Unmanned aerial vehicle route planning method and system based on electromagnetic environment data and three-dimensional map

Abstract

The invention provides an unmanned aerial vehicle route planning method and system based on electromagnetic environment data and a three-dimensional map, and the method comprises the steps: obtaining a three-dimensional live-action image of a target region, and constructing a three-dimensional live-action model; positioning an electromagnetic radiation source; obtaining an electromagnetic radiation signal according to the positioning electromagnetic radiation source, analyzing the characteristics of the electromagnetic radiation signal, and matching a corresponding target electromagnetic radiation source model file; importing the target electromagnetic radiation source model file into the three-dimensional live-action model, and constructing a comprehensive model; and obtaining route planning parameters of the unmanned aerial vehicle, and performing route planning in the comprehensive model according to the route planning parameters to obtain a target route. An electromagnetic radiation source is accurately positioned, a three-dimensional comprehensive model containing the position of the electromagnetic radiation source is constructed, an invisible electromagnetic field is visualized in real time, required electromagnetic information is visually and vividly obtained, a detour path can be planned by using a BP neural network, and a route path can be optimized according to the field intensity of the electromagnetic radiation source. And the safety and effectiveness of unmanned aerial vehicle inspection are improved.

Description

Technical Field

[0001] This application relates to the technical fields of UAV route planning and electromagnetic environment, and particularly relates to a UAV route planning method and system based on electromagnetic environment data and a three-dimensional map. Background Art

[0002] The influence of the electromagnetic environment on the route of an unmanned aerial vehicle (UAV) involves multiple key systems such as navigation, communication, and control. The main dangers include: interfering with the navigation system, for example, strong electromagnetic fields (such as near high-voltage lines, radar stations) or man-made interference (GPS spoofers) can cause the UAV to receive incorrect satellite signals; the communication link is interrupted, and the UAV triggers out-of-control protection (such as returning, hovering, or forced landing). If the GPS is simultaneously interfered, it may be completely disconnected; the flight control system fails, and electromagnetic pulses (such as lightning, military electronic warfare equipment) may burn out the flight control chip or cause abnormal data output from sensors (IMU, barometer). The UAV's attitude calculation is incorrect, resulting in violent shaking or rolling. Interfering with the power system, a strong electric field (such as near high-voltage lines) may interfere with the electronic speed controller (ESC) signal, causing abnormal motor speed and power imbalance. Especially for multi-rotor UAVs, it may directly crash.

[0003] Restricted by the existing technical conditions, it is often difficult to accurately grasp the spatial electromagnetic field distribution in complex terrain and electromagnetic environment, and it is difficult to locate electromagnetic sources. Moreover, the electromagnetic environment is an invisible and intangible existence, which further increases the difficulty of exploration.

[0004] Therefore, designing a UAV route planning method based on electromagnetic environment data and a three-dimensional map to visualize the invisible electromagnetic field in real time, obtain the required electromagnetic information intuitively and vividly, and avoid the electromagnetic environment from affecting the inspection route planning of the UAV is an important research content for those skilled in the art. Summary of the Invention

[0005] In view of this, it is necessary to provide a UAV route planning method based on electromagnetic environment data and a three-dimensional map. By accurately locating electromagnetic radiation sources, constructing a three-dimensional comprehensive model containing the positions of electromagnetic radiation sources, visualizing the invisible electromagnetic field in real time, obtaining the required electromagnetic information intuitively and vividly, it can use a BP neural network to plan a detour path and optimize the flight path according to the field strength of the electromagnetic radiation source, improving the safety and effectiveness of UAV inspection.

[0006] This application provides a method and system for unmanned aerial vehicle (UAV) route planning based on electromagnetic environment data and 3D maps, which includes obtaining 3D real-scene images of a target area and constructing a 3D real-scene model; locating electromagnetic radiation sources; acquiring electromagnetic radiation signals based on the located electromagnetic radiation sources, analyzing the characteristics of the electromagnetic radiation signals, and matching corresponding target electromagnetic radiation source model files; importing the target electromagnetic radiation source model files into the 3D real-scene model to construct a comprehensive model; obtaining UAV route planning parameters, and performing route planning in the comprehensive model according to the route planning parameters to obtain a target route. By accurately locating electromagnetic radiation sources, constructing a 3D comprehensive model including the positions of electromagnetic radiation sources, visualizing invisible electromagnetic fields in real time, intuitively and vividly obtaining the required electromagnetic information, it can use a BP neural network to plan a detour path and optimize the flight path according to the field strength of electromagnetic radiation sources, improving the safety and effectiveness of UAV inspections.

[0007] In a first aspect, an embodiment of this application provides a method for UAV route planning based on electromagnetic environment data and 3D maps, and the method includes:

[0008] S1. Obtain 3D real-scene images of a target area and construct a 3D real-scene model;

[0009] S2. Locate electromagnetic radiation sources based on UAV scattering signals;

[0010] S3. Acquire electromagnetic radiation signals according to the located electromagnetic radiation sources, analyze the characteristics of the electromagnetic radiation signals, and match corresponding target electromagnetic radiation source model files;

[0011] S4. Import the target electromagnetic radiation source model files into the 3D real-scene model to construct a comprehensive model;

[0012] S5. Obtain UAV route planning parameters, and perform route planning in the comprehensive model according to the route planning parameters to obtain a target route.

[0013] Optionally, in an implementation manner of the first aspect of the present invention, the S1. Obtain 3D real-scene images of a target area and construct a 3D real-scene model includes:

[0014] S1.1. The UAV takes high-resolution depth images: By equipping the UAV with a high-resolution camera or an oblique photography device, obtain multi-view and high-overlap image data of the target area. The image data needs to include position and attitude information and positioning data, and at the same time collect lidar point cloud data to assist in obtaining depth information;

[0015] S1.2. Image screening and preprocessing: Screen the high-resolution depth images, eliminate unqualified images, and perform correction, denoising, color homogenization, and enhancement processing on the qualified images;

[0016] S1.3, Image Matching and Adjustment Calculation: Use the SIFT algorithm to extract image feature points, establish corresponding relationships of homologous points through multi-view matching, combine ground control point data, and iteratively optimize through bundle adjustment to calculate the exterior orientation elements of each image, and convert the relative coordinates to the geodetic coordinate system;

[0017] S1.4, Densification of Sparse Point Cloud: Generate an initial three-dimensional point cloud based on the matched homologous points; use the multi-view stereo algorithm for pixel-by-pixel matching to generate a high-density point cloud;

[0018] S1.5, TIN Model Construction and Optimization: Convert the dense point cloud into an irregular triangular network, construct an initial model through the Delaunay triangulation algorithm, adjust the density of the triangular network according to the surface curvature, simplify the triangular patches in flat areas, and retain details in complex areas; repair abnormal triangles to ensure the continuity of the triangular network structure;

[0019] S1.6, Texture Mapping and Model Generation: Automatically select the best texture from multi-view images according to the normal direction and spatial position of the TIN patches; perform color homogenization and seam processing on overlapping textures to ensure color consistency; associate the texture with the geometric model to generate a three-dimensional real-scene model with true geographic coordinates and attribute information.

[0020] Optionally, in an implementation manner of the first aspect of the present invention, the S2, positioning the electromagnetic radiation source based on the scattered signal of the unmanned aerial vehicle, includes:

[0021] Capture the weak signal scattered by the unmanned aerial vehicle, and obtain the Doppler shift curve according to the Doppler shift function;

[0022] Extract the Doppler frequency shift information generated by the unmanned aerial vehicle during movement from the scattered signal according to the Doppler shift curve;

[0023] Combine the real-time flight state of the unmanned aerial vehicle, and use the method of multi-source data fusion for reverse analysis to determine the position of the ground electromagnetic radiation source.

[0024] Optionally, in an implementation manner of the first aspect of the present invention, the determining the position of the ground electromagnetic radiation source includes:

[0025] Establish a three-dimensional space coordinate system, and the carrier frequency of the electromagnetic radiation source arriving at the unmanned aerial vehicle at time i is:

[0026]

[0027] The carrier frequency of the scattered signal of the unmanned aerial vehicle arriving at the ground receiver is:

[0028]

[0029] The Doppler frequency shift function of the electromagnetic radiation source obtained at the ground receiver is:

[0030]

[0031] where f g is the frequency of the electromagnetic radiation source, and c0 represents the propagation speed of electromagnetic waves in air. represents the velocity vector of the UAV with respect to time i, and d p-r (i) represents the unit vector from the electromagnetic radiation source to the position of the UAV, and d p-r (i) represents the unit vector from the position of the ground receiver to the position of the UAV;

[0032] To facilitate obtaining the position of the electromagnetic radiation source, the coordinate system is simplified as follows:

[0033] Taking the ground as the reference plane, with point O as the center and the UAV flight path direction as the X-axis, a three-dimensional coordinate system is established. Among them, assuming that the UAV height is a fixed value h and the speed is a fixed value v, the position of the UAV with respect to time i is expressed as p(x pi , y p ), the position of the ground receiver is expressed as r(x ri , y ri ), and the position of the electromagnetic radiation source is expressed as g(x g , y g ). Then, the simplified Doppler frequency shift function of the electromagnetic radiation source at the i-th moment is:

[0034]

[0035] Substitute multiple groups of data into the simplified Doppler frequency shift function to calculate the position g(x g , y g ) of the electromagnetic radiation source.

[0036] Optionally, in an implementation manner of the first aspect of the present invention, in step S3, the electromagnetic radiation signal is obtained according to the located electromagnetic radiation source, the characteristics of the electromagnetic radiation signal are analyzed, and the corresponding target electromagnetic radiation source model file is matched, including:

[0037] Construct different electromagnetic radiation source model files;

[0038] Obtain the electromagnetic radiation signal according to the located electromagnetic radiation source;

[0039] Extract the characteristics of the electromagnetic radiation signal to obtain the electromagnetic fingerprint characteristics;

[0040] Input the electromagnetic fingerprint characteristics into the electromagnetic radiation source recognition model to obtain the type of the electromagnetic radiation source;

[0041] Obtain the target electromagnetic radiation source model file according to the identified type of the electromagnetic radiation source.

[0042] Optionally, in an implementation manner of the first aspect of the present invention, in step S4, importing the target electromagnetic radiation source model file into the three-dimensional real scene model to construct a comprehensive model includes:

[0043] Importing the three-dimensional real scene model and extracting its geometric topology and texture coordinates;

[0044] Obtaining the electric field strength, potential field function, and radiation power density in the target electromagnetic radiation source model file;

[0045] Normalizing the electric field strength, potential field function, and radiation power density into texture pixel values, matching them with the model vertex coordinates, and generating a lookup table for electromagnetic parameters;

[0046] Converting the electromagnetic parameters into texture colors or transparencies through shader programming or software plug-ins,

[0047] And mapping them to the model surface to obtain a comprehensive model with electromagnetic radiation;

[0048] Wherein the mapping methods include dynamic texture overlay and multi-physical field coupling;

[0049] If it is necessary to display the dynamic change of the electromagnetic field, incorporating the time variable into the texture parameters to generate a dynamic texture sequence; if the mapping method is multi-physical field coupling, in the COMSOL modeling software, directly outputting the electromagnetic simulation field distribution result as a texture map and using the Modo ModelFun tool for automatic mapping;

[0050] Verifying the consistency of the mapping, verifying the physical rationality of the electromagnetic texture through electromagnetic field observation data, and ensuring that the radiation field distribution meets the theoretical expectation;

[0051] Combining and saving the geometric topology, texture coordinates, and the lookup table together.

[0052] Optionally, in an implementation manner of the first aspect of the present invention, in step S5, obtaining the UAV route planning parameters, and performing route planning in the comprehensive model according to the route planning parameters to obtain the target route, includes:

[0053] Judging whether the distance between the UAV and the electromagnetic field source is within the threshold of the safe range through multi-sensor fusion and comprehensive model calculation;

[0054] When the UAV is within the electromagnetic field environment range, combining the electromagnetic field strength vector with the UAV position and attitude data, calculating the resultant force direction of the electromagnetic field on the airframe, and optimizing the flight speed according to the real-time field strength distribution;

[0055] When the UAV is outside the electromagnetic field environment range, using a BP neural network to plan a detour path.

[0056] In a second aspect, an embodiment of the present application provides a UAV route planning system based on electromagnetic environment data and a three-dimensional map, which is applied to the UAV route planning method based on electromagnetic environment data and a three-dimensional map as described in the first aspect, and is characterized by including:

[0057] A three-dimensional real-scene model construction module: acquiring three-dimensional real-scene images of a target area and constructing a three-dimensional real-scene model;

[0058] An electromagnetic radiation source positioning module: positioning an electromagnetic radiation source based on the scattered signals of the UAV;

[0059] A model file matching module: acquiring an electromagnetic radiation signal according to the positioned electromagnetic radiation source, analyzing the characteristics of the electromagnetic radiation signal, and matching a corresponding target electromagnetic radiation source model file;

[0060] A comprehensive model construction module: importing the target electromagnetic radiation source model file into the three-dimensional real-scene model to construct a comprehensive model;

[0061] A target route optimization module, acquiring UAV route planning parameters, and performing route planning in the comprehensive model according to the route planning parameters to obtain a target route.

[0062] In a third aspect, an embodiment of the present application provides an electronic device, which is characterized by including:

[0063] A processor;

[0064] A memory for storing executable instructions of the processor;

[0065] Wherein, when the processor is configured to execute the instructions, it implements the UAV route planning method based on electromagnetic environment data and a three-dimensional map as described in the first aspect.

[0066] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which is characterized in that the computer-readable storage medium stores a program, and the program instructs the device to execute the UAV route planning method based on electromagnetic environment data and a three-dimensional map as described in the first aspect.

[0067] This application provides a method and system for unmanned aerial vehicle (UAV) route planning based on electromagnetic environment data and 3D maps. It acquires 3D real-scene images of the target area and constructs a 3D real-scene model; locates electromagnetic radiation sources; obtains electromagnetic radiation signals based on the located electromagnetic radiation sources, analyzes the characteristics of the electromagnetic radiation signals, and matches the corresponding target electromagnetic radiation source model files; imports the target electromagnetic radiation source model files into the 3D real-scene model to construct a comprehensive model; acquires UAV route planning parameters, and performs route planning in the comprehensive model according to the route planning parameters to obtain the target route. By accurately locating electromagnetic radiation sources, constructing a 3D comprehensive model including the positions of electromagnetic radiation sources, visualizing the invisible electromagnetic field in real time, obtaining the desired electromagnetic information intuitively and vividly, it can use the BP neural network to plan detour paths and optimize the flight path according to the field strength of electromagnetic radiation sources, improving the safety and effectiveness of UAV inspections.

[0068] Beneficial effects:

[0069] (1) Based on the Doppler frequency shift function, accurately locate electromagnetic radiation sources, which can improve the accuracy and precision of UAV route planning based on electromagnetic environment data and 3D maps.

[0070] (2) With the help of 3D visualization technology, visualize the invisible electromagnetic field in real time, and obtain the desired electromagnetic information intuitively and vividly.

[0071] (3) Construct a 3D comprehensive model including the positions of electromagnetic radiation sources, which can use the BP neural network to plan detour paths and optimize the flight path according to the field strength of electromagnetic radiation sources, improving the safety and effectiveness of UAV inspections. Description of the drawings

[0072] Figure 1 It is a schematic flowchart of the UAV route planning method based on electromagnetic environment data and 3D maps provided by an embodiment of this application.

[0073] Figure 2 It is a flowchart of constructing a 3D real-scene model provided by an embodiment of this application.

[0074] Figure 3 It is a schematic diagram of coordinates provided by an embodiment of this application.

[0075] Figure 4 It is an electromagnetic field model diagram provided by an embodiment of this application.

[0076] Figure 5 (a)-5(b) It is a schematic diagram of resultant force and resultant velocity provided by an embodiment of this application.

[0077] Figure 6 It is a schematic diagram of the module of the UAV route planning system based on electromagnetic environment data and 3D maps provided by an embodiment of this application.

[0078] Figure 7 Schematic diagram of an electronic device provided by an embodiment of the present application. Specific implementation manners

[0079] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments.

[0080] It should be noted that in the embodiments of the present application, "at least one" means one or more, and "a plurality" means two or more. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present application belongs. The terms used in the description of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application.

[0081] It should be noted that in the embodiments of the present application, terms such as "first" and "second" are only used for the purpose of distinguishing descriptions, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying order. The features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the embodiments of the present application, words such as "exemplary" or "for example" are used to represent examples, illustrations or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly, the use of words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner.

[0082] All other implementation manners obtained by those of ordinary skill in the art based on the implementation manners in the present application without creative efforts belong to the scope protected by the present application.

[0083] Embodiment 1

[0084] The present application provides a method and system for unmanned aerial vehicle (UAV) route planning based on electromagnetic environment data and 3D maps. It acquires 3D real-scene images of a target area and constructs a 3D real-scene model; locates electromagnetic radiation sources based on the scattered signals of the UAV; obtains electromagnetic radiation signals according to the located electromagnetic radiation sources, analyzes the characteristics of the electromagnetic radiation signals, and matches corresponding target electromagnetic radiation source model files; imports the target electromagnetic radiation source model files into the 3D real-scene model to construct a comprehensive model; obtains UAV route planning parameters, and performs route planning in the comprehensive model according to the route planning parameters to obtain a target route. By accurately locating electromagnetic radiation sources and constructing a 3D comprehensive model containing the positions of electromagnetic radiation sources, it can use a BP neural network to plan a detour path and optimize the flight path according to the field strength of electromagnetic radiation sources, improving the safety and effectiveness of UAV patrols.

[0085] Figure 1 It is a schematic flowchart of a method for UAV route planning based on electromagnetic environment data and 3D maps provided by an embodiment of the present application.

[0086] As Figure 1 shown, a method for UAV route planning based on electromagnetic environment data and 3D maps includes:

[0087] S1. Acquire 3D real-scene images of a target area and construct a 3D real-scene model.

[0088] Figure 2 It is a flowchart for constructing a 3D real-scene model provided by an embodiment of the present application. As Figure 2 shown, it can be understood that in this embodiment, the step S1. Acquire 3D real-scene images of a target area and construct a 3D real-scene model includes:

[0089] S1.1. The UAV takes high-resolution depth images: By equipping the UAV with a high-resolution camera or an oblique photography device, acquire multi-view and high-overlap image data of the target area. The image data needs to include position and attitude information and positioning data, and simultaneously collect lidar point cloud data to assist in obtaining depth information;

[0090] S1.2. Image screening and preprocessing: Screen the high-resolution depth images, eliminate unqualified images, and perform correction, denoising, color homogenization, and enhancement on the qualified images;

[0091] S1.3. Image matching and bundle adjustment calculation: Use the SIFT algorithm to extract image feature points, establish corresponding relationships of homologous points through multi-view matching, combine ground control point data, and iteratively optimize through bundle adjustment to calculate the exterior orientation elements of each image, and convert the relative coordinates to the geodetic coordinate system;

[0092] S1.4, Sparse point cloud densification: Generate an initial 3D point cloud based on matched homologous points; use a multi-view stereo algorithm for pixel-by-pixel matching to generate a high-density point cloud;

[0093] S1.5, TIN model construction and optimization: Convert the dense point cloud into an irregular triangular network, construct an initial model through the Delaunay triangulation algorithm, adjust the density of the triangular network according to the surface curvature, simplify the triangular patches in flat areas, and retain details in complex areas; repair abnormal triangles to ensure the continuity of the triangular network structure;

[0094] S1.6, Texture mapping and model generation: Automatically select the best texture from multi-view images according to the normal direction and spatial position of the TIN patches; perform color homogenization and seam processing on overlapping textures to ensure color consistency; associate the texture with the geometric model to generate a 3D real scene model with true geographic coordinates and attribute information.

[0095] Specifically, use an unmanned aerial vehicle (UAV) equipped with a high-resolution camera or an oblique photography device to obtain multi-view and high-overlap image data of the target area. The image data needs to contain position and attitude information (such as POS data), and in some cases, GNSS positioning (such as RTK technology) is combined to improve the accuracy. Some solutions will also collect lidar point cloud data to assist in obtaining depth information.

[0096] Specifically, screen the original images, and eliminate blurred, occluded, or abnormally illuminated images. The qualified images need to be processed as follows: Geometric correction: Eliminate lens distortion (such as radial distortion, tangential distortion). Denoising and color homogenization: Reduce image noise through algorithms and unify the tone differences under different lighting conditions. Enhancement processing: Improve the image contrast and detail clarity for subsequent feature extraction.

[0097] Specifically, image matching and adjustment calculation: Connection point extraction and matching: Use algorithms such as SIFT and SURF to extract image feature points, and establish the corresponding relationship of homologous points through multi-view matching. Block adjustment: Combine ground control points (GCPs) or POS data, and iteratively optimize through bundle adjustment to calculate the exterior orientation elements (i.e., camera position and attitude parameters) of each image, and convert the relative coordinates to the geodetic coordinate system (such as CGCS2000). Accuracy verification: The planar and elevation residuals after adjustment need to be controlled at the millimeter level (such as 0.12 cm planar residual) to ensure the model accuracy.

[0098] Specifically, sparse point cloud densification includes sparse point cloud generation: Generate an initial 3D point cloud based on matched homologous points. Dense matching: Use a multi-view stereo algorithm (such as PMVS, MVS) for pixel-by-pixel matching to generate a high-density point cloud. Some technologies introduce parallel computing or consistency constraints (such as multi-frame depth map filtering) to improve efficiency and accuracy. Data block processing: Cut and block large-scale point clouds to adapt to the limitations of computing resources.

[0099] Specifically, the construction and optimization of the TIN model include triangulation network generation: converting dense point clouds into irregular triangulation networks (TINs) and constructing an initial model through algorithms such as Delaunay triangulation. Smoothing and optimization: Geometric optimization: Adjusting the density of the triangulation network according to the surface curvature, simplifying triangular patches in flat areas, and retaining details in complex areas. Topological optimization: Repairing abnormal triangles (such as long and narrow triangles) to ensure the continuity of the triangulation network structure. Blocking and LOD construction: Processing large-scale TIN models in blocks and generating multiple levels of detail (LOD) to improve rendering efficiency. Texture mapping and model generation include texture matching: Automatically selecting the best texture (such as the minimum viewing angle or the highest resolution image) from multi-view images according to the normal direction and spatial position of the TIN patches. Texture fusion: Equalizing colors and processing seams for overlapping textures to ensure color consistency. Semantic assignment: Associating textures with geometric models to generate a three-dimensional real-scene model with real geographic coordinates and attribute information.

[0100] S2. Locating the electromagnetic radiation source based on the scattering signal of the unmanned aerial vehicle.

[0101] It can be understood that in this embodiment, the S2, locating the electromagnetic radiation source based on the scattering signal of the unmanned aerial vehicle, includes:

[0102] Capturing the weak signal scattered by the unmanned aerial vehicle and obtaining the Doppler shift curve according to the Doppler shift function;

[0103] Extracting the Doppler frequency shift information generated by the unmanned aerial vehicle during movement from the scattered signal according to the Doppler shift curve;

[0104] Combining the real-time flight state of the unmanned aerial vehicle and using the method of multi-source data fusion for reverse analysis to determine the position of the ground electromagnetic radiation source.

[0105] Specifically, the signal capture technology mainly uses a wideband receiver and a high-sensitivity RF front end, combined with an adaptive filtering algorithm to suppress environmental noise. For low signal-to-noise ratio signals, cyclic stationary analysis can be applied to extract periodic features, or the pseudo-Doppler principle can be used to enhance the detectability of the signal.

[0106] Doppler frequency shift extraction includes time-frequency analysis: Generating a time-frequency spectrogram through short-time Fourier transform (STFT) or wavelet transform to identify the frequency shift curves in the approaching, crossing the baseline, and departing stages of the unmanned aerial vehicle (such as the amplitude / spectrum mutation in Figures (a) and (b)).

[0107] Specifically, the determining of the position of the ground electromagnetic radiation source includes:

[0108] Establishing a three-dimensional space coordinate system, the carrier frequency of the electromagnetic radiation source arriving at the unmanned aerial vehicle at time i is:

[0109]

[0110] The carrier frequency of the UAV scattered to reach the ground receiver is:

[0111]

[0112] The Doppler frequency shift function of the electromagnetic radiation source obtained at the ground receiver is:

[0113]

[0114] Among them, f g is the frequency of the electromagnetic radiation source, c0 represents the propagation speed of electromagnetic waves in the air, represents the velocity vector of the UAV with respect to time i, d p-r (i) represents the unit vector of the electromagnetic radiation source pointing to the UAV position, d p-r (i) represents the unit vector of the ground receiver position pointing to the UAV position;

[0115] In order to facilitate obtaining the position of the electromagnetic radiation source, the coordinate system is simplified. Figure 3 This is the coordinate schematic diagram provided by an embodiment of this application. As Figure 3 shown, specifically:

[0116] Taking the ground as the reference plane, with point O as the center and the UAV flight path direction as the X-axis, a three-dimensional coordinate system is established. Among them, assuming that the UAV height is a fixed value h and the speed is a fixed value v, then the position of the UAV with respect to time i is expressed as p(x pi , y p ), the position of the ground receiver is expressed as r(x ri , y ri ), and the position of the electromagnetic radiation source is expressed as g(x g , y g ). Then the simplified Doppler frequency shift function of the electromagnetic radiation source at the i-th moment is:

[0117]

[0118] Using multiple groups of data to substitute into the simplified Doppler frequency shift function, the position g(x g , y g ) of the electromagnetic radiation source is calculated.

[0119] S3. According to the located electromagnetic radiation source, obtain the electromagnetic radiation signal, analyze the characteristics of the electromagnetic radiation signal, and match the corresponding target electromagnetic radiation source model file.

[0120] It can be understood that in this embodiment, step S3, according to the located electromagnetic radiation source, obtains an electromagnetic radiation signal, analyzes the characteristics of the electromagnetic radiation signal, and matches the corresponding target electromagnetic radiation source model file, including:

[0121] Construct different electromagnetic radiation source model files;

[0122] According to the located electromagnetic radiation source, obtain an electromagnetic radiation signal;

[0123] Extract features from the electromagnetic radiation signal to obtain electromagnetic fingerprint features;

[0124] Input the electromagnetic fingerprint features into an electromagnetic radiation source recognition model to obtain the type of electromagnetic radiation source;

[0125] Obtain the target electromagnetic radiation source model file according to the recognized type of electromagnetic radiation source.

[0126] Specifically, Figure 4 is the electromagnetic field model diagram provided by an embodiment of the present application. As Figure 4 shown, it is a schematic diagram of a three-dimensional electromagnetic field model under ideal conditions. The model is saved and generated as a data file composed of spatial discrete points.

[0127] S4, import the target electromagnetic radiation source model file into the three-dimensional real-scene model to construct a comprehensive model.

[0128] It can be understood that in this embodiment, step S4, importing the target electromagnetic radiation source model file into the three-dimensional real-scene model to construct a comprehensive model, includes:

[0129] Import the three-dimensional real-scene model and extract its geometric topology and texture coordinates;

[0130] Obtain the electric field strength, potential field function, and radiation power density in the target electromagnetic radiation source model file;

[0131] Normalize the electric field strength, potential field function, and radiation power density into texture pixel values, match them with the model vertex coordinates, and generate a lookup table for electromagnetic parameters;

[0132] Through shader programming or software plug-ins, convert the electromagnetic parameters into texture colors or transparencies and map them to the model surface to obtain a comprehensive model with electromagnetic radiation.

[0133] Specifically, when importing a 3D real - scene model and extracting geometric topology and texture coordinates, it is necessary to ensure that the 3D model meets the requirements of a unified coordinate system and elevation accuracy. The model's axis center point is defined as the center point of the bottom surface, and the format must conform to the specification (such as the OSGB format). Use 3D modeling software (such as MultiGen Creator) to import the model, and extract the geometric topology structure and texture coordinates through the processes of dense matching and texture mapping, ensuring that the model structure is complete and there are no problems such as coplanarity and missing seams. The texture should be consistent with the actual landscape, with a resolution of not less than 0.03 m / pixel, a single texture surface not exceeding 1024 pixels, and a color depth of 8 bit.

[0134] When obtaining the target electromagnetic radiation source model parameters, extract parameters such as electric field strength, potential field function, and radiation power density from the radiation source model. These parameters are based on the mathematical model of the radiation source, including characteristics such as position, frequency, polarization mode, and direction function. In the case of a multi - radiation - source scenario, it is necessary to calculate the spatial combined field strength, for example, through the radar radiation source detection range model and the field strength synthesis algorithm. Normalize the parameters and generate an electromagnetic parameter look - up table, normalizing parameters such as electric field strength to texture pixel values from 0 to 255. Referring to the method of quantifying radiation intensity as gray values in infrared scene simulation, use the radiance calculation formula to achieve parameter standardization. Match the normalized data according to the model vertex coordinates to generate an electromagnetic parameter look - up table (LUT). Similar to the process of calculating the electromagnetic radiation spatial field by layer at grid points in the LRBF method, ensure the precise correspondence between the data and the geometric structure. It also includes electromagnetic parameter visualization mapping. Using shader programming or software plugins, through the shaders of graphics libraries such as OpenGL, convert the electromagnetic parameters in the look - up table into RGB colors or transparency values. For example, use interval mapping combined with color rendering capabilities to display the field strength distribution. It is also possible to use CST electromagnetic simulation software for radiation field analysis, directly mapping parameters such as power density and radiation pattern to the model surface, or achieving dynamic rendering by reading the material file through Vega's SensorVision module.

[0135] Specifically, for technical verification and optimization, a verification model can be used. It is necessary to consider the relationship between geometric dimensions and wavelength to avoid a sharp increase in the amount of calculation due to an overly large simulation domain. The model can be simplified using a symmetry plane, and the ray - tracing method can be used to verify the accuracy of the field strength distribution. In complex scenarios, it is necessary to combine propagation loss calculations (such as terrain effects) and received power synthesis algorithms to ensure that the spatio - temporal superposition of electromagnetic environment signals conforms to the actual situation.

[0136] Among them, the mapping methods include dynamic texture overlay and multi - physical - field coupling;

[0137] If it is necessary to display the dynamic changes of the electromagnetic field, the time variable is incorporated into the texture parameters to generate a dynamic texture sequence; if the mapping method is multi-physics field coupling, in the COMSOL modeling software, the electromagnetic simulation field distribution results are directly output as texture maps, and the automatic mapping is carried out using the ModelFun tool of ModFang;

[0138] Verify the consistency of the mapping, and verify the physical rationality of the electromagnetic texture through the electromagnetic field observation data to ensure that the radiation field distribution meets the theoretical expectations;

[0139] Merge the geometric topology, texture coordinates and look-up table and save them together. The saved file can be directly opened to generate the visualization of the three-dimensional model carrying the electromagnetic field when used next time.

[0140] S5. Obtain the UAV route planning parameters, and perform route planning in the comprehensive model according to the route planning parameters to obtain the target route.

[0141] It can be understood that in this embodiment, the step S5, obtaining the UAV route planning parameters, and performing route planning in the comprehensive model according to the route planning parameters to obtain the target route, includes:

[0142] Judge whether the distance between the UAV and the electromagnetic field source is within the threshold of the safe range through multi-sensor fusion and comprehensive model calculation;

[0143] When the UAV is within the electromagnetic field environment range, combine the electromagnetic field strength vector with the UAV position and attitude data, calculate the resultant force direction of the electromagnetic field on the airframe, and optimize the flight speed according to the real-time field strength distribution;

[0144] When the UAV is outside the electromagnetic field environment range, use the BP neural network to plan the bypass path.

[0145] In a complex electromagnetic environment, the sensors and communication links of the UAV may be interfered, and more robust algorithms (such as APSO-BP or DRNN) are required. However, outside the electromagnetic field range, the path planning in the conventional environment has relatively low requirements for real-time performance, and the performance of the BP neural network can meet the requirements at this time.

[0146] The BP neural network inputs the linear motion parameters (such as displacement, angular velocity, etc.) of the drone and external environment information (such as wind direction), and outputs the 4D flight path prediction results. Its three-layer structure (input layer, hidden layer, output layer) can effectively handle non-linear mapping problems, providing a theoretical basis for the detour path planning. A typical BP neural network adopts a three-layer structure of input layer, hidden layer (4 - 13 nodes), and output layer. The input parameters include the motion state of the drone and environmental data, and the output is the displacement and angle of the detour path. To improve efficiency, the BP neural network usually performs offline training first (such as parameter pre-training for the control system of plant protection drones), and then combines real-time data for online adjustment to deal with dynamic obstacles. In the third stage of obstacle avoidance (path planning stage), the BP neural network can be combined with terrain data and GNSS signals to generate the optimal path around static obstacles (such as cranes, bridges).

[0147] Specifically, if the drone is within the electromagnetic field safety threshold, it is necessary to optimize the flight speed to avoid electromagnetic interference. Combining the electric field intensity vector (E) and the magnetic field intensity vector (B), the resultant force direction of the electromagnetic field on the aircraft is calculated through vector superposition. It is also possible to fuse the drone attitude data (estimated by the magnetometer, accelerometer, and gyroscope through gradient descent method or Kalman filter) to correct the resultant force direction in the body coordinate system.

[0148] Specifically, Figure 5 (a)-5(b) are schematic diagrams of the resultant force and resultant velocity provided by an embodiment of the present application. If there are dynamic obstacles, their velocity vectors will affect the calculation of the repulsive force (such as the parallel repulsive force component), thereby indirectly adjusting the movement direction of the drone. As Figure 5 shown, according to the gravitational or repulsive force of the electromagnetic field and the target power of the drone, the total force direction is obtained according to the vector synthesis of forces. Since the total resultant force determines the acceleration direction of the drone, the velocity direction gradually approaches the resultant force direction. According to the velocity of the electromagnetic field and the velocity of the drone, the total velocity direction is obtained according to the vector synthesis of velocities.

[0149] Embodiment 2

[0150] As Figure 6 shown, the present application provides a drone route planning system based on electromagnetic environment data and three-dimensional maps, which is applied to the drone route planning method based on electromagnetic environment data and three-dimensional maps as described in Embodiment 1, including: a three-dimensional real scene model construction module 11, an electromagnetic radiation source positioning module 12, a model file matching module 13, a comprehensive model construction module 14, and a target route optimization module 15.

[0151] It can be understood that in this embodiment, the three-dimensional real scene model construction module 11: obtains the three-dimensional real scene images of the target area and constructs a three-dimensional real scene model.

[0152] It can be understood that in this embodiment, the electromagnetic radiation source positioning module 12 locates the electromagnetic radiation source based on the scattering signals of the UAV.

[0153] It can be understood that in this embodiment, the model file matching module 13 obtains electromagnetic radiation signals according to the located electromagnetic radiation source, analyzes the characteristics of the electromagnetic radiation signals, and matches the corresponding target electromagnetic radiation source model files.

[0154] It can be understood that in this embodiment, the comprehensive model construction module 14 imports the target electromagnetic radiation source model file into the three-dimensional real scene model to construct a comprehensive model.

[0155] It can be understood that in this embodiment, the target flight path optimization module 15 obtains the UAV flight path planning parameters, and performs flight path planning in the comprehensive model according to the flight path planning parameters to obtain the target flight path.

[0156] This application provides a UAV flight path planning method and system based on electromagnetic environment data and three-dimensional maps, which obtains three-dimensional real scene images of a target area and constructs a three-dimensional real scene model; locates electromagnetic radiation sources; obtains electromagnetic radiation signals according to the located electromagnetic radiation sources, analyzes the characteristics of the electromagnetic radiation signals, and matches the corresponding target electromagnetic radiation source model files; imports the target electromagnetic radiation source model files into the three-dimensional real scene model to construct a comprehensive model; obtains the UAV flight path planning parameters, and performs flight path planning in the comprehensive model according to the flight path planning parameters to obtain the target flight path. By accurately locating electromagnetic radiation sources, constructing a three-dimensional comprehensive model including the positions of electromagnetic radiation sources, visualizing invisible electromagnetic fields in real time, intuitively and vividly obtaining the required electromagnetic information, it can use a BP neural network to plan a detour path and optimize the flight path according to the field strength of the electromagnetic radiation source, improving the safety and effectiveness of UAV inspections.

[0157] Figure 7 This is an electronic device provided by an embodiment of this application. As Figure 7 shown, the electronic device at least includes the following parts: a processor 101, a memory 100, a communication interface 103, and a bus 102.

[0158] In an embodiment of this application, the memory 100 is used to store executable instructions of the processor 101, and the processor 101 is configured to implement the device modules for UAV flight path planning based on electromagnetic environment data and three-dimensional maps as Figure 6 shown when executing the instructions.

[0159] In an embodiment of this application, a computer-readable storage medium includes instructions that direct a device to execute the method in the first aspect. For example, the instructions direct the device to execute the method shown in the process steps in Figure 1 .

[0160] The program operating in the electronic device according to an embodiment of the present application may be a program for controlling a central processing unit (CPU) or the like to implement the functions of the above-described embodiments related to a solution of the present invention (a program for causing a computer to function). Then, the information processed by these devices is temporarily stored in a random access memory (RAM) during its processing, and then stored in various ROMs such as a read-only memory (FlashROM), a hard disk drive (HDD), etc., and read, corrected, and written by the CPU as needed.

[0161] It should be noted that a part of the electronic device of the above-described embodiment can also be implemented by a computer. In this case, the program for implementing the control function can be recorded on a computer-readable recording medium, and is implemented by reading the program recorded on the recording medium into the computer and executing it.

[0162] It should be noted that the "computer" mentioned here refers to a computer built in an electronic device, which is a computer including hardware such as an OS and peripheral devices. In addition, the "computer-readable recording medium" refers to a removable medium such as a floppy disk, a magneto-optical disk, a ROM, a CD-ROM, or a storage device such as a hard disk built in a computer.

[0163] Moreover, the "computer-readable recording medium" may include: a medium that dynamically stores a program for a short time, such as a communication line in the case of transmitting a program via a network such as the Internet or a communication line such as a telephone line; a medium that stores a program for a fixed time, such as a volatile memory inside a computer of a server or a client in this case. In addition, the above program may be a program for implementing a part of the above functions, and may also be a program that can implement the above functions by combining with a program already recorded in a computer.

[0164] In addition, the electronic device in the above-described embodiment can also be implemented as an aggregate (device group) composed of multiple devices. Each device constituting the device group may have all or a part of the functions or function blocks of the electronic device of the above-described embodiment. As the device group, it is sufficient to have all the functions or function blocks of the electronic device.

[0165] Those of ordinary skill in the art of the present technology should recognize that the above embodiments are only used to illustrate the present application, rather than to limit the present application. As long as it is within the scope of the spirit of the present application, appropriate changes and variations made to the above embodiments fall within the scope of protection required by the present application.

Claims

1. A method for unmanned aerial vehicle route planning based on electromagnetic environment data and 3D maps, characterized in that The method includes: S1. Obtain a three-dimensional real-scene image of the target area and construct a three-dimensional real-scene model; S2. Locate the electromagnetic radiation source based on the scattered signal of the drone; S3. Obtain the electromagnetic radiation signal according to the located electromagnetic radiation source, analyze the characteristics of the electromagnetic radiation signal, and match the corresponding target electromagnetic radiation source model file; S4. Import the target electromagnetic radiation source model file into the three-dimensional real-scene model to construct a comprehensive model; S5. Obtain the drone route planning parameters, and perform route planning in the comprehensive model according to the route planning parameters to obtain the target route.

2. The method for planning a UAV flight path based on electromagnetic environment data and a 3D map according to claim 1, wherein, The S1, obtain a three-dimensional real-scene image of the target area and construct a three-dimensional real-scene model, includes: S1.

1. The drone captures high-resolution depth images: By carrying a high-resolution camera or oblique photography equipment on the drone, obtain multi-view and high-overlap image data of the target area. The image data needs to include position and attitude information and positioning data. At the same time, collect lidar point cloud data to assist in obtaining depth information; S1.

2. Image screening and preprocessing: Screen the high-resolution depth images, eliminate unqualified images, and perform correction, denoising, color homogenization, and enhancement processing on the qualified images; S1.

3. Image matching and bundle adjustment calculation: Use the SIFT algorithm to extract image feature points, establish corresponding relationships of homologous points through multi-view matching, combine ground control point data, and perform iterative optimization through bundle adjustment to calculate the exterior orientation elements of each image, and convert the relative coordinates to the geodetic coordinate system; S1.

4. Densification of sparse point clouds: Generate an initial three-dimensional point cloud based on the matched homologous points; use the multi-view stereo algorithm to perform pixel-by-pixel matching to generate a high-density point cloud; S1.

5. TIN model construction and optimization: Convert the dense point cloud into an irregular triangular network, construct an initial model through the Delaunay triangulation algorithm, adjust the density of the triangular network according to the surface curvature, simplify the triangular patches in flat areas, and retain details in complex areas; repair abnormal triangles to ensure the continuity of the triangular network structure; S1.

6. Texture mapping and model generation: Automatically select the best texture from multi-view images according to the normal direction and spatial position of the TIN patches; perform color homogenization and seam processing on the overlapping textures to ensure color consistency; associate the texture with the geometric model to generate a three-dimensional real-scene model with real geographic coordinates and attribute information.

3. The method for unmanned aerial vehicle route planning based on electromagnetic environment data and 3D map according to claim 2, characterized in that, The S2, locate the electromagnetic radiation source based on the scattered signal of the drone, includes: Capture the weak signal scattered by the drone, and obtain the Doppler shift curve according to the Doppler shift function; Extract the Doppler frequency shift information generated by the drone during movement from the scattered signal according to the Doppler shift curve; Combined with the real-time flight state of the drone, use the method of multi-source data fusion to perform reverse analysis to determine the position of the ground electromagnetic radiation source.

4. The method for planning an unmanned aerial vehicle flight path based on electromagnetic environment data and a three-dimensional map according to claim 3, wherein, The determining the position of the ground electromagnetic radiation source includes: Establish a three-dimensional space coordinate system. The carrier frequency of the electromagnetic radiation source reaching the drone at time i is: The carrier frequency of the drone scattered to reach the ground receiver is: The Doppler frequency shift function of the electromagnetic radiation source obtained at the ground receiver is: where f g is the frequency of the electromagnetic radiation source, c0 represents the propagation speed of electromagnetic waves in air, denotes the velocity vector of the UAV with respect to time i, d p-r (i) represents the unit vector from the electromagnetic radiation source to the position of the UAV, d p-r (i) represents the unit vector from the position of the ground receiver to the position of the UAV; To facilitate obtaining the position of the electromagnetic radiation source, the coordinate system is simplified as follows: Taking the ground as the reference plane and point O as the center, a three-dimensional coordinate system is established with the UAV flight path direction as the X-axis. Among them, assuming the UAV height is a fixed value h and the speed is a fixed value v, the position of the UAV with respect to time i is expressed as p(x pi , y p ), the position of the ground receiver is expressed as r(x ri , y ri ), and the position of the electromagnetic radiation source is expressed as g(x g , y g ). Then the simplified Doppler frequency shift function of the electromagnetic radiation source at the i-th moment is: Substitute multiple groups of data into the simplified Doppler shift function to calculate the position \(g(x\) g , y g ) of the electromagnetic radiation source.

5. The drone route planning method based on electromagnetic environment data and 3D maps according to claim 4, characterized in that The S3 obtains electromagnetic radiation signals according to the positioned electromagnetic radiation source, analyzes the characteristics of the electromagnetic radiation signals, and matches the corresponding target electromagnetic radiation source model files, including: Construct different electromagnetic radiation source model files; Obtain electromagnetic radiation signals according to the positioned electromagnetic radiation source; Extract features from the electromagnetic radiation signals to obtain electromagnetic fingerprint features; Input the electromagnetic fingerprint features into the electromagnetic radiation source recognition model to obtain the type of electromagnetic radiation source; Obtain the target electromagnetic radiation source model file according to the recognized type of electromagnetic radiation source.

6. A method for unmanned aerial vehicle route planning based on electromagnetic environment data and three-dimensional maps according to claim 5, characterized in that, The S4 imports the target electromagnetic radiation source model file into the 3D real-scene model to construct a comprehensive model, including: Import the 3D real-scene model and extract its geometric topology and texture coordinates; Obtain the electric field strength, potential field function, and radiation power density in the target electromagnetic radiation source model file; Normalize the electric field strength, potential field function, and radiation power density into texture pixel values, match them with the model vertex coordinates, and generate a lookup table of electromagnetic parameters; Through shader programming or software plug-ins, convert the electromagnetic parameters into texture colors or transparencies and map them to the model surface to obtain a comprehensive model with electromagnetic radiation; The mapping methods include dynamic texture overlay and multi-physics field coupling; If it is necessary to display the dynamic changes of the electromagnetic field, incorporate the time variable into the texture parameters to generate a dynamic texture sequence; if the mapping method is multi-physics field coupling, in the COMSOL modeling software, directly output the electromagnetic simulation field distribution result as a texture map, and use the ModModelFun tool for automatic mapping; Verify the consistency of the mapping, verify the physical rationality of the electromagnetic texture through electromagnetic field observation data, and ensure that the radiation field distribution meets the theoretical expectations; Save the geometric topology and texture coordinates together with the lookup table.

7. The drone route planning method based on electromagnetic environment data and 3D maps according to claim 6, wherein The S5 obtains the UAV route planning parameters, and performs route planning in the comprehensive model according to the route planning parameters to obtain the target route, including: Through multi-sensor fusion and comprehensive model calculation, judge whether the distance between the UAV and the electromagnetic field source is within the threshold of the safe range; When the UAV is within the electromagnetic field environment range, combine the electromagnetic field strength vector with the UAV position and attitude data, calculate the resultant force direction of the electromagnetic field on the airframe, and optimize the flight speed according to the real-time field strength distribution; When the UAV is outside the electromagnetic field environment range, use the BP neural network to plan a detour path.

8. A UAV route planning system based on electromagnetic environment data and 3D maps, which is applied to the UAV route planning method based on electromagnetic environment data and 3D maps according to any one of claims 1 to 7, and is characterized in that, Including: 3D real-scene model construction module: Obtain 3D real-scene images of the target area and construct a 3D real-scene model; Electromagnetic radiation source positioning module: Locate the electromagnetic radiation source based on the UAV scattering signal; Model file matching module: Obtain electromagnetic radiation signals according to the positioned electromagnetic radiation source, analyze the characteristics of the electromagnetic radiation signals, and match the corresponding target electromagnetic radiation source model files; Comprehensive model construction module: Import the target electromagnetic radiation source model file into the 3D real-scene model to construct a comprehensive model; The target route optimization module obtains the UAV route planning parameters, and performs route planning in the comprehensive model according to the route planning parameters to obtain the target route.

9. An electronic device, characterized in that, It includes: A processor; A memory for storing instructions executable by the processor; Wherein, when the processor is configured to execute the instructions, it implements the UAV route planning method based on electromagnetic environment data and 3D map according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program, and the program instructs the device to execute the UAV route planning method based on electromagnetic environment data and 3D map according to any one of claims 1 to 7.

Images