Dispelling system and method for interference of specific-frequency electromagnetic waves on target navigation

Abstract

The invention discloses a system and a method for expelling target navigation interfered by electromagnetic waves with specific frequency, and belongs to the technical field of navigation countermeasure. The method comprises the following steps: acquiring a bird nest coordinate set, and acquiring a stress threshold value of a bird category to which each bird nest belongs; calculating noise influence ranges from inside to outside to form a nest region influence range set; recording the model and target coordinates of the target unmanned aerial vehicle; performing space matching on the target coordinate at the current moment and the nest area influence range set, and outputting a judgment result; when the target is out of the influence range of the nest region, the output frequency is the normal interference frequency; when the target is within the influence range of the nest area, judging whether an environmental noise value is greater than a stress threshold value and whether the bird is in the nest, and performing sub-scene judgment and interference execution; in the navigation interference process, screening out the separation boundary points, and generating a predetermined trajectory; and performing navigation interference on the target based on the predetermined trajectory, and terminating the navigation interference when the current coordinate of the target is separated from the bird breeding place area.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of navigation countermeasures, and particularly relates to a specific frequency electromagnetic wave interference target navigation repelling system and method. BACKGROUND

[0002] With the popularization of civilian unmanned aerial vehicles, the unmanned aerial vehicles are increasingly widely applied in fields such as photography and agriculture, but at the same time, the unmanned aerial vehicles pose a serious threat to ecological sensitive areas, especially in the bird breeding period. The current unmanned aerial vehicle countermeasure technology mainly cuts off the navigation or communication link of the unmanned aerial vehicle through radio frequency interference, GPS spoofing and other ways, wherein the mainstream equipment can cover the commonly used frequency band and realize rapid suppression, but in the special scene of the bird breeding ground, there are significant limitations, and it is difficult to balance the repelling effect and the ecological protection demand.

[0003] The traditional countermeasure equipment adopts a broad spectrum or fixed frequency interference mode, without considering the physiological characteristic differences in the bird breeding period, and the stress threshold of different birds to noise is significantly different, and the tolerance of the same bird in the state of nest bird and non-nest bird is different. The existing technology can only realize interference blinding, and cannot guide the unmanned aerial vehicle to leave in a direction, and the disturbed unmanned aerial vehicle may randomly fall, and if it falls into the influence range of other bird nests, the activity area of the nestling or the ecological sensitive belt, secondary interference or even physical injury will be caused. The existing technology takes the whole breeding ground as a single control unit, without differentiating the nest area, and the distribution and sensitive range of different bird nests in the bird breeding ground are different, and the existing equipment may over-intervene in the target far from the bird nest, or may not intervene enough in the target close to the core nest area. SUMMARY

[0004] The purpose of the present application is to provide a specific frequency electromagnetic wave interference target navigation repelling system and method to solve the problems in the prior art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme: In a first aspect, the present application provides a specific frequency electromagnetic wave interference target navigation repelling method, comprising the following steps: Obtain the three-dimensional coordinates of all bird nests in the bird breeding ground to form a bird nest coordinate set; confirm the bird species to which each bird nest belongs through an AI image recognition algorithm, and obtain the stress threshold of noise in the breeding period; take each bird nest coordinate as a center point, calculate the noise influence range from inside to outside based on the stress threshold, and form a nest area influence range set; Scan the target entering the bird breeding area, analyze the target features by using a visual recognition algorithm, match the unmanned aerial vehicle model database, and record the model of the unmanned aerial vehicle; obtain the target coordinates, and update the motion trajectory; The current time target coordinates are matched with the nest influence range set in space, and a judgment result is output; when the target is outside the nest influence range, an electromagnetic wave is generated, the output frequency is a normal interference frequency, and navigation jamming is performed; when the target is within the nest influence range, the current associated nest environment noise value is obtained, whether there is a bird in the corresponding bird nest of the current associated nest is judged through infrared thermal imaging monitoring, and sub-scene judgment and interference execution are performed according to whether the environment noise value is greater than the stress threshold and whether the bird is in the nest; In the process of navigation jamming, the boundary points are screened out, and a predetermined trajectory is generated; the target trajectory is tracked, the target is navigated based on the predetermined trajectory, and when the current coordinates of the target deviate from the bird breeding area, the navigation jamming is terminated.

[0006] In combination with the first aspect, in a first implementation manner of the first aspect of the present application, the bird species to which each bird nest belongs is confirmed by an AI image recognition algorithm, and the stress threshold of noise in the breeding period is obtained, including: The bird nest morphological features and surrounding activity features of potential bird species in the bird breeding area are collected, the bird nest morphological features include the size, shape, material and nesting location characteristics of the bird nest, and the surrounding activity features include the body size, feather color and behavior pattern of the bird; each bird nest morphological image and each bird activity video frame is labeled with a corresponding bird species label, and a feature key area is labeled to form a labeled data set; for bird nest morphological feature recognition, a convolutional neural network architecture is used to extract spatial features of the bird nest through a convolutional layer, and a fully connected layer outputs a morphological feature vector; for surrounding bird activity feature recognition, a hybrid architecture of convolutional neural network combined with LSTM is used, the appearance features of a single-frame bird image are first extracted through a convolutional neural network, and then the dynamic features of bird activity are captured through an LSTM layer to process multi-frame time series data and output an activity feature vector; the morphological feature vector and the activity feature vector are fused according to weights, input into a Softmax classifier, and a bird species probability distribution is output; The labeled data set is divided into a training set, a validation set and a test set for model training and optimization; the bird nest morphological features and surrounding activity features of the bird nest to be identified are input into the model to obtain the bird species of the bird nest to be identified and the stress threshold of noise in the breeding period.

[0007] In combination with the first aspect, in a second implementation manner of the first aspect of the present application, the noise influence range from inside to outside is calculated based on the stress threshold with each bird nest coordinate as a center point, a nest influence range set is formed, including: The surrounding environmental parameters of the bird nest are measured, including the terrain attenuation coefficient k, the meteorological correction factor and the vegetation density correction coefficient ; a point sound source spherical wave attenuation model is used, and the basic formula is: Wherein, L(r) is the noise value at a distance of r meters from the interference source, is the noise source intensity of the interference system when outputting electromagnetic waves of the normal interference frequency, is adjusted according to the model of the unmanned aerial vehicle of different targets, r is the straight-line distance from the interference source to the target point, 20lg(r) is the spherical wave geometric attenuation term, wherein the normal interference frequency is obtained according to the model of the unmanned aerial vehicle, by calling a preset unmanned aerial vehicle model and normal interference frequency mapping table; the basic formula is modified to obtain an actual noise propagation formula, which specifically includes: in the downwind direction, the downwind noise value , in the upwind direction, the upwind noise value , in the vegetation coverage area, the vegetation coverage noise value ; when there are multiple terrains or vegetation around the bird nest, a ring-shaped area centered on the bird nest is divided according to terrain types, and different areas use corresponding k values for calculation; Solve r, so that , wherein represents the lower limit of the stress threshold, r is back calculated by the actual noise propagation formula; a noise meter is deployed at the calculated r, and the interference device is turned on until , the actual noise value noise is measured; when , the k value is increased, r is recalculated, and measurement is performed again until ; the nest area influence ranges of all bird nests are summarized to form a nest area influence range set.

[0008] In combination with the first aspect, in a third implementation manner of the first aspect of the present application, the target entering the bird breeding area is analyzed for target features by using a visual recognition algorithm, a model database of the unmanned aerial vehicle is matched, and the model of the unmanned aerial vehicle is recorded, including: The radar scans the bird breeding area according to a preset flight line, outputs a moving target list in real time, sets a threshold of the motion features of the unmanned aerial vehicle, excludes non-unmanned aerial vehicle targets, pre-processes the collected images to obtain a set of feature images to be recognized, calls a convolutional neural network algorithm to extract features of the set of feature images to be recognized, including structural features, detail features and thermal features of the target, calls a database of features of models of unmanned aerial vehicles, calculates the similarity between the feature vector of the target to be recognized and the standard feature vector of each model in the database by using cosine similarity, matches the model of the unmanned aerial vehicle and records it.

[0009] In combination with the first aspect, in a fourth implementation manner of the first aspect of the present application, the target coordinates at the current moment are spatially matched with the nest area influence range set, and a judgment result is output, including: The Euclidean distance formula is used to calculate the planar distance between the target and the nest center point, and whether the target coordinate falls within the nest is determined based on the nest influence range set. When there is an overlap between the current nest and other nests, it is simultaneously checked whether the target is within the other nest in the overlap. When the target is within multiple nests, it is marked as being in the nest overlap region. The spatial matching is performed on each effective nest in turn according to the order of the nest distribution from the core area to the peripheral area of the breeding ground, and the judgment result is output.

[0010] In combination with the first aspect, in a fifth implementation manner of the first aspect of the present application, when the target is outside the nest influence range, an electromagnetic wave is generated, the output frequency is a normal interference frequency, and navigation interference is performed, including: Based on the current distance between the target and the interference device, the output power of the power amplifier is set in combination with the normal interference frequency, the azimuth angle and the pitch angle of the target relative to the interference device are calculated according to the target coordinate and the position of the interference device, the beam direction of the directional transmitting antenna is adjusted, and the main beam of the antenna is aligned with the target.

[0011] In combination with the first aspect, in a sixth implementation manner of the first aspect of the present application, when the target is within the nest influence range, the environmental noise value of the current associated nest is obtained, and whether there is a bird in the bird nest corresponding to the current associated nest is judged through infrared thermal imaging monitoring, including: The sensor is started to continuously collect noise, real-time receives and stores the environmental noise value per second, and the environmental noise value is preprocessed. The infrared thermal imager is started to collect the static thermal imaging map of the bird nest and the surrounding area, and to judge whether there is a heat source area higher than the environmental temperature in the bird nest area. The dynamic video collection mode is switched to, and it is judged whether the heat source area has dynamic changes, and non-bird heat sources are excluded. When there is a heat source area conforming to the characteristics of birds in the static thermal imaging map, and bird activity characteristics are observed in the dynamic monitoring of the area, it is determined that there is a bird in the bird nest of the current associated nest.

[0012] In combination with the first aspect, in a seventh implementation manner of the first aspect of the present application, the judgment of whether the environmental noise value is greater than the stress threshold and whether the bird is in the nest is performed, and the sub-scene judgment and interference execution are performed, including: In a first sub-scenario, the environmental noise value is greater than the stress threshold and the birds are in the nest, a normal interference frequency corresponding to the target is obtained, and the minimum power capable of suppressing the navigation system of the target is set according to the distance between the target and the interference device; in a second sub-scenario, the environmental noise value is greater than the stress threshold and the birds are in the nest, the frequency is the same as that in the first sub-scenario, and the power is increased on the basis of the power set in the first sub-scenario on the premise that no excessive out-of-band radiation is generated and no legal device in the surrounding area is disturbed; in a third sub-scenario, the environmental noise value is not greater than the stress threshold and the birds are in the nest, an initial interference frequency is set, which is lower than the normal interference frequency corresponding to the target, to ensure that the noise generated by the initial interference signal and the current environmental noise after superposition still does not exceed the stress threshold, and the minimum safe power is set, which is lower than the power in the first scenario, to avoid that the initial power is too high to cause the noise to break through the stress threshold; when the target does not move, the interference frequency is increased by a set proportion; when the superposition still does not exceed the stress threshold, the new frequency is maintained, and when the superposition exceeds the stress threshold, the frequency falls back to the previous frequency and does not continue to increase; the interference frequency after the increase cannot exceed the normal interference frequency corresponding to the target; in a fourth sub-scenario, the environmental noise value is not greater than the stress threshold and the birds are not in the nest, the normal interference frequency corresponding to the target is used, and the power is set to be the power in the first sub-scenario.

[0013] In combination with the first aspect, in an eighth implementation manner of the first aspect of the present application, in the process of performing navigation interference, the disengagement boundary point is screened, and the predetermined trajectory is generated, including: The screening condition includes, in order from high to low according to priority, a point on the boundary of the breeding ground; a straight line distance from the point to the safe drop-off area is relatively shorter among all boundary points; a line connecting the target current position and the point crosses the least number of nest area influence ranges; and the point and the surrounding area have no ecological sensitive area; The boundary line of the breeding ground is decomposed into a plurality of continuous boundary points, the disengagement boundary point is obtained by screening according to the screening condition, the target coordinates are taken as the starting point of the predetermined trajectory, the disengagement boundary point is taken as the end point of the predetermined trajectory, the starting point and the end point are connected to form an initial straight line trajectory, and the trajectory constraint is set, including: not entering any nest area influence range in the whole process; minimizing the total length deviation of the trajectory from the straight line distance between the starting point and the end point; and the trajectory does not cross the protection range of the ecological sensitive area and the legal device; When the initial straight line trajectory completely meets the trajectory constraint, the initial straight line trajectory is taken as the predetermined trajectory; when the initial straight line trajectory crosses any region in the trajectory constraint, a path inflection point that avoids the region and has the least distance increase is found on both sides of the crossed region to form an initial predetermined trajectory; the inflection point is smoothed for the initial predetermined trajectory; whether the initial predetermined trajectory after the smoothing processing is within the executable range of the target is judged in combination with the flight capability of the target, the trajectory inflection point position is further adjusted when it is beyond the executable range, until the trajectory meets the flight capability of the target, and the predetermined trajectory is formed.

[0014] In a second aspect, the application provides a system for expelling a target from navigation interference by specific frequency electromagnetic waves, comprising: The nest area influence range generation module comprises a bird nest coordinate positioning unit, a stress threshold acquisition unit and a nest area influence range generation unit, wherein the bird nest coordinate positioning unit acquires three-dimensional coordinates of all bird nests in the bird breeding area to form a bird nest coordinate set; the stress threshold acquisition unit confirms the bird species to which each bird nest belongs by an AI image recognition algorithm and acquires the stress threshold of noise in the breeding period of the bird species; and the nest area influence range generation unit takes each bird nest coordinate as a center point, calculates the noise influence range from inside to outside based on the stress threshold, and forms a nest area influence range set. The target analysis module comprises a UAV model acquisition unit and a target coordinate acquisition unit, wherein the UAV model acquisition unit scans the target entering the bird breeding area, analyzes the target features by a visual recognition algorithm, matches a UAV model database, and records the model of the UAV; and the target coordinate acquisition unit acquires the target coordinates and updates the motion trajectory of the target. The navigation interference execution module comprises a space matching unit, an out-of-range interference execution unit, an in-range interference data acquisition unit and an in-range interference execution unit, wherein the space matching unit performs space matching between the target coordinates at the current moment and the nest area influence range set and outputs a judgment result; the out-of-range interference execution unit generates electromagnetic waves when the target is outside the nest area influence range, outputs a normal interference frequency, and performs navigation interference; the in-range interference data acquisition unit acquires the environmental noise value of the current associated nest area when the target is in the nest area influence range, judges whether there are birds in the bird nest corresponding to the current associated nest area by infrared thermal imaging monitoring; and the in-range interference execution unit judges whether the environmental noise value is greater than the stress threshold and whether there are birds in the nest, performs sub-scene judgment and interference execution. The predetermined trajectory generation and execution module comprises a predetermined trajectory generation unit and a predetermined trajectory execution unit, wherein the predetermined trajectory generation unit generates a predetermined trajectory by screening a boundary point during navigation interference; and the predetermined trajectory execution unit traces the target trajectory, performs navigation interference on the target based on the predetermined trajectory, and terminates the navigation interference when the current coordinates of the target are out of the bird breeding area.

[0015] Compared with the prior art, the application has the following beneficial effects: 1. The application confirms the bird species of each bird nest by AI image recognition to acquire a specific stress threshold, and calculates the nest area influence range based on the stress threshold; when interference is performed in the nest area, the environmental noise value and the bird state in the nest are further combined to perform interference in different scenes, instead of using a unified interference mode, so that the bird tolerance can be accurately matched, and ecological risks caused by interference signals breaking through the stress threshold can be avoided.

[0016] 2. This invention filters out points that deviate from the boundary and generates a predetermined trajectory during the interference process. By tracking the target trajectory and implementing interference based on the predetermined trajectory, it ensures that the target moves out of the breeding ground along a preset safe path. The target is only allowed to fall after it leaves the breeding ground area, thus completely avoiding the problem of the target randomly falling into sensitive areas such as the nesting area and the chick activity area.

[0017] 3. This invention first constructs a set of nest area influence ranges, and then determines whether the target is inside or outside the nest area by matching the current target coordinates with the set. Then, it executes corresponding interference strategies to achieve differentiated control based on the nest area, taking into account both the effectiveness of interference and the accuracy of control. Attached Figure Description

[0018] Fig. 1 This is a schematic diagram of the steps of a method for driving away navigation targets that are interfered with by electromagnetic waves of a specific frequency, according to the present invention. Fig. 2 This is a system structure diagram of a target navigation deterrent system that uses electromagnetic waves of a specific frequency to interfere with the target, according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example: Figs. 1-2 As shown, the present invention provides a technical solution: like Fig. 1 As shown, this application provides a method for driving away targets whose navigation is interfered with by electromagnetic waves of a specific frequency, including the following steps: Step S100: Obtain the three-dimensional coordinates of all bird nests within the bird breeding grounds to form a set of bird nest coordinates; use AI image recognition algorithms to identify the bird species to which each nest belongs and obtain its stress threshold for noise during the breeding season; take each bird nest coordinate as the center point and calculate the noise impact range from the inside out based on the stress threshold to form a set of nest area impact ranges. Specifically, bird nest morphological features and surrounding activity features of potential bird species in the bird breeding ground are collected, the bird nest morphological features include size, shape, material and nest location characteristics of the bird nest, and the surrounding activity features include body size, feather color and behavior pattern of the bird; a corresponding bird class label is labeled for each bird nest morphological image and each bird activity video frame, and a feature key area is labeled to form a labeled data set; for bird nest morphological feature recognition, a convolutional neural network architecture is adopted, spatial features of the bird nest are extracted through a convolutional layer, and a morphological feature vector is output by a fully connected layer; for surrounding bird activity feature recognition, a hybrid architecture of convolutional neural network combined with LSTM is adopted, appearance features of a single-frame bird image are extracted through a convolutional neural network first, then dynamic features of bird activity are captured through an LSTM layer to process multi-frame time series data, and an activity feature vector is output; the morphological feature vector and the activity feature vector are fused according to weights, input into a Softmax classifier, and a bird class probability distribution is output. The labeled data set is divided into a training set, a validation set and a test set for model training and optimization; bird nest morphological features and surrounding activity features of a bird nest to be identified are collected and input into the model to obtain a bird class of the bird nest to be identified and obtain a stress threshold of the bird nest to noise in a breeding period.

[0021] Further, the bird nest surrounding environment parameters are measured, including a terrain attenuation coefficient k, a meteorological correction factor and a vegetation density correction coefficient ; a point sound source spherical wave attenuation model is adopted, and a basic formula is as follows: wherein, L(r) is a noise value at a distance of r meters from an interference source, is a noise source intensity of an interference system when outputting an electromagnetic wave of a normal interference frequency, is adjusted according to different models of unmanned aerial vehicles, r is a straight-line distance from the interference source to a target point, and 20lg(r) is a spherical wave geometric attenuation term, wherein the normal interference frequency is obtained by calling a preset unmanned aerial vehicle model and normal interference frequency mapping table according to the unmanned aerial vehicle model; the basic formula is corrected to obtain an actual noise propagation formula, which specifically includes: in a downwind direction, a downwind noise value , in an upwind direction, an upwind noise value , and in a vegetation coverage area, a vegetation coverage noise value ; when there are multiple terrains or vegetation around the bird nest, a ring-shaped area centered on the bird nest is segmented according to terrain types, and different areas are calculated by using corresponding k values; r is solved, so that wherein, represents a lower limit of the stress threshold, r is back calculated through the actual noise propagation formula; a noise meter is deployed at the calculated r, and an interference device is turned on to , and an actual noise value noise is measured; when When k is increased, r is recalculated, and the measurement is repeated until ; all nest area influence ranges of the bird nests are summarized to form a nest area influence range set.

[0022] In a specific embodiment, a certain coastal wetland bird breeding ground is taken as the experimental scene, and the target birds are red-bellied sandpipers and spoonbills. 60 morphological samples of each of the two types of birds (red-bellied sandpiper nests with a diameter of 5-7 cm, bowl-shaped, dry branch material, and spoonbill nests with a diameter of 4-6 cm, disc-shaped, reed material) are collected, and 120 bird activity videos of each type are collected (red-bellied sandpipers with a body length of 23-25 cm, gray-brown back, and spoonbills with a body length of 14-16 cm, shovel-shaped beak). A labeled data set is formed (70% training set, 20% validation set, and 10% test set). After model training, the ResNet50 architecture has an accuracy of 96% in identifying nest morphological features, and the CNN+LSTM architecture has an accuracy of 94% in identifying bird activity features. For the 15 bird nests to be identified, the model outputs 10 red-bellied sandpipers (all with a probability of ≥92%) and 5 spoonbills (all with a probability of ≥91%), and the matching ecological database shows that the lower limit of the red-bellied sandpiper hatching period stress threshold is 44 dB, and the lower limit of the spoonbill hatching period stress threshold is 42 dB.

[0023] In the experimental breeding ground, three typical nest areas are selected (No. 1 is a beach terrain, No. 2 is a reed terrain, and No. 3 is a beach-reed mixed terrain). The environmental parameters are measured: the attenuation coefficient k of No. 1 beach terrain is 0.2 dB / m, and the attenuation coefficient k of No. 2 reed terrain is 0.35 dB / m; the wind speed during the experiment is 4 m / s, the weather correction factor f_w in the downwind direction is 0.05 dB / m, and the weather correction factor f_w in the upwind direction is 0.05 dB / m; the vegetation coverage of No. 1 nest area is 20% (f_v=0), the coverage of No. 2 nest area is 70% (f_v=0.06 dB / m), and the coverage of No. 3 nest area is 15% (f_v=0) in the beach part and 65% (f_v=0.05 dB / m) in the reed part.

[0024] Taking No. 1 nest area (red-bellied sandpiper, Tt=44 dB) as an example, the interference source is the normal interference frequency (1561 MHz) for DJI Mavic3, and the corresponding noise source intensity L0 is 83 dB. Substituting into the downwind propagation formula of the beach terrain, it is preliminarily calculated that r=41 meters; at r=41 meters, a noise meter is deployed to measure the actual noise value noise=45.2 dB (>44 dB), and k is increased to 0.22 dB / m to recalculate r=39 meters; the noise is measured again noise=43.8 dB (≤44 dB), and the influence range radius of No. 1 nest area is determined to be 39 meters. According to the reed terrain upwind formula, the influence range radius of No. 2 nest area (spoonbill, Tt=42 dB) is finally determined to be 32 meters, and the actual measured noise value is 41.7 dB.

[0025] Step S200: Scan targets entering bird breeding areas, analyze target features using visual recognition algorithms, match them with the UAV model database, record the UAV model; obtain target coordinates, and update its movement trajectory; Specifically, the radar scans bird breeding areas along a preset route, outputs a list of moving targets in real time, sets a threshold for UAV motion characteristics to exclude non-UAV targets, preprocesses the acquired images to obtain a set of feature images to be identified, calls a convolutional neural network algorithm to extract features from the set of feature images to be identified, including structural features, detailed features, and thermal features of the targets, calls a UAV model feature database, uses cosine similarity to calculate the similarity between the feature vector of the target to be identified and the standard feature vector of each model in the database, matches the UAV model and records it.

[0026] In one specific embodiment, the coastal wetland breeding area in step S100 is used as the experimental scenario. An X-band radar (with a scanning range covering the entire breeding area and a 500-meter buffer zone, an azimuth angle of 360°, and a range accuracy of ±3 meters) is deployed and continuously scans along a circular route. During a certain period, the radar outputs 8 moving targets. The drone motion characteristic thresholds are set (flight speed 2-15 m / s, altitude 5-100 meters, smooth trajectory): 2 floating objects with a speed <2 m / s (balloons, fallen leaves), 1 bird with an altitude >100 meters, and 1 insect with a messy trajectory are excluded. The remaining 4 suspected drone targets trigger the camera to perform linked imaging.

[0027] For four suspected targets, images were captured using a 2-megapixel high-definition camera (1920×1080 resolution, 25fps). After Gaussian filtering for noise reduction and contrast enhancement, a set of feature images to be identified was obtained. A ResNet-18 convolutional neural network was then used to extract features: Target 1 has a 3:1 aspect ratio, a quadcopter structure (structural feature vector dimension 256), a "DJI" logo printed on its fuselage (12 detailed key features), and infrared images showing four thermal highlights from its motors (thermal feature vector dimension 64). The corresponding feature vectors for the other three targets were extracted similarly.

[0028] The system utilizes a drone model feature database (containing standard feature vectors for 10 common models including DJI Mavic 3, Xiaomi FIMIX 8, and Haberson Zino Pro) and calculates cosine similarity. Target 1's feature vector shows 93% similarity to the DJI Mavic 3's standard vector, while its similarity to other models is less than 75%. Target 2 shows 91% similarity to the Xiaomi FIMIX 8, and targets 3 and 4 show 90% and 89% similarity to the Haberson Zino Pro, respectively. The four target models are ultimately recorded as DJI Mavic 3, Xiaomi FIMIX 8, Haberson Zino Pro, and Haberson Zino Pro.

[0029] Step S300: spatially match the target coordinate at the current time with the nest influence range set, output the judgment result; when the target is outside the nest influence range, generate electromagnetic waves, output the frequency as the normal interference frequency, and perform navigation interference; when the target is within the nest influence range, obtain the environmental noise value of the current associated nest, monitor through infrared thermal imaging, determine whether there are birds in the bird nest corresponding to the current associated nest, and perform sub-scene judgment and interference execution. Specifically, the plane distance between the target and the nest center point is calculated by using the Euclidean distance formula, and whether the target coordinate falls within the nest is determined based on the nest influence range set. When there is an overlap between the current nest and other nests, it is simultaneously checked whether the target is within the overlapping other nest. When the target is within multiple nest ranges, it is marked that the target is in the nest overlap region. According to the order of nest distribution from the core area to the peripheral area, spatial matching is performed on each effective nest in turn, and the judgment result is output.

[0030] Further, based on the distance between the target and the interference device at the current time, the output power of the power amplifier is set in combination with the normal interference frequency, the azimuth and elevation of the target relative to the interference device are calculated according to the target coordinate and the position of the interference device, the beam direction of the directional transmitting antenna is adjusted so that the main beam of the antenna is aligned with the target.

[0031] Further, the sensor is started to continuously collect noise, real-time receives and stores the environmental noise value per second, and pre-processes the environmental noise value. The infrared thermal imager is started to collect the static thermal imaging map of the bird nest and the surrounding area, and to determine whether there is a heat source area higher than the environmental temperature in the bird nest area. Switch to the dynamic video acquisition mode to determine whether there is a dynamic change in the heat source area, and to exclude non-bird heat sources. When there is a heat source area conforming to the characteristics of birds in the static thermal imaging map, and bird activity characteristics are observed in the dynamic monitoring of the area, it is determined that there are birds in the bird nest of the current associated nest.

[0032] Further, in the first sub-scenario, the environmental noise value is greater than the stress threshold and the birds are in the nest, the normal interference frequency corresponding to the target is obtained, and the minimum power capable of suppressing the navigation system of the target is set according to the distance between the target and the interference device; in the second sub-scenario, the environmental noise value is greater than the stress threshold and the birds are in the nest, the frequency is the same as that in the first sub-scenario, and the power is increased on the basis of the power set in the first sub-scenario on the premise that no excessive out-of-band radiation is generated and no legal device in the surrounding area is disturbed; in the third sub-scenario, the environmental noise value is not greater than the stress threshold and the birds are in the nest, the initial interference frequency is set, which is lower than the normal interference frequency corresponding to the target, to ensure that the noise generated by the initial interference signal after superposition with the current environmental noise still does not exceed the stress threshold, and the minimum safe power is set, which is lower than the power in the first scenario, to avoid that the initial power is too high to cause the noise to break through the stress threshold; when the target does not move, the interference frequency is increased by a set proportion; when the superposition still does not exceed the stress threshold, the new frequency is maintained, and when it exceeds, the frequency falls back to the previous frequency and does not continue to increase; the interference frequency after the increase cannot exceed the normal interference frequency corresponding to the target; in the fourth sub-scenario, the environmental noise value is not greater than the stress threshold and the birds are not in the nest, the normal interference frequency corresponding to the target is set, which is the power in the first sub-scenario.

[0033] In an embodiment, the DJI Mavic3 identified in step S200 is taken as the target (current coordinates: 120.352°E, 30.121°N), and the No. 1 red-bellied sandpiper nest area (center point coordinates: 120.350°E, 30.120°N, influence range radius 39 meters) and the No. 2 spoon-billed sandpiper nest area (center point coordinates: 120.353°E, 30.122°N, influence range radius 32 meters) are called. The Euclidean distance formula is used to calculate that the distance between the target and the center point of the No. 1 nest area is 28 meters (≤39 meters), the distance between the target and the center point of the No. 2 nest area is 21 meters (≤32 meters), and there is a 15-meter overlapping area between the two nest areas. Finally, it is determined that the target is in the overlapping range of the No. 1 and No. 2 nest areas.

[0034] The interference device is deployed at coordinates 120.348°E, 30.118°N, and the straight-line distance between the target and the interference device is 300 meters. According to the normal interference frequency 1561MHz of the DJI Mavic3, the initial value of the power amplifier output power is set to 22dBm in combination with the distance; the azimuth angle of the target relative to the interference device is calculated to be 18° east of north and the pitch angle is 3°, and the directional antenna beam is adjusted to this angle to ensure that the main beam covers the target.

[0035] For the associated 1st nest area (stress threshold lower limit 44dB), the noise sensor continuously collects 30 seconds of environmental noise values, and the average noise value after preprocessing is 46dB; the infrared thermal imager (resolution 640x512) is started synchronously, and an irregular oval heat source (environmental temperature 25℃) of 39-41℃ is collected in the bird nest area. Dynamic monitoring for 5 minutes observes that the heat source has head rotation and slight standing-up action, and excludes non-bird heat sources such as stones, and determines that there is a bird in the 1st nest area.

[0036] First sub-scenario (noise 46dB>44dB, with birds): use 1561MHz normal interference frequency, set power to 22dBm (just suppress target navigation), after 10 seconds of interference, the target starts to move out of the nest area; Second sub-scenario (noise 46dB>44dB, with birds): frequency remains 1561MHz, power is increased to 25dBm (no out-of-band radiation exceeds the standard), target moving speed is increased from 0.8m / s to 1.2m / s; Third sub-scenario (replace nest area: 3rd nest area noise 42dB≤44dB, with birds): initial frequency 1171MHz (normal frequency 75%), power 13dBm, target no movement, increase to 1239MHz by 5% proportion, superimposed noise 43.5dB≤44dB, maintain the frequency; Fourth sub-scenario (3rd nest area noise 42dB≤44dB, without birds): use 1561MHz normal frequency, set power to 22dBm, after 8 seconds of interference, the target moves away from the boundary point.

[0037] Step S400: In the process of navigation interference, screen the boundary point, generate a predetermined trajectory; track the target trajectory, navigate the target based on the predetermined trajectory, and terminate the navigation interference when the current coordinates of the target are out of the bird breeding area.

[0038] Specifically, the screening conditions include, in order of priority from high to low: the point belongs to the boundary of the breeding ground; the straight-line distance from the point to the safe landing area is relatively shorter among all boundary points; the line connecting the current position of the target and the point passes through the least number of nest area influence ranges; the point and the surrounding area have no ecological sensitive area; The breeding ground boundary line is decomposed into a plurality of continuous boundary points, which are screened according to the screening conditions to obtain the boundary point; the target coordinates are taken as the starting point of the predetermined trajectory, and the boundary point is taken as the end point of the predetermined trajectory, the starting point and the end point are connected to form an initial straight-line trajectory; the trajectory constraints include: not entering any nest area influence range; the total length of the trajectory is minimized with respect to the straight-line distance from the starting point to the end point; the trajectory does not pass through the protection range of the ecological sensitive area and the legal device; When the initial straight trajectory completely meets the trajectory constraint, the initial straight trajectory is taken as the predetermined trajectory; when the initial straight trajectory passes through any region in the trajectory constraint, on both sides of the passing region, a path inflection point that bypasses the region and increases the distance the least is found to form an initial predetermined trajectory; the inflection point is smoothed for the initial predetermined trajectory; in combination with the flight capability of the target, it is judged whether the initial predetermined trajectory after smoothing meets the executable range of the target, when it exceeds, the trajectory inflection point position is further adjusted until the trajectory meets the flight capability of the target to form the predetermined trajectory.

[0039] In a specific embodiment, taking a coastal wetland breeding ground as a scene, the breeding ground boundary is a polygon (including a 500-meter buffer zone), and the safe drop zone is the empty beach on the east side outside the boundary (no nest area, no sensitive area). The boundary line of the breeding ground is decomposed into 50 continuous boundary points, which are screened according to the priority of the screening condition: 20 points with a straight-line distance to the safe drop zone less than 80 meters are preferentially retained, and then 8 points with a line connecting the target current coordinates (120.352°E, 30.121°N) and the point only passing through one nest area are screened, and finally the boundary point (120.358°E, 30.123°N) with no ecological sensitive area within 10 meters is determined as the escape boundary point, which is 72 meters away from the safe drop zone and the line passes through one nest area (only the edge of the No. 2 spoonbill nest area).

[0040] Taking the target current coordinates (120.352°E, 30.121°N) as the starting point and the above escape boundary point as the ending point, an initial straight trajectory is formed, with a length of about 110 meters. Check the trajectory constraint: it is found that the straight trajectory passes through the influence range of the No. 2 nest area (radius 32 meters) at the middle section (120.355°E, 30.122°N) and is close to the chick activity sensitive area (distance 5 meters), which does not meet the constraint of not entering the nest area and sensitive area, and needs to be adjusted.

[0041] Set inflection points on both sides of the initial trajectory passing through the No. 2 nest area: inflection point 1 (120.354°E, 30.120°N) and inflection point 2 (120.356°E, 30.120°N), forming an initial predetermined trajectory of "starting point-inflection point 1-inflection point 2-ending point", with a total length of 125 meters (deviation 13.6% from the original straight distance). Smooth the inflection points: adjust the polyline between inflection point 1 and inflection point 2 to a gentle curve (turning angle from original 60° to 35°), and verify that the adjusted trajectory meets the executable range of the target in combination with the flight capability of the DJI Mavic3 with a maximum turning angle of 40°. The final predetermined trajectory has a total length of 128 meters, with a deviation of 16.4% or less.

[0042] After starting navigation interference, update target coordinates every 5 seconds: when the interference is 10 seconds, the target reaches inflection point 1 (120.354°E, 30.120°N), 20 seconds to reach inflection point 2 (120.356°E, 30.120°N), 35 seconds to reach the escape boundary point (120.358°E, 30.123°N). Continue to monitor for 5 seconds, confirm the target coordinates as (120.359°E, 30.124°N), which has escaped the breeding ground boundary (8 meters away from the boundary), and is moving towards the safe drop zone, immediately terminate the interference, the target moves along the predetermined trajectory without deviation.

[0043] As shown in Fig. 2 The application provides a specific frequency electromagnetic wave interference target navigation driving system, which comprises: The nest area influence range generation module comprises a bird nest coordinate positioning unit, a stress threshold acquisition unit, and a nest area influence range generation unit. The bird nest coordinate positioning unit obtains the three-dimensional coordinates of all bird nests in the bird breeding area to form a bird nest coordinate set. The stress threshold acquisition unit confirms the bird species of each bird nest through an AI image recognition algorithm and obtains the stress threshold of noise during the breeding period. The nest area influence range generation unit takes each bird nest coordinate as the center point and calculates the noise influence range from the inside to the outside based on the stress threshold to form a nest area influence range set. The target analysis module comprises a UAV model acquisition unit and a target coordinate acquisition unit. The UAV model acquisition unit scans the target entering the bird breeding area, analyzes the target features using a visual recognition algorithm, matches the UAV model database, and records the model of the UAV. The target coordinate acquisition unit obtains the target coordinates and updates the motion trajectory. The navigation interference execution module comprises a space matching unit, an out-of-range interference execution unit, an in-range interference data acquisition unit, and an in-range interference execution unit. The space matching unit performs spatial matching between the target coordinates at the current time and the nest area influence range set and outputs the judgment result. The out-of-range interference execution unit generates electromagnetic waves when the target is outside the nest area influence range, outputs the frequency as the normal interference frequency, and performs navigation interference. The in-range interference data acquisition unit obtains the environmental noise value of the current associated nest area when the target is within the nest area influence range, judges whether there are birds in the bird nest corresponding to the current associated nest area through infrared thermal imaging monitoring. The in-range interference execution unit judges whether the environmental noise value is greater than the stress threshold and whether there are birds in the nest, performs sub-scene judgment and interference execution. The predetermined trajectory generation and execution module comprises a predetermined trajectory generation unit and a predetermined trajectory execution unit; the predetermined trajectory generation unit generates a predetermined trajectory by screening out the boundary departure point in the process of navigation interference; the predetermined trajectory execution unit tracks the target trajectory, performs navigation interference on the target based on the predetermined trajectory, and terminates the navigation interference when the current coordinates of the target deviate from the bird breeding area.

[0044] It will be obvious to a person skilled in the art that, without departing from the spirit or essential characteristics of the application, the present application can be implemented in other specific forms. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the description given above, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims should not be construed as limiting the claim concerned.

Claims

1. A method of repelling a target from navigation by a specific frequency electromagnetic wave interference, characterized by, The method comprises the following steps: acquire three-dimensional coordinates of all nests in the bird breeding area to form a nest coordinate set; confirm the bird species of each nest through an AI image recognition algorithm, and acquire the stress threshold of the bird species to noise during the breeding period; take each nest coordinate as a center point, calculate the noise influence range from the inside to the outside based on the stress threshold, and form a nest area influence range set; scan a target entering the bird breeding area, analyze the target features by using a visual recognition algorithm, match a UAV model database, and record the model of the UAV; acquire the target coordinates and update the motion trajectory of the target; perform spatial matching between the target coordinates at the current moment and the nest area influence range set, and output a judgment result; when the target is outside the nest area influence range, generate electromagnetic waves, output a normal interference frequency, and perform navigation interference; when the target is inside the nest area influence range, acquire the environmental noise value of the current associated nest area, judge whether there is a bird in the nest corresponding to the current associated nest area through infrared thermal imaging monitoring, judge whether the environmental noise value is greater than the stress threshold and whether there is a bird in the nest, perform sub-scene judgment and interference execution; in the process of performing navigation interference, screen out the boundary points that are separated, and generate a predetermined trajectory; track the target trajectory, perform navigation interference on the target based on the predetermined trajectory, and terminate the navigation interference when the current coordinates of the target are separated from the bird breeding area.

2. A method of repelling a target from navigation by interfering with a specific frequency electromagnetic wave according to claim 1, characterized in that, The method of confirming the bird species of each nest through the AI image recognition algorithm and acquiring the stress threshold of the bird species to noise during the breeding period comprises the following steps: collect the nest morphological features and surrounding activity features of potential bird species in the bird breeding area, the nest morphological features include the size, shape, material and nest building location characteristics of the nest, and the surrounding activity features include the body size, feather color and behavior pattern of the bird; label the corresponding bird species label for each nest morphological image and each bird activity video frame, label the feature key area, and form a labeled data set; for nest morphological feature recognition, a convolutional neural network architecture is adopted, spatial features of the nest are extracted through a convolutional layer, and a morphological feature vector is output by a full connection layer; for surrounding bird activity feature recognition, a hybrid architecture of a convolutional neural network combined with LSTM is adopted, appearance features of a single-frame bird image are extracted through a convolutional neural network, dynamic features of bird activity are captured by processing multi-frame time series data through an LSTM layer, and an activity feature vector is output; the morphological feature vector and the activity feature vector are fused according to weights, input into a Softmax classifier, and a bird species probability distribution is output; divide the labeled data set into a training set, a validation set and a test set, perform model training and optimization, collect the nest morphological features and surrounding activity features of a to-be-identified nest, input the features into the model, obtain the bird species of the to-be-identified nest, and acquire the stress threshold of the bird species to noise during the breeding period.

3. The method of claim 1, wherein the electromagnetic wave of a specific frequency is a radio frequency wave. The method of taking each nest coordinate as a center point, calculating the noise influence range from the inside to the outside based on the stress threshold, and forming a nest area influence range set comprises the following steps: The bird nest surrounding environment parameters are measured, including terrain attenuation coefficient k, meteorological correction factor and vegetation density correction coefficient The basic formula is as follows: Wherein, L(r) is the noise value at a distance of r meters from the interference source, is the noise source intensity when the interference system outputs electromagnetic waves of normal interference frequency, is adjusted according to the model of different target unmanned aerial vehicles, r is the straight-line distance from the interference source to the target point, and 20lg(r) is the spherical wave geometric attenuation term, wherein the normal interference frequency is obtained by calling a preset normal interference frequency and unmanned aerial vehicle model mapping table according to the model of the unmanned aerial vehicle; the basic formula is corrected to obtain an actual noise propagation formula, specifically including: in the wind direction, the downwind noise value , in the adverse wind direction, the adverse wind noise value , and in the vegetation coverage area, the vegetation coverage noise value When there are multiple terrains or vegetation around the bird nest, a ring-shaped area with the bird nest as the center is divided according to terrain types, and different regions are calculated by using corresponding k values.​ Solve r, so that wherein, represents the lower limit of the stress threshold, and r is deduced by the actual noise propagation formula; a noise meter is deployed at the calculated r, and the interference device is turned on to , the actual noise value noise is measured; when , the value of k is increased, r is recalculated, and measurement is performed again until ; the nest area influence ranges of all bird nests are summarized to form a nest area influence range set.

4. The method of claim 1, wherein the electromagnetic wave of a specific frequency is a radio frequency wave. The method of scanning a target entering the bird breeding area, analyzing the target features by using a visual recognition algorithm, matching a UAV model database, and recording the model of the UAV comprises the following steps: The radar scans the bird breeding area along the preset route, outputs a moving target list in real time, sets a threshold of the motion characteristics of the unmanned aerial vehicle, excludes non-unmanned aerial vehicle targets, pre-processes the collected images to obtain a feature image set to be identified, calls a convolutional neural network algorithm to extract features of the feature image set to be identified, including structural features, detail features and thermal features of the target, calls a database of unmanned aerial vehicle model features, calculates the similarity between the feature vector of the target to be identified and the standard feature vector of each model in the database using cosine similarity, matches the model of the unmanned aerial vehicle and records the model.

5. The method of claim 1, wherein the electromagnetic waves are of a specific frequency. The current time target coordinate is matched with the nest area influence range set in space, and a judgment result is output, including: The plane distance between the target and the nest center point is calculated using the Euclidean distance formula, and whether the target coordinate falls within the nest area is determined based on the nest influence range set. When there is an overlap between the current nest area and other nest areas, it is simultaneously checked whether the target is within the other overlapping nest area. When the target is within the range of multiple nest areas, the target is marked as being in the nest area overlap region. The spatial matching is sequentially performed on each effective nest area in the order of nest distribution from the core area to the peripheral area of the breeding ground, and the judgment result is output.

6. The method of claim 1, wherein the electromagnetic waves are of a specific frequency. When the target is outside the nest influence range, an electromagnetic wave is generated, the output frequency is a normal interference frequency, and navigation interference is performed, including: Based on the distance between the target and the interference device, the output power of the power amplifier is set in combination with the normal interference frequency. According to the target coordinate and the position of the interference device, the azimuth angle and the pitch angle of the target relative to the interference device are calculated, the beam direction of the directional transmitting antenna is adjusted, and the main beam of the antenna is aligned with the target.

7. The method of claim 1, wherein the electromagnetic waves are of a specific frequency. When the target is within the nest influence range, the environmental noise value of the current associated nest area is obtained, and it is judged whether there is a bird in the bird nest corresponding to the current associated nest area through infrared thermal imaging monitoring, including: The sensor is started to continuously collect noise, and the environmental noise value per second is received and stored in real time. The environmental noise value is pre-processed. The infrared thermal imager is started, and the static thermal imaging map of the bird nest and the surrounding area is collected. It is judged whether there is a heat source area higher than the environmental temperature in the bird nest area. Switch to dynamic video acquisition mode to judge whether the heat source area has dynamic changes and exclude non-bird heat sources. When there is a heat source area conforming to the characteristics of birds in the static thermal imaging map, and bird activity characteristics are observed in the dynamic monitoring of the area, it is determined that there is a bird in the bird nest of the current associated nest area.

8. The method of claim 1, wherein the electromagnetic waves are of a specific frequency. It is judged whether the environmental noise value is greater than the stress threshold and whether the bird is in the nest, and the sub-scene judgment and interference execution are performed, including: In a first sub-scenario, the environmental noise value is greater than the stress threshold and the birds are in the nest, a normal interference frequency corresponding to the target is obtained, and the minimum power capable of suppressing the navigation system of the target is set according to the distance between the target and the interference device; in a second sub-scenario, the environmental noise value is greater than the stress threshold and the birds are in the nest, the frequency is the same as that in the first sub-scenario, and the power is increased on the basis of the power set in the first sub-scenario on the premise that no excessive out-of-band radiation is generated and no interference is caused to the surrounding legal devices; in a third sub-scenario, the environmental noise value is not greater than the stress threshold and the birds are in the nest, an initial interference frequency is set, which is lower than the normal interference frequency corresponding to the target, to ensure that the noise generated by the initial interference signal and the current environmental noise after superposition still does not exceed the stress threshold, and the minimum safe power is set, which is lower than the power in the first scenario, to avoid that the initial power is too high to cause the noise to break through the stress threshold; when the target does not move, the interference frequency is increased at a set proportion; when the superposition still does not exceed the stress threshold, the new frequency is maintained, and when the superposition exceeds the stress threshold, the frequency falls back to the previous frequency and does not continue to increase; the interference frequency after the increase cannot exceed the normal interference frequency corresponding to the target; and in a fourth sub-scenario, the environmental noise value is not greater than the stress threshold and the birds are not in the nest, the normal interference frequency corresponding to the target is used, and the power is set to be the power in the first sub-scenario.

9. The method of claim 1, wherein the electromagnetic waves are of a specific frequency. In the process of performing navigation interference, the disengagement boundary point is screened, and a predetermined trajectory is generated, including: The disengagement boundary point is screened, and the screening conditions include, in order from high to low according to priority, a point on the boundary of the breeding ground; a straight line distance from the point to the safe drop-off area is relatively shorter among all boundary points; a line connecting the current position of the target and the point passes through the least number of nest area influence ranges; and there is no ecological sensitive area around the point and the surrounding area; The boundary line of the breeding ground is decomposed into a plurality of continuous boundary points, the disengagement boundary point is obtained by screening according to the screening conditions, the coordinates of the target are taken as the starting point of the predetermined trajectory, and the disengagement boundary point is taken as the end point of the predetermined trajectory, the starting point and the end point are connected to form an initial straight line trajectory, and the trajectory constraint is set, including: not entering any nest area influence range throughout the journey; minimizing the total length of the trajectory and the straight line distance from the starting point to the end point; and the trajectory does not pass through the protection range of the ecological sensitive area and the legal device; When the initial straight line trajectory completely meets the trajectory constraint, the initial straight line trajectory is taken as the predetermined trajectory; when the initial straight line trajectory passes through any region in the trajectory constraint, a path inflection point that avoids the region and has the least distance increase is found on both sides of the region to form an initial predetermined trajectory; the inflection point is smoothed for the initial predetermined trajectory; and whether the initial predetermined trajectory after the smoothing processing is within the executable range of the target is judged in combination with the flight capability of the target, the trajectory inflection point position is further adjusted when it exceeds the executable range, until the trajectory meets the flight capability of the target, and the predetermined trajectory is formed.

10. A repelling system of specific frequency electromagnetic wave interfering with target navigation, using the repelling method of specific frequency electromagnetic wave interfering with target navigation according to any one of claims 1-9, characterized in that, including: The nest area influence range generation module comprises a bird nest coordinate positioning unit, a stress threshold acquisition unit and a nest area influence range generation unit; the bird nest coordinate positioning unit acquires three-dimensional coordinates of all bird nests in the bird breeding area to form a bird nest coordinate set; the stress threshold acquisition unit confirms the bird species to which each bird nest belongs through an AI image recognition algorithm and acquires the stress threshold of the bird species to noise during the breeding period; the nest area influence range generation unit takes each bird nest coordinate as a center point, calculates the noise influence range from the inside to the outside based on the stress threshold, and forms a nest area influence range set; The target analysis module comprises a UAV model acquisition unit and a target coordinate acquisition unit; the UAV model acquisition unit scans the target entering the bird breeding area, analyzes the target features by using a visual recognition algorithm, matches a UAV model database, and records the model of the UAV; the target coordinate acquisition unit acquires the target coordinates and updates the motion trajectory of the target; The navigation interference execution module comprises a space matching unit, an out-of-range interference execution unit, an in-range interference data acquisition unit and an in-range interference execution unit; the space matching unit performs space matching between the target coordinates at the current moment and the nest area influence range set and outputs a judgment result; the out-of-range interference execution unit generates electromagnetic waves when the target is outside the nest area influence range, outputs a normal interference frequency, and performs navigation interference; the in-range interference data acquisition unit acquires the environmental noise value of the current associated nest area when the target is in the nest area influence range, judges whether there are birds in the bird nest corresponding to the current associated nest area through infrared thermal imaging monitoring, and performs sub-scene judgment and interference execution; The predetermined trajectory generation and execution module comprises a predetermined trajectory generation unit and a predetermined trajectory execution unit; the predetermined trajectory generation unit generates a predetermined trajectory by screening a boundary point during navigation interference; the predetermined trajectory execution unit traces the target trajectory, performs navigation interference on the target based on the predetermined trajectory, and terminates the navigation interference when the current coordinates of the target are out of the bird breeding area.

Images