Low-altitude aircraft path planning method with collision warning safety constraint and correction

Abstract

The application discloses a low-altitude aircraft flight path planning method with collision warning safety constraint and correction, and the method comprises the following steps: constructing a flight environment model containing task data and task environment data of the low-altitude aircraft and obtaining an initial flight path; constructing at least two levels of collision warning zones outside obstacles by using a hierarchical collision warning zone model and constructing a collision warning zone safety constraint feedback item, and using the collision warning zone safety constraint feedback item to evaluate and adjust the initial flight path to obtain an initial safety flight path meeting preset safety constraints; using a comprehensive force in a perception potential field to perform position updating iteration correction processing on the initial safety flight path by using a collision warning zone perception potential field correction model; and outputting a final safety smooth flight path. The application combines hierarchical collision zone constraints, risk feedback, potential field correction and safety constraint checking, and can obtain a more safe, smooth and suitable flight path result for actual flight control execution in a complex low-altitude environment.

Description

Technical Field

[0001] This invention relates to the field of intelligent trajectory planning for low-altitude aircraft, and more particularly to a method for trajectory planning of low-altitude aircraft with collision warning safety constraints and corrections. Background Technology

[0002] The low-altitude economy, as an emerging industrial concept, has broad development prospects. Within this economy, low-altitude aircraft, especially unmanned aerial vehicles (UAVs), serve as crucial operational platforms, playing an increasingly vital role in logistics, inspection and monitoring, emergency rescue, and urban management. Improving their autonomous navigation capabilities has gradually become a research hotspot. UAV trajectory planning refers to planning a safe, efficient, and mission-compliant flight path for an aircraft in known or unknown environments, guiding it from its starting point to its target point. The core challenge lies in generating feasible paths in complex three-dimensional environments while balancing local safety, path smoothness, and mission execution efficiency during flight. Therefore, trajectory planning technology, as one of the key technologies for autonomous flight of low-altitude aircraft, has received increasing attention.

[0003] Traditional path planning methods can effectively handle path planning problems in static environments. However, when the environment is highly complex or local risks change significantly, traditional methods struggle to generate flight paths that balance safety and feasibility in a timely manner. In recent years, with the development of deep learning technology, deep reinforcement learning (DRL) has gradually become an important research direction in the field of low-altitude vehicle path planning. Deep reinforcement learning learns optimal decision-making strategies by simulating the interaction between the aircraft and its environment, making it particularly suitable for path planning problems in high-dimensional and complex environments. However, existing path planning methods still have certain limitations in complex low-altitude environments: on the one hand, relying solely on global planning results can lead to problems such as the aircraft flying close to obstacle edges and insufficient local safety margins; on the other hand, the original planned paths often exhibit obvious polygonal characteristics, large local turning angles, and insufficient smoothness, making it difficult to directly meet actual flight control requirements. Therefore, in response to the problems of insufficient local safety constraints, obvious obstacle-following flight phenomenon and poor trajectory smoothness in the existing technology in the complex low-altitude environment, a smart trajectory planning method for low-altitude aircraft based on graded collision zone constraints and potential field correction is proposed, which has become an important research direction in the field of low-altitude aircraft trajectory planning. Summary of the Invention

[0004] The purpose of this invention is to address the technical problems pointed out in the background art by providing a low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections, in order to solve the problems of insufficient local safety margin, obvious obstacle-hugging flight phenomenon, and poor smoothness of the original trajectory faced by low-altitude aircraft when planning their trajectories in complex environments.

[0005] The objective of this invention is achieved through the following technical solution: A method for low-altitude aircraft trajectory planning with collision warning safety constraints and corrections, the method comprising: S1. Construct a flight environment model that includes mission data and mission environment data of the low-altitude aircraft. The mission data includes the mission space region, the starting point position, and the target point position. The mission environment data includes obstacle distribution information. Obtain the initial trajectory based on the mission data of the low-altitude aircraft within the flight environment model. S2. Construct a graded collision warning zone model in the flight environment model. The graded collision warning zone model constructs at least two levels of collision warning zones outside the obstacle and constructs a collision warning zone safety constraint feedback term. Use the collision warning zone safety constraint feedback term to evaluate and adjust the initial trajectory to obtain an initial safe trajectory that meets the preset safety constraints. S3. Construct a collision warning zone perception potential field correction model that includes target guidance terms, obstacle entity repulsion terms, collision warning zone boundary constraint terms, and smoothing constraint terms. The collision warning zone perception potential field correction model uses comprehensive forces in the perception potential field to perform position update iteration correction processing on the initial safe trajectory. S4. When the initial safe track after position update iteration satisfies the preset safety constraints, preset smoothness and position update iteration termination conditions, output the final safe and smooth track.

[0006] To better implement this invention, in method S2, the hierarchical collision warning zone model constructs two levels of collision warning zones outside the obstacle, namely an inner collision warning zone and an outer collision warning zone, with a radius of [missing information] centered on the obstacle. Circular collision warning zone A, which is the inner collision warning zone, is defined with a radius of [missing information] centered on the obstacle. The outer collision warning zone is the circular collision warning zone B, and the outer collision warning zone is the annular collision warning zone formed by subtracting the circular collision warning zone A from the circular collision warning zone B.

[0007] Preferably, the method for obtaining the collision warning zone safety constraint feedback term is as follows: the distance between the pathpoint and the obstacle at time t on the initial trajectory of the low-altitude aircraft is... The local risk parameters of the low-altitude aircraft at the path point at time t are obtained according to the following expression. : Local risk parameters As a collision zone risk indicator quantity corresponding to the path point of a low-altitude aircraft at time t; Safety constraint feedback term of collision warning zone at path point of low-altitude aircraft at time t The expression is as follows: , The penalty coefficient for the outer collision warning zone. This is the penalty coefficient for the inner collision warning zone, and , This is an indicator function.

[0008] Preferably, in method S2, the initial safe trajectory is obtained by adjusting the preset safety constraints of the inner collision warning zone of the collision warning zone to avoid nearby obstacles using the collision warning zone safety constraint feedback term.

[0009] Preferably, the path points in the collision warning zone sensing potential field correction model The combined forces during the potential field correction process The expression is as follows: ,in path point The guiding force of the goal, path point The repulsive force from the obstacle. For the boundary constraints of the collision zone, path point The smoothing constraint force.

[0010] Preferably, the position update iteration method is as follows: S31. Keeping the starting point and target point positions of the initial safe trajectory unchanged, the collision warning zone sensing potential field correction model updates the positions of intermediate path points of the initial safe trajectory using a combination of forces in the sensing potential field. The position update expression is as follows: , Indicates the first Path points during the next iteration Location, Indicates the first Path points during the next iteration Location, Update the step size for the path point. path point The combined forces during the potential field correction process; S32. Perform a preset safety constraint verification on the path points after each iteration of position update. If the preset safety constraint verification fails, roll back and reduce the path point update step size or enhance the collision warning zone boundary constraint item before performing another position update correction. If the preset safety constraint verification passes or the updated path point is located outside the collision warning zone, retain the updated path point. Complete the position update iteration processing of all intermediate path points of the initial safe track.

[0011] Preferably, the target guidance item is used to guide the waypoints to adjust towards the target direction, and the waypoints... Parameters of target guidance The parameter corresponding to the target guide is expressed as follows: ,in The target guiding coefficient, The location of the target point.

[0012] Preferably, the obstacle entity repulsion term is used to suppress the waypoint from approaching the obstacle entity, and the waypoint The parameter of the repulsive force from the obstacle entity is the parameter corresponding to the repulsive force term of the obstacle entity, and the expression is as follows: ,in Obstacles For path points The resulting repulsive force This represents the number of obstacles.

[0013] Preferably, the collision warning zone boundary constraint is used to guide the waypoint away from the inner collision zone and keep it outside the outer warning zone as much as possible. The parameters of the collision zone boundary constraint are... The parameters corresponding to the boundary constraints of the collision warning zone are expressed as follows: ,in Here is the collision zone constraint coefficient. This indicates the correction direction parameter related to the direction of change in collision zone risk.

[0014] Preferably, the smoothing constraint is used to limit the degree of abrupt changes between adjacent path segments, path points Smoothing constraint parameters for The corresponding parameters are expressed as follows: , , respectively path points Adjacent path points, This is the smoothing coefficient.

[0015] Compared with the prior art, the present invention has the following advantages and beneficial effects: (1) The present invention adopts a method of global initial planning, hierarchical safety constraint adjustment and local potential field iterative correction. It retains the accessibility and computational efficiency of the global trajectory planning method, and achieves fine optimization of local high-risk areas through hierarchical early warning zone and perception potential field correction. It solves the problems of insufficient local safety margin, obvious obstacle-following flight phenomenon and poor smoothness of original trajectory faced by low-altitude aircraft when performing trajectory planning in complex environment.

[0016] (2) Based on the initial safe flight path, this invention further constructs a perception potential field correction model that includes target guidance term, obstacle entity repulsion term, collision warning zone boundary constraint term and smoothing constraint term. Among them, the smoothing constraint term effectively eliminates the common broken line features and local large turning angle problems in the original flight path by limiting the degree of abrupt change between adjacent path segments. The target guidance term and obstacle repulsion term together ensure the global reachability and obstacle avoidance capability of the flight path. By iteratively updating the position of the intermediate path points of the flight path through the comprehensive force, the final generated safe and smooth flight path is more in line with the execution requirements of the actual flight control system in terms of geometric continuity, and has higher feasibility and flight quality.

[0017] (3) The present invention constructs a graded collision warning zone model, sets up warning zones of at least two levels (such as inner collision warning zone and outer collision warning zone) outside the obstacle, and introduces a collision warning zone safety constraint feedback term. The safety constraint feedback term can dynamically calculate local risk parameters based on the real-time distance between the aircraft path point and the obstacle, so that the aircraft is subject to a light penalty when approaching the outer warning zone to adjust its trajectory direction in advance, and is subject to a strong penalty and forced avoidance when entering the inner collision zone. This effectively avoids the phenomenon of the aircraft flying close to the edge of the obstacle and significantly improves the local safety margin of the initial trajectory.

[0018] (4) The present invention performs a preset safety constraint verification on the path points after each update. If the verification fails, a rollback is executed and the update step size is dynamically reduced or the boundary constraint term of the collision warning zone is enhanced to correct the path point position. If the verification passes or the updated path point is located outside the collision warning zone, the update result is retained. This not only prevents the local minima or path divergence problem that may occur during the potential field correction process, but also ensures the numerical stability and convergence reliability of the entire iteration process. It can stably generate feasible tracks that meet the dual requirements of safety and smoothness in complex and dense low-altitude obstacle environments. Attached Figure Description

[0019] Figure 1 This is a flowchart of the low-altitude aircraft trajectory planning method of the present invention; Figure 2 This is a block diagram illustrating the principle of safe trajectory generation in the implementation system of the low-altitude aircraft trajectory planning method in the embodiment. Detailed Implementation

[0020] The present invention will be further described in detail below with reference to embodiments: Example like Figure 1 As shown, a low-altitude aircraft trajectory planning method with collision warning safety constraints and corrections is presented, the method comprising: S1. Construct a flight environment model containing mission data and mission environment data for the low-altitude aircraft. Mission data includes the mission space region (i.e., the spatial range from the starting point to the target point), the starting point position, and the target point position. Mission environment data includes obstacle distribution information (the location and size of obstacles within the spatial range). Within the flight environment model, obtain an initial trajectory based on the low-altitude aircraft's mission data. Input the mission environment data and the low-altitude aircraft's mission data into the flight environment model for flight simulation, and then use a global trajectory planning method (such as A) to perform the simulation. The original flight path can be generated using graph search algorithms such as Dijkstra's algorithm, or other existing path planning methods, such as multi-constraint multi-objective global planning methods and geometry-based path planning methods. The flight environment model has a raster map, and the paths are arranged according to the raster positions. Based on the obstacles in the task environment data (the set of obstacles is denoted as...) , Represents obstacle i in the environment. The distribution information (indicating the number of obstacles) corresponds to the setting of grid communication attributes. , This indicates that the location is occupied by an obstacle and is impassable; conversely, it indicates that the location is free space and can be passed by the aircraft. After obtaining the original trajectory, this invention performs local path optimization processes such as local safety margin assessment and optimization, and collision warning zone safety constraints.

[0021] S2. Construct a graded collision warning zone model within the flight environment model. This model establishes at least two levels of collision warning zones outside obstacles and includes safety constraint feedback terms for each zone. This embodiment uses the construction of two-level collision warning zones as an example. The principle is similar for collision warning zones with more than two levels (each level has different warning levels and different handling schemes; for example, constructing three levels would result in inner, middle, and outer collision warning zones. The outer layer is a safety monitoring warning zone; if the low-altitude aircraft's trackpoint is within this zone, a safety monitoring warning is issued, but no task is performed. The middle layer is a yellow collision warning zone; if the low-altitude aircraft's trackpoint is within this zone, a potential safety risk exists, and the collision warning zone's safety constraint feedback terms are used to adjust the risk towards zero. The inner layer is a red collision warning zone; if the low-altitude aircraft's trackpoint is within this zone, a collision safety risk exists, and the local trackpoint is forcibly modified to the red collision warning zone.) In addition, for example, modifying a local trajectory to a new local trajectory that avoids the red collision warning zone (this new local trajectory will be further optimized later), the hierarchical collision warning zone model constructs two levels of collision warning zones outside the obstacle: an inner collision warning zone (the inner collision warning zone is a red collision warning zone; if the low-altitude aircraft's trajectory is in the red collision warning zone, there is a collision safety risk, and the local trajectory is forcibly modified to the red collision warning zone, for example, modifying the local trajectory to a new local trajectory that avoids the red collision warning zone, this new local trajectory will be further optimized later) and an outer collision warning zone (the outer collision warning zone is a yellow collision warning zone; if the low-altitude aircraft's trajectory is in the yellow collision warning zone, there is a potential safety risk, and the collision warning zone safety constraint feedback term is adjusted towards zero risk). A radius of [missing information] is drawn around the obstacle. Circular collision warning zone A, which is the inner collision warning zone, is defined with a radius of [missing information] centered on the obstacle. The outer collision warning zone is the circular collision warning zone B, and the outer collision warning zone is the annular collision warning zone formed by subtracting the circular collision warning zone A from the circular collision warning zone B.

[0022] The initial trajectory is evaluated and adjusted using a collision warning zone safety constraint feedback term to obtain an initial safe trajectory that meets preset safety constraints. In some embodiments, the collision warning zone safety constraint feedback term is obtained as follows: the distance between the path point of the low-altitude aircraft's initial trajectory at time t (the initial trajectory is arranged in a time sequence, and the path point at time t is the path point at time sequence number t) and the obstacle is... , , Represents any point in the obstacle region. This represents the shortest distance from the waypoint to the obstacle at time t on the initial trajectory; the local risk parameters of the low-altitude aircraft at the waypoint at time t are obtained according to the following expression. : Local risk parameters As the collision zone risk indicator quantity corresponding to the path point at time t of the low-altitude aircraft (used to characterize the local safety status of the current path point of the aircraft), the local risk parameters of each obstacle at the path point of the initial trajectory of the low-altitude aircraft at time t are evaluated respectively, and then the local risk parameters are added together to obtain the local risk parameters of the initial trajectory at the path point at time t. , , As a preset risk value for the corresponding situation, a simple example is shown below in this embodiment: , , The initial trajectory of the low-altitude aircraft at time t represents the collision warning zone safety constraint feedback term at the pathpoint. The expression is as follows: , The penalty coefficient for the outer collision warning zone. This is the penalty coefficient for the inner collision warning zone, and , This is the indicator function; it summarizes the collision warning zone safety constraint feedback terms of the initial trajectory at the waypoint at time t, representing the risk of encountering various obstacles at the waypoint at time t.

[0023] This invention utilizes a collision warning zone safety constraint feedback term to adjust the initial trajectory to avoid the inner collision warning zone of nearby obstacles, based on preset safety constraints. The preset safety constraint rules are: first, local trajectories or pathpoints avoid the inner collision warning zone of nearby obstacles; second, local trajectories or pathpoints adjust towards the direction corresponding to the collision warning zone safety constraint feedback term where the risk approaches zero or the risk is reduced. When the aircraft approaches the outer warning zone, it suffers a smaller penalty, guiding it to adjust its trajectory direction in advance. When entering the inner collision zone, it suffers a larger penalty, guiding it to actively move away from high-risk areas. Preferably, for pathpoints entering the outer warning zone, a slight correction method is used to increase the safe distance between them and obstacles; for pathpoints entering the inner collision zone, a forced avoidance method is used to adjust them to a safe area, thus obtaining an initial safe trajectory. Figure 2 To realize the principle diagram of safe trajectory generation in the low-altitude aircraft trajectory planning method system, the entire system framework is divided into three layers, forming a complete safety closed loop from top to bottom: "environmental perception → risk assessment → trajectory feedback", as follows: Environment layer (perception layer): Inputs the drone's own status and information about environmental obstacles; Decision layer (risk assessment layer): Calculates the collision risk assessment between the drone and obstacles; specifically including: Neighborhood Search: A safe neighborhood is defined centered on the drone's current location, and information on all obstacles within this range is searched. Distance Calculation: The real-time distance between the drone and each obstacle is calculated, yielding key indicators such as minimum safe distance and approach speed. Collision Risk Assessment: Based on distance, speed, obstacle type, etc., a quantified collision risk assessment (e.g., low / medium / high risk) is performed, and a risk signal is output to the feedback layer.

[0024] Feedback layer (control and trajectory generation layer): Adjusts actions based on risk assessment results and outputs a safe trajectory. It also feeds back the new trajectory information to the environment layer, forming a closed-loop process of perception, assessment and adjustment for the next round.

[0025] S3. Construct a collision warning zone perception potential field correction model, including target guidance terms, obstacle entity repulsion terms, collision warning zone boundary constraint terms, and smoothing constraint terms. This model uses a combination of forces within the perception potential field to iteratively correct the initial safe trajectory's position update. The path points in the collision warning zone perception potential field correction model... (The path points at time t are sorted by time series, and the path points here are...) The time series is ordered sequentially by waypoints, which correspond one-to-one with the time series, such as waypoints. The combined force of the path point at time t during the potential field correction process The expression is as follows: ,in path point The guiding force of the goal, path point The repulsive force from the obstacle. For the boundary constraints of the collision zone, path point The smoothing constraint applied. The target guidance term is used to guide the waypoints to adjust towards the target direction; the pathpoint... Parameters of target guidance The parameter corresponding to the target guide is expressed as follows: ,in The target guiding coefficient, The location of the target point.

[0026] The obstacle entity repulsion term is used to suppress waypoints from approaching obstacle entities and waypoints. The parameter of the repulsive force from the obstacle entity is the parameter corresponding to the repulsive force term of the obstacle entity, and the expression is as follows: ,in Obstacles For path points The resulting repulsive force The number of obstacles. Repulsive force. The expression is as follows: ,in Indicates the repulsive force coefficient of the obstacle. Indicates obstacles The reference position is used to prevent the waypoint from approaching the obstacle entity.

[0027] The collision warning zone boundary constraint is used to guide waypoints away from the inner collision zone and keep them as far outside the outer warning zone as possible. The parameters of the collision zone boundary constraint are... The parameters corresponding to the collision warning zone boundary constraint terms ensure that the waypoints are not only far from the obstacle entity, but also far from the high-risk boundary area within the obstacle's neighborhood. The expression is as follows: ,in Here is the collision zone constraint coefficient. This represents the corrected direction parameter related to the direction of change in collision zone risk. The smoothing constraint term limits the degree of abrupt changes between adjacent path segments, reducing the polyline characteristics and local abrupt turns in the track, and the waypoint. Smoothing constraint parameters for The corresponding parameters are expressed as follows: , , respectively path points Adjacent path points, This is the smoothing coefficient.

[0028] In some embodiments, the collision warning zone sensing potential field correction model uses a combination of forces in the sensing potential field to perform position update iteration correction processing on the initial safe trajectory. The position update iteration method is as follows: S31. Keeping the starting point and target point positions of the initial safe trajectory unchanged, the collision warning zone sensing potential field correction model updates the positions of intermediate path points of the initial safe trajectory using a combination of forces in the sensing potential field. The position update expression is as follows: , Indicates the first Path points during the next iteration Location, Indicates the first Path points during the next iteration Location, Update the step size for the path point. path point The combined forces during the potential field correction process.

[0029] S32. Verify the preset safety constraints (i.e. the aforementioned preset safety constraints) for the path points after each iteration of position update. If the preset safety constraint verification fails, roll back and reduce the path point update step size or enhance the collision warning zone boundary constraint item before performing another position update correction. If the preset safety constraint verification passes or the updated path point is located outside the collision warning zone, retain the updated path point. Complete the position update iteration processing of all intermediate path points of the initial safe track.

[0030] S4. When the initial safe trajectory after position update iterations satisfies preset safety constraints, preset smoothness, position update iteration termination conditions, and preset safety distance (the path optimization process ends when the change in path point position between two adjacent iterations is less than a preset threshold, or when the number of iterations reaches a preset upper limit), the final safe and smooth trajectory is output. The final safe and smooth trajectory is denoted as... The safe and smooth flight path enhances the local safety margin between the aircraft and obstacles while maintaining the overall accessibility of the original flight path, reduces the risk of obstacle-hugging flight, and improves the continuity and smoothness of the flight path.

[0031] This invention embeds collision zone safety constraint feedback terms into the trajectory generation or evaluation process, enabling the aircraft to adjust its trajectory direction in advance when approaching the outer warning zone and enhance its avoidance capabilities when entering the inner collision zone. This transforms local risk avoidance from post-event judgment into proactive constraints during the planning process, improving the safety of the initial trajectory. Based on the initial safe trajectory, this invention constructs a collision zone perception potential field correction model, unifying target guidance, obstacle entity repulsion, collision zone boundary constraints, and smoothing constraints at intermediate path points. Furthermore, this invention introduces collision zone boundary constraint terms and an updated safety verification mechanism, which can maintain a safe distance between the trajectory and high-risk areas while improving trajectory smoothness. By combining hierarchical collision zone constraints, risk feedback, potential field correction, and safety verification, this invention balances the safety, accessibility, and smoothness of trajectory planning, enabling safer, smoother, and more suitable trajectory results for actual flight control execution in complex low-altitude environments. This provides effective technical support for the autonomous flight of low-altitude aircraft in complex scenarios.

[0032] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for low-altitude aircraft trajectory planning with collision warning safety constraints and corrections, characterized in that: The methods include: S1. Construct a flight environment model that includes mission data and mission environment data of the low-altitude aircraft. The mission data includes the mission space region, the starting point position, and the target point position. The mission environment data includes obstacle distribution information. Obtain the initial trajectory based on the mission data of the low-altitude aircraft within the flight environment model. S2. Construct a graded collision warning zone model in the flight environment model. The graded collision warning zone model constructs at least two levels of collision warning zones outside the obstacle and constructs a collision warning zone safety constraint feedback term. Use the collision warning zone safety constraint feedback term to evaluate and adjust the initial trajectory to obtain an initial safe trajectory that meets the preset safety constraints. S3. Construct a collision warning zone perception potential field correction model that includes target guidance terms, obstacle entity repulsion terms, collision warning zone boundary constraint terms, and smoothing constraint terms. The collision warning zone perception potential field correction model uses comprehensive forces in the perception potential field to perform position update iteration correction processing on the initial safe trajectory. S4. When the initial safe trajectory meets the preset safety constraints, preset smoothness and position update iteration termination conditions, output the final safe and smooth trajectory.

2. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 1, characterized in that: In method S2, the hierarchical collision warning zone model constructs two levels of collision warning zones outside the obstacle: an inner collision warning zone and an outer collision warning zone. These are defined with a radius of [missing information - likely a radius value] centered on the obstacle. Circular collision warning zone A, which is the inner collision warning zone, is defined with a radius of [missing information] centered on the obstacle. The outer collision warning zone is the circular collision warning zone B, and the outer collision warning zone is the annular collision warning zone formed by subtracting the circular collision warning zone A from the circular collision warning zone B.

3. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 2, characterized in that: The method for obtaining the safety constraint feedback term of the collision warning zone is as follows: the distance between the pathpoint and the obstacle at time t is... The local risk parameters of the low-altitude aircraft at the path point at time t are obtained according to the following expression. : Local risk parameters As a collision zone risk indicator quantity corresponding to the path point of a low-altitude aircraft at time t; Safety constraint feedback term of collision warning zone at path point of low-altitude aircraft at time t The expression is as follows: , The penalty coefficient for the outer collision warning zone. This is the penalty coefficient for the inner collision warning zone, and , This is an indicator function.

4. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 3, characterized in that: In method S2, the initial safe trajectory is obtained by adjusting the preset safety constraints of the inner collision warning zone of the collision warning zone to avoid nearby obstacles using the collision warning zone safety constraint feedback term.

5. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 1, characterized in that: The path points in the collision warning zone sensing potential field correction model The combined forces during the potential field correction process The expression is as follows: ,in path point The guiding force of the goal, path point The repulsive force from the obstacle. For the boundary constraints of the collision zone, path point The smoothing constraint force.

6. A low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 1 or 5, characterized in that: The position update iteration method is as follows: S31. Keeping the starting point and target point positions of the initial safe trajectory unchanged, the collision warning zone sensing potential field correction model updates the positions of intermediate path points of the initial safe trajectory using a combination of forces in the sensing potential field. The position update expression is as follows: , Indicates the first Path points during the next iteration Location, Indicates the first Path points during the next iteration Location, Update the step size for the path point. path point The combined forces during the potential field correction process; S32. Perform a preset safety constraint verification on the path points after each iteration of position update. If the preset safety constraint verification fails, roll back and reduce the path point update step size or enhance the collision warning zone boundary constraint items before performing another position update correction. If the preset safety constraint verification passes or the updated path point is located outside the collision warning zone, retain the updated path point. Complete the position update and iteration processing of all intermediate waypoints of the initial safe flight path.

7. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 5, characterized in that: The target guidance item is used to guide the waypoints to adjust towards the target direction, and the waypoints... Parameters of target guidance The parameter corresponding to the target guide is expressed as follows: ,in The target guiding coefficient, The location of the target point.

8. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 5, characterized in that: The obstacle entity repulsion term is used to suppress waypoints from approaching obstacle entities and pathpoints. The parameter of the repulsive force from the obstacle entity is the parameter corresponding to the repulsive force term of the obstacle entity, and the expression is as follows: ,in For obstacles For path points The resulting repulsive force This represents the number of obstacles.

9. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 5, characterized in that: The collision warning zone boundary constraint is used to guide waypoints away from the inner collision zone and keep them as far outside the outer warning zone as possible. The parameters of the collision zone boundary constraint are... The parameters corresponding to the boundary constraints of the collision warning zone are expressed as follows: ,in Here is the collision zone constraint coefficient. This indicates the correction direction parameter related to the direction of change in collision zone risk.

10. The low-altitude aircraft trajectory planning method for collision warning safety constraints and corrections according to claim 5, characterized in that: The smoothing constraint term is used to limit the degree of abrupt changes between adjacent path segments, path points Smoothing constraint parameters for The corresponding parameters are expressed as follows: , , respectively path points Adjacent path points, This is the smoothing coefficient.

Images