Unmanned aerial vehicle autonomous navigation control method in strong electromagnetic environment

Abstract

This invention relates to the field of unmanned aerial vehicle (UAV) navigation technology, specifically to an autonomous navigation control method for UAVs in strong electromagnetic environments. The method includes: acquiring raw electromagnetic interference signals within the flight airspace using a multi-band electromagnetic spectrum sensor array mounted on the UAV, containing broadband power spectral density distribution, pulse interference timestamp sequences, and phase jitter characteristics at specific frequencies; generating an interference source distribution heatmap using an electromagnetic environment feature analyzer, and constructing an electromagnetic situational awareness constraint field by fusing inertial navigation attitude data; dynamically reconstructing the global flight path topology based on this constraint field, generating a set of local flight path segments that avoid high-interference sectors and marking their electromagnetic safety margin levels; and spatially registering the ground feature point image sequence acquired by an optically assisted positioning module with the local flight path segment set to generate composite navigation guidance commands. This method can effectively adapt to strong electromagnetic environments, improving the stability and accuracy of UAV autonomous navigation.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) navigation technology, and in particular to an autonomous navigation control method for UAVs in strong electromagnetic environments. Background Technology

[0002] Currently, most autonomous navigation and control systems for unmanned aerial vehicles (UAVs) employ a combination of inertial navigation, satellite navigation, or single-vision navigation, which can achieve stable navigation in normal electromagnetic environments. However, when facing strong electromagnetic environments, existing technologies often rely on single-frequency electromagnetic sensors to collect interference signals, obtaining only basic interference power information. This information is then combined with a pre-set flight path to complete navigation planning. Some solutions supplement this with optical positioning, but they do not deeply integrate electromagnetic interference information with navigation planning and visual positioning.

[0003] Existing technologies cannot fully capture the complex interference characteristics in strong electromagnetic environments, making it difficult to accurately obtain the spatial distribution, intensity gradient, and frequency band occupancy status of interference sources. They also cannot effectively integrate electromagnetic interference information with UAV inertial navigation attitude data, resulting in a lack of targeted electromagnetic constraints in navigation planning and difficulty in avoiding high-interference areas. Furthermore, existing solutions often use fixed flight paths that cannot be dynamically adjusted according to the real-time electromagnetic environment. Additionally, optical positioning is disconnected from electromagnetic interference constraints, making navigation commands susceptible to electromagnetic interference. This leads to poor navigation stability and insufficient positioning accuracy for UAVs in strong electromagnetic environments, making them prone to flight path deviations and loss of control.

[0004] It is necessary to accurately analyze the complex interference characteristics of strong electromagnetic environments, construct a navigation constraint mechanism that can adapt to electromagnetic environments, realize dynamic adjustment of flight routes, and effectively integrate electromagnetic constraints with visual positioning to solve the problems of UAV navigation being susceptible to interference, unreasonable flight routes, and inaccurate positioning in strong electromagnetic environments. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an autonomous navigation and control method for unmanned aerial vehicles in strong electromagnetic environments.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an autonomous navigation control method for unmanned aerial vehicles (UAVs) in a strong electromagnetic environment, comprising:

[0007] Based on the multi-band electromagnetic spectrum sensor group carried by the UAV, the raw electromagnetic interference signal in the flight airspace is collected. The raw electromagnetic interference signal includes the power spectral density distribution in a wide frequency band, the timestamp sequence of pulse interference, and the phase jitter characteristics at specific frequency points.

[0008] The original electromagnetic interference signal is input into an electromagnetic environment feature analyzer to generate a heat map of interference source distribution under strong electromagnetic environment. The heat map of interference source distribution includes the spatial gradient of interference intensity, the occupancy status of interference frequency band, and the azimuth direction of interference source.

[0009] By fusing the heatmap of the interference source distribution with the current inertial navigation attitude data of the UAV, an electromagnetic situational awareness constraint field is constructed. The electromagnetic situational awareness constraint field is used to limit the available maneuver space of the UAV in subsequent navigation planning.

[0010] Based on the electromagnetic situational awareness constraint field, the pre-set global flight path is dynamically reconstructed to generate a set of local flight path segments that avoid high interference sectors, and the electromagnetic safety margin level is marked on the set of local flight path segments.

[0011] The optical-assisted positioning module on the UAV is activated to collect image sequences of ground feature points. The image sequences are then spatially registered with the set of local flight lines to generate composite navigation and guidance commands that integrate electromagnetic constraints and visual features.

[0012] As a further aspect of the present invention, the original electromagnetic interference signal is input into an electromagnetic environment feature analyzer to generate a heat map of the interference source distribution under the strong electromagnetic environment, including:

[0013] Non-uniform sampling and noise reduction processing is performed on the power spectral density distribution within the wide bandwidth to retain the frequency band segments with energy;

[0014] Cluster analysis was performed on the timestamp sequence of the pulse interference to identify periodic pulse groups and aperiodic transient pulses, and the repetition frequency and duty cycle of the periodic pulse groups were recorded.

[0015] Statistical analysis is performed on the phase jitter characteristics at the specific frequency points, and the root mean square value of the phase noise is calculated, which serves as the basis for judging the stability of narrowband interference.

[0016] Based on the repetition frequency, duty cycle, and root mean square value of phase noise, the spatial domain is divided into grids, and a comprehensive interference index is assigned to each grid cell.

[0017] The comprehensive interference index of all grid cells in the entire airspace is color-coded and spatially interpolated to generate a visualized heatmap of the interference source distribution.

[0018] As a further aspect of the present invention, an electromagnetic situational awareness constraint field is constructed by fusing the interference source distribution heatmap with the current inertial navigation attitude data of the UAV, including:

[0019] The boundary of high-interference areas with a comprehensive interference index exceeding a preset threshold is extracted from the heat map of interference source distribution to form a no-fly envelope.

[0020] The current inertial navigation attitude data of the UAV is read, the current heading angle, pitch angle and roll angle of the UAV are calculated, and the motion envelope in the short term is predicted.

[0021] The motion envelope is subjected to collision detection with the no-fly envelope surface. If a potential intrusion is detected, the minimum deflection angle required to escape the no-fly envelope surface is calculated.

[0022] Using the minimum deflection angle as the constraint radius and the current position of the UAV as the center, a hemispherical electromagnetic situational awareness constraint field is constructed. The inner region of the electromagnetic situational awareness constraint field is defined as the low safety margin region, and the outer region is defined as the high safety margin region.

[0023] As a further aspect of the present invention, based on the electromagnetic situational awareness constraint field, a pre-set global flight path is dynamically reconstructed to generate a set of local flight path segments that avoid high-interference sectors, including:

[0024] The pre-defined global route is discretized into several waypoints, and each waypoint is checked in turn to see if it is located within the high safety margin region of the electromagnetic situational awareness constraint field.

[0025] Waypoints located outside the high safety margin zone are marked as points to be adjusted, and multiple candidate offset points are generated in the horizontal plane with the points to be adjusted as the center.

[0026] Calculate the spatial distance between each candidate offset point and surrounding obstacles and no-fly zones, and select a subset of candidate offset points that meet the safety distance requirements;

[0027] In the subset of candidate offset points, select the candidate offset point with the lowest deviation cost from the global flight path and replace the original point to be adjusted;

[0028] All replaced waypoints and unadjusted waypoints are connected sequentially to form several short straight lines, forming the local route segment set. Based on the average interference index of the traversed area, each short straight line segment is assigned the electromagnetic safety margin level.

[0029] As a further aspect of the present invention, the optical-assisted positioning module mounted on the UAV is activated to acquire image sequences of ground feature points, and the image sequences are spatially registered with the local flight line segment set, including:

[0030] The shutter of the optical auxiliary positioning module is controlled to continuously expose within a fixed time interval to acquire multiple frames of image data containing ground textures;

[0031] A feature point extraction algorithm is performed on the image data to identify corner features with rotation and scale invariance, and the pixel positions of the corner features in the image coordinate system are recorded.

[0032] Extract the geodetic coordinates of the start and end points of the current target route segment from the set of local route segments, and use them as the reference benchmark for registration;

[0033] By utilizing the relative positional relationship of the corner features in the image and combining it with the geodetic coordinates of the reference datum, the pose transformation matrix of the current camera is calculated to obtain the real-time six-degree-of-freedom pose estimation of the UAV relative to the ground.

[0034] The real-time six-degree-of-freedom pose estimation is converted into geographic coordinates, which are used as the current location input for the composite navigation guidance command.

[0035] As a further aspect of the present invention, a feature point extraction algorithm is performed on the image data to identify corner features with rotation and scale invariance, and the pixel positions of the corner features in the image coordinate system are recorded, including:

[0036] Perform Gaussian difference processing on the image data to construct the image scale space;

[0037] In the image scale space, the gray value relationship between each pixel and its neighboring pixels at the same and adjacent scales is detected to locate candidate key points;

[0038] Edge response culling is performed on the candidate key points, the curvature ratio of the key points is calculated, and points that do not meet the curvature ratio threshold are removed.

[0039] Assign orientations to the retained keypoints by calculating the main orientation based on the gradient orientation histogram of the neighboring pixels of the keypoints;

[0040] Based on the scale and principal direction of key points, feature descriptors with rotation and scale invariance are generated.

[0041] The key points corresponding to each feature descriptor are recorded as corner features with rotation and scale invariance, and their two-dimensional coordinates in the original image plane are extracted as the pixel positions.

[0042] As a further aspect of the present invention, it also includes a step of performing anti-interference trajectory tracking control based on the composite navigation guidance command:

[0043] The position error vector and heading error angle are calculated by differentiating the target position in the composite navigation guidance command with the current six-degree-of-freedom pose estimate.

[0044] Read the interference intensity value of the current location in the interference source distribution heat map, and adaptively adjust the weight coefficient of the location error vector in the control law according to the magnitude of the interference intensity value;

[0045] When the interference intensity value exceeds the strong interference threshold, the weight of the position error vector is reduced, and the control response speed to the heading error angle is increased to suppress track drift caused by crosswinds or electromagnetic gusts.

[0046] The weighted position error vector and the adjusted heading error angle are fed into the cascaded PID controller to generate servo deflection commands, which drive the UAV to perform control surface actions.

[0047] While performing rudder surface actions, the position feedback output by the optical auxiliary positioning module is continuously monitored, and the output of the servo deflection command is corrected in a closed loop.

[0048] As a further aspect of the present invention, when the interference intensity value exceeds the strong interference threshold, the weight of the position error vector is reduced, while the control response speed to the heading error angle is improved, including:

[0049] A mapping table between interference intensity and controller parameters is established, wherein the mapping table defines the position loop gain and heading loop gain corresponding to different interference intensity ranges;

[0050] At the start of the control cycle, the mapping table is queried to obtain the position loop gain and heading loop gain corresponding to the interval where the current interference intensity value is located.

[0051] The traditional single control law is switched to a dual-loop independent control law, in which the outer loop position control uses the position loop gain and the inner loop heading control uses the heading loop gain.

[0052] A feedforward compensation term is introduced into the inner loop heading control. The feedforward compensation term is obtained by back-calculation of the azimuth angle of the interference source in the interference source distribution heat map, and is used to offset the expected electromagnetic side thrust in advance.

[0053] The outputs of the two independent control laws are superimposed and merged to generate the final servo deflection command.

[0054] As a further aspect of the present invention, it also includes a real-time evolution learning of the electromagnetic environment and a route replanning step:

[0055] The temporal variation data of the interference source distribution heatmap output by the electromagnetic environment feature analyzer during the flight of the UAV are continuously recorded.

[0056] Fourier transform and time-frequency analysis were performed on the time-series variation data to identify the periodic variation patterns and sudden disturbance modes of the electromagnetic environment.

[0057] When a sudden disturbance pattern is detected, the current route tracking task is paused, and the latest heat map of the interference source distribution is retrieved.

[0058] The dynamic topology reconstruction process is re-executed to generate a new set of local flight segments adapted to the latest electromagnetic situation;

[0059] The new local route segment set is compared with the original global route. If the overlap is less than a threshold, the current route is switched to the new local route segment set to continue the navigation task.

[0060] As a further aspect of the present invention, Fourier transform and time-frequency analysis are performed on the time-series variation data to identify the periodic variation patterns and sudden disturbance modes of the electromagnetic environment, including:

[0061] The time-series change data is divided into several time windows of equal length, and the interference intensity data in each time window is normalized.

[0062] Perform a short-time Fourier transform on each normalized time window of data to generate a time-spectrum graph;

[0063] On the time-spectrum graph, a frequency bandwidth of interest is set, and the curve of energy change over time within the frequency bandwidth is extracted;

[0064] Peak detection is performed on the energy-time curve to identify the moment when energy suddenly increases, and this moment is marked as the suspected start time of sudden disturbance.

[0065] By tracing back the original data stream of the electromagnetic environment feature analyzer before and after the suspected sudden disturbance start time, it is verified whether there is continuous pulse interference or phase locking phenomenon. If so, it is confirmed as the sudden disturbance mode.

[0066] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0067] A multi-band electromagnetic spectrum sensor array is used to collect raw electromagnetic interference signals, including power spectral density distribution, pulse interference timestamp sequences, and phase jitter characteristics at specific frequencies, across a wide bandwidth. An electromagnetic environment feature analyzer generates a heatmap of interference source distribution, incorporating spatial gradients of interference intensity, interference frequency band occupancy, and the azimuth direction of the interference source. This heatmap is then fused with UAV inertial navigation attitude data to construct an electromagnetic situational awareness constraint field, limiting the available maneuver space for navigation planning. This approach comprehensively and accurately captures complex interference characteristics in strong electromagnetic environments, clearly presenting the spatial distribution and intensity variations of interference sources. This allows navigation planning to explicitly avoid high-interference areas, preventing electromagnetic interference from affecting navigation attitude. It addresses the limitations of conventional technologies in accurately sensing electromagnetic situations and the lack of electromagnetic constraints in navigation planning, making navigation planning more targeted and adaptable.

[0068] Based on the electromagnetic situational awareness constraint field, a dynamic topology reconstruction is performed on the preset global route, generating a set of local route segments that avoid high-interference sectors and marking their electromagnetic safety margin levels. The ground feature point image sequence collected by the optically assisted positioning module is spatially registered with the local route segment set to generate a composite navigation guidance command that integrates electromagnetic constraints and visual features. This enables dynamic adjustment of the route, ensuring the electromagnetic safety of the route segments. Simultaneously, by fusing visual positioning and electromagnetic constraints, it compensates for the shortcomings of single electromagnetic navigation or visual navigation, reducing the impact of electromagnetic interference on navigation commands, making navigation guidance more accurate and stable. This solves the problems of conventional technologies where fixed routes cannot adapt to changes in the electromagnetic environment and navigation positioning is susceptible to interference. Attached Figure Description

[0069] Figure 1 This is a flowchart of the autonomous navigation and control method for unmanned aerial vehicles in a strong electromagnetic environment as described in this invention;

[0070] Figure 2 A flowchart for generating a heat map of interference source distribution under strong electromagnetic environment;

[0071] Figure 3 This is a flowchart for dynamic topology reconstruction of the global flight path;

[0072] Figure 4 The diagram shows the effect of interference intensity and track error compensation.

[0073] Figure 5 This is a detection diagram for sudden energy surge events in the electromagnetic environment. Detailed Implementation

[0074] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0075] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0076] See Figure 1 The UAV first uses its onboard multi-band electromagnetic spectrum sensor array to collect raw electromagnetic interference signals within the flight airspace. These signals include a wide-band power spectral density distribution, timestamp sequences of impulse interference, and phase jitter characteristics at specific frequencies. The collected raw electromagnetic interference signals are then processed by an electromagnetic environment feature analyzer to generate a heatmap reflecting the current airspace electromagnetic situation. This heatmap includes the spatial gradient of interference intensity, the occupancy status of interference frequency bands, and the azimuth pointing information of the interference sources. This heatmap is then fused with the UAV's current inertial navigation attitude data provided by the inertial measurement unit to construct an electromagnetic situational awareness constraint field. This constraint field quantifies the impact of electromagnetic interference on flight safety and limits the available maneuver space for the UAV in subsequent navigation planning. Based on this electromagnetic situational awareness constraint field, the pre-defined global flight path is dynamically reconstructed to generate a set of local flight path segments that avoid high-interference sectors. Each flight path segment in this set is then labeled with its electromagnetic safety margin level. The optical-assisted positioning module on the UAV is activated to collect image sequences of ground feature points. The image sequences are then spatially registered with a set of planned local flight path segments to calculate the precise pose information of the UAV. This generates a composite navigation and guidance command that integrates electromagnetic constraints and visual features, driving the UAV to fly along a safe flight path.

[0077] See Figure 2 In one embodiment of the present invention, the specific process of generating a heat map of the interference source distribution and constructing an electromagnetic situational awareness constraint field is involved. The raw electromagnetic interference signal collected by the UAV in the flight area includes the power spectral density distribution from 1 GHz to 6 GHz, a series of pulse arrival timestamps, and phase sampling sequences for the 2.4 GHz and 5.8 GHz frequencies. The multi-band electromagnetic spectrum sensor group transmits the above-mentioned raw electromagnetic interference signal to the airborne electromagnetic environment feature analyzer at a rate of 100 times per second.

[0078] In practical implementation, the electromagnetic environment feature analyzer performs non-uniform sampling noise reduction processing on the power spectral density distribution over a wide bandwidth. The processing sets an energy threshold of three times the mean background noise, retaining only frequency bands with power values ​​exceeding this threshold. For example, significant energy was found in the 1.2GHz to 1.3GHz, 2.4GHz to 2.4835GHz, and 5.725GHz to 5.85GHz frequency bands. In practical implementation, density-based clustering analysis is performed on the timestamp sequences of impulse interference. The clustering algorithm sets a minimum time neighborhood of 10 microseconds and a minimum sample size of 5. The clustering analysis identifies two groups of periodic pulse groups and a large number of discrete aperiodic transient pulses. The repetition frequency of the first group of periodic pulse groups is recorded as 1000 Hz with a duty cycle of 10%, and the repetition frequency of the second group of periodic pulse groups is recorded as 200 Hz with a duty cycle of 50%. In practice, statistical analysis was performed on the phase jitter characteristics of specific frequency points at 2.4GHz and 5.8GHz. The root mean square value of the phase noise at the 2.4GHz frequency point was calculated to be 1.5 degrees, and the root mean square value of the phase noise at the 5.8GHz frequency point was 2.1 degrees. The root mean square value of the phase noise is used as a direct basis for judging the stability of the narrowband interference signal at the corresponding frequency point.

[0079] The airspace is divided into grids based on repetition frequency, duty cycle, and root mean square value of phase noise. The space extending horizontally from 100 to 500 meters below the UAV, covering an area of ​​1 square kilometer, is divided into cubic grid cells with 10-meter sides. A comprehensive interference index is calculated for each grid cell. The calculation formula is expressed as follows:

[0080]

[0081] Where: characters Indicates the overall interference index of the grid cell; character This represents the normalized pulse repetition frequency factor, whose value originates from the repetition frequency of a periodic pulse group; character Represents the normalized pulse duty cycle factor; character Represents the normalized root mean square factor of phase noise; character , , These represent the preset weighting coefficients for the repetition frequency factor, duty cycle factor, and phase noise factor, respectively. In specific implementations, the following settings are used: It is 0.5. It is 0.3. The value is 0.2. This is understandable, as a grid cell located above a suspected source of communication interference might receive a higher overall interference index after the calculation. Values, such as 0.85, represent the overall disturbance index of a relatively clean spatial grid cell. The value may be 0.12. In specific implementation, the comprehensive interference index of all grid cells in the entire airspace is color-coded, with the value mapped from low to high as a gradient from blue to red. A three-dimensional kriging space interpolation algorithm is used to interpolate the discrete grid data to generate a continuous, smooth, and visualized heatmap of interference source distribution. The heatmap clearly shows the spatial gradient of interference intensity, the occupancy status of different frequency bands, and the azimuth of the interference source obtained by estimating the signal arrival direction.

[0082] In some embodiments, an electromagnetic situational awareness constraint field is constructed by fusing a heatmap of interference source distribution with the current inertial navigation attitude data of the UAV. High-interference regions with a comprehensive interference index exceeding a preset threshold of 0.7 are extracted from the interference source distribution heatmap. Grid cells with a comprehensive interference index exceeding the preset threshold of 0.7 are identified as high-threat zones, and the outer surfaces of these cells are connected to form a three-dimensional no-fly envelope. The current inertial navigation attitude data of the UAV is read, and the current heading angle is calculated to be 45 degrees, the pitch angle to be 2 degrees, and the roll angle to be 0.5 degrees. Based on the current velocity and angular velocity of the UAV, the motion envelope for the next 2 seconds is predicted. The motion envelope is an ellipsoid covering possible locations. In a specific implementation, collision detection is performed between the motion envelope and the no-fly envelope. The detection algorithm calculates the minimum distance between the surface of the motion envelope and the no-fly envelope. When the minimum distance is less than zero, it indicates a potential intrusion. It can be understood that if the collision detection algorithm calculates a potential intrusion, the navigation computer calculates the minimum deflection angle required for the UAV's motion envelope to escape the no-fly envelope. Using a minimum deflection angle of 15 degrees as the constraint radius and the current three-dimensional coordinates of the UAV as the center, a hemispherical electromagnetic situational awareness constraint field is constructed in space. The opening of the hemispherical electromagnetic situational awareness constraint field faces the current heading of the UAV. The internal region of the hemispherical electromagnetic situational awareness constraint field is defined as the low safety margin region, and the external region of the hemispherical electromagnetic situational awareness constraint field is defined as the high safety margin region.

[0083] See Figure 3In one embodiment of the present invention, a specific process is involved in dynamically reconstructing the topology of a pre-defined global route to generate a set of local route segments. The global route is adjusted according to an electromagnetic situational awareness constraint field. A pre-defined global route from geographic coordinate point A (116.3°E, 39.9°N, altitude 200m) to point B (116.35°E, 39.92°N, altitude 200m) is discretized into a series of ordered waypoints at 10-meter intervals. The navigation computer sequentially checks whether the three-dimensional coordinates of each waypoint are within the high safety margin zone defined by the electromagnetic situational awareness constraint field. The criterion is whether the comprehensive interference index of the grid cell containing the waypoint coordinates is lower than 0.4. In a specific implementation, the check found that the comprehensive interference indices of the grid cells containing points P5, P6, and P7 in the waypoint sequence were 0.72, 0.68, and 0.65, respectively. These three points were marked as points to be adjusted because they are outside the high safety margin zone.

[0084] In practice, 12 candidate offset points are generated at 30-degree intervals within the horizontal plane of each point to be adjusted, with the center as the target point. The horizontal distance between the candidate offset points and the point to be adjusted is preset to 15 meters. The three-dimensional Euclidean distances between these candidate offset points and known terrain obstacle 3D models, building outlines, and no-fly zone boundaries with a comprehensive interference index exceeding 0.7 are calculated one by one. A minimum safe distance threshold of 5 meters is set in the screening calculation, eliminating all candidate offset points whose distance to any obstacle or no-fly zone is less than 5 meters. The remaining points that meet the safe distance requirement constitute a subset of candidate offset points. To select the best replacement point from the subset of candidate offset points, the navigation computer calculates the deviation cost of each candidate offset point from the original global flight path. The calculation formula is expressed as follows:

[0085]

[0086] Where: characters Indicates the deviation cost of the candidate bias point; character Indicates the shortest perpendicular distance from the candidate offset point to the line segment connecting its original point to be adjusted and its adjacent original waypoint; character This indicates the change in angle (in degrees) between the direction of the newly formed route segment and the direction of the original route segment after inserting the candidate offset point into the route; character With characters These are the distance weighting coefficient and the angle change weighting coefficient, respectively, which are set in the specific implementation. It is 0.7. The value is 0.3. This can be understood as the navigation computer selecting the point with the minimum deviation cost from the subset of candidate bias points. The candidate offset point is used to replace the original point to be adjusted. For example, a candidate point located 12 meters to the left of the original flight path will be removed due to its deviation cost. It was selected with a value of 4.2.

[0087] In some embodiments, all replaced waypoints and unadjusted waypoints are connected by straight line segments in the original order from point A to point B, forming a set of local flight line segments composed of several short straight lines. The navigation system calculates the arithmetic mean of the comprehensive interference indices of all grid cells traversed by each short straight line segment, and assigns an electromagnetic safety margin level based on the magnitude of the average value. For example, flight line segments with an average comprehensive interference index below 0.3 are assigned a "high" electromagnetic safety margin level, those with an average value between 0.3 and 0.5 are assigned a "medium" electromagnetic safety margin level, and those with an average value above 0.5 are assigned a "low" electromagnetic safety margin level. Optionally, the number of candidate offset points generated and the angular interval in the horizontal plane can be adjusted according to computational resources and flight line accuracy requirements; for example, 24 candidate points can be generated with an angular interval of 15 degrees. Optionally, the minimum safe distance threshold can be dynamically set according to the UAV size and flight speed. It is understandable that when calculating the spatial distance between the candidate offset point and surrounding obstacles, the obstacle model can include terrain elevation data, known 3D models of buildings, and real-time detected location information of other aircraft.

[0088] In one embodiment of the present invention, the specific process of activating an optically assisted positioning module for image acquisition, feature extraction, and spatial registration is described. When the UAV flies to a planned route segment within a set of local flight paths, it activates its onboard downward-looking optically assisted positioning module. The module's global shutter is controlled to continuously expose within a fixed 33-millisecond time interval, acquiring five frames of image data containing continuous ground texture information. The resolution of the image data is 1920 pixels multiplied by 1080 pixels. In a specific implementation, a scale-invariant feature transformation (SIN) feature point extraction algorithm is performed on the acquired five frames of image data to identify corner features with rotation and scale invariance. Gaussian difference processing is performed on each frame of image data to construct an image scale space. The image scale space contains five octave layers, and each octave layer contains three scale layers. In a specific implementation, the grayscale value relationship between each pixel and its 26 neighboring pixels at the same and adjacent scales is detected in the image scale space. Pixels whose grayscale values ​​are simultaneously greater than or less than all 26 neighboring pixels are located as candidate keypoints. In practice, edge response elimination is performed on candidate key points. The curvature ratio of the key points is calculated by calculating the eigenvalues ​​of the Hessian matrix at the key point location. Key points with curvature ratios greater than a set threshold of 10 are eliminated to remove edge response points.

[0089] In the implementation, a principal orientation is assigned to the retained keypoints. Based on the gradient orientations of all pixels within a circular neighborhood centered on the keypoint with a radius three times its scale, a 36-bin gradient orientation histogram is constructed, and the orientation corresponding to the peak of the histogram is determined as the principal orientation. In the implementation, based on the scale of the keypoint and the determined principal orientation, the neighborhood of the keypoint is divided into 4x4 sub-regions. Within each sub-region, gradient orientation histograms for eight directions are calculated, ultimately forming a 128-dimensional feature vector. This feature vector generates a feature descriptor with rotation and scale invariance. In the implementation, the two-dimensional coordinates of the keypoint corresponding to each 128-dimensional feature descriptor in the original image plane are extracted and recorded as pixel positions.

[0090] In some embodiments, the geodetic coordinates of the start and end points of the target flight segment that the UAV needs to fly are extracted from the local flight segment set. For example, the start point coordinates are (116.301 degrees east longitude, 39.901 degrees north latitude, 202.5 meters above sea level), and the end point coordinates are (116.304 degrees east longitude, 39.902 degrees north latitude, 201.8 meters above sea level). These start and end point coordinates are used as a reference for spatial registration. In a specific implementation, the pixel position relationship of the multi-diagonal point features successfully matched from consecutive image frames is used, combined with the geodetic coordinates of the reference and the known intrinsic parameter matrix of the optical auxiliary positioning module, to calculate the pose transformation matrix of the current camera. The solution is achieved by solving an optimization problem that minimizes the reprojection error, the objective function of which is... The expression is as follows:

[0091]

[0092] Where: characters Represents the sum of squared reprojection errors of all matching feature points; character Indicates the number of successfully matched feature point pairs; characters Indicates the first The two-dimensional pixel coordinates of a feature point in the current image coordinate system are a two-dimensional vector; the character Indicates and The first match The three-dimensional coordinates of a feature point in the world coordinate system (i.e., the geographic coordinate system) form a three-dimensional vector; a character This represents the pre-calibrated intrinsic parameter matrix of the optical-assisted positioning module; characters This represents the pose transformation matrix to be solved from the world coordinate system to the camera coordinate system, including rotation and translation; characters This represents the projection function from 3D camera coordinates to 2D image coordinates. It can be understood that this is achieved by minimizing the objective function through an iterative optimization algorithm. Finally, the pose transformation matrix is ​​solved. And then from the pose transformation matrix The decomposition yields a real-time six-DOF pose estimate of the UAV relative to the ground, including position (X, Y, Z) and attitude (roll angle, pitch angle, yaw angle).

[0093] In one embodiment of the present invention, a specific process for anti-interference trajectory tracking control based on composite navigation guidance commands is involved. The target position in the composite navigation guidance commands is (116.302 degrees east longitude, 39.905 degrees north latitude, 205 meters above sea level). The current six-degree-of-freedom pose estimation value obtained from the optical auxiliary positioning module is (116.30085 degrees east longitude, 39.90372 degrees north latitude, 207.5 meters above sea level, roll angle 0.5 degrees, pitch angle 1.8 degrees, heading angle 44 degrees). The position components in the target position and the current six-degree-of-freedom pose estimation value are differentially calculated to obtain an eastward position error of +1.2 meters, a northward position error of +1.3 meters, and an azimuth position error of -2.5 meters. The position error vector is synthesized, and the heading error angle between the target heading of 45 degrees and the current heading of 44 degrees is calculated to be +1 degree. The flight controller reads the comprehensive interference index of the grid cell where the UAV's current position is located from the interference source distribution heatmap as the real-time interference intensity value. Based on the magnitude of the interference intensity value, it adaptively adjusts the weighting coefficients of the position error vector in the control law through a preset linear function. The calculation formula is expressed as follows:

[0094]

[0095] Where: characters Represents the weighting coefficients of the position error vector; character This indicates the maximum weighting coefficient, set to 1.0; character This indicates the minimum weighting coefficient, set to 0.3; character This represents the real-time interference intensity value read from the heatmap of interference source distribution; character This indicates the set strong interference threshold, which is set to 0.7. When the real-time interference intensity value... When the value is 0.85 and exceeds the strong interference threshold of 0.7, the weighting coefficient is calculated according to the formula. The weight of the position error vector is reduced to 0.46, thus lowering its weight. Simultaneously, the flight controller increases the response speed coefficient of the heading error angle control loop from 1.0 to 2.0 to improve the control response speed to the heading error angle and suppress track drift caused by electromagnetic interference. In practice, the weighted position error vector and the heading error angle adjusted for the adjusted response speed are fed into a cascade PID controller. The outer position loop of the cascade PID controller receives the weighted position error vector, and the inner heading loop receives the adjusted heading error angle. After proportional, integral, and derivative operations, servo deflection commands are generated for the elevator deflection angle of -2.1 degrees, the aileron deflection angle of +0.8 degrees, and the rudder deflection angle of +0.5 degrees, driving the UAV to execute the corresponding control surface actions. During the control cycle of executing control surface actions, the flight controller continuously monitors the latest position feedback output from the optical auxiliary positioning module, compares the new position feedback with the target position to form a closed loop, and corrects the servo deflection command output for the next control cycle.

[0096] In some embodiments, when the interference intensity value exceeds the strong interference threshold, a specific process is executed to reduce the weight of the position error vector and increase the control response speed of the heading error angle, establishing a mapping table between interference intensity and controller parameters. The mapping table defines the position loop proportional gain and heading loop proportional gain corresponding to different interference intensity ranges. At the beginning of the control cycle, the mapping table is queried to obtain the position loop gain and heading loop gain corresponding to the current interference intensity range. It can be understood that the flight controller switches from a traditional single integrated control law to a dual-loop independent control law with independent position and heading loops. The outer loop position control uses the queried position loop gain, and the inner loop heading control uses the queried heading loop gain. Optionally, the content of the mapping table can be preset offline or updated online through learning. A feedforward compensation term is introduced for the inner loop heading control. The feedforward compensation term is calculated by back-calculating the azimuth angle pointing information of the interference source provided in the interference source distribution heatmap. For example, if the heatmap indicates that the azimuth angle of the main interference source is 120 degrees, the feedforward compensation term calculates an initial rudder offset command to counteract the expected electromagnetic side thrust in advance. In practice, the control quantities output by the dual-loop independent control laws are superimposed and fused to generate the final servo deflection command. Refer to Table 1 for an example of the mapping relationship between disturbance intensity and controller parameters.

[0097] Table 1: Mapping Relationship between Interference Intensity and Controller Parameters

[0098]

[0099] In some embodiments, the process of querying the mapping table is that the flight controller detects the current real-time interference intensity value. The value is 0.85, which lies within the interval [0.7, 1.0], therefore the position loop gain is obtained. =0.5, heading loop gain =2.8. The outer loop position control uses a position loop gain of 0.5 to proportionally calculate the weighted position error vector, while the inner loop heading control uses a heading loop gain of 2.8 to proportionally calculate the heading error angle and superimposes a feedforward compensation term. It can be understood that the feedforward compensation term is calculated based on the angle between the azimuth of the interference source and the UAV's current heading. When the interference source is located on the starboard side of the UAV, the feedforward compensation term outputs a leftward rudder offset command. Optionally, the gain coefficient of the feedforward compensation term can be proportional to the interference intensity value. The commands output by the outer loop position control and the inner loop heading control are superimposed and merged. For example, the outer loop outputs an aileron deflection of +0.5 degrees to correct the position, and the inner loop outputs a rudder deflection of +0.3 degrees with superimposed feedforward compensation of -0.2 degrees, ultimately generating a combined servo deflection command of +0.5 degrees for aileron deflection and +0.1 degrees for rudder deflection.

[0100] See Figure 4 This is a graph showing the effect of interference intensity and trajectory error compensation, illustrating the changes in core performance indicators of UAV anti-interference trajectory tracking control under strong electromagnetic environments. As the interference intensity increases from 0.1 to 0.95, the position error increases continuously from 0.2 meters to 1.5 meters. This indicates that the stronger the electromagnetic interference, the worse the UAV's position tracking accuracy, necessitating a reduction in the position error weight to avoid over-correction. As the interference intensity increases from 0.1 to 0.95, the heading error significantly decreases from 0.8 degrees to 0.05 degrees. This verifies the effectiveness of the control strategy: improving the heading control response speed under strong interference effectively suppresses trajectory drift. In the strong interference zone where the interference intensity exceeds 0.7, the heading error is effectively controlled at an extremely low level, ensuring the UAV's flight stability. The increase in position error is a trade-off for heading stability, representing a proactive performance trade-off.

[0101] In one embodiment of the present invention, a real-time evolution learning and route replanning step of the electromagnetic environment is involved. During the inspection mission of the UAV flying from the starting point S to the destination T, the flight controller continuously records the temporal change data of the interference source distribution heatmap output by the electromagnetic environment feature analyzer every second. A flight mission lasting 300 seconds will generate a temporal change data sequence containing a comprehensive interference index matrix of 300 time points. Fourier transform and time-frequency analysis are performed on the temporal change data to identify the periodic change pattern and sudden disturbance mode of the electromagnetic environment. The temporal change data is divided into multiple time windows of equal length with a window length of 10 seconds and an overlap of 5 seconds. Maximum value normalization is performed on the interference intensity data of all grid cells in each time window to adjust the data amplitude range in each window to the [0,1] interval. In a specific implementation, a short-time Fourier transform is performed on the data of each normalized time window. The short-time Fourier transform uses a Hanning window to generate a time-spectrum diagram that can simultaneously reflect frequency and time information. On the time-spectrum graph, the frequency bandwidth of interest is set to 0 Hz to 5 Hz. The curve of signal energy changing with time within this frequency bandwidth is extracted, and the energy value is obtained by integration.

[0102] In some embodiments, peak detection based on amplitude is performed on the energy-time curve. A detection threshold is set to twice the average value of the energy curve. Points where energy surges exceeding the threshold are identified and marked as suspected sudden disturbance initiation times. The original data stream of the electromagnetic environment feature analyzer within 3 seconds before and after the suspected sudden disturbance initiation time is reviewed to check for continuous pulse interference sequences with intervals less than 100 microseconds, or to check for phase locking at specific frequencies (i.e., a phase change rate consistently below 0.1 degrees / second). If continuous pulse interference or phase locking is present in the original data stream, the event is confirmed as a sudden disturbance mode. It can be understood that if a sudden energy surge event, after backtracking verification, is found to have a pulse train lasting 2 seconds and repeating at a frequency of 2kHz in its corresponding original data stream, this event is confirmed as a sudden disturbance mode. In specific implementations, the energy surge index... The calculation formula can be used to assist in the determination, and the formula is expressed as follows:

[0103]

[0104] Where: characters Indicates the energy surge index; character Indicates the amplitude of a suspected peak in the energy curve; character Indicates the average value of the energy curve within the detection window; character This represents the standard deviation of the energy curve within the detection window. When the energy surge index... When the value is greater than 3.0, a backtracking verification process for the original data stream is triggered.

[0105] In practice, when the detection algorithm confirms the occurrence of a sudden disturbance mode, the flight control computer suspends the UAV's current flight path tracking task, controls the UAV to enter a hovering state, and immediately retrieves the interference source distribution heatmap generated from the latest sensor data from the electromagnetic environment feature analyzer. Using the latest electromagnetic situation reflecting the sudden disturbance as input, the dynamic topology reconstruction process is re-executed. This involves constructing a new electromagnetic situational awareness constraint field based on the new interference source distribution heatmap and replanning the original global flight path to generate a new set of local flight path segments adapted to the latest electromagnetic environment conditions. Optionally, the criterion for triggering flight path replanning can be set to the continuous detection of two or more sudden disturbance modes. The new set of local flight path segments is spatially compared with the original global flight path, and the average Hausdorff distance between each flight path segment in the new set and the corresponding segment of the original flight path is calculated. If this average distance is greater than a preset switching threshold of 20 meters, the overlap between the two is determined to be below the threshold. Understandably, if the overlap is below a threshold, the flight control computer will switch the UAV's current flight command from the original route to a newly generated set of local route segments. The UAV will then exit the hovering state and continue its navigation mission using this new set of local route segments as the path. Optionally, the overlap threshold can be dynamically set based on the safety redundancy of the mission area.

[0106] See Figure 5 This is an electromagnetic environment energy surge event detection graph, showing the change of electromagnetic environment energy over time and the sudden disturbance detection logic based on the energy surge index. The energy curve represents the real-time change of energy value in the electromagnetic environment, fluctuating within the range of 0-1.0, reflecting the normal state of the background electromagnetic environment. The energy surge index is used to quantify the instantaneous change in energy. When the index exceeds the green dashed line (surge threshold), a suspected sudden disturbance detection is triggered. The judgment criteria set for the surge threshold are used to distinguish between normal fluctuations and sudden electromagnetic interference. The graph marks three obvious energy surge events (approximately at 100 seconds, 180 seconds, and 250 seconds), at which time the energy surge index is much higher than the threshold, corresponding to the "sudden disturbance mode" that requires triggering flight path replanning in the patent. It intuitively demonstrates the complete closed loop of real-time electromagnetic environment monitoring and sudden disturbance detection, clearly illustrating how the system accurately identifies sudden interference that poses a threat to UAV navigation from massive amounts of electromagnetic data.

[0107] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for autonomous navigation control of unmanned aerial vehicles in a strong electromagnetic environment, characterized in that, The method includes: Based on the multi-band electromagnetic spectrum sensor group carried by the UAV, the raw electromagnetic interference signal in the flight airspace is collected. The raw electromagnetic interference signal includes the power spectral density distribution in a wide frequency band, the timestamp sequence of pulse interference, and the phase jitter characteristics at specific frequency points. The original electromagnetic interference signal is input into an electromagnetic environment feature analyzer to generate a heat map of interference source distribution under strong electromagnetic environment. The heat map of interference source distribution includes the spatial gradient of interference intensity, the occupancy status of interference frequency band, and the azimuth direction of interference source. By fusing the heatmap of the interference source distribution with the current inertial navigation attitude data of the UAV, an electromagnetic situational awareness constraint field is constructed. The electromagnetic situational awareness constraint field is used to limit the available maneuver space of the UAV in subsequent navigation planning. Based on the electromagnetic situational awareness constraint field, the pre-set global flight path is dynamically reconstructed to generate a set of local flight path segments that avoid high interference sectors, and the electromagnetic safety margin level is marked on the set of local flight path segments. The optical-assisted positioning module on the UAV is activated to collect image sequences of ground feature points, and the image sequences are spatially registered with the set of local flight lines to generate composite navigation guidance commands that integrate electromagnetic constraints and visual features. It also includes the step of performing anti-interference trajectory tracking control based on the composite navigation guidance commands: The position error vector and heading error angle are calculated by differentiating the target position in the composite navigation guidance command with the current six-degree-of-freedom pose estimate. Read the interference intensity value of the current location in the interference source distribution heat map, and adaptively adjust the weight coefficient of the location error vector in the control law according to the magnitude of the interference intensity value; When the interference intensity value exceeds the strong interference threshold, the weight of the position error vector is reduced, while the control response speed to the heading error angle is increased to suppress track drift caused by crosswinds or electromagnetic gusts, including: A mapping table between interference intensity and controller parameters is established, wherein the mapping table defines the position loop gain and heading loop gain corresponding to different interference intensity ranges; At the start of the control cycle, the mapping table is queried to obtain the position loop gain and heading loop gain corresponding to the interval where the current interference intensity value is located. The traditional single control law is switched to a dual-loop independent control law, in which the outer loop position control uses the position loop gain and the inner loop heading control uses the heading loop gain. A feedforward compensation term is introduced into the inner loop heading control. The feedforward compensation term is obtained by back-calculation of the azimuth angle of the interference source in the interference source distribution heat map, and is used to offset the expected electromagnetic side thrust in advance. The outputs of the two independent control laws are superimposed and fused to generate the final servo deflection command; The weighted position error vector and the adjusted heading error angle are fed into the cascaded PID controller to generate servo deflection commands, which drive the UAV to perform control surface actions. While performing rudder surface actions, the position feedback output by the optical auxiliary positioning module is continuously monitored, and the output of the servo deflection command is corrected in a closed loop.

2. The method of claim 1, wherein, The original electromagnetic interference signal is input into an electromagnetic environment feature analyzer to generate a heat map of the interference source distribution under the strong electromagnetic environment, including: Non-uniform sampling and noise reduction processing is performed on the power spectral density distribution within the wide bandwidth to retain the frequency band segments with energy; Cluster analysis was performed on the timestamp sequence of the pulse interference to identify periodic pulse groups and aperiodic transient pulses, and the repetition frequency and duty cycle of the periodic pulse groups were recorded. Statistical analysis is performed on the phase jitter characteristics at the specific frequency points, and the root mean square value of the phase noise is calculated, which serves as the basis for judging the stability of narrowband interference. Based on the repetition frequency, duty cycle, and root mean square value of phase noise, the spatial domain is divided into grids, and a comprehensive interference index is assigned to each grid cell. The comprehensive interference index of all grid cells in the entire airspace is color-coded and spatially interpolated to generate a visualized heatmap of the interference source distribution.

3. The autonomous navigation control method for unmanned aerial vehicles in a strong electromagnetic environment according to claim 2, characterized in that, By fusing the heatmap of the interference source distribution with the current inertial navigation attitude data of the UAV, an electromagnetic situational awareness constraint field is constructed, including: The boundary of high-interference areas with a comprehensive interference index exceeding a preset threshold is extracted from the heat map of interference source distribution to form a no-fly envelope. The current inertial navigation attitude data of the UAV is read, the current heading angle, pitch angle and roll angle of the UAV are calculated, and the motion envelope in the short term is predicted. The motion envelope is subjected to collision detection with the no-fly envelope surface. If a potential intrusion is detected, the minimum deflection angle required to escape the no-fly envelope surface is calculated. Using the minimum deflection angle as a constraint condition, the constraint radius is determined based on the constraint condition. With the current position of the UAV as the center, a hemispherical electromagnetic situational awareness constraint field is constructed. The inner region of the electromagnetic situational awareness constraint field is defined as the low safety margin region, and the outer region is defined as the high safety margin region.

4. The autonomous navigation and control method for unmanned aerial vehicles in a strong electromagnetic environment according to claim 3, characterized in that, Based on the electromagnetic situational awareness constraint field, a dynamic topology reconstruction is performed on the pre-set global flight path to generate a set of local flight path segments that avoid high-interference sectors, including: The pre-defined global route is discretized into several waypoints, and each waypoint is checked in turn to see if it is located within the high safety margin region of the electromagnetic situational awareness constraint field. Waypoints located outside the high safety margin zone are marked as points to be adjusted, and multiple candidate offset points are generated in the horizontal plane with the points to be adjusted as the center. Calculate the spatial distance between each candidate offset point and surrounding obstacles and no-fly zones, and select a subset of candidate offset points that meet the safety distance requirements; In the subset of candidate offset points, select the candidate offset point with the lowest deviation cost from the global flight path and replace the original point to be adjusted; All replaced waypoints and unadjusted waypoints are connected sequentially to form several short straight lines, forming the local route segment set. Based on the average interference index of the traversed area, each short straight line segment is assigned the electromagnetic safety margin level.

5. The autonomous navigation and control method for unmanned aerial vehicles in a strong electromagnetic environment according to claim 4, characterized in that, The optical-assisted positioning module onboard the UAV is activated to acquire image sequences of ground feature points, and the image sequences are spatially registered with the local flight path set, including: The shutter of the optical auxiliary positioning module is controlled to continuously expose within a fixed time interval to acquire multiple frames of image data containing ground textures; A feature point extraction algorithm is performed on the image data to identify corner features with rotation and scale invariance, and the pixel positions of the corner features in the image coordinate system are recorded. Extract the geodetic coordinates of the start and end points of the current target route segment from the set of local route segments, and use them as the reference benchmark for registration; By utilizing the relative positional relationship of the corner features in the image and combining it with the geodetic coordinates of the reference datum, the pose transformation matrix of the current camera is calculated to obtain the real-time six-degree-of-freedom pose estimation of the UAV relative to the ground. The real-time six-degree-of-freedom pose estimation is converted into geographic coordinates, which are used as the current location input for the composite navigation guidance command.

6. The autonomous navigation and control method for unmanned aerial vehicles in a strong electromagnetic environment according to claim 5, characterized in that, A feature point extraction algorithm is performed on the image data to identify corner features that are rotation- and scale-invariant, and the pixel positions of the corner features in the image coordinate system are recorded, including: Perform Gaussian difference processing on the image data to construct the image scale space; In the image scale space, the gray value relationship between each pixel and its neighboring pixels at the same and adjacent scales is detected to locate candidate key points; Edge response culling is performed on the candidate key points, the curvature ratio of the key points is calculated, and points that do not meet the curvature ratio threshold are removed. Assign orientations to the retained keypoints by calculating the main orientation based on the gradient orientation histogram of the neighboring pixels of the keypoints; Based on the scale and principal direction of key points, feature descriptors with rotation and scale invariance are generated. The key points corresponding to each feature descriptor are recorded as corner features with rotation and scale invariance, and their two-dimensional coordinates in the original image plane are extracted as the pixel positions.

7. The autonomous navigation and control method for unmanned aerial vehicles in a strong electromagnetic environment according to claim 1, characterized in that, It also includes real-time evolution learning of the electromagnetic environment and route replanning steps: The temporal variation data of the interference source distribution heatmap output by the electromagnetic environment feature analyzer during the flight of the UAV are continuously recorded. Fourier transform and time-frequency analysis were performed on the time-series variation data to identify the periodic variation patterns and sudden disturbance modes of the electromagnetic environment. When a sudden disturbance pattern is detected, the current route tracking task is paused, and the latest heat map of the interference source distribution is retrieved. The dynamic topology reconstruction process is re-executed to generate a new set of local flight segments adapted to the latest electromagnetic situation; The new local route segment set is compared with the original global route. If the overlap is less than a threshold, the current route is switched to the new local route segment set to continue the navigation task.

8. The autonomous navigation control method for unmanned aerial vehicles in a strong electromagnetic environment according to claim 7, characterized in that, Fourier transform and time-frequency analysis were performed on the time-series variation data to identify the periodic variation patterns and sudden disturbance modes of the electromagnetic environment, including: The time-series change data is divided into several time windows of equal length, and the interference intensity data in each time window is normalized. Perform a short-time Fourier transform on each normalized time window of data to generate a time-spectrum graph; On the time-spectrum graph, a frequency bandwidth of interest is set, and the curve of energy change over time within the frequency bandwidth is extracted; Peak detection is performed on the energy-time curve to identify the moment when energy suddenly increases, and this moment is marked as the suspected start time of sudden disturbance. By tracing back the original data stream of the electromagnetic environment feature analyzer before and after the suspected sudden disturbance start time, it is verified whether there is continuous pulse interference or phase locking phenomenon. If so, it is confirmed as the sudden disturbance mode.

Images