Bird flock oriented repelling path planning method based on Maklink airport intelligent bird repelling system

Abstract

The invention discloses a bird flock oriented repelling path planning method based on a Maklink intelligent airport bird repelling system, and the method comprises the steps: S1, recognizing the surrounding environment of an airport through a radar and photoelectric equipment, carrying out the environment modeling based on a Maklink link diagram theory, and generating a two-dimensional environment model; s2, based on the two-dimensional environment model, planning a repelling path of the airport intelligent bird repelling system by adopting an APF-RRT algorithm, determining path points of bird flock flight, and generating a bird repelling path set; and S3, optimizing the bird repelling path set based on a preset probability function, increasing the probability that the path passes through the safe target area, and generating a guidance repelling path. The guidance bird repelling effect is achieved, it is ensured that the bird flock leaves the airport area according to the expected direction, and therefore the success rate and safety of bird repelling are improved.

Description

Technical Field

[0001] This invention relates to the field of airport safety technology, and in particular to a bird flock guidance and deportation path planning method based on the Maklink airport intelligent bird control system. Background Technology

[0002] Traditional airport bird control methods mainly rely on single technologies such as gas cannons, sonic bird deterrents, laser bird deterrents, bird deterrent windmills, chemical bird deterrents, and bird deterrent robotic eagles. Although these methods can disperse flocks of birds to some extent, the overall effect is not ideal.

[0003] Existing technologies have several limitations. First, most bird control systems lack effective bird detection and identification devices, making it impossible to promptly detect and accurately identify potential threats. Second, these systems lack early warning capabilities and cannot quantify the effectiveness of bird control measures, making it difficult to determine their actual effectiveness. Third, traditional bird control technologies lack intelligent algorithm support, making it impossible to adaptively adjust control strategies based on different bird characteristics and environmental conditions. Furthermore, existing systems typically employ single control methods, lacking specificity and failing to address the diverse behavioral characteristics of different bird species. Finally, due to a lack of big data analysis and deep learning capabilities, these systems cannot learn from historical bird control experience and optimize their strategies. Particularly noteworthy is the general lack of directional control capabilities in existing technologies. They cannot plan and guide flocks of birds away from airport areas along pre-set safe routes, potentially causing them to re-enter the flight zone or runway area during the control process, increasing flight safety risks. Bird behavior is complex and unpredictable; simple deterrence often only temporarily changes the birds' location, rather than providing a long-term solution to guide them away from dangerous areas.

[0004] Therefore, there is an urgent need for a bird flock guidance and deportation path planning method based on the Maklink airport intelligent bird control system. Summary of the Invention

[0005] This invention provides a bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system to solve the above-mentioned problems in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system includes:

[0008] S1: Identify the environment around the airport using radar and optoelectronic equipment, and perform environmental modeling based on Maklink link graph theory to generate a two-dimensional environment model;

[0009] S2: Based on a two-dimensional environment model, the APF-RRT algorithm is used to plan the bird deterrence path of the airport intelligent bird deterrence system, determine the path points of the bird flock, and generate a set of bird deterrence paths.

[0010] S3: Optimize the bird deterrence path set based on a preset probability function, increase the probability of the path passing through a safe target area, and generate a guiding bird deterrence path.

[0011] Step S1 includes:

[0012] S11: Use radar and optoelectronic equipment to map and locate the environment around the airport, and identify actual obstacles and virtual obstacles. Actual obstacles refer to physical entities such as terrain and buildings, while virtual obstacles refer to artificially set up isolation zones for safety reasons.

[0013] S12: Based on Maklink link graph theory, real and virtual obstacles are represented as polygons to generate a two-dimensional environment model.

[0014] Step S2 includes:

[0015] S21: Determine the current position of the flock of birds as the starting point of the path in the two-dimensional environment model;

[0016] S22: The Artificial Potential Field-Fast Expanding Random Tree (APF-RRT) algorithm is used to traverse spatial points and generate a tree-like path structure.

[0017] S23: Based on the tree-like path structure, determine the path points of the bird flock's flight and generate a set of bird deterrence paths.

[0018] Step S3 includes:

[0019] S31: For each path in the bird deterrence path set, calculate the probability that it passes through a preset safe target area. The safe target area refers to an area suitable for bird flocks to fly without affecting flight safety.

[0020] S32: Increase the probability that the path passes through the safe target area by adjusting the probability function parameters;

[0021] S33: Based on the optimized probability distribution, select the optimal guided deportation path.

[0022] Step S11 includes:

[0023] S111: Collect three-dimensional data of the terrain and landforms around the airport using radar equipment;

[0024] S112: Acquire image data of buildings surrounding the airport using optoelectronic equipment;

[0025] S113: Integrate 3D data and image data to identify and locate actual and virtual obstacles, including isolation zones formed by light wall obstacles set up using lasers.

[0026] Step S12 includes:

[0027] S121: Abstract both real and virtual obstacles into convex polygons;

[0028] S122: Based on Maklink link graph theory, connect the vertices of convex polygons with the environment boundary to construct a two-dimensional link graph;

[0029] S123: Based on the two-dimensional link graph, generate a two-dimensional environment model and identify narrow passages for bird flocks to fly.

[0030] Step S22 includes:

[0031] S221: Starting from the path's origin, randomly select adjacent points to extend the path;

[0032] S222: Utilize the artificial potential field (APF) to guide the path extension direction to avoid obstacles;

[0033] S223: Repeat the path expansion until the predetermined spatial range is covered, generating a tree-like path structure.

[0034] Step S23 includes:

[0035] S231: Identify dead zones in a tree-like path structure. Dead zones refer to areas that the path cannot bypass.

[0036] S232: Adjust the path expansion direction based on the probability function to increase the probability that path points will lead to the safe target area;

[0037] S233: Generate a set of bird deterrent routes based on the adjusted path points.

[0038] This also includes:

[0039] S4: By using sound and laser devices to create sound and light interference along the calculated dispersal path, the flock of birds is driven to fly along the planned directional dispersal path.

[0040] Step S4 includes:

[0041] Real-time monitoring of bird activity within the airport area to obtain bird distribution and activity data;

[0042] The sensing device detects environmental parameter information and transmits it to the central processing unit.

[0043] After receiving the bird drive-away command, the central processing unit acquires bird activity data provided by the monitoring equipment, identifies the current location of the bird flock based on the bird activity data, and determines whether the bird flock is within the boundary of the safe target area.

[0044] If the flock of birds is within the boundary of the preset safe area, then sound and laser devices will be activated at specific locations around the flock according to the pre-planned directional dispersal path.

[0045] The central processing unit dynamically adjusts the activation sequence and working intensity of the sound and laser equipment according to the flight speed and direction of the flock of birds, forming a gradient sound and light interference zone on the guiding drive-away path;

[0046] Among them, the frequency of the sound waves emitted by the sound device and the intensity of the beam projected by the laser device are adaptively adjusted according to the characteristics of the bird flock and environmental conditions, forming a gradual driving away channel from the high interference area to the low interference area, so that the bird flock instinctively avoids the high interference area and flies along the preset guiding driving away path.

[0047] The central processing unit tracks the flight path of the flock of birds in real time and automatically adjusts the operating parameters of the sound and light interference equipment based on the deviation between the actual flight status of the flock and the expected path, so as to ensure that the flock of birds safely flies away from the airport area along the planned directional driving path.

[0048] Compared with the prior art, the present invention has the following advantages:

[0049] A method for guiding bird departure paths based on the Maklink intelligent bird control system at airports includes: S1: Identifying the airport's surrounding environment using radar and photoelectric equipment, and modeling the environment based on Maklink link graph theory to generate a two-dimensional environment model; S2: Based on the two-dimensional environment model, using the APF-RRT algorithm to plan the bird departure paths of the intelligent bird control system at airports, determining the path points of the bird flock, and generating a set of bird departure paths; S3: Optimizing the set of bird departure paths based on a preset probability function to increase the probability of the path passing through a safe target area, generating a guiding departure path. This method achieves a guiding bird departure effect, ensuring that the bird flock leaves the airport area in the expected direction, thereby improving the success rate and safety of bird control.

[0050] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention.

[0051] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0052] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0053] Figure 1 This is a flowchart of a bird flock guidance and deportation path planning method based on the Maklink airport intelligent bird control system in an embodiment of the present invention;

[0054] Figure 2 This is a flowchart illustrating the generation of a two-dimensional environment model in an embodiment of the present invention;

[0055] Figure 3 This is a flowchart illustrating the generation of a bird deterrent path set in an embodiment of the present invention. Detailed Implementation

[0056] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0057] The embodiments of the present invention provide, as follows Figure 1 As shown, a bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system includes:

[0058] S1: Identify the environment around the airport using radar and optoelectronic equipment, and perform environmental modeling based on Maklink link graph theory to generate a two-dimensional environment model;

[0059] S2: Based on a two-dimensional environment model, the APF-RRT algorithm is used to plan the bird deterrence path of the airport intelligent bird deterrence system, determine the path points of the bird flock, and generate a set of bird deterrence paths.

[0060] S3: Optimize the bird deterrence path set based on a preset probability function, increase the probability of the path passing through a safe target area, and generate a guiding bird deterrence path.

[0061] The working principle of the above technical solution is as follows: S1: Perform multi-dimensional perception analysis of the environment surrounding the airport to establish an accurate two-dimensional environmental model; wherein, environmental perception is achieved through radar equipment and photoelectric monitoring equipment configured at key locations in the airport. The radar equipment is responsible for detecting dynamic targets and static obstacles around the airport, while the photoelectric equipment provides visualized environmental image information; the perception range covers the area around the airport runway and the main areas of bird activity; environmental modeling is based on Maklink link graph theory, which converts the perceived environmental information into nodes and connections, where nodes represent key locations in the environment, and connections reflect the accessibility and safety between each location; the generation of the two-dimensional environmental model includes: preprocessing and feature recognition of the raw data collected by radar and photoelectric equipment, constructing a node network according to the rules of Maklink graph theory based on the recognition results, and finally forming a two-dimensional topology diagram that reflects the actual environment of the airport.

[0062] S2: Based on the established environmental model, intelligent path planning is implemented to form a systematic bird deterrence path scheme. The path planning uses the APF-RRT algorithm as the core computing engine, which combines the advantages of the artificial potential field method and the fast random tree algorithm. The execution process of the APF-RRT algorithm includes: setting the starting point and target area in the two-dimensional environmental model; calculating the gravitational and repulsive force distribution at each location point using the artificial potential field method; and simultaneously using the fast random tree algorithm for path search and expansion. The determination of bird flock flight path points is based on a comprehensive calculation of the flock's current position, flight speed, and expected deterrence direction. Appropriate spacing is maintained between path points to ensure effective tracking and guidance by the bird deterrence equipment. The generation of the bird deterrence path set is achieved through multiple path planning calculations, each with different initial conditions and constraint parameters, ultimately yielding multiple selectable deterrence paths for subsequent optimization.

[0063] S3: Utilizing a probabilistic optimization mechanism to improve bird deterrence paths and achieve precise guided bird removal. Probabilistic function optimization is a deep processing step based on the set of bird deterrence paths. Each path is evaluated and improved using a pre-defined probability function for safe target areas. The probability function is constructed considering the importance distribution of airport safety areas, the ease of bird removal, and the influence weight of environmental factors. The specific implementation of path optimization includes: calculating the probability value of each path passing through each safety target area, reordering and adjusting the paths according to the probability, and prioritizing paths with a higher probability of passing through high-safety-level areas as the primary deterrence scheme. The final determination of the guided bird removal path comprehensively considers multiple evaluation indicators such as path length, safety, and operability to ensure that the bird deterrence system can effectively guide bird flocks to the predetermined safe area.

[0064] The beneficial effects of the above technical solution are as follows: by comprehensively analyzing the activity patterns of bird flocks and the environmental conditions, the optimal timing for path planning is determined, that is, deep path planning is initiated when the bird flock density reaches the warning threshold or abnormal aggregation occurs. The specific triggering mechanism is to match the corresponding planning strategy and execution parameters from the pre-established decision rule base by monitoring factors such as bird flock distribution density, flight trajectory changes and meteorological conditions.

[0065] In another embodiment, such as Figure 2 As shown, step S1 includes:

[0066] S11: Use radar and optoelectronic equipment to map and locate the environment around the airport, and identify actual obstacles and virtual obstacles. Actual obstacles refer to physical entities such as terrain and buildings, while virtual obstacles refer to artificially set up isolation zones for safety reasons.

[0067] S12: Based on Maklink link graph theory, real and virtual obstacles are represented as polygons to generate a two-dimensional environment model.

[0068] The working principle of the above technical solution is as follows: S11: The airport surrounding environment data acquisition process first utilizes high-precision sensing equipment to conduct a comprehensive scan and mapping of the airport surrounding area. This high-precision sensing equipment includes millimeter-wave ranging radar and high-definition photoelectric detectors. The millimeter-wave ranging radar accurately determines the distance, orientation, and height of target objects by emitting electromagnetic wave signals and receiving reflected signals; the high-definition photoelectric detectors obtain the shape, size, and texture features of objects in the environment through visible light and infrared spectral analysis. During the data acquisition process, the sensing equipment first performs a comprehensive environmental scan of the airport surrounding area to acquire raw point cloud data; then, a data denoising algorithm is used to eliminate data deviations caused by weather, electromagnetic interference, and other factors; next, a 3D reconstruction algorithm is used to convert the point cloud data into a spatial coordinate model; finally, environmental elements are classified and identified, dividing the identified environmental elements into two main types: actual obstacles and virtual obstacles. Actual obstacles refer to objectively existing physical entities around the airport, including physical structures such as terminals, control towers, hangars, communication facilities, fences, and natural terrain; virtual obstacles refer to non-physical safety isolation areas such as no-fly zones, restricted-fly zones, and buffer zones artificially demarcated based on airspace control requirements. During the construction of the environmental recognition model, the system extracts geometric features from actual obstacles, converting the identified obstacle outlines into geometric shapes. Simultaneously, based on the safety distance requirements for different functional areas in airport safety management regulations and combined with the activity patterns of birds at the airport, virtual obstacle boundaries are constructed. The construction of the airport environmental recognition model enables a digital representation of the airport's surrounding environment, laying the foundation for subsequent intelligent bird deterrence path planning.

[0069] S12: Based on the airport environment recognition model, the system simplifies the 3D environment into a 2D planar model using Maklink link graph theory. Maklink link graph theory is a mathematical method for modeling complex spatial environments. It divides the environment into several non-overlapping convex polygonal regions by constructing a series of connecting lines in free space, forming a topological representation of the environment. The system first represents the actual and virtual obstacles identified in step one as polygonal regions on a 2D plane. Each polygon is composed of several vertices connected together, and the spatial connections of the vertices constitute the boundary of the obstacle. Then, the system generates Maklink link lines between adjacent obstacles. These link lines are straight line segments connecting points on the boundaries of different obstacles and do not intersect with any obstacle. Next, the system constructs a feasible path graph based on the generated link lines. This path graph consists of a series of nodes and edges. Nodes represent the midpoints or endpoints of Maklink link lines, and edges represent feasible connections between nodes. To achieve guided bird deterrence, the system deploys an optical bird deterrent device on the boundary of the virtual obstacle region. This device projects high-brightness directional beams, forming a "light wall obstacle." A light barrier is a non-physical visual barrier that, when birds approach, perceives a strong visual stimulus and actively avoids them, thus ensuring that their behavioral paths align with feasible paths in the Maklink link map. The system also comprehensively considers environmental factors such as wind direction and flight altitude, dynamically adjusting the link map to ensure that birds are guided to pre-defined safe areas under different environmental conditions. By constructing a two-dimensional environmental model based on the Maklink link map and the light barrier system, effective monitoring and guided dispersal of bird activity around the airport are achieved, improving the airport's safe operation capabilities.

[0070] The beneficial effects of the above technical solution are as follows: by distinguishing between actual obstacles and virtual obstacles and performing two-dimensional modeling based on Maklink link graph theory, the environment representation is more accurate and computationally efficient, providing an accurate spatial information basis for subsequent path planning and effectively reducing the complexity of environment modeling.

[0071] In another embodiment, such as Figure 3 As shown, step S2 includes:

[0072] S21: Determine the current position of the flock of birds as the starting point of the path in the two-dimensional environment model;

[0073] S22: The Artificial Potential Field-Fast Expanding Random Tree (APF-RRT) algorithm is used to traverse spatial points and generate a tree-like path structure.

[0074] S23: Based on the tree-like path structure, determine the path points of the bird flock's flight and generate a set of bird deterrence paths.

[0075] The working principle of the above technical solution is as follows: S21: The current location of the bird flock is collected in real time by the airport bird monitoring system and transmitted to the path planning system. The bird monitoring system includes comprehensive information input from infrared cameras, radar monitoring stations distributed throughout the airport, and manual observation data. After receiving the bird monitoring data, the system processes the data in real time using a bird flock behavior recognition algorithm to extract the coordinates of the flock's gathering center from the raw monitoring data. The gathering center of the flock refers to the weighted average of the positions of the flock members, and the weights can be dynamically adjusted according to factors such as bird size and population density. The system projects the gathering center of the flock onto a two-dimensional environmental model of the airport, forming the starting node for path planning. The two-dimensional environmental model is a simplified representation of the airport's three-dimensional terrain, including the planarized processing results of key environmental information such as runways, taxiways, buildings, and vegetation areas. The two-dimensional environmental model also marks functional zones related to bird control, such as restricted areas, high-risk areas, and target areas for bird removal. The system determines a bird flock hazard level index based on the flock's species characteristics, behavioral patterns, and current flight activity. The higher the hazard level index, the higher the processing priority assigned to that starting point. When multiple flocks appear simultaneously, the system plans deportation routes sequentially according to the hazard level index, ensuring that the flocks posing the greatest threat to aircraft safety are dealt with first.

[0076] S22: The APF-RRT algorithm is an efficient path planning method that combines the artificial potential field method with the fast expanding random tree algorithm. The artificial potential field method guides path planning by constructing virtual gravitational and repulsive fields, while the fast expanding random tree algorithm explores feasible paths by randomly sampling spatial points and constructing a connected graph. First, the system initializes the path tree structure using the current position of the flock as the root node. In each iteration, the system generates candidate points in the two-dimensional environment space using an adaptive sampling strategy. The adaptive sampling strategy dynamically adjusts the sampling density according to environmental characteristics, with denser sampling points near obstacles or complex areas and sparser sampling points in open areas, improving algorithm efficiency. For each sampling point, the system calculates the connection path from the nearest node in the existing tree to that point and checks whether the path intersects with an obstacle. Obstacle information comes from a two-dimensional environmental model of the airport, including fixed obstacles such as buildings, fences, and equipment, as well as dynamic obstacles such as airport vehicles and temporary construction areas.

[0077] During the path connection verification phase, the system incorporates the concept of an artificial potential field algorithm, assigning different potential fields to various elements in the environment. The target area is given an attractive potential field, guiding the path in that direction; while airport runways, aircraft activity areas, and high-risk areas are given a repulsive potential field, causing the path to move away from these areas. The potential field strength dynamically adjusts with the distance to the object, becoming stronger as the distance increases. By considering the influence of the synthetic potential field force during each path expansion, the system can generate a path tree structure that both avoids restricted areas and approaches the target area. The algorithm also introduces a polar coordinate-based partitioned sampling strategy, dividing the space around the current expansion node into multiple fan-shaped regions. The sampling probability is dynamically adjusted based on the exploration status and environmental characteristics of each region, ensuring that the path tree can cover the entire planning space evenly and effectively. As the number of iterations increases, the path tree gradually expands, forming a tree-like network from the current location of the flock to multiple potential repulsion directions.

[0078] Among them, the drive-off path algorithms based on APF-RRT include:

[0079] Traverse the existing tree nodes and find the node closest to the target point, denoted as node 0;

[0080] Extend horizontally to the right from 0 point to form the OE line segment at the edge of the map. Establish a polar coordinate system with the OE line segment as the polar axis.

[0081] Assigning x using probability functions rand The probability of it appearing in the corresponding region;

[0082] The probability function is executed to obtain x. rand Angle value;

[0083] After selecting the angle, x needs to be adjusted. rand Select the node length. To make x... rand To cover the entire map area, calculations need to be performed for different angles and lengths, x rand The length can be expressed as:

[0084]

[0085] In the formula, rand() represents a random percentage function, OV n This represents the distance from point O to one of the four vertices. When the length exceeds the map boundary, the algorithm is re-executed until the length does not exceed the map boundary.

[0086] S23: After generating the tree-like path structure, the system needs to select the optimal bird deterrence path from numerous possible paths. The system first analyzes the path tree to identify all feasible paths from the root node (the current location of the flock) to the boundary of the safe zone. The safe zone refers to an area far from the core operational area of ​​the airport, where flock migration will not pose a threat to aircraft operations. For each feasible path, the system calculates a path quality score, with scoring factors including path length, smoothness, distance margin from the restricted area, and conformity to the birds' natural migratory habits. Path length reflects deterrence efficiency; shorter paths generally mean a faster deterrence process. Path smoothness measures the frequency and magnitude of path turns; smoother paths better match the natural flight trajectory of the flock. Distance margin from the restricted area ensures that even slight deviations during flight will not lead the flock into the danger zone. Conformity to the birds' natural migratory habits considers factors such as seasonal migration directions and habitat locations, making the deterrence path more aligned with the birds' natural behavioral tendencies.

[0087] After path selection, the system optimizes the chosen path. First, path smoothing is performed, converting the discrete points into a continuous, smooth curve using B-spline or Bézier curve fitting. Then, path simplification is implemented, removing redundant points and retaining key turning points and feature points. Finally, path segmentation divides the overall path into multiple continuous segments, each corresponding to a deterrence phase. For each segment, the system determines a suitable deterrence strategy, including the type, intensity, and timing of the audio-visual bird deterrence equipment. The system dynamically adjusts the parameter configuration of the deterrence equipment based on the bird species, size, and behavioral state, preventing the birds from developing habitual responses to deterrence methods. Ultimately, the system forms a complete set of bird deterrence paths, containing guidance information for the entire bird flock from its current location to the safe area, as well as activation schemes for the deterrence equipment at each path point. The path set also includes alternative paths; if the bird flock's response to the primary recommended path is less than expected, the system can seamlessly switch to an alternative path to continue deterrence. Through this multi-layered, adaptive path planning, the system can effectively drive away flocks of birds in a way that best aligns with their natural behavior, while ensuring airport safety.

[0088] The beneficial effects of the above technical solution are as follows: the tree-like path structure generated by the APF-RRT algorithm can effectively avoid obstacles while ensuring path connectivity. The path generation process has high computational efficiency and spatial coverage, which can meet the real-time path planning needs in the complex environment of airports.

[0089] In another embodiment, step S3 includes:

[0090] S31: For each path in the bird deterrence path set, calculate the probability that it passes through a preset safe target area. The safe target area refers to an area suitable for bird flocks to fly without affecting flight safety.

[0091] S32: Increase the probability that the path passes through the safe target area by adjusting the probability function parameters;

[0092] S33: Based on the optimized probability distribution, select the optimal guided deportation path.

[0093] The working principle of the above technical solution is as follows: S31: The bird deterrent path set refers to multiple candidate paths pre-generated by the system that may be used to guide flocks of birds away from dangerous areas. Each path consists of a series of continuous waypoints, which constitute a complete flight trajectory from the current position of the flock to the safe target area. The safe target area refers to an area suitable for flock flight and does not affect flight safety, usually far away from airport runways, flight paths, and other areas with high aviation activity. The method for calculating the probability of a path passing through the safe target area includes establishing a probability density function model. First, the airport area is divided into grid cells, and each grid cell is assigned a different safety weight value. A high safety weight indicates that the area is suitable for flock flight, such as green areas or water bodies far away from flight paths; a low safety weight indicates that the area is not suitable for flocks to stay, such as the area around the runway or areas with frequent aircraft activity. Based on these grid cells and their weight values, a spatial probability distribution model is constructed, and the kernel density estimation method is used to calculate the cumulative probability of each path crossing the high safety weight area. The probability calculation steps include discretizing the path, decomposing the continuous path into finite path segments; for each path segment, calculating its intersection with each grid cell; calculating the safety probability contribution of the path segment based on the safety weights of the intersecting grid cells; and summing the safety probability contributions of all path segments to obtain the probability of passing through the safe target area along the entire path. The kernel density estimation method uses a Gaussian kernel function to perform a weighted average of the contributions of the grid cells, achieving a transition from discrete grids to a continuous probability distribution.

[0094] S32: Adjusting the probability function parameters is a crucial step in optimizing bird deterrence paths. By dynamically adjusting the core parameters of the probability model, the path is made more likely to pass through safe target areas. Parameter adjustment uses an iterative optimization algorithm, including a combination of gradient ascent and simulated annealing. First, the key parameters to be adjusted are determined, including the safe area weight coefficient, distance decay factor, and track smoothness parameter. The safe area weight coefficient determines the importance of the safe target area in path planning; the distance decay factor controls the degree to which the path deviates from the shortest distance; and the track smoothness parameter affects the turning frequency and angle of the path. In each iteration, the system calculates the probability of passing through the safe target area for all candidate paths based on the current parameters, and then adjusts the parameters according to the gradient direction to increase the probability of passing through. The specific implementation of the iterative optimization algorithm includes calculating the probability value of the path passing through the safe target area under the current parameters; making small parameter changes and observing the trend of probability value changes; adjusting parameters along the direction of increasing probability value; introducing random perturbations to avoid getting trapped in local optima; checking whether the new parameters meet the constraints, such as the path length not exceeding the maximum threshold and the turning angle not exceeding the physiological limits of birds; repeating the above steps until the convergence condition is met or the maximum number of iterations is reached. In addition, the system also considers dynamic environmental factors and adjusts parameters accordingly. For example, it corrects flight energy consumption estimates in path planning based on wind direction and speed data, avoids areas with severe weather conditions, and comprehensively considers seasonal bird behavior characteristics. Through these comprehensive adjustments, the final generated path not only conforms to the natural flight habits of birds but also effectively guides them to safe target areas.

[0095] S33: The optimized probability distribution provides an assessment of the probability that each candidate path will pass through the safe target area. The system selects the optimal guiding deportation path from the candidate path set using a multi-criteria decision-making method. The multi-criteria decision-making comprehensively considers three main aspects: path safety, deportation efficiency, and implementation feasibility.

[0096] The path safety assessment includes the distance between the path and the no-fly zone, the risk of conflict with the aircraft's expected flight path, and the suitability of weather conditions along the path. Based on real-time radar and flight data, the system calculates the spatial and temporal separation of each path from current and expected aviation activity, ensuring that the selected deflection paths do not cause flocks to fly into more dangerous areas.

[0097] The efficiency assessment of bird dispersal considers path length, estimated dispersal time, and energy consumption. The system establishes a bird flight dynamics model, evaluating the actual flight time and energy requirements for each path based on the flight characteristics of different bird species, such as flight speed, turning ability, and altitude change capability. It also considers the curvature and number of turning points of candidate paths to avoid overly complex paths that would be difficult for flocks to follow effectively.

[0098] A feasibility assessment was conducted to analyze the deployment capabilities of bird deterrence equipment and the coverage of bird deterrence signals. The system matched candidate routes with the existing distribution of bird deterrence equipment at the airport to ensure effective coverage of the entire route. For areas with insufficient coverage, the system calculated the optimal location for deploying temporary bird deterrence equipment and evaluated the deployment costs and time.

[0099] The final route selection employs a weighted comprehensive scoring method, summing the scores of each candidate route across three dimensions: safety, efficiency, and feasibility. The weighting coefficients can be dynamically adjusted based on the current airport operational status; for example, safety is prioritized during peak periods, while bird deterrence efficiency is prioritized in emergency situations. The system selects the route with the highest comprehensive score as the final guiding bird deterrence route and generates a detailed activation sequence for bird deterrence equipment to guide the precise implementation of bird deterrence operations.

[0100] The selected optimal bird deterrence path is converted into bird deterrence equipment control commands, which activate bird deterrence devices such as sound, light, and electricity along the path in chronological order, forming a "pressure gradient" perceived by the birds and guiding them to fly along the preset path to a safe target area, effectively reducing the risk of bird strikes and ensuring airport aviation safety.

[0101] The beneficial effects of the above technical solution are as follows: by optimizing the path set through a probability function, the probability of the path passing through a safe target area is increased, thus realizing the directional guidance of bird deterrence paths, which not only ensures the flight safety of aircraft but also reduces unnecessary interference to flocks of birds.

[0102] In another embodiment, step S11 includes:

[0103] S111: Collect three-dimensional data of the terrain and landforms around the airport using radar equipment;

[0104] S112: Acquire image data of buildings surrounding the airport using optoelectronic equipment;

[0105] S113: Integrate 3D data and image data to identify and locate actual and virtual obstacles, including isolation zones formed by light wall obstacles set up using lasers.

[0106] The working principle of the above technical solution is as follows: S111: Airport environmental detection first uses a ground-based radar array deployed around the airport to perform a comprehensive scan of the surrounding area. The radar array uses millimeter-wave technology to emit electromagnetic waves and receive reflected waves. By measuring the round-trip time and phase difference of the signals, it generates precise three-dimensional point cloud data of the terrain around the airport. The terrain data includes ground undulations, height changes, and the location information of fixed obstacles, forming a preliminary spatial contour map. Simultaneously, the system also deploys a high-resolution optoelectronic equipment group, including visible light cameras, infrared thermal imagers, and multispectral imaging equipment, to acquire images of buildings, facilities, and other fixed structures around the airport. The optoelectronic equipment group uses an intelligent rotating gimbal to achieve panoramic shooting and local detail capture, generating an image dataset containing features such as color, texture, and edges. The deployment positions of the optoelectronic equipment are carefully calculated to ensure the integrity of the coverage and monitoring without blind spots, while avoiding data acquisition quality degradation due to mutual interference between devices. During data collection, the system also records environmental condition parameters, including current light intensity, visibility, temperature, and humidity. These parameters are used for calibration and compensation during subsequent data processing to improve imaging quality and accuracy.

[0107] S112: The acquired radar 3D point cloud data and optoelectronic equipment image data are fused through spatiotemporal registration. Spatiotemporal registration maps data collected by different sensors at different times and spatial locations to a unified spatiotemporal reference system, ensuring data consistency and comparability. The system employs a feature point matching algorithm to extract key feature points from radar and image data, such as building corners and unique terrain contours, establishing a correspondence between the two types of data. During the fusion process, a deep learning network is applied to perform semantic segmentation on the point cloud and images, classifying environmental elements into different categories such as buildings, vegetation, roads, and water bodies. The fused data is then processed by a 3D reconstruction algorithm to generate a 3D semantic model of the airport's surrounding environment. Based on this, the system further identifies two types of obstacles: actual obstacles and virtual obstacles. Actual obstacles refer to objectively existing physical entities, such as terminal buildings, control towers, hangars, fences, and other architectural facilities, as well as naturally formed hills, trees, and other terrain features. The system calculates the geometric characteristics of these obstacles, including parameters such as height, width, and inclination, and represents them as polygonal models. Each physical obstacle is assigned a unique identifier, such as buildings B1 and B2, or terrain elements T1 and T2, to facilitate subsequent location and path planning. Virtual obstacles refer to specially designed functional areas, primarily including isolation zones defined by light walls projected by laser equipment. Light walls are a technique that uses high-power lasers to create visible light bands in the air, effectively deterring birds and other flying creatures. The system also converts these virtual obstacles into polygonal representations, such as closed regions formed by connecting vertices. Each virtual obstacle is also assigned a unique identifier, such as virtual regions V1 and V2.

[0108] S113: The system integrates the identified actual and virtual obstacles into a unified spatial coordinate system and constructs an obstacle network model using Maklink graph theory. Maklink graphs are a spatial path planning method that connects obstacle vertices to form a grid, providing feasible pathways for subsequent path planning. During construction, the system first extracts all obstacle vertices, such as the corner points N21, N22, N23, and N24 of actual buildings and the boundary points N1 and N2 of virtual isolation zones. Then, the system analyzes whether the lines connecting adjacent vertices pass through obstacles. If not, it establishes connections between these vertices, forming a Maklink connection graph. This connection graph reflects a network of walkable pathways in space, avoiding all obstacle areas. The system specifically handles virtual isolation zones formed by light wall obstacles. While physically walkable, these areas need to be avoided functionally. Light wall obstacles are achieved through an array of laser emitters deployed around the airport. Each emitter can emit a laser beam of a specific wavelength, forming a visible light curtain in the air. These light curtains strongly stimulate the visual system of birds, causing them to instinctively choose to bypass them. The system adjusts the laser's wavelength, intensity, and flashing frequency based on birds' flight habits and sensitivity to different colors of light, ensuring effective bird deterrence without harming the birds. By incorporating the light barrier area into a Maklink graph model, the system can design optimal paths to guide flocks around the restricted area. The path planning algorithm, based on an improved search method, comprehensively considers path length, energy consumption, and bird flight habits to generate a directional deterrence path from the starting point to the target area. The generated path dynamically adjusts the opening sequence and intensity of the laser light barrier, forming a virtual passage that birds easily perceive and tend to follow, thus guiding the flocks away and ensuring airport operational safety.

[0109] The beneficial effects of the above technical solution are as follows: by integrating multi-source data collected by radar and optoelectronic equipment, it can accurately identify actual obstacles and virtual obstacles, especially the isolation zone formed by the laser light wall, which improves the comprehensiveness and accuracy of environmental perception and provides more reliable environmental data support for path planning.

[0110] In another embodiment, step S12 includes:

[0111] S121: Abstract both real and virtual obstacles into convex polygons;

[0112] S122: Based on Maklink link graph theory, connect the vertices of convex polygons with the environment boundary to construct a two-dimensional link graph;

[0113] S123: Based on the two-dimensional link graph, generate a two-dimensional environment model and identify narrow passages for bird flocks to fly.

[0114] The working principle of the above technical solution is as follows: S121: Various types of physical obstacles are distributed in the airport environment, such as fixed buildings like terminals, control towers, hangars, and apron fences, as well as mobile obstacles such as various vehicles and temporary facilities. Simultaneously, virtual obstacle areas also exist within the airport, such as no-fly zones, safety buffer zones, and high-risk areas—areas that are not physical entities but need to be avoided. To facilitate system processing, these obstacles of different natures must first be geometrically abstracted. Specifically, a polygon approximation algorithm is used to simplify the outline of an obstacle of arbitrary shape into a convex polygon representation. A convex polygon is a polygon where all interior angles are less than 180 degrees, and the line connecting any two points lies inside the polygon. For obstacles with complex shapes, the system first extracts the outline, then identifies key points of the obstacle boundary through feature point detection, and finally uses a convex hull algorithm to connect these feature points to form a convex polygon. The convex hull algorithm is a geometric algorithm used to calculate the smallest convex polygon that can contain all specified points. When there are complex non-convex obstacles in the airport environment, the system decomposes them into a combination of multiple convex polygons. After this step is completed, all obstacles in the airport environment are represented as a standardized set of convex polygons, which facilitates subsequent path planning.

[0115] S122: Constructing a two-dimensional link graph based on Maklink link graph theory. A Maklink link graph is a topological structure used for path planning, forming a network graph reflecting the traversability of the environment by connecting obstacle vertices and boundaries. First, the system identifies the vertices of all convex polygons from step one and uses these vertices as the basic nodes of the link graph. Then, the system sequentially checks the connection between each pair of convex polygon vertices, determining whether the connection intersects with any obstacle. If the connection between two vertices does not intersect with any obstacle, a connection is established between these two points in the link graph. Simultaneously, the system also needs to consider environmental boundaries, i.e., airport boundaries or the edges of the planned area. The system sets a series of uniformly distributed sampling points on the boundary and checks whether the connections between these sampling points and convex polygon vertices are feasible. Feasible connections are also added to the link graph network. During the construction process, the system also calculates the length, direction, and traversal difficulty of each link. The traversal difficulty is determined by comprehensively evaluating factors such as the density of obstacles around the nodes at both ends of the link and the width of the area traversed by the link. The resulting two-dimensional link graph is a network structure containing multiple nodes and connections. Each node represents a feature point in the environment, and each connection represents a potential feasible path.

[0116] S123: Utilizing the two-dimensional link graph constructed in S122, the system further analyzes and identifies the characteristics of bird flight channels in the airport environment. First, the system performs topological analysis on the link graph to identify critical paths and nodes. Critical paths are those connecting important areas in the link graph, such as channels connecting foraging areas and habitats. Critical nodes are intersections connecting multiple paths or important locations in the environment. By calculating the width parameters of each path in the link graph, the system can identify narrow channels in the environment. Narrow channels are areas with obstacles on both sides and relatively small widths; these areas are often essential routes for bird activity and ideal locations for dispersal. The system determines the channel width parameters by comparing the angle and spacing between adjacent link paths. When a path has obstacles on both sides and the spacing is less than a preset threshold, the path is identified as a narrow channel. For identified narrow channels, the system calculates their geometric parameters such as direction, length, and width, and assesses the channel's importance. The importance assessment is based on factors such as the characteristics of the areas connected by the channel, the frequency of bird activity, and the channel's location in the overall environment. Based on the above analysis, the system ultimately establishes a complete two-dimensional environmental model that includes obstacle distribution, feasible passage network, and narrow passage characteristics. This environmental model will serve as the basis for subsequent bird flock dispersal path planning, providing navigation guidance for drones or other bird deterrence equipment. The system can design optimal bird deterrence paths based on the narrow passage distribution in the environmental model, guiding flocks of birds away from the airport area along preset safe routes, thereby reducing the risk of bird strikes and ensuring aviation safety.

[0117] The beneficial effects of the above technical solution are as follows: by abstracting obstacles into convex polygons and constructing a link graph, the complexity of environmental representation is greatly simplified. At the same time, the narrow passages for bird flocks to fly are identified, making path planning more targeted and improving algorithm efficiency and planning quality.

[0118] In another embodiment, step S22 includes:

[0119] S221: Starting from the path's origin, randomly select adjacent points to extend the path;

[0120] S222: Utilize the artificial potential field (APF) to guide the path extension direction to avoid obstacles;

[0121] S223: Repeat the path expansion until the predetermined spatial range is covered, generating a tree-like path structure.

[0122] The working principle of the above technical solution is as follows: Step 1: Path random expansion mechanism. Starting from a set path starting point, the algorithm gradually constructs a path tree by iteratively searching for and connecting new nodes. In each iteration, the algorithm generates random sampling points in the map space based on the current tree structure and environmental state. These random sampling points are distributed in four sector regions using a polar coordinate system, with each region receiving different sampling probability weights based on its exploration level and threat distribution. After the sampling points are generated, the algorithm identifies the node in the path tree closest to that sampling point and attempts to connect the two points to form a new path branch. This random expansion method enables the algorithm to efficiently explore feasible paths in complex environments.

[0123] Step 2: Artificial Potential Field-Guided Obstacle Avoidance. To prevent collisions between the path and obstacles, the algorithm introduces an artificial potential field model as an auxiliary mechanism for path planning. The artificial potential field mainly consists of two components: a repulsive field and a gravitational field. The repulsive field is generated by obstacles, keeping the path away from danger zones; the gravitational field guides the path to extend towards the target direction. When candidate points generated during path expansion are near obstacles, the repulsive field corrects their direction, adjusting the expansion direction to avoid the obstacles. The potential field strength dynamically adjusts with the distance from the obstacle, achieving a safe and smooth obstacle avoidance effect. This method, which combines random sampling and potential field guidance, retains the global exploration capability of the RRT algorithm while enhancing the accuracy of local obstacle avoidance.

[0124] Step 3: Path Iteration Generation and Coverage Determination. The algorithm continuously expands the path tree into unexplored areas by repeatedly executing random expansion and potential field guidance steps. In each iteration, the algorithm dynamically adjusts the sampling probability of each region, prioritizing the exploration of areas with dense threats or low coverage. As the number of iterations increases, the path tree gradually covers the predetermined spatial range, forming a tree-like structure similar to a neural network. The algorithm sets multiple termination conditions, including the path reaching the boundary of the target area, covering all key threat points, and reaching the maximum number of iterations. When the termination conditions are met, the algorithm outputs the final path tree as the action route of the bird control system, achieving effective coverage of bird activity areas in the airport environment.

[0125] Step 4: Dynamic Adaptation and Real-Time Optimization. During path execution, the algorithm can adjust its path planning strategy in real time according to environmental changes. When new obstacles or bird activity areas are detected, the system recalculates the local potential field distribution and replans some path branches if necessary. Through dynamic adjustment of regional sampling probabilities, the algorithm can quickly respond to priority changes and concentrate computing resources on processing critical areas. This adaptive capability makes the APF-RRT algorithm particularly suitable for applications in dynamic environments such as airports, effectively dealing with unpredictable bird behavior patterns and weather condition changes.

[0126] Step 5: Polar Coordinate Partitioning and Efficient Sampling. The algorithm employs a polar coordinate system based on point 0 (the current extended reference point), dividing the space into four sector regions. This partitioning method is more directional than traditional Cartesian coordinate system sampling, balancing the needs of global exploration and local refinement. Within each sector region, the algorithm determines the location of sampling points by randomly generating polar angles and polar radii. The polar radii length is dynamically adjusted according to the region boundaries to avoid generating invalid, distant sampling points.

[0127] The beneficial effects of the above technical solution are as follows: by combining artificial potential field technology to guide the expansion direction of random trees, it not only retains the efficient exploration capability of the RRT algorithm, but also utilizes the obstacle avoidance characteristics of the APF algorithm, effectively solving the problems that the traditional RRT algorithm is prone to getting trapped in local optima and the APF algorithm is prone to getting trapped in local minima, thus improving the reliability of path planning.

[0128] In another embodiment, step S23 includes:

[0129] S231: Identify dead zones in a tree-like path structure. Dead zones refer to areas that the path cannot bypass.

[0130] S232: Adjust the path expansion direction based on the probability function to increase the probability that path points will lead to the safe target area;

[0131] S233: Generate a set of bird deterrent routes based on the adjusted path points.

[0132] The working principle of the above technical solution is as follows: A tree-like path structure refers to a hierarchical path network formed through gradual expansion in an airport environment. This structure starts from the origin and extends outwards to form branches. Dead zones refer to areas where bird deterrent equipment cannot effectively reach or leave due to obstacles or environmental limitations during path planning. Identifying dead zones first requires establishing a grid map of the airport environment, dividing the entire area into several cells. A grid map is a discretized spatial representation method that facilitates path calculation and area analysis by dividing continuous space into regular grid units. Based on the established grid map, accessibility analysis is performed on each cell. Accessibility analysis is the process of assessing whether an effective path exists from the origin to a specific area. The system simulates the bird deterrent equipment extending paths in all directions from the origin, recording the path coverage of each cell. Cells where the path is inaccessible or where a path leads to an inability to return are marked as dead zones. Identifying dead zones requires considering the motion characteristics of the bird deterrent equipment, including parameters such as minimum turning radius, maximum speed, and acceleration. These parameters affect whether the equipment can effectively reach a specific area. During the identification process, the system also extracts environmental topological features, including obstacle distribution, passage width, and open area area. Environmental topological features are key parameters describing the spatial structure of an airport and directly affect the planning results of bird deterrence routes. By analyzing environmental topological features, the system can identify potential path bottlenecks and advantageous passages, providing a basis for subsequent route planning. After blind spot area identification is completed, the system generates a blind spot area map, which contains information such as the location, shape, and area of ​​the blind spot areas, providing an important reference for adjusting the direction of subsequent route expansion.

[0133] After identifying blind spots, the system needs to adjust the path extension direction to ensure that the generated bird-repelling path effectively covers the target area while avoiding blind spots. The path extension direction adjustment employs a probability function-driven random sampling method. This method guides the path towards the safe target area by dynamically adjusting the sampling probability in different directions. A probability function is a mathematical tool used to describe the probability distribution of random events; in this system, it is used to control the directional preference of path extension. The system first establishes a polar coordinate system, with the current path node as the origin and the horizontal direction to the right as the polar axis. The polar coordinate system is a two-dimensional coordinate system that describes the position of a point through angles and distances, suitable for representing a fan-shaped area sampling strategy. The system divides the path extension space into multiple fan-shaped areas, assigning different sampling probability weights to each fan-shaped area. The safe target area refers to the safe area where birds should be driven away, typically open space outside the airport's activity area. For fan-shaped areas in the direction of the safe target area, the system assigns a higher probability weight, increasing the likelihood of the path extending in that direction. Conversely, for fan-shaped areas in the direction of blind spots, a lower probability weight is assigned to reduce the risk of the path entering a blind spot. The calculation of probability weights considers multiple factors, including distance to the safe target area, obstacle density, and coverage of the explored area. By integrating these factors, the system constructs a dynamic probability distribution function, which adjusts in real time as environmental perception information is updated. During path expansion, the system randomly generates sampling points based on the probability distribution function, verifies their validity, and then adds them to the path tree. The path tree is a data structure used to store and manage all feasible path points generated during path planning and their connections. Through this probability-based path expansion method, the system can effectively guide the path towards the safe target area while maintaining random exploration capabilities and avoiding dead zones.

[0134] After adjusting the waypoints, the system needs to generate a complete set of bird deterrence paths based on these waypoints. The bird deterrence path set refers to a set of trajectories that can effectively guide flocks of birds away from the airport area, containing multiple selectable bird deterrence paths. The system first prunes and optimizes the adjusted path tree, removing redundant or inefficient path branches. Pruning optimization is a technique to reduce computational complexity by removing unnecessary path options, improving the system's response speed and resource utilization efficiency. The optimized path tree retains multiple candidate paths leading to the safe target area, forming a path candidate set. The path candidate set is the foundation of the bird deterrence path set, containing multiple feasible paths from the flock's location to the safe target area. The system evaluates each path in the path candidate set, using evaluation metrics including path length, smoothness, obstacle avoidance margin, and guidance effect on flock behavior. After path evaluation, the system sorts the paths according to the evaluation results and selects the optimal paths as the bird deterrence path set. The bird deterrence path set needs to consider the behavioral characteristics of the flock and environmental conditions to ensure that the paths can effectively guide the flock away from the airport area. The system smooths the selected paths, reducing abrupt changes and generating continuous trajectories more suitable for bird deterrence equipment. Smoothing is a technique to reduce path jitter and discontinuities, adjusting path points mathematically to create smoother motion trajectories. After the bird deterrence path set is generated, the system transmits the path information to the bird deterrence equipment, which then executes the bird deterrence task based on this information. During execution, the system monitors the birds' responses in real time, dynamically adjusting the bird deterrence paths as needed. Dynamic adjustment refers to modifying the bird deterrence paths in real time based on the birds' actual reactions and environmental changes, ensuring maximum bird deterrence effectiveness. Through this path-set-based bird deterrence method, the system can guide and drive away flocks of birds, effectively reducing the risk of bird strikes and ensuring airport operational safety.

[0135] The beneficial effects of the above technical solution are as follows: by identifying blind spots in the tree-like path and adjusting the path based on a probability function, it effectively avoids guiding flocks of birds to areas where they cannot fly out, enhances the safety and effectiveness of the bird deterrence process, and improves the success rate of guided bird deterrence.

[0136] In another embodiment, it further includes:

[0137] S4: By using sound and laser devices to create sound and light interference along the calculated dispersal path, the flock of birds is driven to fly along the planned directional dispersal path.

[0138] The working principle of the above technical solution is as follows: After detecting bird activity within the airport area, the system first collects bird activity data, including the flock's current location, flight direction, flight altitude, flock size, and bird species identification information. Data collection methods include real-time video monitoring via high-definition cameras distributed throughout the airport area; the images captured by the cameras are transmitted to a central processing unit for analysis. The airport area also has a radar system deployed, capable of detecting high-altitude or long-distance bird flocks that are difficult to see with the naked eye. The bird activity information detected by the radar system is fused with the data collected by the video monitoring system. After bird activity data is collected, the central processing unit performs intelligent analysis. The system conducts a risk assessment of the current bird activity based on the airport operation area map and a historical bird activity database. The airport operation area map divides the airport area into different security levels according to importance and air traffic density; for example, runways and takeoff / landing areas are the highest security level areas, taxiways are the second highest security level areas, and aprons are the medium security level areas. The historical bird activity database records the activity patterns of various bird flocks and their response characteristics to dispersal measures under different seasons, time periods, and weather conditions. After the risk assessment results are determined, the system combines current flight schedules, weather conditions, and bird behavior characteristics to construct a bird-guided dispersal path network using the Maklink algorithm. Based on the constructed network, the system calculates the optimal dispersal path, which should avoid important airport operating areas and guide the flock to a safe area. Once the optimal dispersal path is determined, the system converts the path information into control commands for the audio-visual devices. The control commands include the start time, operating intensity, interference mode, and stop time for each sound and laser device. The sound devices are programmed to play sounds of specific frequencies, such as the sounds of bird predators, bird warning calls, or ultrasound of specific frequencies. The laser devices are set to emit visual stimulus beams at specific locations and times, with the beam's color, intensity, and flashing frequency adjusted according to the visual sensitivity characteristics of the target birds. The deployment of the audio-visual devices follows the gradient enhancement principle, setting higher intensity interference in areas where the flock is expected to leave and lowering the interference intensity in areas where the flock is expected to fly towards, thus creating an interference gradient to guide the flock along the predetermined path. When the system executes the dispersal operation, it first activates the audio-visual devices near the flock's current location, generating initial interference to prompt the flock to take off and begin moving. Subsequently, the system activates other audio-visual devices deployed along the predetermined dispersal path, gradually guiding the flock of birds to fly along the planned route. The operating status of each audio-visual device is controlled by the system in real time and dynamically adjusted based on the actual movement of the flock. The flock position monitoring unit continuously tracks the flock's movement and feeds back the latest position information to the central processing unit. If the flock deviates from the predetermined path, the system immediately adjusts the operating parameters of the audio-visual devices in the relevant area, increasing the interference intensity in the direction of deviation and weakening the interference intensity in the correct direction, guiding the flock back to the predetermined path.Throughout the entire dispersal process, the system collects data on the birds' responses to different disturbance methods. This data is used to update the historical bird activity database and optimize future dispersal strategies. Once the flock successfully flies away from the danger zone and reaches a safe area, the system gradually shuts down the audio-visual equipment, completing a full guided dispersal operation. The system also generates a dispersal effectiveness evaluation report, recording information such as the time of the dispersal, bird species, number, initial location, flight path, and final location, serving as a basis for subsequent optimization of system parameters.

[0139] The beneficial effects of the above technical solution are as follows: the introduction of sound and light interference methods to guide the flock of birds to fly along the planned path, the seamless integration of theoretical path planning and actual bird driving methods, forming a closed-loop control system, which greatly improves the actual effect of guiding and driving away the flock of birds and ensures the implementation of the bird driving solution.

[0140] In another embodiment, step S4 includes:

[0141] Real-time monitoring of bird activity within the airport area to obtain bird distribution and activity data;

[0142] The sensing device detects environmental parameter information and transmits it to the central processing unit.

[0143] After receiving the bird drive-away command, the central processing unit acquires bird activity data provided by the monitoring equipment, identifies the current location of the bird flock based on the bird activity data, and determines whether the bird flock is within the boundary of the safe target area.

[0144] If the flock of birds is within the boundary of the preset safe area, then sound and laser devices will be activated at specific locations around the flock according to the pre-planned directional dispersal path.

[0145] The central processing unit dynamically adjusts the activation sequence and working intensity of the sound and laser equipment according to the flight speed and direction of the flock of birds, forming a gradient sound and light interference zone on the guiding drive-away path;

[0146] Among them, the frequency of the sound waves emitted by the sound device and the intensity of the beam projected by the laser device are adaptively adjusted according to the characteristics of the bird flock and environmental conditions, forming a gradual driving away channel from the high interference area to the low interference area, so that the bird flock instinctively avoids the high interference area and flies along the preset guiding driving away path.

[0147] The central processing unit tracks the flight path of the flock of birds in real time and automatically adjusts the operating parameters of the sound and light interference equipment based on the deviation between the actual flight status of the flock and the expected path, so as to ensure that the flock of birds safely flies away from the airport area along the planned directional driving path.

[0148] The working principle of the above technical solution is as follows: Bird activity data is transmitted in real time to the central processing unit, including key information parameters such as the spatial coordinates of the flock, flight altitude, activity density, movement speed, and flight direction. Environmental parameter sensing devices simultaneously collect environmental element data such as current temperature, humidity, wind direction and speed, visibility, and atmospheric pressure. This environmental element data is of significant reference value for analyzing bird behavior characteristics. Upon receiving a bird dispersal command, the central processing unit immediately initiates a location determination program. This program compares the current location information of the bird flock with the pre-set boundary parameters of the safe target area in the system to perform spatial relationship calculations, thereby determining whether the bird flock has entered the control area requiring dispersal. The safe target area refers to the airspace frequently used by aircraft take-off and landing activities within a specific area around the airport runway. When the system confirms that the bird flock is within the boundary of the safe target area, it automatically activates a preset guided dispersal scheme. This scheme is based on avian behavioral principles and designs multiple dispersal path templates for different bird species. The system automatically selects the optimal dispersal strategy based on the currently identified bird species. The central processing unit calculates the optimal deterrence path based on the bird flock's location and airport area map. This path typically avoids sensitive areas, guiding the flock away along a safe passage. The deterrence actuator consists of an acoustic deterrence device and a laser deterrence device. The acoustic device emits sound waves in specific frequency bands, including predator warnings, distress calls, and ultrasonic signals; the laser deterrence device precisely projects beams of light with varying brightness and patterns. The central processing unit dynamically calculates and controls the activation timing and intensity of the acoustic and optical devices based on the bird flock's flight parameters, ensuring the devices activate in a pre-planned spatial sequence, creating a gradient acoustic and optical interference zone along the guide path. The acoustic frequency is adjusted according to the bird's auditory sensitivity curve, selecting the frequency range most effective for specific bird species; the wavelength and intensity of the laser beam are adaptively adjusted based on ambient lighting conditions and the visual characteristics of the target birds, forming a gradual deterrence channel from high-interference to low-interference areas. A gradual dispersal channel refers to a spatial region with unevenly distributed audio-visual stimulation intensity created around a flock of birds. High-interference zones are positioned in directions where the birds should not fly, while low-interference zones point in safe directions the system hopes the flock will migrate in. This utilizes birds' natural instinct to seek advantage and avoid harm, guiding them to actively choose low-interference zones. During the dispersal process, the central processing unit continuously receives bird position feedback data and calculates the deviation between the actual flight trajectory and the expected path in real time. When the deviation exceeds a preset threshold, the system immediately adjusts the operating parameters of the relevant dispersal equipment, including increasing or decreasing the intensity of audio-visual stimulation in specific directions, and reconstructing the gradient interference zone to ensure the flock continues to fly along the corrected directional dispersal path.The entire expulsion process is continuously monitored until the flock of birds has completely left the safe target area. The system then automatically generates an expulsion event report, recording core data such as bird species identification results, expulsion path, time taken, and expulsion success rate, providing data support for subsequent expulsion strategy optimization.

[0149] The beneficial effects of the above technical solution are as follows: by real-time monitoring and dynamic adjustment of the parameters of the audio-visual equipment, a gradient interference zone and a gradual repulsion channel are formed. By utilizing the instinctive behavioral characteristics of birds, they are guided to autonomously avoid high interference areas. This not only ensures the repulsion effect but also minimizes the stress damage to the bird flock, demonstrating the organic combination of science and technology with ecological protection.

[0150] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from the spirit and scope of this invention.

Claims

1. A bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system, characterized in that, include: S1: Identify the environment around the airport using radar and optoelectronic equipment, and perform environmental modeling based on Maklink link graph theory to generate a two-dimensional environment model; S2: Based on a two-dimensional environment model, the APF-RRT algorithm is used to plan the bird deterrence path of the airport intelligent bird deterrence system, determine the path points of the bird flock, and generate a set of bird deterrence paths. S3: Optimize the bird deterrence path set based on a preset probability function, increase the probability of the path passing through a safe target area, and generate a guiding bird deterrence path.

2. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 1, characterized in that, Step S1 includes: S11: Use radar and optoelectronic equipment to map and locate the environment around the airport, and identify actual obstacles and virtual obstacles. Actual obstacles refer to physical entities such as terrain and buildings, while virtual obstacles refer to artificially set up isolation zones for safety reasons. S12: Based on Maklink link graph theory, real and virtual obstacles are represented as polygons to generate a two-dimensional environment model.

3. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 1, characterized in that, Step S2 includes: S21: Determine the current position of the flock of birds as the starting point of the path in the two-dimensional environment model; S22: The Artificial Potential Field-Fast Expanding Random Tree (APF-RRT) algorithm is used to traverse spatial points and generate a tree-like path structure. S23: Based on the tree-like path structure, determine the path points of the bird flock's flight and generate a set of bird deterrence paths.

4. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 1, characterized in that, Step S3 includes: S31: For each path in the bird deterrence path set, calculate the probability that it passes through a preset safe target area. The safe target area refers to an area suitable for bird flocks to fly without affecting flight safety. S32: Increase the probability that the path passes through the safe target area by adjusting the probability function parameters; S33: Based on the optimized probability distribution, select the optimal guided deportation path.

5. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 2, characterized in that, Step S11 includes: S111: Collect three-dimensional data of the terrain and landforms around the airport using radar equipment; S112: Acquire image data of buildings surrounding the airport using optoelectronic equipment; S113: Integrate 3D data and image data to identify and locate actual and virtual obstacles, including isolation zones formed by light wall obstacles set up using lasers.

6. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 2, characterized in that, Step S12 includes: S121: Abstract both real and virtual obstacles into convex polygons; S122: Based on Maklink link graph theory, connect the vertices of convex polygons with the environment boundary to construct a two-dimensional link graph; S123: Based on the two-dimensional link graph, generate a two-dimensional environment model and identify narrow passages for bird flocks to fly.

7. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 3, characterized in that, Step S22 includes: S221: Starting from the path's origin, randomly select adjacent points to extend the path; S222: Utilize the artificial potential field (APF) to guide the path extension direction to avoid obstacles; S223: Repeat the path expansion until the predetermined spatial range is covered, generating a tree-like path structure.

8. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 3, characterized in that, Step S23 includes: S231: Identify dead zones in a tree-like path structure. Dead zones refer to areas that the path cannot bypass. S232: Adjust the path expansion direction based on the probability function to increase the probability that path points will lead to the safe target area; S233: Generate a set of bird deterrent routes based on the adjusted path points.

9. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 1, characterized in that, Also includes: S4: By using sound and laser devices to create sound and light interference along the calculated dispersal path, the flock of birds is driven to fly along the planned directional dispersal path.

10. The bird flock guidance and dispersal path planning method based on the Maklink airport intelligent bird control system according to claim 9, characterized in that, Step S4 includes: Real-time monitoring of bird activity within the airport area to obtain bird distribution and activity data; The sensing device detects environmental parameter information and transmits it to the central processing unit. After receiving the bird drive-away command, the central processing unit acquires bird activity data provided by the monitoring equipment, identifies the current location of the bird flock based on the bird activity data, and determines whether the bird flock is within the boundary of the safe target area. If the flock of birds is within the boundary of the preset safe area, then sound and laser devices will be activated at specific locations around the flock according to the pre-planned directional dispersal path. The central processing unit dynamically adjusts the activation sequence and working intensity of the sound and laser equipment according to the flight speed and direction of the flock of birds, forming a gradient sound and light interference zone on the guiding drive-away path; Among them, the frequency of the sound waves emitted by the sound device and the intensity of the beam projected by the laser device are adaptively adjusted according to the characteristics of the bird flock and environmental conditions, forming a gradual driving away channel from the high interference area to the low interference area, so that the bird flock instinctively avoids the high interference area and flies along the preset guiding driving away path. The central processing unit tracks the flight path of the flock of birds in real time and automatically adjusts the operating parameters of the sound and light interference equipment based on the deviation between the actual flight status of the flock and the expected path, so as to ensure that the flock of birds safely flies away from the airport area along the planned directional driving path.

Images