Low-altitude aircraft air traffic management method and system

Abstract

The invention discloses a low-altitude aircraft air traffic management method and system, and relates to the technical field of aircraft air traffic management, and the method comprises the following steps: a traffic management system formulates the flight path, speed, course and formation of each unmanned plane according to a preset flight plan and performance demands before the flight performance begins; by monitoring the flight state of the unmanned aerial vehicle in real time and combining external disturbances such as wind speed and airflow, the path and speed of the unmanned aerial vehicle deviating from the trajectory can be dynamically adjusted, trajectory deviation caused by wind speed change is effectively prevented, the flight formation is kept stable, the relative distance is prevented from being too close or disordered, and the collision risk is reduced. Meanwhile, the deep learning model intelligently evaluates the flight consistency deviation, identifies the unmanned aerial vehicle deviating from the trajectory, and adjusts the path in time when the deviation exceeds a threshold value, thereby ensuring the flight safety, preventing the collision caused by the trajectory deviation, and guaranteeing the stability and safety of the group flight.

Description

Technical Field

[0001] The present invention relates to the technical field of aircraft air traffic management, and in particular to a method and system for low-altitude aircraft air traffic management. Background Art

[0002] Low-altitude aircraft air traffic management refers to the technology and management methods for coordinating, dispatching, monitoring and controlling various aircraft (such as drones, air taxis, aircraft, etc.) flying in low-altitude airspace (usually from the ground to an altitude of 1,000 meters). With the increasing use of low-altitude aircraft, especially in the fields of logistics, express delivery, agriculture, drone performances, etc., the number of low-altitude aircraft has increased sharply, and the volume and complexity of air traffic have also increased. Therefore, the safe management and efficient dispatch of low-altitude airspace are particularly important. The low-altitude aircraft air traffic management system usually involves real-time positioning of aircraft, flight path planning, collision prevention, airspace division, flight flow control, and communication and coordination with other aircraft and ground control centers. Its core goal is to ensure that aircraft can fly safely and orderly in low-altitude airspace, avoid collisions, interference or other safety hazards, and maximize the utilization efficiency of airspace, especially in scenes such as drone performances, where multiple aircraft need to be precisely coordinated to ensure the smooth and safe performance.

[0003] The prior art has the following deficiencies:

[0004] In the prior art, when drone groups perform flight performances, the air traffic management system usually predicts future wind speed changes based on wind speed data and sets a fixed flight speed accordingly. However, due to differences in wind speeds in different areas, wind speed changes may cause the flight trajectory of drones to deviate, especially when flying in groups, some drones may be affected by wind speeds to varying degrees. Fixed flight speeds cannot adapt to these changes, so failure to adjust speed and path in time will have serious consequences.

[0005] In open airspace, it is common to see differences in wind speed at different altitudes and locations. Sudden gusts or turbulent airflow may occur in some areas, causing individual drones to deviate from their flight paths. If the drones in a group cannot adjust their flight status in time according to the wind speed changes, the relative distance between the groups may be broken and the formation may be disrupted, thereby increasing the risk of collision. In particular, when the group flight speed fails to adapt to the local wind speed, the overall stability decreases and the aircraft cannot maintain a safe distance. If a drone deviates from its trajectory due to changes in wind speed, it may collide with a neighboring aircraft, causing damage or loss of control.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The purpose of the present invention is to provide a method and system for air traffic management of low-altitude aircraft, which can dynamically adjust the path and speed of drones that deviate from the trajectory by real-time monitoring of the flight status of drones and combining external disturbances such as wind speed and airflow, effectively preventing trajectory deviation caused by wind speed changes, thereby maintaining the stability of the flight formation, avoiding relative distances that are too close or disordered, and reducing the risk of collision. At the same time, the deep learning model intelligently evaluates the flight consistency deviation, identifies drones that deviate from the trajectory, and adjusts the path in time when the deviation exceeds the threshold, ensuring flight safety, preventing collisions caused by trajectory deviations, and ensuring the stability and safety of group flight, so as to solve the problems in the above-mentioned background technology.

[0008] In order to achieve the above object, the present invention provides the following technical solution: a method for air traffic management of low-altitude aircraft, comprising the following steps:

[0009] Before the flight show begins, the traffic management system determines the flight path, speed, heading and formation of each drone based on the preset flight plan and performance requirements;

[0010] During the flight show, the flight status parameters of each drone are monitored in real time to provide basic data for subsequent flight data analysis and adjustment;

[0011] Pre-process the flight data acquired in real time to ensure the accuracy and reliability of subsequent feature extraction and analysis;

[0012] Extract key features that reflect the current UAV’s deviation from the predetermined trajectory from the pre-processed flight data. Analyze the extracted key features under the detection window to quantify the flight consistency of the UAV group and evaluate whether the overall flight status maintains the predetermined formation and coordination.

[0013] The quantified flight consistency features are input into the pre-trained deep learning model to intelligently evaluate the flight status changes of each drone;

[0014] According to the evaluation results of the deep learning model, the current flight status of the UAV is divided into predetermined trajectory flight and deviated trajectory flight;

[0015] For drones that are in the predetermined trajectory flight state, continue to perform at the preset flight speed to ensure stable flight;

[0016] For drones that are assessed to have deviated from their flight trajectory, their flight speed is dynamically adjusted based on the prediction results of the deep learning model to bring them back to the predetermined trajectory and avoid collision accidents.

[0017] Preferably, key features reflecting the deviation of the current UAV from the predetermined trajectory are extracted from the preprocessed flight data, and the extracted features include the curvature of the UAV flight path and the dynamic balance state of the UAV during flight. Under the detection window, the extracted curvature of the UAV flight path and the dynamic balance state of the UAV during flight are analyzed to generate a track curvature change factor and a dynamic balance deviation factor, respectively. The flight consistency of the UAV group is quantified by the track curvature change factor and the dynamic balance deviation factor, and whether the overall flight state maintains the predetermined formation and coordination is evaluated.

[0018] Preferably, the specific steps of analyzing the curvature of the UAV flight path under the detection window to generate the track curvature change factor are as follows:

[0019] First, continuous path points are extracted from the flight trajectory of the UAV and the obtained path points are calibrated as p i , p i ={p1, p2, p3, ..., p n}, where p i is the i-th path point, n is the total number of path points, and each path point p i ={x i ,y i 、z i}, {x i ,y i 、z i} is the three-dimensional coordinate position of the i-th path point. The direction vector between any two adjacent path points is calculated. The local curvature is calculated through the direction vector. The calculation expression is as follows:

[0020] ,

[0021] In the formula, is the path point p i and path point p i+1 The direction vector between is the path point p i and path point p i-1 is the direction vector between them, |·| is the magnitude of the direction vector, κ i is the local curvature;

[0022] To consider the change in the time dimension, the local curvature κ is calculated by dividing the time interval of each pair of adjacent path points by i Normalized according to the time interval, the time-weighted curvature change is obtained, and the calculation expression is as follows:

[0023] ,

[0024] In the formula, κ i+1 is the local curvature of the next path point, Δti , i+1 is the time interval between adjacent path points, indicating that path point p i and path point p i+1 The time difference between i is the curvature change;

[0025] In order to further reflect the change of track curvature during the entire flight, the curvature change is weighted and summed to generate the track curvature change factor. The generation formula is as follows:

[0026] ,

[0027] Where Trajectory Curvature is the track curvature variation factor, w i is the path segment weight factor used to weight the contribution of each path segment to the total curvature variation factor.

[0028] Preferably, the specific steps of analyzing the dynamic balance state of the UAV during flight in the detection window to generate the dynamic balance deviation factor are as follows:

[0029] First, analyze the change of the UAV's flight attitude. By calculating the change amplitude of the attitude, the attitude deviation of the UAV during flight is obtained. The calculation expression is as follows:

[0030] Δθ j (t)=|θ j (t)-θ j ref (t)|

[0031] , where θ j (t) is the jth attitude angle of the UAV at time t, θ j ref (t) is the jth attitude angle of the UAV on the predetermined trajectory followed at time t, Δθ j (t) is the attitude deviation, which represents the deviation value of the j-th attitude angle of the UAV at time t, |θ j (t)-θ j ref (t)| is the jth attitude angle θ of the UAV at time t. j (t) and the jth attitude angle θ of the UAV on the predetermined trajectory followed at time t j ref The absolute deviation of (t);

[0032] After obtaining the attitude deviation of the drone, calculate the change trend of the deviation, that is, the attitude change rate. The calculation expression is as follows:

[0033] ,

[0034] In the formula, is the attitude deviation change rate, Δt is the time interval, θ j (t-Δt) is the posture deviation at the previous moment;

[0035] External disturbances will directly affect the flight stability of the drone. In order to accurately reflect the impact of external disturbances on dynamic balance, the attitude deviation is compensated through environmental data, and the true dynamic balance state after the disturbance is calculated. The calculation expression is as follows:

[0036] Δθ comp (t) = Δθ j (t)-f(W(t),A(t)),

[0037] Where f(W(t),A(t)) is the external disturbance compensation function, W(t) is the wind speed data, A(t) is the airflow data, Δθ comp (t) is the posture deviation after compensation;

[0038] Finally, the dynamic balance deviation factor is generated according to the compensated posture deviation. The generation formula is as follows:

[0039] ,

[0040] In the formula, Dynamic Balance is the dynamic balance deviation factor, T is the total time window duration, is the absolute value of the attitude deviation change rate.

[0041] Preferably, after analyzing the extracted features, the track curvature change factor and dynamic balance deviation factor generated after the analysis are input into a pre-trained deep learning model, and a flight consistency deviation index is generated by the deep learning model. The flight state changes of each UAV can be evaluated by the flight consistency deviation.

[0042] Preferably, the flight consistency deviation index generated by the pre-trained deep learning model when the flight state change of each drone is intelligently evaluated is compared with the pre-set reference threshold of the flight consistency deviation index to divide the current drone flight state. The division steps are as follows:

[0043] If the flight consistency deviation index is greater than a preset flight consistency deviation index reference threshold, the current UAV flight state is classified as off-track flight;

[0044] If the flight consistency deviation index is less than or equal to a preset flight consistency deviation index reference threshold, the current UAV flight state is classified as a predetermined trajectory flight.

[0045] Preferably, for a drone that is assessed to be off-trajectory, the specific steps of dynamically adjusting its flight speed based on the prediction results of the deep learning model to return it to the predetermined trajectory and avoid collision accidents are as follows:

[0046] After confirming that the drone has deviated from the trajectory, the deep learning model is used to dynamically predict and adjust the flight speed. The flight speed is predicted based on historical flight data, current flight status, and external environment to ensure that the drone can return to the predetermined trajectory. The expression is as follows:

[0047] ,

[0048] Where V preset is the preset speed of the drone, Flight Consistency is the flight consistency deviation index, Flight ref is the reference threshold of the flight consistency deviation index, Flight max is the maximum value of the flight consistency deviation index, ΔV max is the maximum flight speed adjustment range, V adjusted is the adjusted flight speed;

[0049] In addition to adjusting the flight speed, correcting the flight path is also a key step in restoring the intended trajectory. According to the prediction results of the deep learning model, the heading and flight path of the drone are adjusted to ensure that it flies in the intended direction. The calculation expression is as follows:

[0050] ,

[0051] In the formula, θ preset is the preset heading angle, λ is the heading adjustment weight factor, which is used to control the sensitivity of heading adjustment, Δθ max is the maximum heading adjustment angle, θ adjusted is the adjusted heading angle;

[0052] By acquiring the UAV's flight data in real time and comparing it with the preset trajectory, the flight consistency deviation index Flight is continuously updated. Consistency , and correct the state of the aircraft based on the feedback results to avoid over-adjustment or under-adjustment due to system lag. The feedback control strategy is adopted to ensure a smooth transition during the adjustment process. The formula is as follows:

[0053] F feedback =α adjust ·V adjusted +β adjust ·θ adjusted

[0054] , where α adjust is the flight speed adjustment weight factor, βadjust is the heading angle adjustment weight factor, F feedback is the feedback correction value.

[0055] A low-altitude aircraft air traffic management system, comprising a flight plan and path setting module, a real-time flight status monitoring module, a flight data preprocessing module, a feature extraction and consistency analysis module, an intelligent evaluation module, a flight status classification module, a stable flight maintenance module, and a dynamic adjustment and trajectory regression module;

[0056] Flight planning and path setting module: before the flight show begins, the traffic management system formulates the flight path, speed, heading and formation of each drone according to the preset flight plan and performance requirements;

[0057] Real-time flight status monitoring module: During the flight performance, the flight status parameters of each drone are monitored in real time to provide basic data for subsequent flight data analysis and adjustment;

[0058] The flight data preprocessing module preprocesses the flight data acquired in real time to ensure the accuracy and reliability of subsequent feature extraction and analysis;

[0059] The feature extraction and consistency analysis module extracts key features that reflect the current UAV’s deviation from the predetermined trajectory from the pre-processed flight data. Under the detection window, the extracted key features are analyzed to quantify the flight consistency of the UAV group and evaluate whether the overall flight status maintains the predetermined formation and coordination.

[0060] The intelligent evaluation module inputs the quantified flight consistency features into the pre-trained deep learning model to intelligently evaluate the flight status changes of each drone;

[0061] The flight status classification module divides the current UAV flight status into scheduled trajectory flight and off-trajectory flight according to the evaluation results of the deep learning model;

[0062] The stable flight maintenance module continues to perform at the preset flight speed for the drone in the predetermined trajectory flight state to ensure its stable flight;

[0063] The dynamic adjustment and trajectory regression module dynamically adjusts the flight speed of drones that are assessed to have deviated from their flight trajectory based on the prediction results of the deep learning model, so that they return to the predetermined trajectory and avoid collision accidents.

[0064] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0065] The present invention can dynamically adjust the flight path and speed of a drone that has deviated from its trajectory by monitoring the drone's flight status in real time and combining it with external disturbances such as wind speed data and air turbulence. This adjustment can effectively prevent the drone from deviating from its trajectory due to its inability to adapt to wind speed changes, thereby avoiding the phenomenon of too close relative distance or disordered formation during group flight. By accurately controlling the drone's flight status, the stability of the group can be maintained, ensuring that each drone maintains a safe distance, and significantly reducing the risk of collision caused by flight trajectory deviation.

[0066] The present invention uses a deep learning model to intelligently evaluate flight consistency deviations, quickly identify drones that deviate from the predetermined trajectory, and adjust flight parameters based on the deviations. When the flight consistency deviation index exceeds the set threshold, potential collision risks are automatically identified, and accurate path adjustments are made to drones that deviate from their trajectory. This mechanism ensures timely detection of abnormal behaviors during flight, prevents collision risks between drones due to deviations from their trajectories, and thus ensures the safety of group flight. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0068] Figure 1 The present invention is a method flow chart of a low-altitude aircraft air traffic management method.

[0069] Figure 2 The present invention is a module schematic diagram of a low-altitude aircraft air traffic management system. DETAILED DESCRIPTION

[0070] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0071] The present invention provides Figure 1 A low-altitude aircraft air traffic management method shown includes the following steps:

[0072] Before the flight show begins, the traffic management system determines the flight path, speed, heading and formation of each drone based on the preset flight plan and performance requirements;

[0073] This step ensures that all drones have clear flight instructions and a coordinated action plan when the show begins.

[0074] Design detailed flight paths and formations based on the theme of the performance, the venue environment, and the needs of the audience. For example, design drones to form specific patterns (such as letters, graphics) or perform specific actions (such as rotation, up and down fluctuations) in the air. Set the initial flight speed and heading for each drone to ensure that the entire group can perform the predetermined flight actions in a coordinated and consistent manner. Through the traffic management system, ensure that each drone is in the correct position at the beginning of the performance to avoid the risk of collision in the initial stage.

[0075] During the flight show, the flight status parameters of each drone are monitored in real time to provide basic data for subsequent flight data analysis and adjustment;

[0076] Each drone is equipped with high-precision sensors (such as GPS, inertial measurement unit IMU, barometer, etc.) to collect flight data in real time. The collected data is transmitted to the air traffic management system in real time through wireless communication technology (such as Wi-Fi, 4G / 5G, dedicated communication frequency bands). Ensure that the data of all drones are synchronized in time to facilitate subsequent overall analysis and coordination adjustments.

[0077] Pre-process the flight data acquired in real time to ensure the accuracy and reliability of subsequent feature extraction and analysis;

[0078] The purpose of cleaning and formatting the flight data acquired in real time is to ensure the accuracy and reliability of subsequent feature extraction and analysis. During the data cleaning process, it is necessary to remove or correct abnormal data points caused by sensor failure, communication interference and other factors to avoid these abnormal values ​​affecting the analysis results. Data formatting is to unify the flight data from different sensors and devices into a consistent format for subsequent processing and analysis. Time synchronization ensures that all data are arranged in chronological order and maintains the temporal consistency of the data, thereby avoiding analysis errors caused by time deviation and ensuring the accuracy of flight data in the spatial and temporal dimensions.

[0079] Extract key features that reflect the current UAV’s deviation from the predetermined trajectory from the pre-processed flight data. Analyze the extracted key features under the detection window to quantify the flight consistency of the UAV group and evaluate whether the overall flight status maintains the predetermined formation and coordination.

[0080] The key features reflecting the deviation of the current UAV from the predetermined trajectory are extracted from the preprocessed flight data. The extracted features include the curvature of the UAV flight path and the dynamic balance state of the UAV during flight. Under the detection window, the extracted curvature of the UAV flight path and the dynamic balance state of the UAV during flight are analyzed to generate the track curvature change factor and the dynamic balance deviation factor respectively. The flight consistency of the UAV group is quantified by the track curvature change factor and the dynamic balance deviation factor, and whether the overall flight state maintains the predetermined formation and coordination is evaluated.

[0081] The greater the curvature of a drone's flight path, the more it usually indicates that the drone has deviated from its intended trajectory. During flight, the curvature of the flight path usually reflects the degree of change in the aircraft's heading. When external factors such as wind speed or airflow changes occur, especially when the wind speed difference is large, the aircraft may be forced to change its heading to cope with these disturbances. If a drone does not adjust its flight attitude and speed in time when encountering sudden gusts or air turbulence, its flight trajectory will deviate irregularly, resulting in an increase in the curvature of the path. Especially in group flight, excessive bending of a drone may break the formation of the group, causing the distance between aircraft to become too close or chaotic, which increases the risk of collision. Wind speed changes or unstable airflow usually affect the stability of the aircraft, especially during flight, when the drone needs to constantly adjust its heading to maintain stable flight. If these adjustments cannot be completed quickly or accurately, the curvature of the flight path will increase dramatically, reflecting that the aircraft has failed to fly according to the intended route.

[0082] The specific steps for analyzing the curvature of the UAV flight path under the detection window to generate the track curvature change factor are as follows:

[0083] First, continuous path points are extracted from the flight trajectory of the UAV and the obtained path points are calibrated as p i , p i ={p1, p2, p3, ..., p n}, where p i is the i-th path point, n is the total number of path points, and each path point p i ={x i ,y i 、z i}, {x i ,y i 、z i} is the three-dimensional coordinate position of the i-th path point. The direction vector between any two adjacent path points is calculated. The local curvature is calculated through the direction vector. The calculation expression is as follows:

[0084] ,

[0085] In the formula, is the path point p i and path point p i+1 The direction vector between represents the direction from path point p i To the path point p i+1 Direction, is the path point p i and path point p i-1 The direction vector between represents the direction from path point p i To the path point p i-1 is the direction of the vector, |·| is the magnitude of the direction vector, κ i is the local curvature, indicating that the path point p i The local curvature of the neighborhood measures how curved the path is at that point;

[0086] This step calculates the curvature of the path, reflecting the local change of the path at a certain point. When the UAV's heading changes drastically, the local curvature will increase, indicating that the aircraft deviates from the planned path.

[0087] The modulus of a direction vector, |·|, is also called the magnitude or length of a vector. The modulus of a vector reflects the magnitude or strength of the vector. In the calculation of track curvature, the role of the modulus is to measure the geometric relationship between the two direction vectors to ensure that the calculation of the curvature reflects the actual curvature of the path. Specifically, when calculating the curvature of the path, the modulus of the vector is used to normalize the magnitude of the direction vector to remove the influence of the path length, thereby ensuring that the curvature is only related to the shape and curvature of the path, and is not affected by the length or size of the path itself. By standardizing the vector modulus, it can be ensured that the curvature calculation only reflects the curvature changes of the path, and is not dependent on the scale of the flight path, which is particularly important for evaluating the degree of trajectory deviation of the drone.

[0088] To consider the change in the time dimension, the local curvature κ is calculated by dividing the time interval of each pair of adjacent path points by i Normalized according to the time interval, the time-weighted curvature change is obtained, and the calculation expression is as follows:

[0089] ,

[0090] In the formula, κ i+1 is the local curvature of the next path point, Δt i , i+1 is the time interval between adjacent path points, indicating that path point p i and path point p i+1 The time difference between i is the curvature change, indicating that the path point p i and path point p i+1 The curvature change between is the change in the curvature of the path between two path points;

[0091] This step reflects how quickly the vehicle's path changes. If the change in curvature is very drastic and occurs over a short period of time, then Δκ i The value of will be large, indicating that the aircraft may deviate rapidly from the planned track due to external factors (such as wind speed changes). By introducing time differences, the dynamic changes of the UAV in a short period of time can be accurately evaluated, thereby improving the ability to respond to emergencies.

[0092] In order to further reflect the change of track curvature during the entire flight, the curvature change is weighted and summed to generate the track curvature change factor. The generation formula is as follows:

[0093] ,

[0094] Where Trajectory Curvature is the track curvature variation factor, w i is the path segment weight factor used to weight the contribution of each path segment to the total curvature variation factor.

[0095] By weighted summation, a comprehensive measure of the curvature change during the entire flight process is obtained. Curvature A large value indicates that the drone has experienced a large path deviation during the flight, further verifying the trajectory deviation of the drone. This helps to conduct a comprehensive analysis of the deviation trend during the entire flight process.

[0096] The larger the performance value of the track curvature change factor generated after analyzing the curvature of the UAV's flight path under the detection window, the higher the performance value usually indicates that the current UAV has deviated from the predetermined trajectory, and vice versa, it means that the UAV is on the predetermined trajectory. Track curvature is a key indicator to measure the curvature of the flight path. When the flight path of the UAV turns or bends sharply, the curvature value will increase. External factors such as wind speed changes and air turbulence may cause the aircraft to deviate from the original route, resulting in a significant change in the curvature of the track. At this time, the performance value of the track curvature change factor will increase significantly, reflecting that the aircraft is experiencing an unstable flight state and deviates from the set flight trajectory. When the value of the track curvature change factor is small or remains stable, it means that the path of the aircraft changes relatively smoothly, and it can usually maintain the predetermined trajectory well, and the degree of deviation is low.

[0097] A dramatic fluctuation in the balance state of a drone during flight usually indicates that the aircraft has deviated from the intended trajectory. The balance state of an aircraft mainly relies on its attitude control system to maintain flight stability, including angles such as pitch, roll, and yaw. When external factors, such as wind speed changes or air turbulence, affect the drone, it may cause the drone's flight attitude to fluctuate dramatically, manifested as rapid changes in the aircraft's pitch, roll, or yaw angles. These changes indicate that the drone is adjusting its flight attitude to cope with environmental disturbances, which may cause the aircraft to deviate from the intended path. Especially in the case of strong gusts or unstable airflow, the aircraft's balance control system must respond quickly to avoid excessive deviation of the aircraft. However, if the external disturbance is too strong or the control system does not respond enough, the aircraft's attitude may be difficult to return to a stable state, resulting in a continuous deviation from the track. This dramatic fluctuation in attitude not only affects flight stability, but may also lead to collisions with other drones or obstacles. Therefore, a dramatic fluctuation in the balance state is an important signal that the aircraft has deviated from the intended trajectory, especially in an environment where the external airflow changes dramatically.

[0098] The specific steps for analyzing the dynamic balance state of the drone during flight in the detection window to generate the dynamic balance deviation factor are as follows:

[0099] First, analyze the changes in the flight attitude of the drone (such as pitch angle, roll angle, and yaw angle). By calculating the change amplitude of the attitude, the attitude deviation of the drone during flight is obtained. The calculation expression is as follows:

[0100] Δθ j (t)=|θ j (t)-θ j ref (t)|

[0101] , where θ j (t) is the jth attitude angle of the UAV at time t, θ j ref (t) is the jth attitude angle of the UAV on the predetermined trajectory followed at time t, Δθ j (t) is the attitude deviation, which represents the deviation value of the j-th attitude angle of the UAV at time t, |θ j (t)-θ j ref (t)| is the jth attitude angle θ of the UAV at time t. j (t) and the jth attitude angle θ of the UAV on the predetermined trajectory followed at time t j ref The absolute deviation of (t);

[0102] This step quantifies the dynamic equilibrium state of the aircraft at each moment by calculating the instantaneous deviation of the attitude. If the deviation value is large, it means that the drone has failed to maintain a stable attitude during this period, indicating that the drone may be deviating from the predetermined trajectory.

[0103] After obtaining the attitude deviation of the drone, calculate the change trend of the deviation, that is, the attitude change rate. By analyzing the time change of the attitude deviation, it is possible to identify whether the aircraft continues to remain unstable or whether it can quickly restore stability. The calculation expression is as follows:

[0104] ,

[0105] In the formula, is the attitude deviation change rate, which indicates the instantaneous change rate of the attitude deviation of the drone at time t. It is usually used to measure the speed of deviation change. Δt is the time interval, θ j (t-Δt) is the attitude deviation at the previous moment, which indicates the deviation of the j-th attitude angle of the UAV at the previous moment (t-Δt);

[0106] This step helps analyze whether posture deviation is increasing. The value increases significantly, indicating that the attitude deviation of the UAV has an accelerating trend. The UAV may continue to deviate from the predetermined trajectory, indicating that dynamic adjustment is needed.

[0107] External disturbances, such as wind speed changes and airflow turbulence, will directly affect the flight stability of the drone. In order to accurately reflect the impact of external disturbances on dynamic balance, the attitude deviation is compensated through environmental data (such as real-time wind speed and airflow information) to calculate the true dynamic balance state after the disturbance. The calculation expression is as follows:

[0108] Δθ comp (t) = Δθ j (t)-f(W(t),A(t))

[0109] , where f(W(t),A(t)) is the external disturbance compensation function, which is used to adjust the attitude deviation according to real-time environmental data (such as wind speed and airflow). W(t) is the wind speed data, which represents the wind speed data at the location of the drone at time t. Wind speed is a vector that includes the size and direction of the wind. A(t) is the airflow data, which represents the airflow data at the location of the drone at time t. Airflow is the state of air flow, usually including turbulence, updraft, downdraft, etc. Δθ comp (t) is the compensated attitude deviation, which is the attitude deviation after external disturbance compensation, indicating the deviation of the UAV from the reference trajectory at time t, after deducting the influence of external disturbances such as wind speed and airflow;

[0110] Finally, the dynamic balance deviation factor is generated according to the compensated attitude deviation. This factor can comprehensively consider the influence of the aircraft's attitude change, change rate, and external disturbance on flight stability. The generation formula is as follows:

[0111] ,

[0112] In the formula, Dynamic Balance is the dynamic balance deviation factor, T is the total time window duration, is the absolute value of the attitude deviation change rate.

[0113] The larger the performance value of the dynamic balance deviation factor generated by analyzing the dynamic balance state of the drone during flight under the detection window, the more it indicates that the current drone deviates from the predetermined trajectory. The dynamic balance deviation factor mainly measures the stability of the drone's attitude during flight, especially the changes in pitch, roll and yaw angles. When the flight attitude of the drone fluctuates greatly or changes dramatically, it means that it is affected by external disturbances (such as wind speed changes or air turbulence), and the flight control system fails to effectively maintain flight stability, causing the aircraft to deviate from the predetermined trajectory. This deviation usually causes the heading and path of the aircraft to no longer be consistent with the original trajectory. Therefore, the value of the dynamic balance deviation factor increases, reflecting that the attitude control of the aircraft is abnormal and deviates from the expected route. On the contrary, if the performance value of the dynamic balance deviation factor is small, it means that the attitude control of the aircraft is stable, it is not affected by external disturbances or these disturbances have been successfully compensated, so the aircraft maintains a flight state consistent with the predetermined trajectory.

[0114] The quantified flight consistency features are input into the pre-trained deep learning model to intelligently evaluate the flight status changes of each drone;

[0115] After analyzing the extracted features, the track curvature change factor and dynamic balance deviation factor generated after analysis are input into the pre-trained deep learning model. The flight consistency deviation index is generated by the deep learning model, and the flight state changes of each UAV can be evaluated through the flight consistency deviation.

[0116] A pre-trained deep learning model refers to a deep learning model that is trained using a large amount of flight data and related flight state features (such as track curvature change factors and dynamic balance deviation factors) in a drone group flight performance system to ensure that the model can perform real-time prediction and evaluation during actual flight. The training process usually involves using historical flight data, including various flight states (such as flight consistency, flight trajectory, and aircraft attitude, etc.) and environmental data (such as wind speed, airflow changes, etc.) to teach the deep learning model to recognize and understand patterns and features that may cause deviations from the trajectory during flight. During training, the model will learn how to map the input flight data features to the corresponding flight states (such as whether it deviates from the trajectory, flight stability, etc.) through the method of supervised learning, and continuously adjust its parameters through the back-propagation algorithm to minimize the error between the predicted value and the actual flight result. This method enables the deep learning model to acquire the ability to handle complex nonlinear relationships and to cope with the ever-changing environmental factors during flight.

[0117] The trained deep learning model can evaluate new flight data in real time. Especially in actual flight performances, the model can intelligently evaluate whether the flight status of the drone meets the predetermined trajectory and flight consistency requirements based on the real-time input features such as the track curvature change factor and the dynamic balance deviation factor. If the model evaluates that the flight status has deviated, it will output a flight consistency deviation index, which is used to quantify the flight consistency of the current aircraft. The higher the flight consistency deviation index, the greater the deviation of the aircraft's flight status from the predetermined trajectory. Conversely, it means that the aircraft's status is stable and remains within the predetermined trajectory. This intelligent evaluation system based on the deep learning model can adjust the flight strategy in real time and dynamically to ensure the stability and safety of group flight. Through the self-adjustment and optimization of deep learning, the system can adapt to different flight environments and changing factors, so as to better cope with complex flight performance tasks and avoid collisions or other flight accidents of drones.

[0118] The deep learning model is not limited here, and can realize the trajectory curvature change factor Trajectory Curvature and Dynamic Balance Deviation Factor Dynamic Balance Perform comprehensive analysis to generate the Flight Consistency Deviation Index Consistency In order to realize the technical solution of the present invention, the present invention provides a specific generation method;

[0119] Flight consistency deviation index Consistency The generation formula is as follows:

[0120] ,

[0121] In the formula, f a 、f b Trajectory is the trajectory curvature change factor Curvature and Dynamic Balance Deviation Factor Dynamic Balance Preset proportionality factor, and f a 、f b Both are greater than 0.

[0122] It can be seen from the flight consistency deviation index that the larger the performance value of the track curvature change factor generated after analyzing the curvature of the UAV's flight path under the detection window, the larger the performance value of the dynamic balance deviation factor generated after analyzing the dynamic balance state of the UAV during flight under the detection window, and the larger the performance value of the flight consistency deviation index generated when the flight state change of each UAV is intelligently evaluated by the pre-trained deep learning model, the larger the performance value, which indicates that the current UAV is deviating from the predetermined trajectory, otherwise, it indicates that the current UAV is flying on the predetermined trajectory.

[0123] According to the evaluation results of the deep learning model, the current flight status of the UAV is divided into predetermined trajectory flight and deviated trajectory flight;

[0124] The flight consistency deviation index generated by the pre-trained deep learning model when intelligently evaluating the flight state changes of each drone is compared with the pre-set reference threshold of the flight consistency deviation index to divide the current drone flight state. The division steps are as follows:

[0125] If the flight consistency deviation index is greater than a preset flight consistency deviation index reference threshold, the current UAV flight state is classified as off-track flight;

[0126] If the flight consistency deviation index is less than or equal to a preset flight consistency deviation index reference threshold, the current UAV flight state is classified as a predetermined trajectory flight.

[0127] Predetermined track flight means that the drone strictly follows the flight path and heading set by the system during flight, and maintains a stable flight state according to the predetermined flight plan. Off-track flight means that during the flight of the drone, due to external environmental factors (such as wind speed changes, air turbulence), it is unable to accurately fly along the predetermined route, and there are obvious yaw, stall or path deviation.

[0128] For drones that are in the predetermined trajectory flight state, continue to perform at the preset flight speed to ensure stable flight;

[0129] For drones that are in a predetermined trajectory flight state, the purpose of continuing to perform at a preset flight speed is to ensure that the aircraft can stably maintain the predetermined flight path and avoid deviation or instability of the trajectory due to speed adjustment or unnecessary intervention. The preset flight speed is usually accurately calculated based on the performance of the drone, weather conditions and mission requirements, and can ensure the smooth operation of the aircraft in different environments. By maintaining a fixed speed, the drone can maintain a consistent flight state with other drones in group flight, ensure the stability of the formation, reduce the risk of collision with other drones, and improve the coordination and accuracy of the performance. In addition, maintaining a consistent flight speed can also help the system provide reference data when evaluating and adjusting the status of other aircraft, so that the air traffic management system can make real-time adjustments and optimizations more effectively.

[0130] For drones that are assessed to be off-track, their flight speed is dynamically adjusted based on the prediction results of the deep learning model to return them to the predetermined trajectory and avoid collision accidents.

[0131] For drones that are assessed to be off-track, the specific steps to dynamically adjust their flight speed based on the prediction results of the deep learning model to return them to the predetermined trajectory and avoid collision accidents are as follows:

[0132] After confirming that the drone has deviated from its trajectory, the deep learning model is used to dynamically predict and adjust the flight speed. The flight speed is predicted based on historical flight data, current flight status, and external environment (such as wind speed and airflow disturbance) to ensure that the drone can return to the predetermined trajectory. The expression is as follows:

[0133] ,

[0134] Where V preset is the preset speed of the drone, Flight Consistency is the flight consistency deviation index, Flight ref is the reference threshold of the flight consistency deviation index, Flight max is the maximum value of the flight consistency deviation index, ΔV max is the maximum flight speed adjustment range, which is predicted by the deep learning model based on historical flight data and external environment changes. adjusted is the adjusted flight speed;

[0135] Through this formula, according to the flight consistency deviation index Flight Consistency The flight speed is dynamically adjusted based on the changes in the flight path to ensure that the drone gradually returns to the planned route.

[0136] In addition to adjusting the flight speed, correcting the flight path is also a key step in restoring the intended trajectory. According to the prediction results of the deep learning model, the heading and flight path of the drone are adjusted to ensure that it flies in the intended direction. The adjustment range is based on the flight consistency deviation index Flight Consistency Flight consistency deviation index reference threshold ref The difference is dynamically adjusted to ensure that excessive correction does not cause drastic fluctuations in the aircraft attitude. The calculation expression is as follows:

[0137] ,

[0138] In the formula, θ preset is the preset heading angle, λ is the heading adjustment weight factor, which is used to control the sensitivity of heading adjustment, Δθ max is the maximum heading adjustment angle, θ adjusted is the adjusted heading angle;

[0139] Through this adjustment, the heading of the aircraft can be more accurately pointed to the predetermined trajectory and matched with the adjusted flight speed to ensure the consistency of group flight.

[0140] By acquiring the UAV's flight data in real time and comparing it with the preset trajectory, the flight consistency deviation index Flight is continuously updated. Consistency , and correct the state of the aircraft based on the feedback results to avoid over-adjustment or under-adjustment due to system lag. The feedback control strategy is adopted to ensure a smooth transition during the adjustment process. The formula is as follows:

[0141] F feedback =α adjust ·V adjusted +β adjust ·θ adjusted

[0142] , where α adjust is the flight speed adjustment weight factor, β adjust is the heading angle adjustment weight factor, F feedback It is the feedback correction value, which is used to dynamically evaluate the correction effect of the flight status.

[0143] This feedback mechanism ensures the accuracy and stability of the drone when it returns to the predetermined trajectory, avoids flight instability caused by excessive adjustment, ensures that the drone gradually returns and stabilizes on the predetermined trajectory, and reduces the risk of collision.

[0144] Flight speed adjustment weight factor α adjust It is a factor that determines the flight speed adjustment to the overall feedback correction value (F feedback) contribution. It indicates the relative importance of adjusting the flight speed in correcting the flight trajectory in adjusting the flight speed and heading angle. adjust When the value of is high, the system will rely more on adjusting the flight speed to achieve trajectory correction, especially when the wind speed changes greatly or the flight speed of the drone fluctuates significantly, so that the flight speed can be quickly changed to quickly restore to the predetermined trajectory. Conversely, a lower α adjust This means that the adjustment of flight speed has less impact on feedback correction and relies more on the correction of heading angle.

[0145] Heading angle adjustment weight factor β adjust Determines the overall feedback correction value F of the heading angle adjustment feedback ) is the contribution of the flight direction to the flight speed in the trajectory correction process. adjust When β is larger, the system tends to return to the predetermined trajectory by adjusting the heading angle. Especially when the flight path is greatly curved, the system will give priority to changing the heading angle to restore the flight direction. adjust This means that the aircraft relies more on adjusting the flight speed to correct its trajectory, and the adjustment of the heading angle is relatively weak.

[0146] For drones that are assessed to be off-track, the flight speed is dynamically adjusted based on the prediction results of the deep learning model. The main function is to return the drone to the predetermined trajectory through precise speed control, thereby avoiding collision accidents caused by deviation from the flight trajectory. In a complex flight environment, drones may be affected by factors such as wind speed changes, airflow disturbances, and heading adjustments, causing the flight path to deviate. When the Flight Consistency Deviation Index Flight Consistency When the predetermined reference threshold is exceeded, the flight status of the drone is identified as abnormal and immediate measures need to be taken to correct it. At this time, based on the prediction results of the deep learning model, the best solution for adjusting the flight speed can be intelligently predicted based on the current status information of the aircraft (such as heading angle, speed, path curvature, etc.) and external environmental factors (such as wind speed, airflow, etc.).

[0147] Through learning and training from historical data, the deep learning model can identify complex flight status patterns and make dynamic adjustments based on real-time flight data. By adjusting the flight speed, the flight trajectory of the drone can be precisely controlled to avoid stability problems caused by excessive deviation from the path or too fast or too slow flight speed. In addition, dynamic adjustment of the flight speed can also ensure that the relative distance of other drones in the group flight is maintained, avoiding the overall flight formation disorder and collision risk caused by the deviation of a single drone. The core role of this step is to ensure that the drone group can complete the flight performance in a coordinated manner through intelligent evaluation and real-time response to the flight status of the drone, ensuring safety and avoiding possible flight accidents.

[0148] The present invention can dynamically adjust the flight path and speed of a drone that has deviated from its trajectory by monitoring the drone's flight status in real time and combining it with external disturbances such as wind speed data and air turbulence. This adjustment can effectively prevent the drone from deviating from its trajectory due to its inability to adapt to wind speed changes, thereby avoiding the phenomenon of too close relative distance or disordered formation during group flight. By accurately controlling the drone's flight status, the stability of the group can be maintained, ensuring that each drone maintains a safe distance, and significantly reducing the risk of collision caused by flight trajectory deviation.

[0149] The present invention uses a deep learning model to intelligently evaluate flight consistency deviations, quickly identify drones that deviate from the predetermined trajectory, and adjust flight parameters based on the deviations. When the flight consistency deviation index exceeds the set threshold, potential collision risks are automatically identified, and accurate path adjustments are made to drones that deviate from their trajectory. This mechanism ensures timely detection of abnormal behaviors during flight, prevents collision risks between drones due to deviations from their trajectories, and thus ensures the safety of group flight.

[0150] The present invention provides Figure 2 A low-altitude aircraft air traffic management system shown includes a flight plan and path setting module, a real-time flight status monitoring module, a flight data preprocessing module, a feature extraction and consistency analysis module, an intelligent evaluation module, a flight status classification module, a stable flight maintenance module, and a dynamic adjustment and trajectory regression module;

[0151] Flight planning and path setting module: before the flight show begins, the traffic management system formulates the flight path, speed, heading and formation of each drone according to the preset flight plan and performance requirements;

[0152] Real-time flight status monitoring module: During the flight performance, the flight status parameters of each drone are monitored in real time to provide basic data for subsequent flight data analysis and adjustment;

[0153] The flight data preprocessing module preprocesses the flight data acquired in real time to ensure the accuracy and reliability of subsequent feature extraction and analysis;

[0154] The feature extraction and consistency analysis module extracts key features that reflect the current UAV’s deviation from the predetermined trajectory from the pre-processed flight data. Under the detection window, the extracted key features are analyzed to quantify the flight consistency of the UAV group and evaluate whether the overall flight status maintains the predetermined formation and coordination.

[0155] The intelligent evaluation module inputs the quantified flight consistency features into the pre-trained deep learning model to intelligently evaluate the flight status changes of each drone;

[0156] The flight status classification module divides the current UAV flight status into scheduled trajectory flight and off-trajectory flight according to the evaluation results of the deep learning model;

[0157] The stable flight maintenance module continues to perform at the preset flight speed for the drone in the predetermined trajectory flight state to ensure its stable flight;

[0158] The dynamic adjustment and trajectory regression module dynamically adjusts the flight speed of drones that are assessed to have deviated from their flight trajectory based on the prediction results of the deep learning model, so that they return to the predetermined trajectory and avoid collision accidents.

[0159] A low-altitude aircraft air traffic management method provided by an embodiment of the present invention is implemented through the above-mentioned low-altitude aircraft air traffic management system. The specific methods and processes of a low-altitude aircraft air traffic management system are detailed in the embodiment of the above-mentioned low-altitude aircraft air traffic management method, which will not be repeated here.

[0160] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0161] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0162] It should be noted that, in this article, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0163] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0164] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0165] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0166] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0168] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0169] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for air traffic management of low-altitude aircraft, characterized in that: The following steps are involved: Before the flight show begins, the traffic management system determines the flight path, speed, heading and formation of each drone based on the preset flight plan and performance requirements; During the flight show, the flight status parameters of each drone are monitored in real time to provide basic data for subsequent flight data analysis and adjustment; Pre-process the flight data acquired in real time to ensure the accuracy and reliability of subsequent feature extraction and analysis; Extract key features that reflect the current UAV’s deviation from the predetermined trajectory from the pre-processed flight data. Analyze the extracted key features under the detection window to quantify the flight consistency of the UAV group and evaluate whether the overall flight status maintains the predetermined formation and coordination. The quantified flight consistency features are input into the pre-trained deep learning model to intelligently evaluate the flight status changes of each drone; According to the evaluation results of the deep learning model, the current flight status of the UAV is divided into predetermined trajectory flight and off-trajectory flight; For drones that are in the predetermined trajectory flight state, continue to perform at the preset flight speed to ensure stable flight; For drones that are assessed to be flying off-trail, their flight speed is dynamically adjusted based on the prediction results of the deep learning model to bring them back to the predetermined trajectory and avoid collision accidents.

2. A low-altitude aircraft air traffic management method according to claim 1, characterized in that: The key features reflecting the deviation of the current UAV from the predetermined trajectory are extracted from the preprocessed flight data. The extracted features include the curvature of the UAV flight path and the dynamic balance state of the UAV during flight. Under the detection window, the extracted curvature of the UAV flight path and the dynamic balance state of the UAV during flight are analyzed to generate the track curvature change factor and the dynamic balance deviation factor respectively. The flight consistency of the UAV group is quantified by the track curvature change factor and the dynamic balance deviation factor, and whether the overall flight state maintains the predetermined formation and coordination is evaluated.

3. A low altitude aircraft air traffic management method according to claim 2, characterized in that: The specific steps for analyzing the curvature of the UAV flight path under the detection window to generate the track curvature change factor are as follows: First, continuous path points are extracted from the flight trajectory of the UAV and the obtained path points are calibrated as p i , p i ={p1, p2, p3, ..., p n }, where p i is the i-th path point, n is the total number of path points, and each path point p i ={x i ,y i 、z i }, {x i ,y i 、z i } is the three-dimensional coordinate position of the i-th path point. The direction vector between any two adjacent path points is calculated. The local curvature is calculated through the direction vector. The calculation expression is as follows: ， In the formula, is the path point p i and path point p i+1 The direction vector between is the path point p i and path point p i-1 is the direction vector between them, |·| is the magnitude of the direction vector, κ i is the local curvature; To consider the change in the time dimension, the local curvature κ is calculated by dividing the time interval of each pair of adjacent path points by i Normalized according to the time interval, the time-weighted curvature change is obtained, and the calculation expression is as follows: ， In the formula, κ i+1 is the local curvature of the next path point, Δt i , i+1 is the time interval between adjacent path points, indicating that path point p i and path point p i+1 The time difference between i is the curvature change; In order to further reflect the change of track curvature during the entire flight, the curvature change is weighted and summed to generate the track curvature change factor. The generation formula is as follows: ， Where Trajectory Curvature is the track curvature variation factor, w i is the path segment weight factor used to weight the contribution of each path segment to the total curvature variation factor.

4. A low-altitude aircraft air traffic management method according to claim 2, characterized in that: The specific steps for analyzing the dynamic balance state of the drone during flight in the detection window to generate the dynamic balance deviation factor are as follows: First, analyze the change of the UAV's flight attitude. By calculating the change amplitude of the attitude, the attitude deviation of the UAV during flight is obtained. The calculation expression is as follows: ， In the formula, θ j (t) is the jth attitude angle of the UAV at time t, is the jth attitude angle of the UAV on the predetermined trajectory at time t, Δθ j (t) is the attitude deviation, which represents the deviation value of the j-th attitude angle of the UAV at time t. is the jth attitude angle θ of the drone at time t j (t) and the jth attitude angle of the UAV on the predetermined trajectory followed at time t The absolute deviation of After obtaining the attitude deviation of the drone, calculate the change trend of the deviation, that is, the attitude change rate. The calculation expression is as follows: ， In the formula, is the attitude deviation change rate, Δt is the time interval, θ j (t-Δt) is the posture deviation at the previous moment; External disturbances will directly affect the flight stability of the drone. In order to accurately reflect the impact of external disturbances on dynamic balance, the attitude deviation is compensated through environmental data, and the true dynamic balance state after the disturbance is calculated. The calculation expression is as follows: Δθ comp (t)=Δθ j (t)-f(W(t),A(t)), Where f(W(t),A(t)) is the external disturbance compensation function, W(t) is the wind speed data, A(t) is the airflow data, Δθ comp (t) is the posture deviation after compensation; Finally, the dynamic balance deviation factor is generated according to the compensated posture deviation. The generation formula is as follows: ， In the formula, Dynamic Balance is the dynamic balance deviation factor, T is the total time window duration, is the absolute value of the attitude deviation change rate.

5. A low-altitude aircraft air traffic management method according to claim 2, characterized in that: After analyzing the extracted features, the track curvature change factor and dynamic balance deviation factor generated after analysis are input into the pre-trained deep learning model. The flight consistency deviation index is generated by the deep learning model, and the flight state changes of each UAV can be evaluated through the flight consistency deviation.

6. A low altitude aircraft air traffic management method according to claim 5, characterized in that: The flight consistency deviation index generated by the pre-trained deep learning model when intelligently evaluating the flight state changes of each drone is compared with the pre-set reference threshold of the flight consistency deviation index to divide the current drone flight state. The division steps are as follows: If the flight consistency deviation index is greater than a preset flight consistency deviation index reference threshold, the current UAV flight state is classified as off-track flight; If the flight consistency deviation index is less than or equal to a preset flight consistency deviation index reference threshold, the current UAV flight state is classified as a predetermined trajectory flight.

7. A low-altitude aircraft air traffic management method according to claim 6, characterized in that: For drones that are assessed to be off-track, the specific steps to dynamically adjust their flight speed based on the prediction results of the deep learning model to return them to the predetermined trajectory and avoid collision accidents are as follows: After confirming that the drone has deviated from the trajectory, the deep learning model is used to dynamically predict and adjust the flight speed. The flight speed is predicted based on historical flight data, current flight status, and external environment to ensure that the drone can return to the predetermined trajectory. The expression is as follows: ， Where V preset is the preset speed of the drone, Flight Consistency is the flight consistency deviation index, Flight ref is the reference threshold of the flight consistency deviation index, Flight max is the maximum value of the flight consistency deviation index, ΔV max is the maximum flight speed adjustment range, V adjusted is the adjusted flight speed; In addition to adjusting the flight speed, correcting the flight path is also a key step in restoring the intended trajectory. According to the prediction results of the deep learning model, the heading and flight path of the drone are adjusted to ensure that it flies in the intended direction. The calculation expression is as follows: ， In the formula, θ preset is the preset heading angle, λ is the heading adjustment weight factor, which is used to control the sensitivity of heading adjustment, Δθ max is the maximum heading adjustment angle, θ adjusted is the adjusted heading angle; By acquiring the UAV's flight data in real time and comparing it with the preset trajectory, the flight consistency deviation index Flight is continuously updated. Consistency , and correct the state of the aircraft based on the feedback results to avoid over-adjustment or under-adjustment due to system lag. The feedback control strategy is adopted to ensure a smooth transition during the adjustment process. The formula is as follows: F feedback =a adjust ·V adjusted +b adjust ·i adjusted , In the formula, α adjust is the flight speed adjustment weight factor, β adjust is the heading angle adjustment weight factor, F feedback is the feedback correction value.

8. A low-altitude aircraft air traffic management system, used to implement a low-altitude aircraft air traffic management method as described in any one of claims 1 to 7, characterized in that: It includes flight planning and path setting module, real-time flight status monitoring module, flight data preprocessing module, feature extraction and consistency analysis module, intelligent evaluation module, flight status classification module, stable flight maintenance module and dynamic adjustment and trajectory regression module; Flight planning and path setting module: before the flight show begins, the traffic management system formulates the flight path, speed, heading and formation of each drone according to the preset flight plan and performance requirements; Real-time flight status monitoring module: During the flight performance, the flight status parameters of each drone are monitored in real time to provide basic data for subsequent flight data analysis and adjustment; The flight data preprocessing module preprocesses the flight data acquired in real time to ensure the accuracy and reliability of subsequent feature extraction and analysis; The feature extraction and consistency analysis module extracts key features that reflect the current UAV’s deviation from the predetermined trajectory from the pre-processed flight data. Under the detection window, the extracted key features are analyzed to quantify the flight consistency of the UAV group and evaluate whether the overall flight status maintains the predetermined formation and coordination. The intelligent evaluation module inputs the quantified flight consistency features into the pre-trained deep learning model to intelligently evaluate the flight status changes of each drone; The flight status classification module divides the current UAV flight status into scheduled trajectory flight and off-trajectory flight according to the evaluation results of the deep learning model; The stable flight maintenance module continues to perform at the preset flight speed for the UAV in the predetermined trajectory flight state to ensure its stable flight; The dynamic adjustment and trajectory regression module dynamically adjusts the flight speed of drones that are assessed to have deviated from their flight trajectory based on the prediction results of the deep learning model, so that they return to the predetermined trajectory and avoid collision accidents.

Images