river water body cross-section model;

[0029] Step S2: Acquire the reflected radar signal; calculate the water flow velocity of the reflected radar signal to obtain radar detected water flow velocity data;

[0030] Step S3: Acquire a river water body image set; annotate the river water body image set with key points to generate water body image key points; map the water body flow velocity of the river water body image set based on the water body image key points to generate image detection water body flow velocity data;

[0031] Step S4: performing water velocity measurement calculation by fusing the reflected radar signal and the river water image set according to the radar detected water velocity data and the image detected water velocity data, and generating fused measured water velocity data; performing standard measurement process inversion on the radar detected water velocity data and the image detected water velocity data according to the fused measured water velocity data, and obtaining standard radar detection process data and standard image detection process data;

[0032] Step S5: extracting the intersection height of the standard radar detection process data and the standard image detection process data to generate standard flow measurement height data; extracting the intersection coordinates of the standard radar detection process data and the standard image detection process data to obtain standard flow measurement coordinate data;

[0033] Step S6: Define the safe height for the fused measured water flow velocity data to obtain the safe flight height data; construct a closed-loop control model for the flow measurement standard height data, flow measurement standard coordinate data and safe flight height data, so as to achieve a safe UAV control task; transmit the river water cross-section diagram and the fused measured water flow velocity data to the ground station in real time, so as to achieve the UAV flow measurement task.

[0034] The present invention realizes the three-dimensional perception of river water bodies by combining multi-source information such as drone positioning data, water depth data, and underwater terrain data, and fully obtains relevant data on water flow velocity by using multiple measurement methods such as reflected radar signals and river water body images. Through the fusion calculation of radar detection and image detection water flow velocity data, more accurate and reliable fusion measurement water flow velocity data is generated, and the standard detection process data is inverted based on the fusion data, which improves the accuracy and reliability of the measurement results. The safe flight altitude is defined according to the fusion measurement data, and a closed-loop control model is constructed to ensure that the drone completes the flow measurement task within a safe range. The river water body cross-section diagram and flow velocity data are transmitted to the ground station in real time to provide timely and reliable hydrological information for decision support. The full process automation of drone flow measurement operations is realized by integrating hydrological monitoring, remote sensing measurement, and intelligent control technologies. The advantages of various technical means are fully utilized to realize the comprehensive collection, high-precision fusion, and intelligent control of flow measurement data, which greatly improves the flow measurement efficiency and data quality.

[0035] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a multi-factor UAV flow measurement method of the present invention. In this example, the multi-factor UAV flow measurement method includes the following steps:

[0036] Step S1: obtaining a designated flow measurement river channel, a vertical water depth data set, an underwater terrain data set, and a real-time positioning data of a UAV; performing a water depth remote sensing stereo reconstruction on the vertical water depth data set and the underwater terrain data set based on the real-time positioning data of the UAV to obtain a river water body cross-section model;

[0037] In the embodiment of the present invention, after obtaining the flight qualification of the relevant UAV, the UAV is deployed to perform the measurement task over the designated flow measurement river channel, and the real-time positioning data of the UAV is obtained by using the high-precision GNSS and inertial navigation system (INS) carried. The water depth data of multiple vertical positions are measured using sonar equipment to form a vertical water depth data set. At the same time, underwater terrain data is collected using a laser radar (LiDAR) or a multi-beam echo sounder to generate an underwater terrain data set. The real-time positioning data of the UAV is combined with the vertical water depth data set and the underwater terrain data set, and a three-dimensional reconstruction algorithm is used to perform three-dimensional remote sensing reconstruction of the water depth to generate a river water body cross-section model.

[0038] Step S2: Acquire the reflected radar signal; calculate the water flow velocity of the reflected radar signal to obtain radar detected water flow velocity data;

[0039] In the embodiment of the present invention, after obtaining the flight qualification of the relevant drone, a Doppler radar is installed on the drone to transmit electromagnetic wave signals and receive signals reflected from the water surface. The frequency offset of the reflected signal is analyzed and the Doppler effect principle is applied to calculate the flow velocity on the water surface. The received radar signal is filtered and denoised using a data processing algorithm to ensure the accuracy of the flow velocity data. The generated radar-detected water flow velocity data is transmitted to the ground station in real time through a wireless communication module for subsequent data fusion and analysis.

[0040] Step S3: Acquire a river water body image set; annotate the river water body image set with key points to generate water body image key points; map the water body flow velocity of the river water body image set based on the water body image key points to generate image detection water body flow velocity data;

[0041] In the embodiment of the present invention, a high-definition camera on a drone is used to capture high-resolution images of the river to form a river water image set. Computer vision technology and image processing algorithms are used to analyze and process the images, and key points in the water body, such as water surface ripples and floating objects, are marked. By tracking the movement trajectory of these key points in the image, the water flow velocity is calculated using the optical flow method or feature point matching algorithm. The calculation results are integrated into image detection water flow velocity data to ensure that the flow velocity information of each key point is accurate.

[0042] Step S4: performing water velocity measurement calculation by fusing the reflected radar signal and the river water image set according to the radar detected water velocity data and the image detected water velocity data, and generating fused measured water velocity data; performing standard measurement process inversion on the radar detected water velocity data and the image detected water velocity data according to the fused measured water velocity data, and obtaining standard radar detection process data and standard image detection process data;

[0043] In the embodiment of the present invention, the Kalman filter data fusion algorithm is used to fuse the water velocity data detected by radar and the water velocity data detected by image, and the advantages of two different measurement methods are used to generate more accurate fused measurement water velocity data. Based on the fused measurement results, the standard measurement process of radar detection and image detection is inverted to extract the standard process data of each detection method. Through these standard data, the consistency and repeatability of different measurement methods are ensured, and the reliability of the measurement results is improved.

[0044] Step S5: extracting the intersection height of the standard radar detection process data and the standard image detection process data to generate standard flow measurement height data; extracting the intersection coordinates of the standard radar detection process data and the standard image detection process data to obtain standard flow measurement coordinate data;

[0045] In an embodiment of the present invention, the standard radar detection process data and the standard image detection process data are subjected to intersection analysis to extract the height data of each detection point in the flow measurement process. Statistical analysis methods, such as mean and variance calculations, are used to determine the standard height of each detection point. Standard flow measurement height data are generated, which can reflect the high consistency of the flow measurement process under different environments and conditions, and ensure the accuracy of the measurement results. Intersection coordinates of the standard radar detection process data and the standard image detection process data are extracted, and the two sets of data are matched by timestamps and spatial coordinates to find overlapping measurement points. For each matching point, the coordinate information of the radar and the image is extracted, and the average value is calculated to generate the standard flow measurement coordinate data.

[0046] Step S6: Define the safe height for the fused measured water flow velocity data to obtain the safe flight height data; construct a closed-loop control model for the flow measurement standard height data, flow measurement standard coordinate data and safe flight height data, so as to achieve a safe UAV control task; transmit the river water cross-section diagram and the fused measured water flow velocity data to the ground station in real time, so as to achieve the UAV flow measurement task.

[0047] In the embodiment of the present invention, based on the fusion measurement of water flow velocity data, the safe flight altitude that the UAV should maintain during the flow measurement task is defined to avoid collision with the water surface and other obstacles. In combination with the flow measurement standard altitude data and the flow measurement standard coordinate data, a closed-loop control model is constructed, and a PID control algorithm is used to achieve real-time adjustment of the UAV flight path to ensure that the UAV maintains a safe flight altitude and maintains the path selection of the standard path during the flow measurement process. Wireless communication technology is used to instantly transmit the river water cross-section diagram and the fusion measurement water flow velocity data to the ground station, which can monitor the progress of the UAV flow measurement task in real time and analyze and store the data.

[0048] Preferably, step S1 comprises the following steps:

[0049] Step S11: screening the river channel according to the flow measurement plan to obtain the designated flow measurement river channel;

[0050] Step S12: Collecting real-time positioning data of the drone through GNSS;

[0051] Step S13: using a UAV ground penetrating radar to measure the vertical depth of the designated flow measurement channel to obtain a vertical depth data set;

[0052] Step S14: performing underwater topography exploration of the designated flow measurement channel by using a UAV ground penetrating radar to generate an underwater topography data set;

[0053] Step S15: performing corresponding water body splicing on the vertical water depth data set and the underwater terrain data set to generate river water depth data;

[0054] Step S16: Performing three-dimensional reconstruction of the river channel three-dimensional water depth data based on the real-time positioning data of the UAV to obtain a cross-sectional view of the river channel water body.

[0055] The benefit of the present invention lies in the targeted selection of river channels suitable for flow measurement operations, laying the foundation for subsequent work. The real-time positioning data of the drone is obtained through GNSS, providing key input for the reconstruction of three-dimensional water depth data. The vertical water depth and underwater topography are measured using the drone's ground-penetrating radar to obtain high-precision measured data of the river water body. The vertical water depth and underwater topography data are correspondingly spliced ​​to generate three-dimensional data that fully reflects the water depth of the river water body. Based on the drone positioning data, the three-dimensional water depth data is stereoscopically reconstructed to obtain a model diagram that intuitively reflects the cross-sectional morphology of the river water body. The detection advantages of the drone are fully utilized to obtain detailed three-dimensional data of the river water body, providing accurate basic water body data for subsequent radar flow velocity measurement and image flow velocity extraction, ensuring the reliability and accuracy of the entire flow measurement process.

[0056] In the embodiment of the present invention, according to the predetermined flow measurement plan, the target river channel is first screened. The screening process is based on the geographical location of the river channel, the river hydrological characteristics, the historical hydrological data and the importance of the flow measurement of the river channel. GIS (Geographic Information System) software is used to analyze and compare each candidate river channel, and finally determine the designated flow measurement river channel for the flow measurement task. The selection of the river channel must meet the requirements of the flow measurement plan, including water flow, water quality, water depth and terrain characteristics. A drone is deployed over the designated flow measurement river channel, and the global navigation satellite system (GNSS) module carried by the drone is started to collect the positioning data of the drone in real time. The GNSS module uses the signals of multiple satellites for positioning, and improves the positioning accuracy through the differential GPS (DGPS) technology to ensure that the error of the acquired positioning data is within the centimeter range. The drone is equipped with a ground penetrating radar system to measure the water depth at the predetermined vertical position on the designated flow measurement river channel. The ground penetrating radar system calculates the water depth of each vertical position by emitting electromagnetic waves and receiving signals reflected from the riverbed. The measurement results are recorded and summarized to generate a vertical water depth data set. In order to ensure the accuracy of the data, the vertical line measurement must cover multiple key points of the river channel to form dense data sampling. The ground penetrating radar system carried by the drone is used to conduct a comprehensive underwater terrain exploration of the designated flow measurement river channel. The ground penetrating radar obtains detailed data of the underwater terrain by scanning the riverbed. The detected terrain data is processed and analyzed to generate an underwater terrain data set. Data processing includes noise removal, error correction and data smoothing. The acquired vertical line water depth data set is spliced ​​with the underwater terrain data set. The two data sets are aligned using geographic coordinates, and the data at each vertical line position is interpolated and fitted to generate continuous river water depth data. Cubic spline interpolation or Kriging interpolation methods are used. The positioning data collected in real time by the drone is used to perform three-dimensional stereo reconstruction of the spliced ​​river water depth data. The water depth data is converted into a three-dimensional model of the river channel through remote sensing technology and three-dimensional modeling algorithms. A three-dimensional stereo image of the river channel is constructed using TIN (irregular triangulated network) or raster data format. Finally, a cross-sectional view of the river water body is generated.

[0057] Preferably, step S16 comprises the following steps:

[0058] Step S161: performing coordinate point grid cutting on the three-dimensional water depth data of the river channel to generate water depth grid units of the river channel;

[0059] Step S162: coordinate alignment is performed on the real-time positioning data of the UAV and the grid unit of the river depth to obtain alignment coordinate data;

[0060] Step S163: performing coordinate co-reference data fusion on the real-time positioning data of the UAV and the three-dimensional water depth data of the river channel based on the aligned coordinate data to generate corresponding fused three-dimensional data;

[0061] Step S164: reconstructing a three-dimensional model of the corresponding fused three-dimensional data to obtain a three-dimensional model of the river channel;

[0062] Step S165: Perform two-dimensional cross-section cutting on the three-dimensional model of the river channel to generate a cross-sectional diagram of the river channel water body.

[0063] The benefit of the present invention lies in that the three-dimensional water depth data of the river channel is cut into grid units, providing more detailed basic data for subsequent coordinate alignment and data fusion. The real-time positioning data of the drone is aligned with the water depth grid unit to ensure the spatial consistency of the two types of data. Based on the aligned coordinate data, the drone positioning information and the three-dimensional water depth data are fused to generate more complete corresponding three-dimensional data. Using the fused three-dimensional data, a three-dimensional model of the river channel is reconstructed to more intuitively reflect the topographic features of the river channel. Through two-dimensional cross-sectional cutting, a cross-sectional view of the river water body is extracted from the three-dimensional model to provide necessary water body geometry information for subsequent flow velocity measurement and analysis. The effective fusion of drone positioning data and three-dimensional water depth data is achieved, and a high-precision three-dimensional model of the river channel and a cross-sectional view of the water body are generated. These basic data provide reliable input for the downstream radar flow velocity measurement and image flow velocity extraction steps, ensuring the accuracy and reliability of the entire flow measurement process.

[0064] In an embodiment of the present invention, the acquired three-dimensional water depth data of the river channel is processed and converted into a regular coordinate point grid. The water depth data of the river channel is gridded according to a certain spatial resolution by using Delaunay triangulation or quadtree segmentation algorithm. Each grid unit represents the water depth information in a specific area, ensuring that the resolution of the grid unit can fully reflect the subtle changes in the river channel. The real-time positioning data of the drone is aligned with the water depth grid unit data of the river channel. First, the drone positioning data and the grid unit data are unified into the same coordinate system through coordinate conversion. Then, the real-time positioning point of the drone is mapped to the corresponding water depth grid unit by using the nearest neighbor interpolation or bilinear interpolation method. The real-time positioning data of the drone and the three-dimensional water depth data of the river channel are fused by using the aligned coordinate data. The information of the two data sources is integrated by using the weighted average and Kalman filter data fusion algorithm to generate a more accurate three-dimensional data set. In the fusion process, the data redundancy and conflict problems are mainly handled to ensure that the fused data has high accuracy and consistency. The fused three-dimensional data is processed to construct a three-dimensional model of the river channel. Use 3D reconstruction algorithms, such as Marching Cubes or Poisson Surface Reconstruction, to convert the fused 3D data into a 3D mesh model. During the processing, data smoothing and noise removal are performed to ensure the accuracy and detail retention of the model. Perform 2D cross-section cutting on the generated 3D model of the river. Use cutting algorithms, such as plane cutting or surface cutting, to cut the 3D model according to the predetermined vertical position to generate a 2D cross-section diagram. Each cross-section diagram shows the water depth and terrain features at a specific location. During the cutting process, ensure that the cutting surface is accurately aligned with the model to avoid data distortion. Finally, a cross-section diagram of the river water body is generated.

[0065] Preferably, step S2 comprises the following steps:

[0066] Step S21: performing reflection radar extraction on the designated flow measurement river channel to obtain reflection radar signals;

[0067] Step S22: performing pulse compression on the reflected radar signal to generate a high-quality reflected radar signal;

[0068] Step S23: performing Doppler frequency shift estimation on the high-quality reflected signal to generate signal frequency shift data;

[0069] Step S24: Calculate the water flow velocity of the high-quality reflected radar signal according to the signal frequency shift data to obtain radar-detected water flow velocity data.

[0070] The benefit of the present invention lies in that the reflection radar is used to detect the designated flow measurement river channel, obtain the reflection radar signal, and provide basic data for the subsequent water body flow velocity measurement. The obtained reflection radar signal is pulse compressed to enhance the signal quality and improve the accuracy of subsequent processing. Based on the high-quality reflection radar signal, the frequency shift data that can reflect the flow state of the water body is obtained through Doppler frequency shift estimation. Using the signal frequency shift data and combining with the radar principle, the flow velocity data of the water body is calculated to provide important flow velocity information for the entire flow measurement process. The advantage of radar as an active sensor is fully utilized, and accurate water body flow velocity data is finally obtained through effective processing of the reflected signal. These flow velocity data are combined with the aforementioned three-dimensional model of the river channel and the water body cross-section information, which provides a reliable reference and supplement for the subsequent image flow velocity extraction step, further ensuring the accuracy of the entire flow measurement process.

[0071] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:

[0072] Step S21: performing reflection radar extraction on the designated flow measurement river channel to obtain reflection radar signals;

[0073] In an embodiment of the present invention, a drone equipped with a synthetic aperture radar (SAR) is deployed over a designated flow measurement river channel, and the drone is hovered over the river channel and scanned laterally along the river channel to capture the reflected radar signal of the entire cross section. This method ensures that detailed information at different locations of the river channel can be obtained. The SAR system transmits a high-frequency microwave signal at each hovering point and receives the signal reflected from the surface of the water body. The signal reflected from the surface of the river channel is received for data collection. The use of high-frequency microwave signals ensures that even small fluctuations on the surface of the water body can be captured. The collected reflected radar signal is transmitted to the ground station via a wireless communication system for further processing. The entire process is monitored in real time to ensure the integrity and accuracy of the signal.

[0074] Step S22: performing pulse compression on the reflected radar signal to generate a high-quality reflected radar signal;

[0075] In the embodiment of the present invention, the collected reflected radar signal is input into the pulse compression processing system, and the signal is processed using the matched filtering technology. The matched filtering technology enhances the time resolution and signal-to-noise ratio of the signal by performing a convolution operation on the received echo signal and the transmitted reference signal. By eliminating multipath interference and noise, the quality of the signal is improved, and a high-quality reflected radar signal is generated. The processed signal can more clearly reflect the subtle characteristics of the river water body.

[0076] Step S23: performing Doppler frequency shift estimation on the high-quality reflected signal to generate signal frequency shift data;

[0077] In the embodiment of the present invention, the Doppler frequency shift estimation algorithm is used to analyze the high-quality reflected radar signal. The Doppler frequency shift is the change in signal frequency caused by the movement of the target. By detecting this frequency change, the movement speed of the target can be inferred. In the specific operation, the short-time Fourier transform (STFT) or autoregressive spectrum estimation method is used to perform frequency domain analysis on the signal and extract the frequency shift information. The generated signal frequency shift data will reflect the flow velocity change of the river water body.

[0078] Step S24: Calculate the water flow velocity of the high-quality reflected radar signal according to the signal frequency shift data to obtain radar-detected water flow velocity data.

[0079] In the embodiment of the present invention, based on the signal frequency shift data, the radar flow velocity estimation model is applied to calculate the flow velocity of the water body. The estimation model uses the relationship between the Doppler frequency shift and the target speed through the formula Where v is the water velocity, f d is the Doppler frequency shift, λ is the wavelength of the radar signal, and the velocity of the water body is accurately calculated. The calculation results are sorted out to generate radar-detected water velocity data, providing accurate water velocity information for subsequent hydrological analysis and velocity monitoring.

[0080] Preferably, step S3 comprises the following steps:

[0081] Step S31: collecting water body images of a designated flow measurement river channel to obtain a river channel water body image set;

[0082] Step S32: performing multi-temporal image registration on the river water body image set to generate an aligned water body image;

[0083] Step S33: marking key points of the aligned water body image to generate key points of the water body image;

[0084] Step S34: performing sparse optical flow estimation on key points of the water body image to obtain estimated flow velocity data;

[0085] Step S35: performing dense optical flow calculation on the aligned water body image set to obtain image pixel motion vectors;

[0086] Step S36: Perform velocity vector mapping on the estimated velocity data based on the image pixel motion vector to generate image detected water body velocity data.

[0087] The benefit of the present invention lies in that by collecting water body images of the flow measurement river channel, image information of the river channel water body is obtained, providing a data basis for subsequent image flow velocity analysis. The collected multi-phase water body images are aligned to ensure the consistency of spatial positions, laying the foundation for subsequent key point marking and optical flow calculation. Key points are marked on the aligned water body images to provide key information for sparse optical flow estimation, thereby obtaining preliminary water body flow velocity data. Based on the key points of the water body image, the motion vector of the pixel scale of the whole image is further obtained through dense optical flow calculation, reflecting a more detailed water flow state. The flow velocity data obtained by the sparse optical flow estimation is fused with the pixel motion vector calculated by the dense optical flow to generate more accurate and reliable image detection water body flow velocity data. The water body image information collected by the drone is fully utilized, and accurate water body flow velocity data is finally obtained through a series of processing such as image registration, key point marking, and optical flow estimation. These flow velocity data are combined with the radar measurement flow velocity data obtained above, further improving the accuracy and reliability of the entire flow measurement process.

[0088] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:

[0089] Step S31: collecting water body images of a designated flow measurement river channel to obtain a river channel water body image set;

[0090] In an embodiment of the present invention, a drone is deployed on a designated flow measurement river channel, equipped with a high-definition camera and an image acquisition system, and photographs are taken at multiple key points along the river channel. The drone automatically collects river water images at different angles and heights according to a preset flight path and time interval. During the image acquisition process, it is ensured that the shooting range covers the entire river basin to avoid blind spots. All collected image data are transmitted to the ground station in real time via a wireless communication system to generate a complete river water image set for subsequent processing.

[0091] Step S32: performing multi-temporal image registration on the river water body image set to generate an aligned water body image;

[0092] In an embodiment of the present invention, the acquired river water body image set is processed, and the multi-temporal image registration technology is used to align the images taken at different times and different viewing angles. The key feature points in each image are extracted using an image feature matching algorithm, such as SIFT (Scale Invariant Feature Transform) or ORB (Rapid Feature Points). Then, the transformation matrix between the images is calculated using the RANSAC (Random Sampling Consensus) algorithm, and the registration operation is performed. Finally, an aligned water body image is generated to ensure that the subsequently processed images have a consistent reference coordinate system.

[0093] Step S33: marking key points of the aligned water body image to generate key points of the water body image;

[0094] In an embodiment of the present invention, the aligned water body image is annotated with key points. Artificial intelligence image recognition technology is used in combination with a deep learning model (such as a CNN convolutional neural network) to automatically identify and annotate the key points of the water body in the image. These key points usually include river boundaries and water features with obvious characteristic water flows. For parts that are difficult to identify automatically, a manual assisted annotation method is used to ensure the accurate position of each key point. After the annotation is completed, an annotation data set containing the key points of the water body image is generated.

[0095] Step S34: performing sparse optical flow estimation on key points of the water body image to obtain estimated flow velocity data;

[0096] In the embodiment of the present invention, a sparse optical flow estimation algorithm is used to perform motion estimation on the key points of the annotated water body image. The Lucas-Kanade optical flow algorithm is used to calculate the flow velocity of the water body by analyzing the displacement of the key points between consecutive frames. The algorithm establishes a small window near each key point, tracks its motion path in subsequent frames, and estimates its velocity vector. The motion data of all key points are integrated to generate preliminary estimated flow velocity data, which provides a basis for further flow velocity analysis.

[0097] Step S35: performing dense optical flow calculation on the aligned water body image set to obtain image pixel motion vectors;

[0098] In the embodiment of the present invention, a dense optical flow calculation method is used to comprehensively analyze the aligned water body image set. The Farneback optical flow algorithm or the Pyramidal Lucas-Kanade optical flow algorithm is used to estimate the motion of each pixel in the image and generate a dense pixel motion vector field. The algorithm processes the image layer by layer by establishing a multi-scale pyramid structure to ensure that large-scale motion and subtle motion are captured simultaneously. The image pixel motion vector finally generated can fully reflect the flow characteristics of the water body.

[0099] Step S36: Perform velocity vector mapping on the estimated velocity data based on the image pixel motion vector to generate image detected water body velocity data.

[0100] In an embodiment of the present invention, the image pixel motion vector obtained by dense optical flow calculation is mapped and integrated with the previous sparse optical flow estimated flow velocity data. The sparse key point flow velocity data is interpolated using a bilinear interpolation or weighted average algorithm to expand it to the entire pixel motion vector field. By comparison and correction, the consistency and accuracy of the estimated flow velocity data and the dense optical flow calculation results are ensured. Finally, complete image detection water body flow velocity data is generated to provide accurate river water body flow velocity information for hydrological analysis and flow velocity monitoring.

[0101] Preferably, step S4 comprises the following steps:

[0102] Step S41: performing a difference comparison between the water body flow velocity data detected by radar and the water body flow velocity data detected by image to obtain detection difference data;

[0103] Step S42: performing fusion water velocity measurement calculation on the reflected radar signal and the river water image set based on the detection difference data to generate fusion measured water velocity data;

[0104] Step S43: performing a standard value search on the radar detected water body flow velocity data and the image detected water body flow velocity data according to the fused measured water body flow velocity data, to obtain standard radar detection data and standard image detection data;

[0105] Step S44: performing measurement process inversion on the standard radar detection data and the standard image detection data to obtain standard radar detection process data and standard image detection process data.

[0106] The invention is beneficial in that the difference between the water velocity data detected by radar and the water velocity data detected by image is compared, the detection difference between the two methods is identified, and a reference is provided for subsequent data fusion. Based on the detection difference data, a comprehensive water velocity measurement calculation is performed by fusing the radar signal and the water body image, and more accurate and reliable fusion measurement water velocity data is generated. Using the fusion measurement water velocity data as a standard, the radar detection data and the image detection data are calibrated and standardized to obtain standard radar detection data and image detection data. For the standard radar detection data and image detection data, the corresponding detection process data is further inverted to provide a basis for the subsequent accurate analysis of the characteristics and advantages and disadvantages of the two measurement methods. It embodies the idea of ​​multi-source information fusion, gives full play to the respective advantages of the two measurement methods of radar and image, and finally generates more accurate and reliable water velocity data through a series of operations such as data comparison, fusion calculation, and standardized processing. These fusion measurement results and their inversion data provide important support for evaluating the performance of the entire flow measurement method, and help to further optimize and improve the entire flow measurement process.

[0107] In the embodiment of the present invention, the radar detection water body flow velocity data and the image detection water body flow velocity data are preprocessed respectively to ensure the uniformity of data format. The error analysis method is used to compare the difference between the two sets of data. By calculating the velocity difference of the two data sets at the same position, an error distribution diagram is generated. The error distribution is analyzed using statistical analysis software to determine the size and distribution law of the error. The generated detection difference data includes the mean value, variance and specific value of the error changing with position. Using the detection difference data and combining the Kalman filter algorithm, the reflected radar signal and the river water body image set are fused to calculate the water body velocity. First, the radar signal and image data are converted into a unified coordinate system and time synchronization is performed. Then, based on the Kalman filter algorithm, the detection difference data is input, and the velocity values ​​of the two data sources are gradually corrected and fused to reduce the error. Through multiple iterative calculations, the final fused measurement water body flow velocity data is generated. The data has higher accuracy and reliability and can be used for subsequent hydrological analysis and monitoring. According to the fused measured water velocity data, the original radar detection water velocity data and image detection water velocity data are corrected and calibrated using the standard value retrieval algorithm. First, the fused measurement data is used as a benchmark to calculate the standard velocity value at each position. Then, the radar data and image data are fitted using the least squares method or other optimization algorithms to correct their deviations and obtain standard radar detection data and standard image detection data. This process ensures that all measurement data are consistent with the fusion results and improves the accuracy and consistency of the overall data. The measurement process is inverted for the calibrated standard radar detection data and standard image detection data. Using an inversion algorithm such as minimum entropy inversion or Bayesian inversion, the measurement data is converted into data for each step in the original measurement process. The specific operation includes: inputting the standard data into the inversion model, and gradually restoring each parameter and step in the measurement process through model calculation. The generated standard radar detection process data and standard image detection process data record in detail all the details from the initial measurement to the final data generation.

[0108] Preferably, the calculation of water velocity measurement by fusing the reflected radar signal and the river water image set based on the detection difference data includes:

[0109] Performing spatiotemporal feature recognition on the detection difference data to obtain spatiotemporal feature data of speed difference;

[0110] Based on the spatiotemporal characteristic data of velocity difference, feature matching is performed on the reflected radar signal and the river water image set to generate a difference corresponding feature matrix;

[0111] According to the difference corresponding feature matrix, the reflected radar signal and the river water body image set are collaboratively de-differenced and fused to obtain collaborative fusion data;

[0112] Reconstruct the water flow field of the collaborative fusion data to generate a simulated water flow field;

[0113] Estimate the water velocity of the simulated water flow field and obtain the fused measured water velocity data

[0114] The benefit of the present invention lies in that the spatiotemporal feature recognition of the detection difference data can more comprehensively analyze the difference characteristics of the two measurement methods of radar and image under different time and space conditions, and provide a targeted basis for subsequent data fusion. Based on the spatiotemporal feature data of speed difference, the corresponding relationship between the reflected radar signal and the water body image is established by feature matching, which lays the foundation for the next step of collaborative de-difference fusion. Using the difference corresponding feature matrix, the collaborative data fusion between the reflected radar signal and the water body image set can better eliminate the inherent differences between the two measurement methods and obtain more accurate collaborative fusion data. The water flow field of the collaborative fusion data is reconstructed to generate a simulated water flow field, which can more intuitively reflect the flow state of the actual water body and provide support for further velocity estimation. Based on the reconstructed simulated water flow field, a more comprehensive and reliable fusion measurement of water velocity data is finally obtained, which gives full play to the advantages of the two measurement methods of radar and image. It reflects the in-depth analysis of the detection difference, and eliminates the inherent differences between the two measurement methods by feature matching and collaborative fusion, and obtains more accurate and reliable water velocity data. These fusion results not only include the measurement data itself, but also involve the simulated water flow field, providing more comprehensive information support for subsequent hydrological analysis.

[0115] In an embodiment of the present invention, the detection difference data is processed, and the difference data is first converted into a spatiotemporal data set and analyzed using a spatiotemporal feature extraction algorithm. A spatiotemporal decomposition method, such as principal component analysis (PCA) or a temporal convolutional neural network (TCN), is applied to extract the main features in the time series and spatial distribution. The specific operation includes time series modeling of the data, extracting the time characteristics of the flow velocity change, and analyzing the spatial distribution characteristics through a spatial filter. The identified features are sorted into velocity difference spatiotemporal feature data. The velocity difference spatiotemporal feature data is feature matched with the reflected radar signal and the river water body image set. First, feature points or feature areas, such as edge points and corner points, are extracted from the reflected radar signal and the water body image. Then, these features are compared using a feature matching algorithm, such as SURF (Speeded Robust Features) or ORB (Rapid Rotation Invariant Features). By calculating the similarity between the feature points, a difference corresponding feature matrix is ​​generated, which records the corresponding feature points in the radar signal and the image and their matching degree. The difference corresponding feature matrix is ​​used to perform collaborative data fusion on the reflected radar signal and the river water body image set. The difference between radar signal and image data is corrected by weighted average or least square method. The specific operation includes: applying the characteristic matrix to the corresponding parts of radar signal and image data, weighted fusion of data according to the matching degree, and removing or reducing the difference in data. The water flow field is reconstructed by the synergistic fusion data, and the flow field reconstruction algorithm, such as grid generation and fluid dynamics simulation, is used. First, the synergistic fusion data is converted into the input data of the flow field model, including velocity field and pressure field. Then, the flow field simulation and grid division are performed based on these data using fluid simulation software to reconstruct the water flow field. The distribution and change of the flow field are calculated by numerical simulation method to generate a simulated water flow field. The flow velocity is estimated based on the generated simulated water flow field. The flow field is analyzed in detail by applying the flow velocity calculation algorithm, such as finite volume method (FVM) or finite element method (FEM). According to the velocity distribution in the flow field model, the flow velocity values ​​in different areas are calculated. The specific operation includes gridding the flow field, calculating the flow velocity for each grid unit, and integrating the flow velocity data of each area. The final fused measured water velocity data accurately reflects the flow velocity conditions in the river channel and provides reliable data support for hydrological analysis and monitoring.

[0116] Preferably, step S5 comprises the following steps:

[0117] Step S51: extracting the standard height of the standard radar detection process data to obtain radar detection standard height data;

[0118] Step S52: extracting the standard height of the standard image detection process data to obtain the image detection standard height data;

[0119] Step S53: Repeat the flight altitude recognition for the radar detection standard altitude data and the image detection standard altitude data to generate flow measurement standard altitude data;

[0120] Step S54: extracting standard coordinates from the standard radar detection process data and the standard image detection process data to obtain radar monitoring standard coordinate data and image monitoring standard coordinate data;

[0121] Step S55: performing intersection coordinate positioning on the radar monitoring standard coordinate data and the image monitoring standard coordinate data to obtain the flow measurement standard coordinate data.

[0122] The invention is beneficial in that the standard height data of the ground penetrating radar detection process data and the standard image detection process data are respectively extracted and measured, laying the foundation for subsequent height recognition and data fusion. By repeatedly identifying the flight height of the radar detection standard height data and the image detection standard height data, the final flow measurement standard height data can be determined. This standard height data is crucial to the entire flow measurement process because it directly affects the accuracy of the velocity measurement calculation. After determining the standard height data, the errors and differences in height between the radar and image measurement methods can be better evaluated and corrected, thereby further optimizing the entire flow measurement method. The determination of the standard height data provides basic parameters for subsequent hydrological analysis based on the fusion measurement of water body flow velocity data, which helps to improve the reliability of the analysis results. The monitoring standard coordinates are respectively extracted from the standard radar detection data and the standard image detection data, and then the flow measurement standard coordinate data is obtained through intersection positioning, which can ensure the spatial positioning accuracy of the flow measurement data and lay the foundation for the subsequent water body flow velocity field analysis and water body parameter calculation. These steps highlight the strict control and optimization of the measurement parameters, make full use of the advantages of the two detection methods of radar and image, and ensure the accuracy and reliability of the final flow measurement results. It reflects the in-depth analysis and optimization of the entire flow measurement process, provides reliable basic data support for subsequent hydrological analysis, and is an important component of the multi-factor UAV flow measurement method.

[0123] In an embodiment of the present invention, measurement data is extracted from standard radar detection process data. The water surface height is calculated using the difference in reflection time in the radar echo signal, and the echo signal is converted into height information. The specific operation includes using a radar signal processing algorithm, such as delay line estimation, to determine the distance of each measurement point. By converting the distance data into height data, combined with the calibration model, the standard height of each measurement point in the radar detection process is extracted. Finally, radar detection standard height data is generated. The image information in the standard image detection process data is processed to extract the measurement height data. First, the water surface height in the image is marked by image processing technology. The image coordinate system is aligned with the actual height system using an image calibration tool. The position of the water surface is accurately determined by applying an image analysis algorithm, such as edge detection or morphological processing. The water surface height in the image is converted into actual standard height data through known camera calibration parameters and actual measurement conditions. The obtained image detection standard height data characterizes the water body height information extracted during the image detection process. The radar detection standard height data is compared and analyzed with the image detection standard height data. First, the two sets of data are sorted and aligned with the same spatial resolution and time point. Data fusion algorithms, such as Kalman filtering or weighted averaging, are applied to process the height information in the two sets of data. Repeated flight height data are identified by comparing and fusing the height values ​​in the two detection data. These repeated flight height data are used to determine the standard height range. Finally, standard flow measurement height data is generated. Standard coordinates are extracted from standard radar detection process data and standard image detection process data. First, each measurement point is located using high-precision GNSS to obtain accurate spatial coordinates. Radar data and image data are processed using Python programming language and Pandas library respectively to extract the coordinate information of each measurement point. During data processing, duplicates and outliers are removed to ensure data accuracy. The processed coordinate data are stored as radar monitoring standard coordinate data and image monitoring standard coordinate data respectively. Intersection coordinates are located for radar monitoring standard coordinate data and image monitoring standard coordinate data. First, the two sets of coordinate data are imported into the Python environment, and the NumPy and SciPy libraries are used for data processing. Through set operations, the intersection coordinates of the two sets of data are found. The specific operations include converting the coordinate data into a set format and extracting repeated coordinate points using the intersection operation of the set. The extracted intersection coordinates are verified and cleaned to remove existing errors, and finally generate standard coordinate data for flow measurement.

[0124] Preferably, step S6 comprises the following steps:

[0125] Step S61: performing turbulence characteristic analysis on the fused measured water velocity data to obtain water velocity field data;

[0126] Step S62: performing local wind shear calculation on the water body velocity field data to generate water body surface wind shear intensity distribution data;

[0127] Step S63: defining a safe altitude for the wind shear intensity distribution data on the water surface based on a preset UAV wind tolerance threshold, and obtaining safe flight altitude data;

[0128] Step S64: constructing a closed-loop control model for the flow measurement standard height data, the flow measurement standard coordinate data and the safe flight height data to generate a closed-loop control model;

[0129] Step S66: dynamically controlling the flight path of the flow measuring UAV using the closed-loop control model, thereby achieving a safe UAV control task;

[0130] Step S67: The river water cross-section diagram and the fused measured water velocity data are instantly transmitted to the ground station, thereby realizing the UAV flow measurement task.

[0131] The invention is beneficial in that by analyzing the turbulence characteristics of the fused measured water velocity data, the water velocity field data is obtained, and the local wind shear intensity distribution on the water surface is further calculated, so that the meteorological environment of the measured river can be more comprehensively understood, which is helpful to evaluate the safety and feasibility of the UAV flying over the river. According to the preset UAV wind tolerance threshold, combined with the wind shear intensity distribution data on the water surface, the safe UAV flight altitude is determined, which can effectively avoid the UAV from being subjected to excessive wind impact under severe meteorological conditions and ensure flight safety. The flow measurement standard height data, the flow measurement standard coordinate data and the safe flight altitude data are integrated to construct a closed-loop control model, and the UAV flight altitude is adjusted in real time by using the model to ensure that the UAV flies at the optimal measurement altitude, which can not only improve the measurement accuracy, but also effectively avoid safety hazards. The river water cross-section diagram and flow velocity data collected in real time are transmitted to the ground station for real-time analysis and processing. By integrating the water cross-section diagram and flow velocity data, the hydrological conditions of the entire river can be more comprehensively reflected. The multi-factor UAV flow measurement system can greatly improve the accuracy and reliability of flow measurement data while ensuring the safety of the measurement process by comprehensively considering turbulence characteristics, wind shear intensity, and safe flight altitude factors, and adopting technical means such as closed-loop control.

[0132] In an embodiment of the present invention, velocity data points are extracted from the fused measured water body velocity data. The velocity data is processed using a turbulence analysis method, such as vortex intensity analysis or flow field irregularity calculation. An appropriate algorithm, such as Fourier transform or wavelet transform, is selected to calculate turbulence characteristic indicators, such as turbulence intensity or turbulence energy spectrum. Through these analyses, the velocity field characteristics of each point in the water body are obtained to form water body velocity field data. The wind shear intensity of each point is calculated based on the water body velocity field data. The water body velocity field data is combined with the ground wind speed data using a wind shear calculation model. A local wind shear calculation algorithm, such as a wind shear gradient algorithm, is applied to convert the velocity change at each point on the water body surface into wind shear intensity. A spatial distribution map of wind shear intensity is generated. According to a preset drone wind tolerance threshold, the wind shear intensity distribution data on the water body surface is compared with the threshold. The intensity value in the wind shear intensity distribution data is used to determine whether the wind tolerance of the drone is exceeded. The area exceeding the threshold is defined as an unsafe area, and the minimum flight altitude within the safe area is determined. Through computer programs or control systems, these data are converted into safe flight altitude data. These data indicate the safe flight altitude requirements of drones under different wind shear intensities. . A closed-loop control model is constructed for the standard flow measurement altitude data, standard flow measurement coordinate data, and safe flight altitude data, and Simulink is used for modeling and simulation. First, the three sets of data are imported into the Simulink environment and integrated using the data interface module. A closed-loop control system is formed by constructing a PID controller and a state feedback controller. Among them, the construction of the altitude adjustment module is based on the standard flow measurement altitude data and the safe flight altitude data. First, the altitude data is discretized by the interpolation method to form a reference curve for altitude adjustment. Then, the PID (proportional-integral-differential) controller is used to adjust the flight altitude. The PID controller calculates the error between the current altitude and the reference altitude, and performs proportional, integral, and differential operations to generate an altitude adjustment instruction. By adjusting the parameters of the PID controller (proportional gain, integral time, and differential time), the accuracy and response speed of the altitude adjustment are ensured; the construction of the path optimization module is based on the standard flow measurement coordinate data. First, an optimized flight path is generated through a path planning algorithm (such as the Dijkstra algorithm). The path planning algorithm uses the standard coordinate data of the flow measurement to calculate the shortest or optimal path from the starting point to the target point of the drone. Then, the state feedback controller is used to dynamically adjust the flight path of the drone based on the real-time position information. The state feedback controller continuously adjusts the path to ensure that the drone flies according to the predetermined optimal path and can quickly respond and adjust the path when encountering obstacles or emergencies.The model contains an altitude adjustment module and a path optimization module. By adjusting the flight altitude and path of the drone in real time, the safety of flight and the accuracy of the flow measurement task are ensured. The flight altitude of the drone is continuously adjusted to maintain it within the standard altitude range, while avoiding areas with excessive wind shear. The flight path of the drone is adjusted to achieve the standard drone flow measurement path selection. The closed-loop control model is used to dynamically control the flight altitude path of the flow measurement drone. First, the model code is embedded in the flight control system of the drone. The real-time operating system (RTOS) is used to ensure the real-time response capability of the control system. Real-time flight status data and environmental data are obtained through onboard sensors and GPS systems. The control system performs dynamic calculations based on the closed-loop control model and adjusts the flight altitude and path of the drone in real time. Ensure that the drone flies safely in areas with high wind shear intensity and accurately completes the flow measurement task. During the flight, the control system continuously updates and optimizes the flight parameters to ensure the efficiency and safety of the flight mission. The river water cross-section diagram and the fused measured water velocity data are transmitted to the ground station in real time using wireless communication technology. The communication module carried by the drone transmits real-time data to the ground station through the 4G / 5G network or satellite communication system. The ground station is equipped with a high-performance computer and data receiving system, and uses the TCP / IP protocol for data transmission and processing. The received data is processed and stored initially to generate a real-time river water cross-section diagram and flow velocity distribution diagram. The ground station displays this data through data visualization software, monitors the execution of the UAV's flow measurement mission in real time, and adjusts and optimizes the flight strategy in a timely manner.

[0133] In this specification, a multi-factor UAV flow measurement system is also provided, which is used to execute the multi-factor UAV flow measurement method as described above, and the multi-factor UAV flow measurement system includes:

[0134] The cross-section construction module is used to obtain the designated flow measurement river channel, vertical water depth data set, underwater terrain data set and drone real-time positioning data; based on the drone real-time positioning data, the vertical water depth data set and underwater terrain data set are reconstructed by water depth remote sensing to obtain the river water body cross-section model;

[0135] The radar flow measurement module is used to obtain the reflected radar signal; the water flow velocity is calculated based on the reflected radar signal to obtain the radar detected water flow velocity data;

[0136] The image flow measurement module is used to obtain a river water body image set; mark key points of the river water body image set to generate water body image key points; map the water body flow velocity of the river water body image set based on the water body image key points to generate image detection water body flow velocity data;

[0137] The fusion flow measurement module is used to perform water velocity measurement calculation on the reflected radar signal and the river water image set according to the radar detected water velocity data and the image detected water velocity data, and generate fused measured water velocity data; perform standard measurement process inversion on the radar detected water velocity data and the image detected water velocity data according to the fused measured water velocity data, and obtain standard radar detection process data and standard image detection process data;

[0138] The flow measurement height calculation module is used to extract the intersection height of the standard radar detection process data and the standard image detection process data to generate the flow measurement standard height data; extract the intersection coordinates of the standard radar detection process data and the standard image detection process data to obtain the flow measurement standard coordinate data;

[0139] The UAV control module is used to define the safe height for the fused measured water flow velocity data to obtain the safe flight height data; to construct a closed-loop control model for the flow measurement standard height data, flow measurement standard coordinate data and safe flight height data, thereby realizing the safe UAV control task; and to transmit the river water cross-section diagram and the fused measured water flow velocity data to the ground station in real time, thereby realizing the UAV flow measurement task.

[0140] The invention is beneficial in that it integrates multiple data sources such as drone positioning data, water depth data, reflection radar data, and water body image data, which makes up for the limitations of a single data source and improves the accuracy and reliability of the final water body flow velocity measurement. Based on the drone positioning data, the vertical water depth and underwater terrain data are three-dimensionally reconstructed, and a detailed river water body cross-section model is established, which provides a more accurate geospatial reference for subsequent flow velocity calculation. By analyzing the reflection radar signal and the water body image set, water body flow velocity data of two different dimensions are obtained, and then a more reliable final flow velocity result is obtained through fusion calculation. Inverting the fused measurement results back to the standard radar detection process and image detection process helps to further optimize and standardize the data acquisition link and improve data quality. A closed-loop control model is constructed according to the standard flow measurement height, standard flow measurement coordinate data, and safe flight height, which can dynamically adjust the flight height and path of the drone to ensure the safety of the flow measurement operation. The measurement results are transmitted to the ground in a timely manner, which can realize real-time monitoring and visual display of the river hydrological conditions, and provide support for related applications.

[0141] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0142] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A multi-factor UAV flow measurement method, characterized in that: The following steps are involved: Step S1: obtaining a designated flow measurement river channel, a vertical water depth data set, an underwater terrain data set, and a real-time positioning data of a UAV; performing a water depth remote sensing stereo reconstruction on the vertical water depth data set and the underwater terrain data set based on the real-time positioning data of the UAV to obtain a river water body cross-section model; Step S2: Acquire the reflected radar signal; calculate the water flow velocity of the reflected radar signal to obtain radar detected water flow velocity data; Step S3: Acquire a river water body image set; annotate the river water body image set with key points to generate water body image key points; map the water body flow velocity of the river water body image set based on the water body image key points to generate image detection water body flow velocity data; Step S4: performing water velocity measurement calculation by fusing the reflected radar signal and the river water image set according to the radar detected water velocity data and the image detected water velocity data, and generating fused measured water velocity data; performing standard measurement process inversion on the radar detected water velocity data and the image detected water velocity data according to the fused measured water velocity data, and obtaining standard radar detection process data and standard image detection process data; Step S5: extracting the intersection height of the standard radar detection process data and the standard image detection process data to generate the standard flow measurement height data; extracting the intersection coordinates of the standard radar detection process data and the standard image detection process data to obtain the standard flow measurement coordinate data; wherein, extracting the intersection height of the standard radar detection process data and the standard image detection process data to generate the standard flow measurement height data includes the following steps: The standard radar detection process data is measured and the standard height is extracted to obtain the radar detection standard height data; The standard height is measured and extracted from the standard image detection process data to obtain the image detection standard height data; Repeat the flight altitude recognition of radar detection standard altitude data and image detection standard altitude data to generate flow measurement standard altitude data; Step S6: Define the safe height for the fused measured water flow velocity data to obtain the safe flight height data; construct a closed-loop control model for the flow measurement standard height data, flow measurement standard coordinate data and safe flight height data, so as to achieve a safe UAV control task; transmit the river water cross-section diagram and the fused measured water flow velocity data to the ground station in real time, so as to achieve the UAV flow measurement task.

2. The multi-factor UAV flow measurement method according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: screening the river channel according to the flow measurement plan to obtain the designated flow measurement river channel; Step S12: Collecting real-time positioning data of the drone through GNSS; Step S13: using a UAV ground penetrating radar to measure the vertical depth of the designated flow measurement channel to obtain a vertical depth data set; Step S14: performing underwater topography exploration of the designated flow measurement channel by using a UAV ground penetrating radar to generate an underwater topography data set; Step S15: performing corresponding water body splicing on the vertical water depth data set and the underwater terrain data set to generate river water depth data; Step S16: Performing three-dimensional reconstruction of the river channel three-dimensional water depth data based on the real-time positioning data of the UAV to obtain a cross-sectional view of the river channel water body.

3. The multi-factor UAV flow measurement method according to claim 2 is characterized in that: Step S16 includes the following steps: Step S161: performing coordinate point grid cutting on the three-dimensional water depth data of the river channel to generate water depth grid units of the river channel; Step S162: coordinate alignment is performed on the real-time positioning data of the UAV and the grid unit of the river depth to obtain alignment coordinate data; Step S163: performing coordinate co-reference data fusion on the real-time positioning data of the UAV and the three-dimensional water depth data of the river channel based on the aligned coordinate data to generate corresponding fused three-dimensional data; Step S164: reconstructing a three-dimensional model of the corresponding fused three-dimensional data to obtain a three-dimensional model of the river channel; Step S165: Perform two-dimensional cross-section cutting on the three-dimensional model of the river channel to generate a cross-sectional diagram of the river channel water body.

4. The multi-factor UAV flow measurement method according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: performing reflection radar extraction on the designated flow measurement river channel to obtain reflection radar signals; Step S22: performing pulse compression on the reflected radar signal to generate a high-quality reflected radar signal, specifically, improving the quality of the signal by eliminating multipath interference and noise to generate a high-quality reflected radar signal; Step S23: performing Doppler frequency shift estimation on the high-quality reflected signal to generate signal frequency shift data; Step S24: Calculate the water flow velocity of the high-quality reflected radar signal according to the signal frequency shift data to obtain radar-detected water flow velocity data.

5. The multi-factor UAV flow measurement method according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: collecting water body images of a designated flow measurement river channel to obtain a river channel water body image set; Step S32: performing multi-temporal image registration on the river water body image set to generate an aligned water body image; Step S33: marking key points of the aligned water body image to generate key points of the water body image; Step S34: performing sparse optical flow estimation on key points of the water body image to obtain estimated flow velocity data; Step S35: performing dense optical flow calculation on the aligned water body image set to obtain image pixel motion vectors; Step S36: Perform velocity vector mapping on the estimated velocity data based on the image pixel motion vector to generate image detected water body velocity data.

6. The multi-factor UAV flow measurement method according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing a difference comparison between the water body flow velocity data detected by radar and the water body flow velocity data detected by image to obtain detection difference data; Step S42: performing fusion water velocity measurement calculation on the reflected radar signal and the river water image set based on the detection difference data to generate fusion measured water velocity data; Step S43: performing a standard value search on the radar detected water body flow velocity data and the image detected water body flow velocity data according to the fused measured water body flow velocity data, to obtain standard radar detection data and standard image detection data; Step S44: performing measurement process inversion on the standard radar detection data and the standard image detection data to obtain standard radar detection process data and standard image detection process data.

7. The multi-factor UAV flow measurement method according to claim 6, characterized in that: The calculation of water velocity measurement based on the reflected radar signal and the river water image set based on the detection difference data includes: Performing spatiotemporal feature recognition on the detection difference data to obtain spatiotemporal feature data of speed difference; Based on the spatiotemporal characteristic data of velocity difference, feature matching is performed on the reflected radar signal and the river water image set to generate a difference corresponding feature matrix; According to the difference corresponding feature matrix, the reflected radar signal and the river water body image set are collaboratively de-differenced and fused to obtain collaborative fusion data; Reconstruct the water flow field of the collaborative fusion data to generate a simulated water flow field; The water velocity is estimated for the simulated water flow field to obtain fused measured water velocity data.

8. The multi-factor UAV flow measurement method according to claim 1, characterized in that: Step S5 includes the following steps: Step S51: extracting the standard height of the standard radar detection process data to obtain radar detection standard height data; Step S52: extracting the standard height of the standard image detection process data to obtain the image detection standard height data; Step S53: Repeat the flight altitude recognition for the radar detection standard altitude data and the image detection standard altitude data to generate flow measurement standard altitude data; Step S54: extracting standard coordinates from the standard radar detection process data and the standard image detection process data to obtain radar monitoring standard coordinate data and image monitoring standard coordinate data; Step S55: performing intersection coordinate positioning on the radar monitoring standard coordinate data and the image monitoring standard coordinate data to obtain the flow measurement standard coordinate data.

9. The multi-factor UAV flow measurement method according to claim 1, characterized in that: Step S6 includes the following steps: Step S61: performing turbulence characteristic analysis on the fused measured water velocity data to obtain water velocity field data; Step S62: performing local wind shear calculation on the water body velocity field data to generate water body surface wind shear intensity distribution data; Step S63: defining a safe altitude for the wind shear intensity distribution data on the water surface based on a preset UAV wind tolerance threshold, and obtaining safe flight altitude data; Step S64: constructing a closed-loop control model for the flow measurement standard height data, the flow measurement standard coordinate data and the safe flight height data to generate a closed-loop control model; Step S66: dynamically controlling the flight path of the flow measuring UAV using the closed-loop control model, thereby achieving a safe UAV control task; Step S67: The river water cross-section diagram and the fused measured water velocity data are instantly transmitted to the ground station, thereby realizing the UAV flow measurement task.

10. A multi-factor UAV flow measurement system, characterized in that: Used to execute the multi-factor UAV flow measurement method according to claim 1, the multi-factor UAV flow measurement system comprises: The cross-section construction module is used to obtain the designated flow measurement river channel, vertical water depth data set, underwater terrain data set and drone real-time positioning data; based on the drone real-time positioning data, the vertical water depth data set and underwater terrain data set are reconstructed by water depth remote sensing to obtain the river water body cross-section model; The radar flow measurement module is used to obtain the reflected radar signal; the water flow velocity is calculated based on the reflected radar signal to obtain the radar detected water flow velocity data; The image flow measurement module is used to obtain a river water body image set; mark key points of the river water body image set to generate water body image key points; map the water body flow velocity of the river water body image set based on the water body image key points to generate image detection water body flow velocity data; The fusion flow measurement module is used to perform water velocity measurement calculation on the reflected radar signal and the river water image set according to the radar detected water velocity data and the image detected water velocity data, and generate fused measured water velocity data; perform standard measurement process inversion on the radar detected water velocity data and the image detected water velocity data according to the fused measured water velocity data, and obtain standard radar detection process data and standard image detection process data; The flow measurement height calculation module is used to extract the intersection height of the standard radar detection process data and the standard image detection process data to generate the flow measurement standard height data; extract the intersection coordinates of the standard radar detection process data and the standard image detection process data to obtain the flow measurement standard coordinate data; The UAV control module is used to define the safe height for the fused measured water flow velocity data to obtain the safe flight height data; to construct a closed-loop control model for the flow measurement standard height data, flow measurement standard coordinate data and safe flight height data, thereby realizing the safe UAV control task; and to transmit the river water cross-section diagram and the fused measured water flow velocity data to the ground station in real time, thereby realizing the UAV flow measurement task.

Images