Unmanned aerial vehicle autonomous homing control method imitating protopigeon homing behavior

Abstract

The invention discloses an unmanned aerial vehicle autonomous homing control method imitating a protopigeon homing behavior. The method comprises the implementation steps of 1, scene initialization; step 2, rasterizing an environment map; step 3, A * algorithm global path planning; 4, designing a parameterized path based on polynomial interpolation; step 5, guiding a path based on a simulated protopigeon homing behavior; and step 6, obtaining a homing path and a homing result of the unmanned aerial vehicle. The method is a complete unmanned aerial vehicle autonomous homing method, and can meet the autonomous homing and obstacle avoidance tasks of the unmanned aerial vehicle in a relatively complex environment; transition of path planning from the front end to the rear end is achieved, smoothing of the speed in the process is completed, and track abrupt change caused by abrupt change of the speed in transition states of different path sections is avoided. An unmanned aerial vehicle autonomous control scheme is established, real-time performance is good, adaptability is high, and good adaptation to the environment can be achieved.

Description

Technical Field

[0001] The invention relates to a control method for autonomous homing of an unmanned aerial vehicle (UAV) imitating the homing behavior of a pigeon, and belongs to the field of autonomous control of unmanned aerial vehicles. Background Art

[0002] At present, the use of drones for search and homing has attracted widespread attention. In real life, drones have been used for post-disaster search and rescue and target search and strike. In the foreseeable future, drone delivery services will also be developed. While performing related tasks, drones need to reach a designated location and face various challenges brought by complex environments, such as obstacles such as trees or wires that hinder the normal flight of drones. In related environments, drones need to have the ability to autonomously search to adapt to various complex terrains and environmental features. Autonomous homing allows drones to adapt to diverse environments, avoid collisions with environmental obstacles, and safely navigate to the target location.

[0003] The problem of drone homing has been widely discussed, including path planning, path tracking, etc. However, the relevant discussions are mainly focused on the top-level application design. In terms of specific drone control, how to implement the control method to the drone control is still insufficient. By parameterizing the planned path in time, forming a dynamically feasible path for the drone is a good way to sink the top-level path design to the specific control of the drone. Through the time-parameterized path design, the formed path has a series of advantages such as strong feasibility and flexible adjustment, which is a better technical route for the future drone homing control.

[0004] At present, the commonly used obstacle avoidance methods for UAVs include a series of methods represented by the artificial potential field method. This method is easy to use and widely used, but it lacks certain intelligence and is prone to falling into the dilemma of local optimality when facing complex environments. Another main method is the speed obstacle method, but it has high computational complexity and large computational overhead.

[0005] When returning to their nests, pigeons demonstrate excellent navigation performance and path optimization capabilities. Integrating the homing behavior of original pigeons into the autonomous search process of drones can effectively improve the intelligence of drone search, and then improve the adaptability and flexibility of drones, so that they can better adapt to the actual mission scenarios of drones, flexibly adjust the drone's search behavior, and realize autonomous search of drones.

[0006] In summary, the present invention aims at the current problem of autonomous navigation of drone clusters in complex environments, takes into account obstacles in the environment, and adopts a control method for autonomous navigation of drones that imitates the homing behavior of pigeons to autonomously control drones. This method uses the polynomial interpolation method to smooth the original planned path of the drone and realize the back-end control of the drone during the navigation process. At the same time, inspired by the homing behavior of pigeons, by imitating the homing behavior of pigeons, the drone can avoid environmental obstacles and reach the target. This method is highly practical and adaptable, and can realize autonomous navigation of drones. Summary of the invention

[0007] 1. Purpose of the invention:

[0008] The present invention discloses a control method for autonomous homing of a drone imitating the homing behavior of a pigeon, and aims to provide a method for improving the autonomous capability level of the drone, realizing the autonomous homing of the drone in a complex environment, and being able to actively avoid obstacles in the environment. The method aims to improve the collision avoidance capability of the drone in various tasks in different environments, and designs a drone control method matching the method, thereby improving the adaptability and autonomy of the drone in a complex environment, and improving the autonomous capability level of the drone.

[0009] 2. Technical solution:

[0010] The present invention aims at the problem of autonomous homing of a UAV in an environment and invents a control method for autonomous homing of a UAV that imitates the homing behavior of a pigeon. The flow chart is as follows: Figure 1 As shown, the specific implementation steps are as follows:

[0011] Step 1: Scene initialization

[0012] Model the drone as a double integral system as shown below:

[0013]

[0014] Among them, p, v, and u are the position, velocity, and acceleration vectors of the drone, respectively.

[0015] Subsequently, the environment map is initialized, initial obstacles are set in the environment, and parameters such as the shape, size, and center point position of each obstacle are set.

[0016] Step 2: Rasterize the environment map

[0017] According to the size of the drone, the radius of the drone is selected as the expansion parameter to expand the obstacles. Then, the scene map is divided into grids according to the set grid map resolution. Finally, the distribution of obstacles in the environment in the grid is checked. If there is an obstacle in a certain grid, the value of the grid is set to 1; otherwise, it is set to 0.

[0018] Step 3: A* algorithm global path planning

[0019] According to the grid map obtained in step 2, the current grid position of the drone is used as the planning starting point, and the grid position of the target to be reached is used as the end point. The A* algorithm is used to design the original path from the starting point to the end point. The path is represented by a series of discrete points on the path, that is, the global path: Γ p =[p1,p2…,p n ], where p1 is the initial position of the drone, p n =p a is the target point of the UAV and also the final path point of the UAV. i (i=2,3,…,n-1) is the middle path point of the UAV.

[0020] Among them, the objective function g(p) of the A* algorithm is the Euclidean distance between the map grid and the target. The specific calculation formula is as follows:

[0021] g(p)=||p a -p|| (2)

[0022] Among them, p is the coordinate currently traversed by the A* algorithm, p a is the coordinates of the target.

[0023] Step 4: Parameterized path design based on polynomial interpolation

[0024] According to the global path Γ obtained in step 3 p , parameterize the global path into a feasible global path, and refine the parameterized path through polynomial interpolation. The specific steps are as follows:

[0025] S41. Preliminary design of time parameterized path

[0026] Based on the global path designed in step 3, time parameterization is performed. First, a parameterized time t1=0 is designed for the initial position of the drone, and then the maximum speed v of the drone is calculated. m , for the path point sequence (i.e., global path) Γ p Each spatial path point (hereinafter referred to as “path point”) designed in the above formula is set with a reasonable expected arrival time, and then an initial time parameterized path Γ is obtained. t =[(t1,p1),(t2,p2),…,(t n ,p n )]. The temporal parameters of the i-th path point are calculated as follows:

[0027]

[0028] In order to facilitate the subsequent path interpolation processing, the time difference between every two path points is a constant value.

[0029] S42, speed parameterized path design

[0030] According to the obtained time parameterization path Γ t Set the desired speed v at the end of the path n =0, and then the backward difference method is used to calculate the expected speed at each path point except the end point, that is, the parameterized path speed. The specific calculation method is as follows:

[0031]

[0032] where t i ,p i are the expected arrival time and path position of the i-th time parameterized path point respectively.

[0033] By parameterizing the path speed, we get the speed parameterized path Γ v =[(t1,p1,v1),(t2,p2,v2),…,(t n ,p n ,v n )].

[0034] S43, Parameterized Path Polynomial Interpolation

[0035] First, the time base sequence of the time parameterized path of the UAV is refined according to the observation interval of the UAV. Then, the speed parameterized path Γ obtained in step S42 is v , using the polynomial fitting method, interpolate multiple time points in each path, determine the speed of each interpolation time point in the time base sequence, and smooth the speed change of the drone. The specific design method is as follows:

[0036] For the speed parameterized path Γ with n path points designed in S42 v =[(t1,p1,v1),(t2,p2,v2),…,(t n ,p n ,v n )], design a series of interpolation time points t between each two moments i,j , and together with the expected arrival time sequence in the time parameterized path, it serves as the time base sequence T′=[t1,t 1,1 ,t 1,2 ,…,t2,t 2,1 ,…,t n ]=[t′1,t′2,…,t′ n ]. Assuming that there are N interpolation points between every two times, there is a corresponding relationship (5):

[0037]

[0038] Among them, t i,j The calculation formula is as follows:

[0039]

[0040] Where N is the number of interpolation points between every two time points.

[0041] The linear interpolation method is also used to obtain the path point position coordinates p′ corresponding to each interpolation moment in each time base sequence i , and the global path Γ p The path point sequence in the interpolation path is taken together as the path point sequence P′=[p1,p 1,1 ,p 1,2 ,…,p2,p 2,1 ,…,p n ]=[p′1,p′2,…,p′ n ]. The specific calculation formula is as follows:

[0042]

[0043] p′ i With p i,j The corresponding relationship is as follows:

[0044]

[0045] Then, the path Γ is parameterized according to the original velocity v , use the cubic polynomial interpolation method to calculate the velocity sequence V′=[v′1,v′2,…,v′ n ], the specific calculation method is as follows:

[0046] First, select the velocity parameterization path Γ v The first four time points t1, t2, t3, t4 are grouped into three, and the cubic polynomial velocity fitting functions f1(t) and f2(t) are calculated for each three time points. Each cubic polynomial function has the following form: f(t) = a3t 3 +a2t 2 +a1t+a0.

[0047] The speed at each time point should be equal to the speed designed in the speed parameterization path, and the first-order derivative (acceleration) at the intersection of the two polynomials should be equal. Based on the above conditions, the following 8 equations are listed:

[0048]

[0049] Through the above 8 equations, the 8 coefficients of the two polynomial functions are obtained, and the expected speed at the corresponding moment is calculated according to the moment of each point in the time base sequence T′. The specific calculation formula is as follows:

[0050]

[0051] For the velocity parameterized path Γ v The i-th point (i≥5) of , let the velocity fitting function satisfied by the i-1-th point be f i-3 (t), has been solved, then the time t(t i-1 <t≤t i ) Speed ​​fitting function f over time period i (t):

[0052]

[0053] And calculate each moment t′ in the time base series through the above fitting function j Speed:

[0054] v′ j =f i-2 (t′ j )(t i-1 <t′ j ≤t i ) (12)

[0055] Combining the interpolation results T′, P′ and V′ of the above time [Formula (5) (6)], position [Formula (7), (8)], and velocity [Formula (10), (12)], we can get the interpolated path Γ = [(t′1, p′1, v′1), (t′2, p′2, v′2), …, (t′ n ,p′ n ,v′ n )], which is the final interpolation path.

[0056] Step 5: Path guidance to simulate the homing behavior of pigeons

[0057] First, based on the interpolated path Γ designed in step 4, the basic control items for autonomous homing of the UAV are designed. The specific calculation is as follows:

[0058] u b =K p (p′ i -p)+K v (v′ i -v) (13)

[0059] Among them, K p , K vare position and speed control gains respectively; p i , v i are the position and velocity corresponding to each segment in the time parameterized path; p and v are the position and velocity vectors of the UAV respectively; i represents the number of the time parameterized path where the UAV is currently located.

[0060] Then, based on the homing behavior of pigeons, the control correction items for the drone homing are designed. The specific steps are as follows:

[0061] When returning to their nests, pigeons will follow the relevant principles of homing. Mapping it to the search process of the drone mainly includes two aspects: First, the pigeon nest has an attractive effect on the pigeon flock, which is mapped to the tendency of the drone to the target point. Second, when returning to their nests, pigeons will avoid terrain obstacles or attacks by predators, which is mapped to the drone's avoidance behavior of obstacles in the environment. Based on the above two aspects, the path guidance correction designed to simulate the homing behavior of pigeons is as follows:

[0062] S51, Nest attraction item

[0063] The UAV greedily moves forward to the target point, and the corresponding control correction is:

[0064]

[0065] Among them, u a is the acceleration correction of the UAV moving towards the target, p a is the position vector of the target point, p is the position vector of the drone, K a It is the acceleration gain of the nest attraction term.

[0066] S52, predation and obstacle avoidance

[0067] When the UAV faces obstacles or attackers in the environment, it generates predation and obstacle avoidance acceleration in order to avoid them. The present invention is inspired by the process of pigeons avoiding hawks, and realizes autonomous obstacle avoidance control of the UAV.

[0068] First, we use sampling to traverse each grid in the grid map. For grids with obstacles, we record the center coordinates of the grid as the point cloud coordinates of the obstacle. By integrating the point cloud coordinates of obstacles in the entire map, we can get the point cloud information of obstacles in the environment.

[0069] Using all the obstacle point cloud information, the obstacle avoidance items for the drone are calculated as follows:

[0070]

[0071] Among them, u d Represents the total acceleration of the drone due to obstacle avoidance, K srepresents the acceleration gain generated by the drone to avoid the sth obstacle point cloud, R c Indicates the safe distance of the drone, p s Represents the position vector of the sth obstacle point cloud.

[0072] S53, UAV simulated pigeon homing motion status update

[0073] Based on the above principles of S51 and S52, the control quantity of the UAV can be designed. The total acceleration of the UAV is u as shown below. By substituting it into the UAV control model, the motion state of the UAV can be updated.

[0074] u=u b +u a +u d (17)

[0075] Step 6: The path and results of the drone search

[0076] According to the designed UAV path and the control method of UAV autonomous homing, the UAV is controlled to reach the specified location, and the path and results of the UAV homing are displayed.

[0077] 3. Advantages and effects:

[0078] The present invention proposes a control method for autonomous homing of a UAV imitating the homing behavior of pigeons. The present invention mainly has the following advantages: First, a complete autonomous homing method for a UAV is invented, covering the front-end and back-end technical solutions of the autonomous homing of a UAV. The method is novel, comprehensive and practical, and can meet the autonomous homing and obstacle avoidance tasks of a UAV in a relatively complex environment; Second, based on the polynomial interpolation method, the planned original path is interpolated, the transition of path planning from the front end to the back end is realized, and the speed is smoothed in the process, avoiding the sudden change of trajectory caused by the sudden change of speed in the transition state of different path segments; Third, by imitating the homing behavior of pigeons, an autonomous control scheme for a UAV is established to guide the UAV to a designated target, which has the advantages of good real-time performance and strong adaptability, and can achieve good adaptation to the environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 Flowchart of a control method for autonomous homing of a UAV that imitates the homing behavior of a pigeon

[0080] Figure 2 The scene that the drone is autonomously searching for

[0081] Figure 3 Scene raster map

[0082] Figure 4 The initial feasible path found by the drone

[0083] Figure 5 Obstacle point cloud map

[0084] Figure 6 Autonomous trajectory of drones DETAILED DESCRIPTION

[0085] The following is a process of autonomous homing of a drone in a specific environment to verify the effectiveness of the autonomous homing of a drone and the method thereof proposed in the present invention. The simulation environment configuration of this example is Savior Y7000P, and the software is Matlab2021.

[0086] Step 1: Scene initialization

[0087] Set the initial coordinates of the drone to [0,0,10], and approximate the drone to a circle with a radius of 0.5. Set the target of the drone to [10,6,10]. At the same time, set the environment to have three obstacles, namely: a circular obstacle with a center coordinate of [5,5] and a radius of 2; a rectangular obstacle with a center coordinate of [8,3], a length and width of 2.5 and 2 respectively; a rectangular obstacle with a center coordinate of [7,7], a length and width of 2.

[0088] Step 2: Rasterize the environment map

[0089] The obstacle is expanded, and the length, width or radius of the obstacle plus the radius of the drone are used as the new obstacle parameters and the map is updated. The resolution of the length and width of the map is set to 0.3, and a raster map is generated.

[0090] Step 3: A* algorithm global path planning

[0091] In the grid map, the A* algorithm is used to plan the global path and obtain a series of path points.

[0092] Step 4: Parameterized path design based on polynomial interpolation

[0093] S41. Preliminary design of time parameterized path

[0094] The time of the initial path point is designed to be 0, the time interval between each two path points is designed to be 0.1, and the parameterized time corresponding to each path point is calculated according to formula (3).

[0095] S42, speed parameterized path design

[0096] The velocity of each point on the parameterized path is calculated according to equation (4).

[0097] S43, parameterized path velocity polynomial interpolation

[0098] For the obtained parameterized path, the interpolation time and corresponding coordinates are calculated according to equations (5)-(8), and then the speed corresponding to each interpolation time is calculated according to equations (9)-(12), and the final interpolation path is obtained.

[0099] Step 5: Path guidance to simulate the homing behavior of pigeons

[0100] First, the basic control items of the UAV's autonomous homing at the current moment are calculated according to formula (13), and then the path guidance correction of the UAV's homing behavior of the original pigeon is calculated.

[0101] S51, Nest attraction item

[0102] The control correction of the nest attraction term is calculated according to equation (14).

[0103] S52, predation and obstacle avoidance

[0104] Traverse the grid map, sample the point cloud information of map obstacles, and then calculate the drone's obstacle avoidance correction term according to equations (15)-(16).

[0105] S53, UAV simulated pigeon homing motion status update

[0106] Calculate the control amount of the UAV homing according to formula (17), bring it into the UAV model (1), and update the state of the UAV.

[0107] Step 6: The path and results of the drone search

[0108] According to the designed UAV path and the control method of UAV autonomous homing, the UAV is controlled to reach the specified location, and the path and results of the UAV homing are displayed.

[0109] Figure 2 This is the scene diagram that the drone autonomously searches for in this example. Figure 3 This is the rasterized environment map in this example. Figure 4 is the original feasible path planned by the A* algorithm. Figure 5 is the obstacle point cloud map, Figure 6 The trajectory of the drone is represented by the discrete points in the figure. The drone finally avoids obstacles in the environment and successfully reaches the target point, completing the drone's autonomous search mission.

Claims

1. A control method for autonomous homing of a drone that simulates the homing behavior of a pigeon, characterized by: The specific implementation steps of this method are as follows: Step 1: Scene initialization Model the drone as a double integral system as shown below: Among them, p, v, and u are the position, velocity, and acceleration vectors of the UAV respectively; Then, initialize the environment map, set the initial obstacles in the environment, and set the shape, size, and center point position parameters of each obstacle; Step 2: Rasterize the environment map According to the size of the drone, the radius of the drone is selected as the expansion parameter to expand the obstacles; then, the scene map is divided into grids according to the set grid map resolution; finally, the distribution of obstacles in the environment in the grid is checked. If there is an obstacle in a certain grid, the value of the grid is set to 1; otherwise, it is set to 0; Step 3: A* algorithm global path planning According to the grid map obtained in step 2, the current grid position of the drone is used as the planning starting point, and the grid position of the target to be reached is used as the end point. The A* algorithm is used to design the original path from the starting point to the end point. The path is represented by a series of discrete points on the path, that is, the global path: Γ p =[p1,p2…,p n ], where p1 is the initial position of the drone, p n =p a is the target point of the UAV and also the final path point of the UAV. i (i=2,3,…,n-1) is the middle path point of the drone; Among them, the objective function g(p) of the A* algorithm is the Euclidean distance between the map grid and the target. The specific calculation formula is as follows: g(p)=||p a -p|| (2) Among them, p is the coordinate currently traversed by the A* algorithm, p a is the coordinates of the target; Step 4: Parameterized path design based on polynomial interpolation According to the global path Γ obtained in step 3 p , parameterize the global path into a feasible global path, and refine the parameterized path by polynomial interpolation; Step 5: Path guidance to simulate the homing behavior of pigeons First, based on the interpolated path Γ designed in step 4, the basic control items for autonomous homing of the UAV are designed; then, based on the homing behavior of pigeons, the control correction items for homing of the UAV are designed; Step 6: The path and results of the drone search According to the designed UAV path and the control method of UAV autonomous homing, the UAV is controlled to reach the specified location, and the path and results of the UAV homing are displayed.

2. The method according to claim 1, characterized in that: The step 4 specifically includes the following sub-steps: S41. Preliminary design of time parameterized path Based on the global path designed in step 3, time parameterization is performed. First, a parameterized time t1=0 is designed for the initial position of the UAV, and then the maximum speed v of the UAV is calculated. m , for the path point sequence, i.e. the global path Γ p Each spatial path point designed in the above is referred to as a "path point" below. Set a reasonable expected arrival time to obtain an initial time parameterized path Γ t =[(t1,p1),(t2,p2),…,(t n ,p n )]; The temporal parameters of the i-th path point are calculated as follows: In order to facilitate the subsequent path interpolation processing, the time difference between every two path points is a constant value; S42, speed parameterized path design According to the obtained time parameterization path Γ t Set the desired speed v at the end of the path n =0, and then the backward difference method is used to calculate the expected speed at each path point except the end point, that is, the parameterized path speed. The specific calculation method is as follows: Among them, t i ,p i are the expected arrival time and path position of the i-th time parameterized path point respectively; By parameterizing the path speed, we can get the speed parameterized path Γ v =[(t1,p1,v1),(t2,p2,v2),…,(t n ,p n ,v n )]; S43, Parameterized Path Polynomial Interpolation First, the time base sequence of the time parameterized path of the UAV is refined according to the observation interval of the UAV; then, the speed parameterized path Γ obtained in step S42 is v , the polynomial fitting method is used to interpolate multiple time points in each path, determine the speed of each interpolation time point in the time base sequence, and smooth the speed change of the drone; the specific design method is as follows: For the speed parameterized path with n path points designed in S42 Γ v =[(t1,p1,v1),(t2,p2,v2),…,(t n ,p n ,v n )], design a series of interpolation time points t between each two moments i,j , and together with the expected arrival time sequence in the time parameterized path, it serves as the time base sequence T′=[t1,t 1,1 ,t 1,2 ,…,t2,t 2,1 ,…,t n ]=[t′1,t′2,…,t′ n ]; Assuming that there are N interpolation points between every two times, there is a corresponding relationship (5): Among them, t i,j The calculation formula is as follows: Where N is the number of interpolation points between every two time points; The linear interpolation method is also used to obtain the path point position coordinates p corresponding to each interpolation moment in each time base sequence. i ′, and the global path Γ p The path point sequence in the interpolation path is taken together as the path point sequence P′=[p1,p 1,1 ,p 1,2 ,…,p2,p 2,1 ,…,p n ]=[p1′,p′2,…,p′ n ]; the specific calculation formula is as follows: p i ′ and p i,j The corresponding relationship is as follows: Then, the path Γ is parameterized according to the original velocity v , use the cubic polynomial interpolation method to calculate the velocity sequence V′=[v1′,v′2,…,v′ n ], the specific calculation method is as follows: First, select the velocity parameterization path Γ v The first four time points t1, t2, t3, t4 are grouped into three, and the cubic polynomial velocity fitting functions f1(t) and f2(t) are calculated for each three time points; each cubic polynomial function has the following form: f(t) = a3t 3 +a2t 2 +a1t+a0; The speed at each time point should be equal to the speed designed in the speed parameterization path, and the first-order derivatives of the two polynomials at the intersection should be equal; based on the above conditions, the following 8 equations are listed: Through the above 8 equations, the 8 coefficients of the two polynomial functions are obtained, and the expected speed at the corresponding time is calculated according to the time of each point in the time base sequence T′; the specific calculation formula is as follows: For the velocity parameterized path Γ v The i-th point (i≥5) of , let the velocity fitting function satisfied by the i-1-th point be f i-3 (t), has been solved, then the time t(t i-1 <t≤t i ) Speed ​​fitting function f over time period i (t): And calculate each moment t′ in the time base series through the above fitting function j Speed: v′ j =f i-2 (t′ j )(t i-1 <t′ j ≤t i ) (12) Combining the above interpolation results of time, position and speed T′, P′ and V′, we can get the interpolated path Γ = [(t′1, p′1, v′1), (t′2, p′2, v′2), …, (t′ n ,p′ n ,v′ n )], which is the final interpolation path.

3. The method according to claim 1, characterized in that: In step 5, based on the interpolation path Γ designed in step 4, the basic control items for autonomous homing of the UAV are designed; the specific calculation is as follows: u b =K p (p′ i -p)+K v (v′ i -v) (13) Among them, K p , K v are position and speed control gains respectively; p i , v i are the position and velocity corresponding to each segment in the time parameterized path; p and v are the position and velocity vectors of the UAV respectively; i represents the number of the time parameterized path where the UAV is currently located.

4. The method according to claim 1, characterized in that: Based on the homing behavior of pigeons, the control correction items for drone homing are designed in step 5. The specific steps are as follows: The path guidance modification designed to simulate the homing behavior of the original pigeon is as follows: S51, Nest attraction item The UAV greedily moves forward to the target point, and the corresponding control correction is: Among them, u a is the acceleration correction of the UAV moving towards the target, p a is the position vector of the target point, p is the position vector of the drone, K a is the acceleration gain of the nest attraction term; S52, predation and obstacle avoidance When the UAV faces obstacles or attackers in the environment, it generates predation and obstacle avoidance acceleration in order to avoid them. The present invention is inspired by the process of pigeons avoiding hawks, and realizes autonomous obstacle avoidance control of the UAV. First, use the sampling method to traverse each grid in the grid map. For the grid with obstacles, record the center coordinates of the grid as the point cloud coordinates of the obstacle; by integrating the point cloud coordinates of the obstacles in the entire map, the point cloud information of the obstacles in the environment is obtained; Using all the obstacle point cloud information, the obstacle avoidance items for the drone are calculated as follows: Among them, u d Represents the total acceleration of the drone due to obstacle avoidance, K s represents the acceleration gain generated by the drone to avoid the sth obstacle point cloud, R c Indicates the safe distance of the drone, p s Represents the position vector of the sth obstacle point cloud; S53, UAV simulated pigeon homing motion status update Based on the principles of S51 and S52 above, the control quantity of the UAV is designed. The total acceleration of the UAV is u as shown below. Substitute it into the UAV control model, that is, update the UAV motion state; in=in b +in a +in d (17).

Images