Bird strike defense precision protection and bird repelling method and system based on aircraft operation position

Abstract

This application discloses a precise bird strike prevention and control method and system based on aircraft flight location. The method collects real-time aircraft flight data and bird flock activity data, processes the data to extract features, and uses an LSTM model to predict the future trajectory of the bird flock and generate a position probability cloud map. Simultaneously, it dynamically constructs a three-dimensional safety channel centered on the aircraft based on different flight phases; then, it calculates the collision risk probability between the aircraft's future position and the bird flock probability cloud, compares it with a preset threshold to classify the risk level, and finally triggers graded bird deterrence measures based on the risk level. This application achieves precise protection "wherever the flight goes, protection follows," solving the problems of incomplete bird strike prevention and resource waste in existing technologies. It constructs a comprehensive, dynamic, and targeted bird strike prevention system, effectively improving the intelligence level and efficiency of airport bird strike prevention.

Description

Technical Field

[0001] This invention relates to the field of aviation safety technology, and more specifically, to a bird strike defense method and system based on the aircraft's operating position for precise bird deterrence. Background Technology

[0002] Bird strikes are a significant threat to aviation safety, especially during takeoff and landing, when bird activity highly overlaps with flight paths, easily leading to bird strike incidents. Currently, airport bird control primarily relies on timed or manual control methods, lacking the ability to respond to real-time aircraft positions. This results in low efficiency, wasted resources, and blind spots during flight intervals. During the critical takeoff and landing phases, if bird control equipment fails to activate in time before the aircraft approaches, birds may be suddenly startled and flee erratically, increasing the probability of bird strikes. Furthermore, if the selection and placement of bird control equipment are not aligned with the aircraft's operational location, ground-based deterrence devices can easily cause sudden disturbances and disorderly escapes from birds on the ground, significantly increasing the risk of bird strikes during low-altitude takeoff and landing. Therefore, the selection and standardized placement of bird control equipment should be tailored to the aircraft's operational location.

[0003] Furthermore, traditional systems lack the ability to perceive and predict real-time bird activity, and cannot dynamically adjust protection strategies based on bird behavior. Currently, airport bird control equipment mainly operates on a timed basis. This timed-driven mode is a crude and static control method, and its core drawback is that the fixed operating sequence cannot match the dynamic variables of the airport (flight takeoffs and landings, real-time bird activity, environmental changes), violating the core requirements of precise bird control and safety management. Specific professional drawbacks are as follows:

[0004] The system is disconnected from the dynamics of aircraft takeoff and landing, amplifying the safety risks of bird strikes. This mode, triggered / shut down at fixed times, cannot be integrated with real-time flight schedules or aircraft proximity signals, creating two major safety vulnerabilities: First, the equipment may not be activated during critical takeoff and landing phases, resulting in a gap in bird control and allowing flocks of birds to linger on the runway and approach area. Second, if the equipment is accidentally triggered when an aircraft approaches, the sudden sound and light stimuli can startle ground birds, causing them to fly erratically and directly threatening low-altitude aircraft, contradicting the original purpose of bird control.

[0005] Unable to adapt to real-time bird activity dynamics, the effectiveness of bird dispersal is polarized. Bird activity fluctuates dynamically due to seasonal migration, weather changes, food source distribution, and diurnal time. The fixed frequency, duration, and intensity of the timed pattern result in insufficient dispersal efforts during periods of high bird activity, failing to effectively disperse flocks; while excessive operation during periods of low bird activity is neither necessary for prevention and control nor does it cause ineffective disturbance, leading to overall low dispersal efficiency.

[0006] This can easily lead to bird adaptation (habituation), permanently eliminating the bird deterrent effect. Fixed-timed sound, light, ultrasonic, and gas cannon deterrent signals cause birds to develop stable conditioned reflexes, gradually adapting to the stimulus pattern until they eventually show no stress response to the device's deterrent actions. Once behavioral habituation occurs, the device will become completely ineffective, and this adaptation is irreversible, significantly increasing the difficulty of subsequent bird deterrent modifications.

[0007] Waste of equipment resources drives up overall operation and maintenance costs. Continuous operation when not needed for epidemic prevention and control will accelerate the aging of equipment hardware, consume electricity and consumables (such as gas and sound-generating components), and increase the manpower and material costs of daily inspection, maintenance and replacement. At the same time, static operation cannot achieve equipment load balancing, which can easily cause local equipment overload damage and increase the operation and maintenance costs throughout the entire life cycle.

[0008] Without coordinated and collaborative capabilities, it is difficult to build a comprehensive prevention and control system. The timed mode operates independently in a closed loop and cannot be linked with radar bird detection systems, flight information systems, or airport security systems. It lacks the closed-loop capability of accurate detection, intelligent analysis, and targeted bird removal. It cannot achieve multi-device coordinated deployment and differentiated bird removal by region, remaining at a single-point, extensive traditional prevention and control model, which cannot meet the requirements of modern airport smart bird control construction.

[0009] There is a lack of emergency response capabilities for sudden bird incidents. In response to emergency scenarios such as migratory bird flocks passing through, large-scale bird gatherings, and sudden bird incidents attracted by food sources, the timed mode cannot respond immediately or strengthen bird control. It can only wait for the bird to be triggered at a fixed time, which can easily lead to missing the best window of opportunity for intervention and causing regional bird strike safety hazards.

[0010] Combining civil aviation bird strike prevention technical specifications with aircraft operational safety logic, a planar prevention and control model that relies solely on ground-based bird deterrence equipment and fails to consider aircraft spatial trajectories for selection and deployment suffers from fundamental defects such as airspace coverage gaps, reverse safety risks, and ineffective prevention and control. It cannot fundamentally construct a bird strike defense system. Specific professional issues are as follows:

[0011] Vertical control coverage is fragmented, leaving core risk airspace completely out of control. Ground-based bird deterrence equipment (sonic waves, gas cannons, visual deterrence, ground traps, etc.) is highly effective at the 0-30 meter ground level, only able to control the activities of terrestrial birds; while the core airspace with high bird strike incidence during aircraft takeoff, climb, approach, and landing is the 30-500 meter low-altitude flight path. This area is the main space for migratory flocks and medium-sized birds to circle and cross, and ground equipment has no high-altitude deterrence capability, directly creating a vertical control gap and failing to effectively intercept bird strike risks during critical takeoff and landing phases.

[0012] Disturbing birds in the opposite direction directly exacerbates the immediate risk of bird strikes. Strong noises, explosions, and bright lights from ground equipment can cause ground birds to suddenly take flight and flee in disarray, forcing low-risk flocks from runways and lawns into the core flight paths of aircraft takeoffs and landings, transforming them into a direct collision hazard. Especially during critical flight windows, this reverse disturbance significantly increases the probability of bird strikes, completely contradicting the core objective of bird strike prevention.

[0013] Mismatch between spatial deployment and risk zoning leads to a lack of targeted prevention and control capabilities. Airports are divided into extremely high-risk areas such as runway end safety zones, approach surfaces, and takeoff / climb surfaces, as well as low-risk areas such as taxiways, aprons, and green belts, based on bird strike risk. Failure to consider aircraft spatial location when deploying equipment, and simply laying it out on the ground, results in insufficient control density in high-risk areas and redundant waste of resources in low-risk areas. This prevents precise control of key risk points and leads to a severe imbalance in the overall allocation of defense resources.

[0014] This can easily lead to spatial habituation in birds, rendering existing defense equipment ineffective over the long term. Fixed ground-based equipment exhibits highly fixed stimulus patterns and effective ranges, causing birds to quickly develop adaptive spatial avoidance behaviors—actively operating above the effective altitude of the ground equipment and congregating in the takeoff and descent zones of aircraft. Once these irreversible behavioral habits are formed, existing ground-based equipment will completely lose its deterrent effect, and the bird strike defense system will gradually collapse.

[0015] The lack of a three-dimensional, layered prevention and control system makes it impossible to address bird threats from all angles. Bird activity exhibits a strict vertical stratification: the ground layer consists of foraging birds such as passerines; the mid-to-low altitude layer is occupied by migratory birds and birds of prey; and the high altitude layer is home to long-distance migratory flocks. Ground-based equipment can only handle terrestrial birds and has no control over high-risk bird flocks crossing at mid-to-low altitudes. It cannot construct a three-dimensional, layered defense loop of "ground clearing, low-altitude interception, and high-altitude early warning," resulting in a natural lack of a prevention and control dimension.

[0016] Unable to integrate with intelligent bird strike prevention systems, its dynamic defense capability is zero. Modern airport bird strike prevention requires the coordinated response of ADS-B flight trajectories, bird-detecting radar spatial data, and bird-repelling equipment. It needs to trigger targeted bird removal based on the real-time spatial position of the aircraft and the three-dimensional coordinates of the bird flock. Single ground equipment lacks spatial adaptability and cannot achieve dynamic control of "flight approach - precise bird removal." It can only operate statically and is completely incompatible with the technical standards and safety requirements of intelligent airport bird strike prevention.

[0017] In summary, this model can only disperse birds in the shallow ground, which cannot cover the core risk airspace for aircraft take-off and landing, and may even create bird strike hazards. It is a superficial and ineffective approach to bird strike prevention and cannot solve the problem at its root. Summary of the Invention

[0018] The main objective of this invention is to provide a precise bird strike defense method and system based on the aircraft's operating position, so as to at least solve the problems of incomplete bird strike defense and superficial treatment in the existing technology, and to build a comprehensive, dynamic and targeted bird strike defense system.

[0019] To achieve the above objectives, a precise bird strike prevention and bird control method and system based on the aircraft's flight position is provided.

[0020] In a first aspect, the present invention provides a precise bird strike defense and bird control method based on the aircraft's operational position, the command method comprising:

[0021] Real-time acquisition of aircraft flight data and bird activity data in the target area; data processing of flight data and activity data to obtain standard flight data and standard activity data; extraction of flight features and bird features from standard flight data and standard activity data.

[0022] A bird flock trajectory prediction model is constructed. The bird flock trajectory prediction model is used to predict the centroid position of the bird flock at future time and generate a probability cloud map of the bird flock's position at future time.

[0023] Preset flight conditions, match flight conditions according to standard flight data to obtain different flight stages of the aircraft, preset multiple protection thresholds, match multiple protection thresholds based on different flight stages to build a dynamic safety channel centered on the aircraft.

[0024] Obtain the aircraft's first position data at the current moment from standard flight data, calculate the aircraft's second position data at the future moment based on the first position data, and calculate the probability of collision risk between the aircraft and the flock of birds based on the second position data and the center of mass position.

[0025] A preset collision risk level threshold is set, and the risk level at the current moment is obtained by comparing the collision risk probability with the collision risk level threshold. Different bird deterrence measures are taken in the dynamic safety passage according to different risk levels.

[0026] Specifically, real-time data collection includes aircraft flight data and bird activity data in the target area, including:

[0027] The system utilizes ADS-B receivers, radar, and visual positioning arrays at the airport to collect real-time flight data of aircraft; and bird detection radar, infrared thermal imagers, visible light imagers, and acoustic listening arrays to collect real-time activity data of bird flocks in the target area.

[0028] Specifically, flight data and activity data are processed to obtain standard flight data and standard activity data. Flight features and flock features are extracted from the standard flight data and standard activity data, including:

[0029] Outliers in the flight data are removed and physical constraints are introduced into the flight data to obtain standard flight data;

[0030] The α-β-γ filtering algorithm is used to extract the trajectory features of the aircraft from the standard flight data, and the altitude and vertical velocity features of the aircraft are obtained from the standard flight data as flight features.

[0031] The DBSCAN density clustering algorithm is used to aggregate activity data from discrete point states into population states to obtain standard activity data.

[0032] Based on standard activity data, the bird flock's center location, size, walking radius, and average velocity vector are calculated as flock characteristics.

[0033] Specifically, a bird flock trajectory prediction model is constructed. This model is used to predict the centroid position of the flock at future times and generate a probability cloud map of the flock's future positions. This includes:

[0034] A bird flock trajectory prediction model is constructed based on the LSTM model. The climate characteristics at the current moment are obtained, and the bird flock characteristics and climate characteristics are input into the bird flock trajectory prediction model to obtain the centroid position of the bird flock at the future moment.

[0035] A three-dimensional Gaussian kernel is superimposed at the centroid to obtain a position probability cloud map.

[0036] Specifically, preset flight conditions are used to match flight conditions with standard flight data to obtain different flight phases for the aircraft, including:

[0037] Acquire altitude, vertical velocity, and horizontal velocity data from standard flight data;

[0038] Preset flight conditions, including multiple altitude thresholds, multiple vertical velocity thresholds, and multiple horizontal velocity thresholds;

[0039] Multiple altitude thresholds, multiple vertical rate thresholds, and multiple horizontal rate thresholds are matched based on altitude data, vertical rate data, and horizontal rate data to obtain different flight phases of the aircraft.

[0040] Specifically, multiple altitude thresholds, multiple vertical velocity thresholds, and multiple horizontal velocity thresholds are matched based on altitude data, vertical velocity data, and horizontal velocity data to obtain different flight stages of the aircraft, including:

[0041]

[0042] in, Represents height data, Represents vertical velocity data. This represents horizontal velocity data.

[0043] Specifically, multiple protection thresholds are preset, and these thresholds are matched based on different flight phases to construct a dynamic safety channel centered on the aircraft, including:

[0044] Multiple protection thresholds are preset, including the warning time window of the dynamic safety channel, the horizontal half-axis and the vertical half-axis;

[0045] When the aircraft is taxiing on the ground, the warning time window for the dynamic safety passage, the lateral half-axis, and the vertical half-axis are 10s, 70m, and 15m, respectively.

[0046] When the aircraft is in the takeoff phase, the warning time window, the horizontal half-axis and the vertical half-axis of the dynamic safety passage are 12s, 100m and 40m respectively.

[0047] When the aircraft is in the climb phase, the warning time window for the dynamic safety passage, the lateral half-axis, and the vertical half-axis are 15s, 120m, and 60m, respectively.

[0048] When the aircraft is in the cruise phase, the warning time window for the dynamic safety passage, the lateral half-axis, and the vertical half-axis are 15s, 150m, and 50m, respectively.

[0049] When the aircraft is in the approach phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety passage are 8s, 110m and 50m respectively.

[0050] When the aircraft is in the landing phase, the warning time window, lateral half-axis and vertical half-axis of the dynamic safety passage are 12s, 80m and 20m respectively.

[0051] Specifically, the probability of a collision between the aircraft and the flock of birds is calculated based on the second position data and the centroid position, including:

[0052] For each future moment, generate 10 dynamic secure channel masks;

[0053] Determine whether a flock of birds is within a safe area based on the location of its centroid;

[0054] Determine whether the position of the centroid satisfies the ellipsoidal inequality;

[0055] The probability of collision between an aircraft and a flock of birds is calculated based on a dynamic safety channel mask and a position probability cloud map.

[0056] Specifically, a preset collision risk level threshold is used to compare the collision risk probability with the collision risk level threshold to obtain the risk level at the current moment, including:

[0057] The preset collision risk level thresholds are 0.2 and 0.5;

[0058] When the probability of collision risk is less than 0.2, the risk level at the current moment is determined to be low risk.

[0059] When 0.2 ≤ collision risk probability < 0.5, the risk level at the current moment is determined to be medium risk;

[0060] When the probability of collision risk is ≥0.5, the risk level at the current moment is determined to be high risk.

[0061] Secondly, this invention provides a bird strike defense precision protection bird control system based on aircraft operating position, wherein the identification system is applied to the command method of the first aspect, and the command system includes:

[0062] The data acquisition and processing unit is used to collect the flight data of the aircraft and the activity data of the bird flock in the target area in real time, process the flight data and activity data to obtain standard flight data and standard activity data, and extract flight features and bird flock features from the standard flight data and standard activity data.

[0063] The bird flock trajectory prediction unit is connected to the data acquisition and processing unit. The bird flock trajectory prediction unit is used to construct a bird flock trajectory prediction model, use the bird flock trajectory prediction model to predict the centroid position of the bird flock at future time, and generate a probability cloud map of the bird flock's position at future time.

[0064] The protected area generation unit is connected to the bird flock trajectory prediction unit. The protected area generation unit is used to preset flight conditions, match flight conditions according to standard flight data to obtain different flight stages of the aircraft, preset multiple protection thresholds, and match multiple protection thresholds based on different flight stages to construct a dynamic safety channel centered on the aircraft.

[0065] The collision risk prediction unit is connected to the protection zone generation unit. The collision risk prediction unit is used to obtain the first position data of the aircraft at the current moment from the standard flight data, calculate the second position data of the aircraft at the future moment based on the first position data, and calculate the collision risk probability between the aircraft and the flock of birds based on the second position data and the center of mass position.

[0066] The risk rating and graded response control unit is connected to the collision risk prediction unit. The risk rating and graded response control unit is used to preset the collision risk level threshold, compare the collision risk probability with the collision risk level threshold to obtain the risk level at the current moment, and take different bird deterrence measures in the dynamic safety passage according to different risk levels.

[0067] This application provides a precise bird strike prevention and bird control method and system based on aircraft flight location. The method collects real-time flight data of the aircraft and bird activity data of the target area. After data processing to extract flight and bird flock features, it uses an LSTM model to predict the future trajectory of the bird flock and generate a position probability cloud map. At the same time, it constructs a dynamic safety channel according to different flight stages of the aircraft, and then calculates the collision risk probability between the future position of the aircraft and the centroid of the bird flock. The probability is compared with a preset risk level threshold to determine the risk level, thereby taking graded bird control measures for different risk levels, achieving precise early warning and differentiated response to bird strike risks. Attached Figure Description

[0068] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0069] Figure 1 A flowchart illustrating a precise bird-repelling method for bird strike defense based on aircraft location, provided in this application;

[0070] Figure 2 A schematic diagram of the overall algorithm flow for a precise bird-repelling method based on aircraft location for bird strike defense provided in this application;

[0071] Figure 3 This application provides a connection diagram for a bird strike defense precision protection bird deterrence system based on aircraft operating position. Detailed Implementation

[0072] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0073] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein.

[0074] In this invention, the terms "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0075] This application provides a precise bird strike prevention and control method and system based on aircraft flight location. The method collects aircraft flight data and bird activity data in real time, and extracts features through data cleaning and clustering. Then, it uses an LSTM model to predict the future trajectory of the bird flock and generate a spatial location probability cloud map. At the same time, it dynamically adjusts the protection threshold according to the flight stage of the aircraft to build an adaptive safety channel. On this basis, it assesses the collision risk by calculating the degree of overlap between the future position of the aircraft and the probability cloud of the bird flock, and compares the risk probability with a preset threshold to classify it into three levels: low, medium and high. Finally, it triggers differentiated bird control measures according to the risk level, realizing precise early warning and hierarchical control of bird strike risk based on aircraft flight status.

[0076] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0077] Figure 1 A flowchart illustrating a precise bird-repelling method for bird strike defense based on aircraft location, as provided in this application, is shown below. Figure 1 As shown, this embodiment provides a precise bird-scaring method for bird strike defense based on aircraft location. The command method includes:

[0078] Real-time acquisition of aircraft flight data and bird activity data in the target area; data processing of flight data and activity data to obtain standard flight data and standard activity data; extraction of flight features and bird features from standard flight data and standard activity data.

[0079] A bird flock trajectory prediction model is constructed. The bird flock trajectory prediction model is used to predict the centroid position of the bird flock at future time and generate a probability cloud map of the bird flock's position at future time.

[0080] Preset flight conditions, match flight conditions according to standard flight data to obtain different flight stages of the aircraft, preset multiple protection thresholds, match multiple protection thresholds based on different flight stages to build a dynamic safety channel centered on the aircraft.

[0081] Obtain the aircraft's first position data at the current moment from standard flight data, calculate the aircraft's second position data at the future moment based on the first position data, and calculate the probability of collision risk between the aircraft and the flock of birds based on the second position data and the center of mass position.

[0082] A preset collision risk level threshold is set, and the risk level at the current moment is obtained by comparing the collision risk probability with the collision risk level threshold. Different bird deterrence measures are taken in the dynamic safety passage according to different risk levels.

[0083] This application provides a precise bird strike prevention and bird control method based on aircraft flight location. This method collects aircraft flight data and bird activity data in real time, extracts features after data processing, uses an LSTM model to predict the future trajectory of the bird flock and generates a position probability cloud map. At the same time, it dynamically constructs a safety channel according to different flight stages of the aircraft, and then calculates the collision risk probability between the future position of the aircraft and the centroid of the bird flock. By comparing it with a preset risk level threshold, the risk level is determined, and graded bird control measures are taken for different risk levels to achieve precise early warning and differentiated response to bird strike risks.

[0084] Figure 2 A schematic diagram of the overall algorithm flow for a precise bird-repelling method based on aircraft flight position, as provided in this application, is shown below. Figure 2 The diagram shows the overall algorithm flow of a bird strike defense precision protection bird deterrence method based on aircraft operating position provided in this embodiment.

[0085] This method achieves accurate prediction of bird flock movements through multi-source sensor data fusion and deep learning prediction models; it significantly improves the pertinence and accuracy of risk assessment by dynamically adjusting safety passages based on aircraft operation phases; it adopts a graded response mechanism to take bird deterrence measures according to risk levels, ensuring flight safety while avoiding resource waste caused by excessive bird deterrence; the entire method forms a complete closed-loop command system for bird strike defense, effectively improving the intelligence level and efficiency of airport bird strike prevention.

[0086] Specifically, real-time data collection includes aircraft flight data and bird activity data in the target area, including:

[0087] The system utilizes ADS-B receivers, radar, and visual positioning arrays at the airport to collect real-time flight data of aircraft; and bird detection radar, infrared thermal imagers, visible light imagers, and acoustic listening arrays to collect real-time activity data of bird flocks in the target area.

[0088] This application provides a precise bird-repelling and bird-strike defense method based on aircraft operating position. This method integrates multi-source detection equipment at the airport, using an ADS-B receiver, radar, and visual positioning array to collect the aircraft's first position data and flight data in real time. At the same time, it uses bird detection radar, infrared thermal imager, visible light imager, and acoustic listening array to obtain the third position data and bird activity data of the target area, thereby achieving a comprehensive and three-dimensional perception of the aircraft's operating status and the bird flock's activity pattern.

[0089] First, establish a unified spatiotemporal reference: connect all sensors to the Precise Time Protocol (PTP) network to ensure that the time synchronization error is ≤10ms; at the same time, convert all collected geographic coordinates into the Northeast Sky (ENU) rectangular coordinate system with the airport reference point as the origin, so as to provide a unified spatiotemporal reference for multi-source data fusion.

[0090] Secondly, multi-source acquisition of aircraft flight data is conducted: ADS-B receivers deployed at airports receive 1090ES broadcast signals to collect real-time first position data and navigation data such as the aircraft's ICAO address, latitude and longitude, altitude, ground speed, heading, and vertical speed; radar is used to supplement the acquisition of trajectory information for general aviation aircraft, military aircraft, or targets whose ADS-B signals have been lost without ADS-B equipment, providing point or track data; and a visual positioning array composed of high-definition infrared-visible dual-mode cameras deployed along the runway is used to identify aircraft taxiing on the ground and in the low-altitude segment (altitude <100m) to obtain their precise position information.

[0091] Secondly, multi-source data collection of bird activity is carried out: bird detection radars deployed at both ends of the runway and in ecological hotspots are used to collect real-time data on the three-dimensional position, speed, number, and population density of bird flocks in the target area; dual-light-visible light and infrared thermal imagers are used to identify birds roosting on the ground during the day and at night; and an acoustic listening array is used to run a bird voiceprint recognition model to help determine bird species and activity levels and obtain bird activity data.

[0092] This method employs multi-sensor fusion technology, overcoming the limitations of single detection methods and significantly improving the comprehensiveness and reliability of data acquisition. Through the coordinated use of visual and acoustic arrays, it enhances the accuracy of bird flock activity identification and anti-interference capabilities. The synchronous acquisition of aircraft and bird flock data provides a precise data foundation for subsequent trajectory prediction and risk assessment, thereby providing stronger data support for bird strike defense decisions.

[0093] Specifically, flight data and activity data are processed to obtain standard flight data and standard activity data. Flight features and flock features are extracted from the standard flight data and standard activity data, including:

[0094] Outliers in the flight data are removed and physical constraints are introduced into the flight data to obtain standard flight data;

[0095] The α-β-γ filtering algorithm is used to extract the trajectory features of the aircraft from the standard flight data, and the altitude and vertical velocity features of the aircraft are obtained from the standard flight data as flight features.

[0096] The DBSCAN density clustering algorithm is used to aggregate activity data from discrete point states into population states to obtain standard activity data.

[0097] Based on standard activity data, the bird flock's center location, size, walking radius, and average velocity vector are calculated as flock characteristics.

[0098] This application provides a precise bird-repelling method for bird strike defense based on aircraft flight position. This method removes outliers from flight data and introduces physical constraints to obtain standard flight data. Then, it uses an α-β-γ filtering algorithm to extract the trajectory features of the aircraft and obtains altitude and vertical velocity features from the first position data. At the same time, it uses the DBSCAN density clustering algorithm to aggregate discrete bird flock activity data into a group state to obtain standard activity data, and calculates features such as the center position of the bird flock, the group size, the dispersion radius, and the average velocity vector based on this.

[0099] First, outlier removal is performed on the collected raw aircraft flight data. Outliers with abrupt changes in speed or altitude, such as data exceeding the aircraft's physical performance limits, are identified and removed. Simultaneously, physical constraints are introduced for secondary filtering. For example, constraints such as a maximum climb rate ≤15m / s and a maximum speed range are set based on aircraft performance to remove data that does not conform to the laws of flight physics, thus obtaining smooth, physically reasonable standard flight data.

[0100] Secondly, an α-β-γ filtering algorithm is used to smooth the trajectory and extract features from the standard flight data. This filtering algorithm estimates the target's position, velocity, and acceleration using a recursive method, as shown in the following formula:

[0101] ;

[0102] ;

[0103] ;

[0104] in, , and These represent the aircraft position, velocity, and acceleration (trajectory characteristics) estimated after filtering at time k, respectively. , and These represent the position, velocity, and acceleration at time k, predicted based on the state at time k-1, respectively. This represents the position observation value at time k (position in standard flight data). Indicates the sampling interval. , and This represents the filter gain coefficient, which is set empirically to α=0.5, β=0.4, and γ=0.1.

[0105] The DBSCAN (Density-Based Spatial Clustering of Applications with Noise) density clustering algorithm is used to process discrete point cloud activity data collected by bird detection equipment. This process aggregates discrete individual detection points into group states, yielding standardized activity data. The core discrimination criterion of the DBSCAN algorithm is:

[0106]

[0107] in, This indicates a point p as the center and a radius of . The number of data points contained in the neighborhood of . This represents the minimum cluster size threshold (set to 3). This represents the neighborhood radius, which is dynamically adjusted based on the resolution of detection equipment such as radar.

[0108] After identifying each bird flock using DBSCAN clustering, the following characteristic parameters of the bird flocks were calculated based on the clustering results:

[0109] The formula for calculating the centroid of the flock is:

[0110]

[0111] in, Indicates the centroid of the flock. Indicates the number of individuals in a flock of birds. Represents the three-dimensional spatial coordinates of the i-th individual in the flock.

[0112] Size characteristics (flock size): That is, the number of individuals contained in a cluster.

[0113] The formula for calculating the scatter radius characteristic is:

[0114]

[0115] in, This represents the maximum distance from a stray individual in a flock to the center, used to describe the spatial spread of the flock.

[0116] The formula for calculating the average velocity vector is:

[0117]

[0118] in, Represents the average velocity vector. This represents the velocity vector of the i-th individual in the flock, used to describe the overall movement trend of the flock.

[0119] This method effectively improves the accuracy and physical rationality of flight data through a dual mechanism of outlier removal and physical constraints; the application of the α-β-γ filtering algorithm enables smooth tracking and feature extraction of aircraft trajectories; the DBSCAN clustering algorithm can adaptively identify the flock aggregation state, overcoming the shortcomings of traditional methods in handling the fuzzy boundaries of flocks; and the finally extracted multidimensional flock features provide comprehensive and accurate input parameters for subsequent trajectory prediction and risk analysis.

[0120] Specifically, a bird flock trajectory prediction model is constructed. This model is used to predict the centroid position of the flock at future times and generate a probability cloud map of the flock's future positions. This includes:

[0121] A bird flock trajectory prediction model is constructed based on the LSTM model. The climate characteristics at the current moment are obtained, and the bird flock characteristics and climate characteristics are input into the bird flock trajectory prediction model to obtain the centroid position of the bird flock at the future moment.

[0122] A three-dimensional Gaussian kernel is superimposed at the centroid to obtain a position probability cloud map.

[0123] This application provides a precise bird-scaring method for bird strike defense based on aircraft flight position. This method constructs a bird flock trajectory prediction model based on an LSTM model, inputting the current bird flock characteristics and climate characteristics into the model to predict the centroid position of the bird flock in the future. Subsequently, a three-dimensional Gaussian kernel is superimposed on the predicted centroid position to generate a spatial probability cloud map of the bird flock in the future. The probability density of the bird flock at any point in space is calculated based on the standard deviation between the centroid position and the Gaussian kernel, thereby realizing a probabilistic expression of the future movement trend of the bird flock.

[0124] First, a bird flock trajectory prediction model is constructed based on a Long Short-Term Memory (LSTM) network. The LSTM model can effectively capture the temporal dependencies of bird flock movements and is suitable for dynamic trajectory prediction.

[0125] Construct the input feature vector. Sample at fixed time intervals (e.g., every 5 seconds), and set the sliding window length to 6 (i.e., using historical data from the past 30 seconds) to construct the input feature vector at the current time t. :

[0126]

[0127] in, This represents the input feature vector. , and This represents the three-dimensional spatial coordinates of the centroid of the flock at the current moment (derived from the extracted features of the flock's center position). , and This represents the components of the average velocity vector of the flock in the three directions at the current moment (derived from the extracted average velocity vector features). This represents the temperature data (climate characteristics) at the current moment. This indicates the wind speed and direction data (climate characteristics) at the current moment.

[0128] The aforementioned feature vectors are input into the bird flock trajectory prediction model, and the model outputs the predicted position of the bird flock's centroid at a future time (e.g., within the next 30 seconds). The prediction step size can be set according to actual needs, and is usually used to generate the predicted position of discrete time points in seconds.

[0129] Because flock movement has an inherent randomness, a single point prediction cannot fully express the uncertainty of its future position. Therefore, the centroid position at each predicted future moment... At that point, a three-dimensional Gaussian kernel is superimposed to generate a probability cloud map of the bird flock's position in space at that moment.

[0130] For any future moment Flocks of birds at any point in space The probability density function at point is defined as:

[0131]

[0132] in, This represents the probability density value of a flock of birds appearing at spatial point r at a future time, with a value ranging from [0, 1]. Represents the three-dimensional coordinates of any point in space. This represents the centroid position of the flock at a future time, as predicted by the LSTM model. This represents the Euclidean distance from the spatial point r to the predicted centroid location. The standard deviation of the three-dimensional Gaussian kernel is used to control the diffusion of the probability distribution and reflects the uncertainty of the spatial dispersion of bird flocks. It is calibrated based on actual observation data, and is typically taken as an empirical value of 10m or dynamically adjusted according to the size characteristics of the bird flock. This represents an exponential function.

[0133] The LSTM model in this method can effectively capture the time-series dependence of bird flock movement, and the combination of climate feature input significantly improves the accuracy of trajectory prediction. The uncertainty of bird flock location is expressed in the form of location probability cloud map, which overcomes the limitation of traditional point prediction that cannot reflect the range of bird flock dispersal. The introduction of three-dimensional Gaussian kernel enables the system to quantify the probability of bird flock existence at various points in space, providing a more scientific probability measurement basis for subsequent collision risk assessment.

[0134] Specifically, preset flight conditions are used to match flight conditions with standard flight data to obtain different flight phases for the aircraft, including:

[0135] Acquire altitude, vertical velocity, and horizontal velocity data from standard flight data;

[0136] Preset flight conditions, including multiple altitude thresholds, multiple vertical velocity thresholds, and multiple horizontal velocity thresholds;

[0137] Multiple altitude thresholds, multiple vertical rate thresholds, and multiple horizontal rate thresholds are matched based on altitude data, vertical rate data, and horizontal rate data to obtain different flight phases of the aircraft.

[0138] This application provides a precise bird strike defense method based on the aircraft's flight position. This method obtains key parameters such as the aircraft's altitude, vertical speed, and horizontal speed from standard flight data, and presets flight conditions including multiple altitude thresholds, vertical speed thresholds, and horizontal speed thresholds. Then, it matches real-time flight data with these thresholds to accurately identify the different flight stages the aircraft is currently in.

[0139] From the preprocessed standard flight data, obtain the following three key motion parameters for the current time t:

[0140] Altitude data: The aircraft's current vertical altitude from the airport reference point or mean sea level, in meters (m); Vertical rate data: The instantaneous rate of change of the aircraft's altitude, indicating the speed of climb or descent, in meters per second (m / s), with positive values ​​indicating climb and negative values ​​indicating descent; Horizontal rate data: The aircraft's horizontal speed relative to the ground, in knots (kt) or meters per second (m / s).

[0141] Preset flight conditions are defined by establishing a set of threshold rules to distinguish different flight phases. These flight conditions include multiple altitude thresholds, multiple vertical velocity thresholds, and multiple horizontal velocity thresholds. Specific threshold settings are as follows: Altitude thresholds: 10m, 30m, 100m, 500m; Vertical velocity thresholds: ±1m / s, ±2m / s, ±3m / s (where positive and negative signs distinguish between climb and descent); Horizontal velocity threshold: 30kt (approximately 15.4m / s).

[0142] The real-time acquired altitude, vertical velocity, and horizontal velocity data are logically matched with multiple preset thresholds, and the current flight phase of the aircraft is determined based on the following piecewise function:

[0143]

[0144] in, Represents height data, Represents vertical velocity data. This represents horizontal velocity data.

[0145] This method employs a multi-dimensional threshold matching mechanism, comprehensively utilizing multiple indicators such as altitude, vertical velocity, and horizontal velocity, significantly improving the accuracy and reliability of flight phase identification. By pre-setting clear threshold conditions, it achieves standardization and automation of flight phase identification, providing a precise time window basis for the subsequent construction of dynamic safety channels. This method is simple and efficient, capable of responding to changes in aircraft operating status in real time, ensuring that bird strike prevention measures remain highly synchronized with the actual operating status of the aircraft.

[0146] Specifically, multiple altitude thresholds, multiple vertical velocity thresholds, and multiple horizontal velocity thresholds are matched based on altitude data, vertical velocity data, and horizontal velocity data to obtain different flight stages of the aircraft, including:

[0147]

[0148] in, Represents height data, Represents vertical velocity data. This represents horizontal velocity data.

[0149] This application provides a precise bird strike prevention and bird control method based on aircraft position. This method utilizes real-time acquired aircraft altitude data (ht) and vertical velocity data (ht). The horizontal velocity data (vt) is substituted into a preset piecewise function to make conditional judgments: when ht < 10m and vt < 30kt, it is determined to be ground taxiing; when ht ∈ [10, 100]m and dhdt > 2m / s, it is takeoff; when ht ∈ [100, 500]m and dhdt > 3m / s, it is climb phase; when ht > 500m and dhdt < 2m / s, it is cruise phase; when ht < 500m and dhdt < -2m / s, it is approach phase; and when ht < 30m and dhdt < 1m / s, it is landing phase, thereby accurately identifying the current flight phase of the aircraft.

[0150] This method achieves highly automated and refined identification of flight phases through quantitative threshold combinations and segmented logic. The judgment conditions for each phase are closely combined with the actual operating characteristics of the aircraft, such as distinguishing between the takeoff and climb phases, making the phase division more professional and practical. The clear mathematical expressions facilitate real-time calculation and decision-making by the computer system, providing accurate and reliable phase inputs for subsequent dynamic adjustment of the safety passage, and significantly improving the response targeting of the bird strike defense system.

[0151] Specifically, multiple protection thresholds are preset, and these thresholds are matched based on different flight phases to construct a dynamic safety channel centered on the aircraft, including:

[0152] Multiple protection thresholds are preset, including the warning time window, horizontal half-axis, and vertical half-axis of the dynamic security channel;

[0153] When the aircraft is taxiing on the ground, the warning time window for the dynamic safety passage, the lateral half-axis, and the vertical half-axis are 10s, 70m, and 15m, respectively.

[0154] When the aircraft is in the takeoff phase, the warning time window, the horizontal half-axis and the vertical half-axis of the dynamic safety passage are 12s, 100m and 40m respectively.

[0155] When the aircraft is in the climb phase, the warning time window for the dynamic safety passage, the lateral half-axis, and the vertical half-axis are 15s, 120m, and 60m, respectively.

[0156] When the aircraft is in the cruise phase, the warning time window for the dynamic safety passage, the lateral half-axis, and the vertical half-axis are 15s, 150m, and 50m, respectively.

[0157] When the aircraft is in the approach phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety passage are 8s, 110m and 50m respectively.

[0158] When the aircraft is in the landing phase, the warning time window, lateral half-axis and vertical half-axis of the dynamic safety passage are 12s, 80m and 20m respectively.

[0159] This application provides a precise bird strike prevention and bird control method based on the aircraft's operating position. This method presets differentiated protection thresholds for different flight stages of the aircraft, and sets specific warning time windows, lateral half-axis and vertical half-axis parameters for each stage: 10 seconds, 70 meters and 15 meters for the ground taxiing stage, 12 seconds, 100 meters and 40 meters for the takeoff stage, 15 seconds, 120 meters and 60 meters for the climb stage, 15 seconds, 150 meters and 50 meters for the cruise stage, 8 seconds, 110 meters and 50 meters for the approach stage, and 12 seconds, 80 meters and 20 meters for the landing stage, thereby constructing a three-dimensional safety channel centered on the aircraft and dynamically adjusted according to the flight stage.

[0160] An ellipsoid is adopted as the basic geometric shape of the space protection zone. The dynamic safety passage is a three-dimensional ellipsoidal protection zone centered on the aircraft's current position and extending along the flight direction; its size and shape are dynamically adjusted according to the aircraft's flight phase. The three semi-axes of the ellipsoid are defined as follows:

[0161] Major semi-axis a: Axial radius along the flight direction (heading), determining the forward extension distance of the passageway; Lateral semi-axis b: Horizontal lateral radius perpendicular to the flight direction, determining the left and right width of the passageway; Vertical semi-axis c: Radius in the vertical direction, determining the vertical height range of the passageway.

[0162] Based on the identified current flight phase of the aircraft, corresponding protection parameters are matched from a preset protection threshold table. These protection thresholds include the warning time window, the lateral half-axis, and the vertical half-axis. The specific matching relationships are shown in Table 1.

[0163] Table 1 Matching Flight Phase with Protection Thresholds

[0164] Flight phase (s) Lateral half-axis b (m) Vertical semi-axis c (m) Ground sliding 10 70 15 Takeoff 12 100 40 Climbing phase 15 120 60 cruise 15 150 50 Approach 8 110 50 landing 12 80 20

[0165] Let the aircraft's current flight state be:

[0166] Location coordinates: That is, the current three-dimensional spatial position of the aircraft;

[0167] Ground speed: , that is, the magnitude of the aircraft's horizontal speed;

[0168] Heading angle: , which is the direction angle of the aircraft's horizontal flight, is used to determine the direction of the major axis of the ellipsoid.

[0169] Based on the matched warning time window, the semi-major axis of the dynamic safety channel is calculated using the following formula:

[0170]

[0171] in, This represents the major semi-axis of the dynamic safety passage, measured in meters (m). This indicates the aircraft's ground speed, measured in meters per second (m / s). This indicates the warning time window, measured in seconds (s), representing how far in advance the system needs to begin protection.

[0172] The horizontal and vertical half-axis are obtained directly from the threshold table above, and determine the width and height range of the channel, respectively.

[0173] Construct a dynamic safety passage ellipsoidal with the aircraft's current position as the center and the heading direction as the main axis. For any point in space, transform it to a coordinate system centered on the aircraft:

[0174]

[0175] in, Indicates the transformed coordinates. Represents the original coordinates. This represents a rotation matrix used to rotate the global coordinates to a coordinate system with the aircraft's heading as the main axis, so that the x-axis is aligned with the aircraft's flight direction, the y-axis is horizontal, and the z-axis is vertical.

[0176] Then point The condition for being located within a dynamic safety passage is that the ellipsoidal inequality must be satisfied:

[0177]

[0178] in, , and Point The three coordinate components in the aircraft heading coordinate system Indicates the major semi-axis (forward extension distance). This indicates the horizontal half-axis (left and right width). This indicates the vertical half-axis (vertical height).

[0179] The spatial region that satisfies the above inequalities is the dynamic safety corridor centered on the aircraft. This corridor moves with the aircraft and adaptively adjusts its size according to changes in the flight phase, achieving precise coverage of the critical protected airspace around the aircraft.

[0180] This method fully considers the differences in the sensitivity of aircraft to bird strike risk at different operational stages. By dynamically adjusting the size of the safety passage and the warning time, the protection range is highly matched with the actual operational status of the aircraft. The parameters for each stage are set based on the aircraft's performance and risk characteristics. For example, the lateral range is expanded during the high-speed cruise phase, and vertical protection is strengthened during the critical take-off and landing phases, which significantly improves the pertinence and accuracy of risk assessment. This hierarchical dynamic protection mechanism can avoid excessive warnings while ensuring safety, and achieve the optimal allocation of protection resources.

[0181] Specifically, the probability of a collision between the aircraft and the flock of birds is calculated based on the second position data and the centroid position, including:

[0182] For each future moment, generate 10 dynamic secure channel masks;

[0183] Determine whether a flock of birds is within a safe area based on the location of its centroid;

[0184] Determine whether the position of the centroid satisfies the ellipsoidal inequality;

[0185] The probability of collision between an aircraft and a flock of birds is calculated based on a dynamic safety channel mask and a position probability cloud map.

[0186] This application provides a precise bird strike defense method based on aircraft location. The method generates 10 dynamic safety channel masks for each future moment. First, it determines whether the bird flock is within the safe area based on the centroid position of the flock, and then verifies the geometric constraints using ellipsoidal inequalities. Finally, it integrates the dynamic safety channel masks with the probability cloud map of the bird flock's position to accurately calculate the collision risk probability between the aircraft and the bird flock.

[0187] Assuming the aircraft maintains a constant speed for a short period after the current time t (e.g., within the next 10 seconds), then the aircraft at future times... Second position data The calculation formula is:

[0188]

[0189] in, This represents the aircraft's current position data (three-dimensional coordinates). The three-dimensional velocity components representing the aircraft's current moment can be obtained from standard flight data. Indicates the time step, in seconds (s). This indicates the predicted aircraft at future moments. The location coordinates.

[0190] For each future moment, there is a different prediction time step. Generate 10 dynamic secure channel masks in seconds. Each mask It is a three-dimensional binary space matrix, representing the time... At that time, to predict the aircraft's position The spatial area covered by the central dynamic safety passage. For any point in space... At any given moment The criterion for determining whether something belongs to a safe passage is whether the ellipsoidal inequality is satisfied.

[0191] First, obtain the centroid position of the flock of birds at the current moment. And use a bird flock trajectory prediction model to obtain future moments. Predicted centroid position of bird flock .

[0192] For each future moment Calculate the relative vector of the flock's centroid with respect to the aircraft's predicted position at that moment:

[0193]

[0194] Then, this relative vector is transformed into a coordinate system with the aircraft's heading as the principal axis, resulting in... The condition for determining whether the centroid of the flock is located within the dynamic safety passage at that moment is that the ellipsoidal inequality is satisfied:

[0195]

[0196] in, These represent the major, lateral, and vertical semi-axis of the dynamic safety passage at that moment, respectively, determined based on the aircraft's flight phase at that time. This represents the three components of the relative vector in the aircraft heading coordinate system.

[0197] If the above inequality is satisfied, it indicates that the centroid of the flock is located within the dynamic safety passage at that moment, and the mask... The value at this spatial point is True (or 1); otherwise, it is False (or 0).

[0198] By performing spatiotemporal coupling analysis between the dynamic safety passage mask and the probability cloud map of the bird flock's position at the corresponding time, the collision risk probability between the aircraft and the bird flock is calculated. The bird flock at future time... The spatial location probability cloud map is obtained from the step "Constructing the bird flock trajectory prediction model", and its probability density at any point $r$ in space is: For each future moment Calculate the collision risk probability at that moment. :

[0199]

[0200] in, Indicates time The probability cloud map of bird flock locations, i.e., the probability density matrix of each point in space. Indicates time Dynamic security channel mask matrix, This represents element-wise multiplication (Hadamard product), meaning that only the probability values ​​of spatial points where the mask is True (the flock of birds is inside the safe passage) are retained, while the probability values ​​of points where the mask is False are set to zero. This indicates taking the maximum probability value within the intersection region.

[0201] For all predicted future moments seconds, if any Make (This threshold can be adjusted according to actual needs), then it is determined to be an event requiring intervention, triggering subsequent bird deterrence strategies. The final collision risk probability can be defined as the maximum risk value across all future moments:

[0202]

[0203] Through the above steps, a quantitative risk assessment was achieved by spatiotemporally coupling the uncertainty of the bird flock's location (location probability cloud map) with the future trajectory of the aircraft (dynamic safety channel), providing a scientific basis for subsequent graded response.

[0204] This method employs a multi-judgment mechanism (regional judgment, ellipsoidal inequality) to progressively screen risk scenarios, significantly improving the reliability and accuracy of risk assessment. By generating multiple safety channel masks, it effectively addresses the uncertainty of aircraft trajectories. The calculation method that combines position probability cloud maps with masks can quantify the probability of bird flocks existing within safety channels, avoiding the limitations of traditional binary judgment. This multi-dimensional probability calculation method provides a scientific and precise quantitative basis for subsequent risk classification and decision-making.

[0205] Specifically, a preset collision risk level threshold is used to compare the collision risk probability with the collision risk level threshold to obtain the risk level at the current moment, including:

[0206] The preset collision risk level thresholds are 0.2 and 0.5;

[0207] When the probability of collision risk is less than 0.2, the risk level at the current moment is determined to be low risk.

[0208] When 0.2 ≤ collision risk probability < 0.5, the risk level at the current moment is determined to be medium risk;

[0209] When the probability of collision risk is ≥0.5, the risk level at the current moment is determined to be high risk.

[0210] This application provides a precise bird strike prevention and bird control method based on the aircraft's operating position. The method uses two preset collision risk level thresholds of 0.2 and 0.5 to compare the calculated collision risk probability with the thresholds: when the probability is less than 0.2, it is judged as low risk; when the probability is greater than or equal to 0.2 and less than 0.5, it is judged as medium risk; and when the probability is greater than or equal to 0.5, it is judged as high risk, thereby realizing the classification of bird strike risk at the current moment.

[0211] The intensity and scope of bird control measures are dynamically adjusted based on the risk level. First, it is determined whether there are any flights in operation. If not, all equipment is placed in standby mode. If there are flights, a tiered response is implemented based on the flight phase and risk level.

[0212] In scenarios without bird detection equipment: The system directly queries the preset "flight phase-equipment combination" strategy table. For example, when the aircraft enters the "takeoff" phase, the system immediately activates omnidirectional acoustic wave, gas cannon, air cannon, low-altitude laser, ultrasonic wave, and directional acoustic wave on both sides of the runway; when entering the "climb" phase, it also activates high-altitude gas cannon, high-altitude omnidirectional acoustic wave, and bird deterrent launcher. The equipment follows the principle of "early activation and transit shutdown" to ensure that the aircraft is under effective protection throughout the entire flight.

[0213] In scenarios with bird detection equipment: Based on the above, the system integrates bird data to calculate conflict risks and dynamically fine-tunes the equipment's operation to achieve a better energy efficiency ratio.

[0214] The activation methods for bird deterrence devices in takeoff and landing scenarios are as follows:

[0215] Takeoff scene:

[0216] The activation order is ground equipment → low-altitude equipment → high-altitude equipment (activated 15-30 seconds in advance along the takeoff direction).

[0217] Once the tail of the aircraft leaves the equipment's coverage radius, the system will shut down after a 2-second delay, while equipment in unaffected areas will remain on standby.

[0218] Landing scene:

[0219] When the aircraft descends to ≤250m and enters the glide slope, the activation sequence is as follows: the equipment areas to be flown over are activated sequentially, and the equipment is deactivated one by one after the aircraft has flown over the equipment coverage area. When the aircraft is taking off, when the aircraft is at the end of the runway preparing for takeoff, the activation sequence is as follows: the equipment areas to be flown over are activated sequentially, and the equipment is automatically deactivated after the aircraft's altitude reaches ≥200m.

[0220] Bird deterrent launchers, being classified as hazardous equipment, must meet both of the following independent conditions before they can be fired:

[0221] Condition 1: Aircraft position authorization (first control)

[0222] The aircraft is located within a geofence ≤ 500 m from the launcher.

[0223] Condition 2: Bird situation confirmation authorization (second control)

[0224] There are valid bird targets within the protected area, meaning the current bird strike risk level is high.

[0225] In summary, as shown in Table 2, the correspondence between flight phases and response strategies is as follows:

[0226] Table 2 Correspondence between Flight Phases and Response Strategies

[0227] Flight phase Enable device type Control Mode Aircraft takeoff phase Ground taxiing and takeoff ≤50 meters Omnidirectional sound waves, gas cannons, air cannons, low-altitude lasers, ultrasound, directional sound waves Ground + Low-altitude + All-area Climb (50–200m) High-altitude gas cannons, high-altitude air cannons, high-altitude omnidirectional sonic cannons, laser-based sonic cannons, directional sonic cannons, bird deterrent projectile launchers High altitude + all-area aircraft landing phase Approach ≤250m High-altitude gas cannons, high-altitude air cannons, high-altitude omnidirectional sonic cannons, laser-based sonic cannons, directional sonic cannons, bird deterrent projectile launchers High altitude + all-area Landing Omnidirectional sound waves, gas cannons, air cannons, low-altitude lasers, directional sound waves, and ultrasound. Ground + Low-altitude + All-area High-risk events All relevant equipment + bird deterrent ammunition (dual control authorization) Full-scale synchronous high-intensity response

[0228] This method employs a dual-threshold grading mechanism, transforming continuous risk probability values ​​into discrete risk levels. This makes risk assessment results more intuitive and easier to understand, facilitating commanders' rapid comprehension of the current situation. The three-level risk rating covers a complete gradient from safe to dangerous, providing a clear decision-making basis for subsequent differentiated bird control measures. The selection of thresholds of 0.2 and 0.5 balances the sensitivity and specificity of risk, enabling timely warnings of potential risks while avoiding frequent false alarms due to excessively low thresholds, thus achieving a balance between the accuracy and practicality of risk warnings.

[0229] Figure 3 A connection diagram of a precision bird-repelling system for bird strike defense based on aircraft location provided in this application is shown below. Figure 3 As shown, this embodiment provides a bird strike defense precision protection bird control system based on aircraft operating position. The command system includes:

[0230] The data acquisition and processing unit is used to collect the flight data of the aircraft and the activity data of the bird flock in the target area in real time, process the flight data and activity data to obtain standard flight data and standard activity data, and extract flight features and bird flock features from the standard flight data and standard activity data.

[0231] The bird flock trajectory prediction unit is connected to the data acquisition and processing unit. The bird flock trajectory prediction unit is used to construct a bird flock trajectory prediction model, use the bird flock trajectory prediction model to predict the centroid position of the bird flock at future time, and generate a probability cloud map of the bird flock's position at future time.

[0232] The protected area generation unit is connected to the bird flock trajectory prediction unit. The protected area generation unit is used to preset flight conditions, match flight conditions according to standard flight data to obtain different flight stages of the aircraft, preset multiple protection thresholds, and match multiple protection thresholds based on different flight stages to construct a dynamic safety channel centered on the aircraft.

[0233] The collision risk prediction unit is connected to the protection zone generation unit. The collision risk prediction unit is used to obtain the first position data of the aircraft at the current moment from the standard flight data, calculate the second position data of the aircraft at the future moment based on the first position data, and calculate the collision risk probability between the aircraft and the flock of birds based on the second position data and the center of mass position.

[0234] The risk rating and graded response control unit is connected to the collision risk prediction unit. The risk rating and graded response control unit is used to preset the collision risk level threshold, compare the collision risk probability with the collision risk level threshold to obtain the risk level at the current moment, and take different bird deterrence measures in the dynamic safety passage according to different risk levels.

[0235] This application provides a precise bird strike defense and bird control system based on aircraft location. The system consists of a data acquisition and processing unit, a bird flock trajectory prediction unit, a protected area generation unit, a collision risk prediction unit, and a risk rating and graded response control unit connected in sequence. The data acquisition and processing unit is responsible for collecting real-time data on the aircraft and bird flocks and extracting features. The bird flock trajectory prediction unit predicts the future centroid position of the bird flock and generates a position probability cloud map. The protected area generation unit constructs a dynamic safety channel based on the aircraft's flight phase. The collision risk prediction unit calculates the probability of collision between the future position of the aircraft and the bird flock. The risk rating and graded response control unit finally classifies the risk level by comparing it with preset thresholds and triggers graded bird control measures, thus forming a complete closed-loop bird strike defense command system.

[0236] The system adopts a modular design, with clear division of labor among functional units and close data flow, ensuring high efficiency and maintainability of system operation. It forms a complete closed loop from data acquisition to risk response, realizing intelligent management of the entire bird strike prevention process. The standardized interface design between units facilitates system upgrades and functional expansion. Through the collaborative work of multiple units, the complex bird strike prevention problem is decomposed into sub-tasks that can be independently optimized, significantly improving the overall performance and reliability of the system and providing a systematic technical solution for airport bird strike prevention.

Claims

1. A precise bird-repelling method for bird strike defense based on aircraft flight position, characterized in that, The command methods include: Real-time acquisition of aircraft flight data and bird activity data in the target area; data processing of the flight data and activity data to obtain standard flight data and standard activity data; extraction of flight features and bird features from the standard flight data and standard activity data. A bird flock trajectory prediction model is constructed, and the bird flock trajectory prediction model is used to predict the centroid position of the bird flock at future time and generate a probability cloud map of the bird flock's position at future time. Preset flight conditions, match the flight conditions according to the standard flight data to obtain different flight stages of the aircraft, preset multiple protection thresholds, match multiple protection thresholds based on different flight stages to construct a dynamic safety channel centered on the aircraft; The aircraft's first position data at the current moment is obtained from the standard flight data. The aircraft's second position data at a future moment is calculated based on the first position data. The probability of collision between the aircraft and the flock of birds is calculated based on the second position data and the center of mass position. A preset collision risk level threshold is set, and the risk level at the current moment is obtained by comparing the collision risk probability with the collision risk level threshold. Different bird deterrence measures are taken in the dynamic safety passage according to different risk levels.

2. The bird strike defense and precise bird control method based on aircraft flight position according to claim 1, characterized in that, The real-time acquisition of aircraft flight data and bird activity data in the target area includes: The flight data of the aircraft is collected in real time using an ADS-B receiver, radar, and visual positioning array at the airport; the activity data of the bird flock in the target area is collected in real time using a bird detection radar, infrared thermal imager, visible light imager, and acoustic listening array.

3. The bird strike defense and precise bird control method based on aircraft flight position according to claim 1, characterized in that, The process of processing the flight data and activity data to obtain standard flight data and standard activity data, and extracting flight features and flock features from the standard flight data and standard activity data, includes: Outliers in the flight data are removed and physical constraints are introduced into the flight data to obtain the standard flight data; The trajectory features of the aircraft are extracted from the standard flight data using an α-β-γ filtering algorithm, and the altitude and vertical velocity features of the aircraft are obtained from the standard flight data as the flight features. The DBSCAN density clustering algorithm is used to aggregate the activity data from discrete point states into group states to obtain the standard activity data. Based on the standard activity data, the bird flock's center position characteristics, size characteristics, walking radius characteristics, and average speed vector are calculated as the bird flock characteristics.

4. The bird strike defense and precise bird control method based on aircraft flight position according to claim 1, characterized in that, The construction of the bird flock trajectory prediction model, which uses the model to predict the centroid position of the bird flock at future times and generate a probability cloud map of the bird flock's future positions, includes: The bird flock trajectory prediction model is constructed based on the LSTM model. The climate characteristics at the current moment are obtained, and the bird flock characteristics and the climate characteristics are input into the bird flock trajectory prediction model to obtain the centroid position of the bird flock at the future moment. A three-dimensional Gaussian kernel is superimposed at the centroid location to obtain the position probability cloud map.

5. The bird strike defense and precise bird control method based on aircraft flight position according to claim 1, characterized in that, The preset flight conditions, which are matched with the standard flight data to obtain different flight phases of the aircraft, include: Acquire altitude, vertical velocity, and horizontal velocity data from the standard flight data; The flight conditions are preset, wherein the flight conditions include multiple altitude thresholds, multiple vertical velocity thresholds, and multiple horizontal velocity thresholds; Based on the altitude data, the vertical rate data, and the horizontal rate data, multiple altitude thresholds, multiple vertical rate thresholds, and multiple horizontal rate thresholds are matched to obtain different flight phases of the aircraft.

6. The bird strike defense and precise bird control method based on aircraft flight position according to claim 5, characterized in that, The step of matching multiple altitude thresholds, multiple vertical rate thresholds, and multiple horizontal rate thresholds with the altitude data, the vertical rate data, and the horizontal rate data to obtain different flight phases of the aircraft includes: in, This indicates the height data. This represents the vertical rate data. This refers to the horizontal rate data.

7. The bird strike defense and precise bird control method based on aircraft flight position according to claim 1, characterized in that, The preset multiple protection thresholds, matched based on different flight phases, construct a dynamic safety channel centered on the aircraft, including: Multiple protection thresholds are preset, including the warning time window, horizontal half-axis, and vertical half-axis of the dynamic security channel; When the aircraft is in the ground taxiing phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety passage are 10s, 70m and 15m respectively. When the aircraft is in the takeoff phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety channel are 12s, 100m and 40m respectively. When the aircraft is in the climb phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety channel are 15s, 120m and 60m respectively. When the aircraft is in the cruise phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety channel are 15s, 150m and 50m respectively. When the aircraft is in the approach phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety channel are 8s, 110m and 50m respectively. When the aircraft is in the landing phase, the warning time window, the lateral half-axis and the vertical half-axis of the dynamic safety passage are 12s, 80m and 20m respectively.

8. The bird strike defense and precise bird control method based on aircraft flight position according to claim 1, characterized in that, The step of calculating the collision risk probability between the aircraft and the flock of birds based on the second position data and the centroid position includes: For each of the future moments, generate 10 dynamic secure channel masks; Determine whether the flock of birds is within a safe area based on the centroid location; Determine whether the position of the centroid satisfies the ellipsoidal inequality; The collision risk probability between the aircraft and the flock of birds is calculated based on the dynamic safety channel mask and the position probability cloud map.

9. The bird strike defense and precise bird control method based on aircraft flight position according to claim 1, characterized in that, The preset collision risk level threshold is used to compare the collision risk probability with the collision risk level threshold to obtain the risk level at the current moment, including: The preset collision risk level thresholds are 0.2 and 0.5; When the collision risk probability is less than 0.2, the risk level at the current moment is determined to be low risk. When 0.2 ≤ the collision risk probability < 0.5, the risk level at the current moment is determined to be medium risk; When the collision risk probability is ≥0.5, the risk level at the current moment is determined to be high risk.

10. A bird strike defense precision protection bird control system based on aircraft flight position, wherein the command system is applied to the command method according to any one of claims 1 to 9, and the command system comprises: A data acquisition and processing unit is used to acquire the flight data of the aircraft and the activity data of the bird flock in the target area in real time, perform data processing on the flight data and the activity data to obtain standard flight data and standard activity data, and extract the flight features and the bird flock features from the standard flight data and the standard activity data. A flock trajectory prediction unit is connected to the data acquisition and processing unit. The flock trajectory prediction unit is used to construct the flock trajectory prediction model, use the flock trajectory prediction model to predict the flock characteristics to obtain the centroid position of the flock at future times, and generate the position probability cloud map of the flock at future times. A protected area generation unit is connected to the bird flock trajectory prediction unit. The protected area generation unit is used to preset the flight conditions, match the flight conditions according to the standard flight data to obtain different flight stages of the aircraft, preset multiple protection thresholds, and match multiple protection thresholds based on different flight stages to construct a dynamic safety channel centered on the aircraft. A collision risk prediction unit is connected to the protection zone generation unit. The collision risk prediction unit is used to obtain the first position data of the aircraft at the current moment from the standard flight data, calculate the second position data of the aircraft at a future moment based on the first position data, and calculate the collision risk probability between the aircraft and the flock of birds based on the second position data and the center of mass position. A risk rating and graded response control unit is connected to the collision risk prediction unit. The risk rating and graded response control unit is used to preset the collision risk level threshold, compare the collision risk probability with the collision risk level threshold to obtain the risk level at the current moment, and take different bird deterrence measures in the dynamic safety passage according to different risk levels.

Images