Unmanned aerial vehicle cluster path on-the-way autonomous obstacle avoidance method based on behavior constraint

Abstract

The invention discloses an unmanned aerial vehicle cluster path on-the-way autonomous obstacle avoidance method based on behavior constraint, and the method comprises the steps: pre-designing a flight path, with a synergistic effect, of each single body of an unmanned aerial vehicle cluster, so that the flight path of an unmanned aerial vehicle avoids most fixed obstacles; meanwhile, the unmanned aerial vehicle clusters share respective space region division information established by a local coordinate system, the unmanned aerial vehicle individuals sense the existence of temporary obstacles by using sensors, and the obstacle avoidance measures are taken according to the space region division information, the rules of the obstacle avoidance measures and a preset flight path; according to the method, the original relative position and the preset flight path in the unmanned aerial vehicle cluster can be rapidly recovered after the temporary obstacle is encountered and obstacle avoidance is completed, which is equivalent to rapid recovery of an ideal state, so that autonomous obstacle avoidance of the unmanned aerial vehicles is realized, and the mutual influence of the unmanned aerial vehicle cluster in the flight process can be reduced in the mode.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and more specifically, to an autonomous obstacle avoidance method for an unmanned aerial vehicle cluster during flight based on behavior constraints. Background Art

[0002] In the current era of rapid technological development, unmanned aerial vehicles (UAVs) have been widely used in many fields such as military and civilian due to their unique advantages. UAVs can be seen actively participating in various scenarios, from military reconnaissance and battlefield attacks to logistics distribution, agricultural plant protection, and film shooting. The emergence of the concept of UAV clusters has brought a qualitative leap to their applications. Multiple UAVs working together can complete more complex and arduous tasks, greatly improving work efficiency and task execution capabilities.

[0003] However, with the continuous expansion of UAV cluster application scenarios and the increase in task complexity, the safety issues of UAVs during flight have become increasingly prominent, and obstacle avoidance technology has become a key bottleneck. In a complex real-world environment, UAVs face various obstacles, including both relatively fixed obstacles such as buildings, mountains, and high-voltage towers in terms of position and shape, and dynamic and changing temporary obstacles such as birds, other UAVs, and suddenly emerging vehicles.

[0004] In an urban environment with dense buildings, UAVs need to accurately avoid fixed obstacles such as high-rise buildings and communication towers during flight. A slight mistake may lead to a collision accident, resulting in damage to the UAV or even causing a safety accident. When conducting surveying and mapping or agricultural operations in mountainous areas with complex terrain and undulating mountains, UAVs not only need to avoid mountain peaks but also deal with birds that may suddenly appear, which poses extremely high requirements for their obstacle avoidance capabilities.

[0005] Most traditional obstacle avoidance technologies are designed for single UAVs and have many deficiencies in dealing with the obstacle avoidance problems of UAV clusters. Moreover, the obstacle avoidance algorithms of single UAVs are difficult to take into account the coordination and mutual influence among multiple UAVs in the cluster. When multiple UAVs face obstacles simultaneously, chaos and collisions are likely to occur. Summary of the Invention

[0006] In order to achieve autonomous obstacle avoidance of an unmanned aerial vehicle (UAV) cluster during the flight path, the present invention provides an autonomous obstacle avoidance method for a UAV cluster based on behavior constraints. The flight paths of each individual of the UAV cluster are pre-designed to have a synergistic effect, so that the flight paths of the UAVs avoid most of the fixed obstacles. At the same time, the UAV cluster shares the information on the spatial region division established in its respective local coordinate systems. Sensors are used to enable individual UAVs to sense the presence of temporary obstacles. Combining the spatial region division information, the rules of obstacle avoidance measures, and the preset flight paths, after encountering a temporary obstacle and completing obstacle avoidance, the UAV quickly resumes its original relative position in the UAV cluster and the preset flight path, which is equivalent to quickly restoring the ideal state, thereby achieving autonomous obstacle avoidance of the UAVs, and this method can reduce the mutual influence of the UAV cluster during flight.

[0007] The technical solution of the present invention is described as follows:

[0008] An autonomous obstacle avoidance method for a UAV cluster based on behavior constraints, which obtains the geographical information of the mission movement area and determines the attribute information of the fixed obstacles in the mission movement area;

[0009] Based on the above geographical information and fixed obstacle information, the flight paths of the UAV cluster are designed;

[0010] Multiple state determination points are set in the flight path, and the state parameters of the UAVs at each state determination point are preset;

[0011] A local coordinate system of a single UAV is established with the UAV as the center, and the environment around the UAV is divided into multiple regions according to the mission requirements and the safety distance. The UAV cluster shares the information on the local spatial region division;

[0012] According to the preset flight path, the regions divided according to the local coordinate system, and the preset behavior constraint rules, the existing flight motion vectors of the UAVs are corrected, and the UAVs complete autonomous obstacle avoidance.

[0013] In the above autonomous obstacle avoidance method for a UAV cluster based on behavior constraints, the preset state parameters of the UAVs include the position coordinates, flight altitude, flight speed, flight acceleration, and attitude angle of the UAVs, and state constraints regarding the state parameters are established for each UAV in the UAV cluster when it is at its respective corresponding state determination point.

[0014] In the above autonomous obstacle avoidance method for a UAV cluster based on behavior constraints, when flying from one state determination point to an adjacent state determination point, the UAV adjusts its own state parameters according to the trajectory planning algorithm, and the trajectory planning algorithm is a calculation method regarding multiple state parameters established by using polynomial interpolation based on a three-dimensional rectangular coordinate system.

[0015] The above-mentioned method for autonomous obstacle avoidance during the path of an unmanned aerial vehicle (UAV) cluster based on behavior constraints establishes a local coordinate system for a single UAV centered on the UAV, and divides the environment around the UAV into

[0016] (1) Prohibited area, an area where the UAV must never enter;

[0017] (2) Warning area, when an obstacle enters the warning area, the UAV takes obstacle avoidance measures. If the obstacle is a UAV of the same UAV cluster or a communicable device / facility, a warning is sent to the other party at the same time;

[0018] (3) Straight flight area, the area where the UAV flies normally;

[0019] (4) Buffer area, a transition area between different areas, used to smooth the switching of state constraints between various areas.

[0020] Furthermore, the local area of the UAV is divided into multiple concentric spherical shell areas. Taking the centroid of the UAV as the origin O and the radii as r1, r2, r3,..., r n (r1 < r2 < r3 < … < r n ), the area between adjacent spherical shells constitutes a divided space; let the buffer width be Δr, then the range of the i-th divided space is [r i , r i + Δr], and the range of the (i + 1)-th divided space is [r i + Δr, r i+1 .

[0021] Furthermore, the UAVs in the same UAV cluster share their respective position information and maintain communication, so that any UAV in the same UAV cluster can obtain the local area division information of the remaining UAVs, construct a set of the same local space areas of the UAV cluster, and any UAV in the UAV cluster implements flight actions according to the set of local space areas.

[0022] In the above-mentioned method for autonomous obstacle avoidance during the path of an unmanned aerial vehicle (UAV) cluster based on behavior constraints, the obstacles include fixed obstacles and temporary obstacles. During the design process of the flight path, the fixed obstacles are sensed and marked according to the satellite and geographical system, and the temporary obstacles are obtained through the sensors carried by the UAV itself.

[0023] In the above-mentioned method for autonomous obstacle avoidance during the path of an unmanned aerial vehicle (UAV) cluster based on behavior constraints, the autonomous obstacle avoidance follows the following behavior constraint rules:

[0024] (1) Away rule, based on the existing flight motion vector of the UAV, an obstacle avoidance correction vector pointing away from the temporary obstacle is added;

[0025] (2) Return rule: When there is no obstacle to avoid within the warning area of the UAV, according to the preset flight path, add a return correction vector to adjust the state parameters of the UAV itself, so that the UAV itself returns to the preset flight path;

[0026] (3) Overall rule: When the difference between the flight motion vector of a UAV and the average flight motion vector of the UAV cluster exceeds the set threshold, or the distance of the UAV's position relative to the centroid of the UAV cluster exceeds the set threshold, based on the existing flight motion vector of the UAV, add an overall correction vector to make the difference between the flight motion vector of the UAV and the average flight motion vector of other UAV clusters within the set threshold, and the distance of the UAV's position relative to the centroid of the UAV cluster within the set threshold.

[0027] Furthermore, the away rule and the return rule are two mutually exclusive settings, and the priority of the overall rule is lower than that of the away rule. During the execution of the return rule, if the flight state of the UAV simultaneously meets the overall rule, the formed return correction vector and the overall correction vector are superimposed on the existing flight motion vector of the UAV at the same time.

[0028] Furthermore, the overall correction vector formed due to the difference between the flight motion vector of the UAV and the average flight motion vector of the UAV cluster exceeding the set threshold is

[0029] The overall correction vector formed due to the distance of the UAV's position relative to the centroid of the UAV cluster exceeding the set threshold is

[0030] For the present invention according to the above solution, its beneficial effects are as follows:

[0031] 1. By means of satellite, GPS and other geographic information systems, collect the geographic data of the mission movement area, and use professional analysis tools to extract the attribute information such as the coordinates, shape, and size of fixed obstacles from these data, providing accurate basic data for subsequent flight path planning, enabling planners to clearly know the obstacle distribution in the mission area. Based on these precise information, a more reasonable and safe flight path can be planned, effectively reducing the possibility of collision between the UAV and fixed obstacles, and initially improving the flight safety. When setting the state determination points, the path nodes can be planned more scientifically according to the positions of fixed obstacles. By setting state determination points near fixed obstacles, the state parameters of the UAV can be strictly restricted, avoiding danger due to approaching obstacles, and further improving the flight safety and stability.

[0032] 2. Select multiple key positions on the planned flight path as state determination points, such as the starting point, turning points, end point, etc. For each state determination point, specify in detail the state parameters of the UAV, such as position coordinates, flight altitude, flight speed, flight acceleration, and attitude angles, and set reasonable value ranges for these parameters as state constraints. Divide the entire flight mission into multiple controllable subtask stages, making the flight process of the UAV more controllable and stable. Based on the state constraints of the state determination points, the trajectory planning algorithm can calculate the flight trajectory between adjacent state determination points more accurately, enabling the UAV to obtain smooth flight trajectory information and further improving the flight smoothness.

[0033] 3. Construct a local coordinate system with the current position of each UAV as the origin, determine the axis directions according to the mission requirements and flight directions, and divide the environment around the UAV into a prohibited area, a warning area, a straight flight area, and a buffer area, which standardizes the relative positions and flight behaviors of the UAVs in the UAV cluster and greatly reduces the collision risk between UAVs. By dividing the areas, the UAVs can perceive potential dangers in advance and make corresponding adjustments, enhancing the safety and coordination of the cluster flight.

[0034] 4. When the UAV changes its flight state due to obstacle avoidance, the return rule in the obstacle avoidance rules enables it to quickly return to a flight trajectory close to the original state according to the preset parameters of the state determination points, ensuring the coherence and accuracy of the flight mission. If a temporary obstacle is encountered near a certain state determination point, after the UAV executes the avoidance rule and moves away, the return rule will make it adjust its flight state towards that state determination point to ensure that the state constraints of that point can be finally met, so that the task execution accuracy is not affected too much, improving the controllability of the entire flight process.

[0035] 5. The local coordinate system and area division provide a spatial judgment basis for the obstacle avoidance rules, and the obstacle avoidance rules further improve its functions. In the local coordinate system, the UAV can more accurately judge the relative position relationships between itself and other UAVs and obstacles. At the same time, the setting of the buffer area cooperates with the obstacle avoidance rules, enabling the UAV to smoothly adjust its state when switching between different areas and avoiding flight instability caused by rule switching, enhancing the safety and coordination of the UAV cluster flight. Brief Description of the Drawings

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0037] Figure 1 It is a schematic flowchart structure diagram of the present invention.

[0038] Figure 2 is Figure 1 a partially enlarged view of

[0039] Figure 3 is Figure 1 a partially enlarged view of Specific embodiments

[0040] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0041] An autonomous obstacle avoidance method for an unmanned aerial vehicle (UAV) cluster during path based on behavior constraints. The specific obstacle avoidance design is as Figure 1 , Figure 2 , Figure 3 shown. Obtain the geographical information of the mission movement area, and determine the attribute information of the fixed obstacles in the mission movement area; based on the above geographical information and fixed obstacle information, design the flight path of the UAV cluster; set multiple state determination points in the flight path, and preset the state parameters of the UAVs at each state determination point; establish a local coordinate system of a single UAV with the UAV as the center, and divide the environment around the UAV into multiple regions according to the mission requirements and safety distance. The UAV cluster shares the local space region division information; according to the preset flight path, the regions divided according to the local coordinate system, and the preset behavior constraint rules, correct the existing flight motion vectors of the UAVs, and the UAVs complete autonomous obstacle avoidance.

[0042] Before the start of the UAV cluster mission, determine the flight path of the UAV cluster in advance according to the mission content. Obtain the geographical information of the mission movement area, obtain the geographical information within the mission geographical area through a geographical information system such as a satellite or GPS, and determine the attribute information of the fixed obstacles in the mission movement area through the geographical information, including the coordinates, shapes, sizes, etc. of the fixed obstacles, which is used to help plan a safe flight path.

[0043] The fixed obstacles are determined by the mission of the UAV cluster. All items that the UAVs must avoid in a specific area belong to fixed obstacles (sometimes certain geographical areas are set as fixed obstacles, and some areas where flight is not suitable can also be considered fixed obstacles).

[0044] Based on the above geographical information and fixed obstacle information, design the flight path of the UAV cluster. The flight mission of the UAVs will have an initial point and a termination point. The flight path is a path from the initial point to the termination point, while avoiding all known fixed obstacles.

[0045] The planning of the flight path is a complex process that requires comprehensive consideration of various factors. This process can refer to the path planning of transmission lines. Generally, the flight path of the UAV needs to be as short as possible, making full use of the flight flexibility to improve the mission efficiency. At the same time, the flight path should avoid known fixed obstacles to ensure the safety of the UAV. In addition, the flight path should also meet the performance limitations of the UAV, such as the maximum speed, maximum acceleration, etc. In practical applications, the planning of the flight path usually requires the combination of multiple algorithms and technologies.

[0046] After the flight path is confirmed, in order to meet the requirements of improving the flexibility, safety and mission completion efficiency of the flight path, multiple state determination points are pre-set in the determined flight path, a set of state determination points is established, and the state of the UAV is set at each state determination point. These state determination points are key positions on the flight path, such as the starting point, the ending point, the turning point, the inspection point, etc. At each state determination point, the UAV needs to meet the pre-set state parameters, which include the position coordinates, flight altitude, flight speed, flight acceleration and attitude angle of the UAV.

[0047] State constraints are set for each state determination point, that is, the state parameters of the UAV meet the range within the pre-set state parameters. The state constraints of the UAV ensure specific flight conditions of the UAV. For example, the flight altitude constraint ensures that the UAV flies within a safe altitude range to avoid collisions with ground obstacles or other aircraft. At the same time, the flight altitude also affects the signal transmission quality and energy consumption of the UAV. Therefore, it needs to be reasonably optimized on the premise of meeting the mission requirements; the flight speed constraint can prevent the UAV from speeding in certain areas, thus avoiding the inability to avoid obstacles in time due to excessive speed. It can also appropriately increase the flight speed in relatively open geographical environments, or reduce the speed in complex environments or when approaching the target area to ensure safety and precise operation; the heading angle constraint can ensure that the UAV flies in a specific direction and avoid deviating from the predetermined path; the attitude angle constraint, such as the attitude angle constraints at the starting point and the ending point, can ensure that the UAV remains stable during the take-off and landing phases and can fly along the predetermined flight path. By reasonably setting state constraints, the adaptability and safety of the UAV in complex environments can be improved.

[0048] The state determination points and state constraints of the drone are equivalent to dividing the flight mission of the drone, forming multiple consecutive flight mission branches, thereby making the flight mission of the drone more controllable and stable. The drone can fly precisely along the preset trajectory, greatly improving the accuracy of mission execution. In addition, once the drone detects an anomaly during flight, such as a sensor failure or strong airflow interference, it can quickly adjust its flight state according to the state constraints of the adjacent state determination points, hover within a safe range or take emergency landing measures to avoid equipment damage or safety accidents caused by loss of control. Each drone flies according to the state determination points and state constraints on its respective flight path, and can maintain an accurate spacing and relative position, providing strong guarantee for the coordinated flight of multiple drones.

[0049] The coordinates of the drone represent the position of the drone in space through a three-dimensional rectangular coordinate system. Here, the coordinate values are all converted from the longitude and latitude information provided by the satellite and the drone system through specific coordinate conversion formulas. Once the drone has coordinates, its position coordinates and flight altitude can be obtained through the coordinates. The speed and acceleration of the drone are measured by sensors installed on the drone, such as airspeed tubes, accelerometers, gyroscopes, etc. The sensed values of the sensors are calculated and converted into the flight speed, flight acceleration, etc. of the drone. By monitoring the feedback of the speed and acceleration of these sensors, the power system of the drone is adjusted to control the flight speed and flight acceleration of the drone to meet the state constraints. The attitude angles include pitch angle, roll angle and yaw angle.

[0050] Setting multiple state determination points is to further improve the planning accuracy and reliability of the flight path of the drone swarm, ensuring that the drone can accurately reach the preset state at each state determination point. Further, in order to achieve this effect, the drone is set with a trajectory planning algorithm. The trajectory planning algorithm is a calculation method based on polynomial interpolation. Given two adjacent state determination points and the state constraints of the state determination points, the flight trajectory function between the two state determination points can be represented by a polynomial function.

[0051] The flight trajectory function x(t) of the drone in the x direction is

[0052] The flight trajectory function y(t) of the drone in the y direction is

[0053] The flight trajectory function z(t) of the drone in the z direction is

[0054] Where t is time, a0, a1, a2,..., a m 、b1、b2、...、b n 、c1、c2、...、cp are the polynomial coefficients of each flight trajectory function, and these polynomial coefficients can be determined by P i , P i+1 the position coordinates, flight speed, flight acceleration, flight altitude and other state constraints of two adjacent state determination points (such as learning models, simulation and other methods can be used). In application, similar polynomial interpolation methods can also be used for planning. In this way, it can be ensured that the UAV smoothly adjusts its state during the flight from one state determination point to the next state determination point to meet the preset requirements at each state determination point, so as to achieve the precise planning and reliable execution of the entire flight path.

[0055] Establish a local coordinate system for each UAV. The origin of the local coordinate system is set as the current position of the UAV, and the coordinate axis directions are defined according to the mission requirements and the flight direction of the UAV.

[0056] The local coordinate system of the UAV is x = x0 + x′e x , y = y0 + y′e y , z = z0 + z′e z , where (x0, y0, z0) are the coordinates of the three-dimensional rectangular coordinate system of the UAV (established through the satellite and the UAV system), (x′, y′, z′) are the coordinates of the local coordinate system of the UAV, and (e x , e y , e z ) are the coordinate axis directions of the local coordinate system.

[0057] According to the local coordinate systems established with each UAV as the center, and according to the mission requirements and safety distances, the environment around the UAVs is divided into multiple regions, which are respectively

[0058] (1) Prohibited area, the area that the UAV must never enter. For the entire UAV cluster, the area closest to itself for all UAVs is the prohibited area, and no UAV can enter the prohibited area of other UAVs;

[0059] (2) Warning Area: To a certain extent, it is an area where drones are prohibited from flying into. However, there is a possibility of accidents. Under the condition of preset rules with a high design priority level, drones can enter this area (for example, in order to avoid collisions, damages and other measures with a high priority level, a drone can enter the warning area of other drones. The no-entry area is set because it is very difficult for a drone to complete obstacle avoidance actions in this spatial area and the probability of collision is very high, so entry is prohibited). When an obstacle enters this area of a certain drone, a warning needs to be issued (if the obstacle is a drone in the same drone cluster or a device / facility that can communicate, etc.) and obstacle avoidance measures need to be taken. For any drone in the drone cluster, if a drone enters its own warning area, a warning will be sent to the drone entering the warning area and preparations will be made to take obstacle avoidance measures. In this area, the flight speed, flight acceleration and attitude angle of all drones are set with restrictions and strictly controlled within a certain range. The drone flying into this area needs to quickly react and adjust its own state to appropriately leave the adjustment space and time for the drone entering the warning area to make preparations for avoidance and avoid a collision between the two drones;

[0060] (3) Straight-Flight Area: An area where drones can fly normally. Drones fly according to the preset state parameters;

[0061] (4) Buffer Zone: A transition area between different regions, used to smooth the switching of state constraints between regions. Usually, when entering this area, a drone needs to adjust its own state, including flight acceleration, flight speed and attitude angle, to leave as much adjustment space and time as possible.

[0062] Generally, the local area of a drone is divided into multiple concentric spherical shell areas. With the centroid of the drone as the origin O and radii r1, r2, r3,..., r n (r1 < r2 < r3 < … < r n ), the area between adjacent spherical shells constitutes a divided space. To prevent the drone from becoming unstable when switching between different divided spaces, a buffer zone is set between adjacent divided spaces. Let the buffer zone width be Δr, then the range of the i-th divided space is [r i , r i + Δr], and the range of the (i + 1)-th divided space is [r i + Δr, r i+1 .

[0063] The UAV swarm shares location information with each other and maintains communication. Therefore, any UAV in the same UAV swarm can obtain the local area division information of the other UAVs and construct a set of the same local space areas of the UAV swarm. According to the local area division situation, the flight of the UAV follows the laws of the area. For example, no UAV will fly into the prohibited area of other UAVs. When a UAV enters the warning area due to uncontrollable factors such as obstacle avoidance, the UAV that enters the warning area will send an alarm and take corresponding obstacle avoidance measures.

[0064] During the flight of the UAV along the preset flight path, in addition to fixed obstacles, there will also be temporary obstacles (the division of fixed obstacles and temporary obstacles is basically based on the flight path design. Fixed obstacles are obstacles for which the perception and avoidance actions have been designed in the flight path, and temporary obstacles are obstacles that appear on the flight path after the flight path has been designed, including obstacles that have been sensed through satellite and geographical systems). Fixed obstacles refer to obstacles whose positions and shapes remain unchanged during the flight mission, and temporary obstacles refer to obstacles that suddenly appear on the predetermined flight path of the UAV swarm during flight and have not been obtained through satellite and other geographical systems before the mission starts. Therefore, solid obstacles are pre-sensed and marked through satellites and geographical systems, and the perception and avoidance actions have basically been completed when designing the flight path, while temporary obstacles are obtained through real-time sensing by the sensors carried by the UAV itself.

[0065] The sensors carried by the UAV itself are one or more of lidar, cameras, radars, ultrasonic sensors, etc. Lidar determines the distance and position of obstacles by emitting laser beams and measuring the time of the reflected light, and can provide high-precision three-dimensional point cloud data, which is suitable for detecting and identifying obstacles in complex environments; cameras detect obstacles through image recognition technology and use image processing algorithms such as edge detection and feature extraction to identify and distinguish the shapes and positions of obstacles; millimeter-wave radars determine the relative speed of obstacles by emitting electromagnetic waves in the millimeter-wave frequency band and measuring the frequency difference between the transmitted signal and the received signal, and then determine the distance of temporary obstacles, which is suitable for long-distance detection and perception of obstacles moving at high speeds, and has good detection performance under adverse weather conditions (such as fog, rain, snow); ultrasonic sensors determine the distance of obstacles by emitting ultrasonic waves and measuring the time of the reflected wave, which is suitable for detecting close-range obstacles and has the characteristics of low cost and easy installation.

[0066] The unmanned aerial vehicle (UAV) relies on the sensors carried by itself to obtain information about the surrounding environment, including the positions and flight states of other UAVs, as well as the position information of obstacles. When a temporary obstacle (including no-fly zones, warning zones formed by other UAVs, etc., that is, the understanding of obstacles should be the set of areas where the UAV cannot come into contact with them, including physical obstacles that can be detected by sensors and artificially set no-fly / do-not-recommend-flying areas obtained through communication) enters the warning zone of its locally divided space, the UAV takes obstacle avoidance measures. The obstacle avoidance measures will follow the following behavior constraint rules.

[0067] (1) Away rule: The biggest change in the obstacle avoidance measures is to change the flight motion vector of the UAV. The UAV avoiding the obstacle will, based on the principle of staying away, add an obstacle avoidance correction vector (including motion direction and speed) pointing away from the temporary obstacle to the existing flight motion vector. This obstacle avoidance correction vector is the reverse vector of the motion vector of the temporary obstacle entering the warning zone of the UAV. Regardless of whether the entry state of the temporary obstacle into the warning zone is continuous or discontinuous, the motion vector of the centroid or center of the continuous part is calculated as the motion vector of the temporary obstacle entering the warning zone of the UAV, and the reverse vector is taken as the obstacle avoidance correction vector.

[0068] (2) Return rule: When there are no obstacles to avoid within the warning zone of the UAV, according to the preset flight path, a return correction vector is added to adjust the state parameters of the UAV itself, so that it returns to the original flight path. (Auxiliary trajectory planning algorithm)

[0069] (3) Global rule: When the difference between the flight motion vector of a UAV and the average flight motion vector of the UAV cluster exceeds the set threshold, or the distance of the UAV's position relative to the centroid of the UAV cluster exceeds the set threshold, an overall correction vector is added to the existing flight motion vector (if it is a problem with the flight motion vector, the overall correction vector is where is the overall correction vector of UAV i, is the flight motion vector of UAV i, k1 is an adjustment coefficient used to adjust the speed of change of the flight motion vector, is the average flight motion vector of the UAV cluster; if it is a distance problem, the overall correction vector is is the overall correction vector of UAV i, is the flight motion vector of UAV i, k2 is an adjustment coefficient used to adjust the speed of approaching the centroid of the UAV cluster, is the position of the centroid of the UAV cluster, (which is the location of the drone i), such that the difference between the flight motion vector of the drone and the average flight motion vector of other drone clusters is within a set threshold range, and the distance of the location of the drone from the centroid of the drone cluster is within a set threshold range (if both problems exist, then the two overall correction vectors are superimposed).

[0070] For the above three rules, the away rule and the regression rule are two mutually exclusive settings and do not interfere with each other per se. However, the priority of the overall rule is lower than that of the away rule, that is, the overall rule is not considered during the execution of the away rule. During the execution of the regression rule, if the flight state of the drone simultaneously meets the overall rule, then the formed regression correction vector and the overall correction vector are superimposed on the existing flight motion vector of the drone at the same time.

[0071] The above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An autonomous obstacle avoidance method for an unmanned aerial vehicle (UAV) cluster during path navigation based on behavior constraints, characterized in that, Obtain the geographical information of the mission movement area and determine the attribute information of the fixed obstacles within the mission movement area; Based on the above geographical information and fixed obstacle information, design the flight path of the UAV swarm; Set multiple state determination points in the flight path and preset the state parameters of the UAVs at each state determination point; Establish a local coordinate system for a single UAV centered on the UAV, and divide the environment around the UAV into multiple regions according to the mission requirements and safety distance. The UAV swarm shares the local space region division information; According to the preset flight path, the regions divided according to the local coordinate system, and the preset behavior constraint rules, correct the existing flight motion vectors of the UAVs, and the UAVs complete autonomous obstacle avoidance.

2. The method for autonomous obstacle avoidance during the path of a UAV cluster based on behavior constraints according to claim 1, wherein The preset state parameters of the UAVs include the position coordinates, flight altitude, flight speed, flight acceleration, and attitude angles of the UAVs, and establish state constraints on the state parameters when each UAV in the UAV swarm is at its corresponding state determination point.

3. The method for autonomous obstacle avoidance during the path of a UAV cluster based on behavior constraints according to claim 1, wherein When flying from one state determination point to an adjacent state determination point, the UAV adjusts its own state parameters according to the trajectory planning algorithm. The trajectory planning algorithm is a calculation method based on the three-dimensional rectangular coordinate system using polynomial interpolation for multiple state parameters.

4. The autonomous obstacle avoidance method for an unmanned aerial vehicle cluster during path based on behavior constraints according to claim 1, characterized in that Establish a local coordinate system for a single UAV centered on the UAV, and divide the environment around the UAV into (1) Prohibited area, an area where the UAV must never enter; (2) Warning area, when there is an obstacle entering the warning area, the UAV takes obstacle avoidance measures. If the obstacle is a UAV of the same UAV swarm or a communicable device / facility, a warning is sent to each other at the same time; (3) Straight flight area, the area where the UAV flies normally; (4) Buffer area, a transition area between different regions, used to smooth the switching of state constraints between regions.

5. The method for autonomous obstacle avoidance during the path of an unmanned aerial vehicle cluster based on behavior constraints according to claim 4, characterized in that, The local area of the drone is divided into multiple concentric spherical shell areas. Taking the centroid of the drone as the origin O, with radii r1, r2, r3,..., r n (r1 < r2 < r3 < … < r n ), the area between adjacent spherical shells constitutes a divided space; let the buffer width be Δr, then the range of the i-th divided space is [r i , r i + Δr], and the range of the (i + 1)-th divided space is [r i + Δr, r i+1 .

6. The autonomous obstacle avoidance method for an unmanned aerial vehicle cluster during path based on behavior constraints according to claim 4, characterized in that, The UAVs in the same UAV swarm share their respective position information and maintain communication, so that any UAV in the same UAV swarm can obtain the local area division information of the remaining UAVs, construct a set of the same local space regions of the UAV swarm, and any UAV in the UAV swarm implements flight actions according to the set of local space regions.

7. The autonomous obstacle avoidance method for an unmanned aerial vehicle (UAV) cluster during path based on behavior constraints according to claim 1, characterized in that The obstacles include fixed obstacles and temporary obstacles. During the design process of the flight path, the fixed obstacles are sensed and marked according to the satellite and geographical system. The temporary obstacles are obtained through the sensors carried by the UAV itself.

8. The method for autonomous obstacle avoidance during the path of an unmanned aerial vehicle cluster based on behavior constraints according to claim 1, wherein The autonomous obstacle avoidance follows the following behavior constraint rules: (1) Away rule, based on the existing flight motion vector of the UAV, add an obstacle avoidance correction vector pointing away from the temporary obstacle; (2) Regression rule, when there is no obstacle to avoid within the warning area of the UAV, according to the preset flight path, add a regression correction vector to adjust the state parameters of the UAV itself, so that the UAV itself returns to the preset flight path. (3) Overall rule: When the difference between the flight motion vector of a drone and the average flight motion vector of the drone cluster exceeds the set threshold, or the distance of the drone's position relative to the centroid of the drone cluster exceeds the set threshold, an overall correction vector is added to the existing flight motion vector of the drone, so that the difference between the flight motion vector of the drone and the average flight motion vector of other drone clusters is within the set threshold, and the distance of the drone's position relative to the centroid of the drone cluster is within the set threshold.

9. The method for autonomous obstacle avoidance during the path of an unmanned aerial vehicle cluster based on behavior constraints according to claim 8, wherein The away rule and the return rule are two mutually exclusive settings. The priority of the overall rule is lower than that of the away rule. During the execution of the return rule, if the flight state of the drone simultaneously meets the overall rule, the formed return correction vector and the overall correction vector are superimposed on the existing flight motion vector of the drone at the same time.

10. The method for autonomous obstacle avoidance during the path of an unmanned aerial vehicle cluster based on behavior constraints according to claim 8, wherein The overall correction vector formed by the difference between the flight motion vector of the drone and the average flight motion vector of the drone cluster exceeding the set threshold is The overall correction vector formed because the distance of the UAV's position relative to the centroid of the UAV cluster exceeds the set threshold is

Images