Unmanned aerial vehicle bidirectional collaborative recovery method and recovery system based on multi-source positioning

Abstract

The invention discloses an unmanned aerial vehicle bidirectional collaborative recovery method and recovery system based on multi-source positioning, and relates to the technical field of unmanned aerial vehicle positioning and control. Multi-source sensing systems such as a laser radar and a visual sensor are integrated into a recovery system, so that the recovery system with external sensing capability and bidirectional interaction capability is constructed. Comprising the steps that the UAV sends a recovery request to a ground end; the ground end sends a ground end position and recovery channel parameters to the unmanned aerial vehicle; planning an expected recovery track, and flying to a recovery channel based on the planned track; and the ground end carries out real-time positioning on the unmanned aerial vehicle through the multi-source positioning system and sends a track correction instruction to the unmanned aerial vehicle according to a positioning result, so that the unmanned aerial vehicle flies according to the track correction instruction, and meanwhile, the recovery system carries out attitude adjustment according to a real-time positioning result. The unmanned aerial vehicle can autonomously plan the flight path, and the recovery system can guide the unmanned aerial vehicle to realize accurate docking, so that the recovery success rate and the attitude matching precision are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle positioning and control, and in particular to a unmanned aerial vehicle bidirectional collaborative recovery system and a recovery method based on multi-source positioning. Background Art

[0002] In recent years, small and medium-sized UAVs have been widely used in military reconnaissance, logistics, emergency rescue, environmental monitoring and other fields due to their advantages of low cost, high mobility, flexible deployment and strong mission adaptability. In order to improve their operating efficiency and reuse capabilities, efficient, safe and controllable UAV recovery technology after the mission is completed has gradually become a research hotspot.

[0003] At present, the mainstream recovery methods of fixed-wing UAVs include: taxiing landing, parachute buffering, net bag collision capture, horizontal rope interception, skyhook recovery, etc. The above recovery methods usually adopt a recovery strategy in which the UAV is the active party and the recovery system is the passive party, that is, the recovery system passively receives the UAV, and the UAV is the only active control party and undertakes all navigation, attitude adjustment and docking tasks. Although this one-way control mode is simple in structure, it also has several problems.

[0004] First, the drone's own positioning system mostly relies on positioning devices such as GPS and RTK. Due to positioning errors, drift accumulation and environmental interference, it is difficult to maintain high-precision trajectory control in actual missions; and the traditional recovery method has high requirements for the recovery path and attitude accuracy, which often leads to docking failures or structural damage. Secondly, the existing recovery system lacks perception and active adjustment capabilities, and cannot quickly adapt according to the current state of the drone and environmental changes, resulting in poor dynamic responsiveness and insufficient robustness of the system. In addition, the closed-loop feedback mechanism of the one-way control system is not sound. If the drone deviates from the predetermined recovery channel, it often needs to retake off and restart the process, resulting in energy waste and mission delays. In severe cases, there is even a risk of recovery failure or collision. Summary of the invention

[0005] In view of the above problems, the present invention proposes a two-way collaborative recovery method and system for unmanned aerial vehicles (UAVs) based on multi-source positioning. By integrating multi-source perception systems such as lidar and vision sensors into the recovery system, a recovery system with external perception and two-way interaction capabilities is constructed. Through the multi-source fusion positioning method, the recovery system can obtain the high-precision real-time pose of the UAV. At the same time, the recovery system can send trajectory correction commands to the UAV flight control system according to the positioning error, thereby realizing a two-way collaborative control mechanism for the external system to assist the flight control in trajectory correction, effectively alleviating the accuracy pressure of the UAV navigation system and improving the recovery success rate. Generally speaking, the present invention can solve the problems existing in the prior art, such as one-way control, inaccurate positioning, weak attitude matching ability, and poor environmental adaptability, through multi-sensor data fusion, multi-source perception positioning, and a UAV collaborative recovery scheme supporting two-way interaction.

[0006] The technical solution of the present invention is as follows: It includes the following steps:

[0007] S1: The UAV sends a recovery request and the current flight parameters of the UAV to the ground end; the flight parameters include the longitude, latitude, altitude, speed, heading, remaining flight time, and flight attitude of the UAV.

[0008] S2: After receiving the recovery request from the UAV, the ground end sends the ground end position and recovery channel parameters to the UAV.

[0009] S3: The UAV plans an expected recovery trajectory based on the ground end position, recovery channel parameters, and the current position and heading of the UAV, and flies towards the recovery channel based on the planned trajectory.

[0010] S4: When the UAV enters the recovery channel, it sends a docking request to the ground end; the ground end real-time locates the UAV through the multi-source positioning system, and sends a trajectory correction command to the UAV according to the positioning result, so that the UAV flies according to the trajectory correction command. At the same time, the recovery system adjusts the attitude according to the real-time positioning result to meet the docking requirements with the UAV.

[0011] S5: When the UAV reaches the recovery distance threshold, the ground end judges the recovery conditions; if the recovery conditions are not met, the UAV performs a re-flight and repeats the recovery process; if the recovery conditions are met, the UAV continues to maintain its heading until the recovery is completed.

[0012] The multi-source positioning system includes a plurality of perception sources arranged around the recovery system, and the perception sources include vision cameras and lidar sensors.

[0013] After the UAV enters the recovery channel, the ground end real-time locates the UAV through the multi-source positioning system, which specifically includes the following steps:

[0014] Synchronize the point cloud data and image data obtained from each perception source in terms of time;

[0015] Perform coordinate transformation on the point cloud data to obtain the projected point cloud in the image coordinate system. Denote a point detected by the lidar in the environment The coordinate in the lidar coordinate system is and its coordinate in the visual camera coordinate system is and there is where and are the rotation matrix and translation matrix from the lidar coordinate system to the visual camera coordinate system respectively. The coordinate of the projected point cloud in the image coordinate system can be expressed as:

[0016] In the formula, represents the horizontal direction in the image, represents the vertical direction in the image, and K represents the transformation matrix;

[0017] Based on the deep learning model, perform target recognition on the unmanned aerial vehicle in the image data, and screen out the pixel range of the target unmanned aerial vehicle in the image, denoted as In the formula, , are the horizontal and vertical coordinate positions of the target in the image, , are the horizontal coordinate area and vertical coordinate area occupied by the target respectively. Screen the projected point cloud coordinates obtained, and the screening conditions are shown in the following formula:

[0018] ;

[0019] Fuse and calculate the processed data from different perception sources. Denote as the representation of the target point cloud data generated by the lidar using a four-dimensional coordinate vector. Set the following formula:

[0020] , which is the conversion relationship between the lidar coordinate system and the lidar coordinate system. Among them, and are the rotation relationship and translation relationship respectively. The coordinate vector of each lidar point cloud in the lidar coordinate system can be expressed as: , is the number of point clouds generated by the Xth lidar;

[0021] Obtain the spliced point cloud data in the lidar coordinate system For each point after point cloud stitching is completed define its neighborhood:

[0022] where is the set of point clouds after stitching, is the neighborhood radius;

[0023] If , is the quantity threshold, then the point is a core point, and all density-reachable points are expanded into a cluster to obtain several clustering clusters Let the cluster with the largest number of points be the target UAV cluster, that is , then the average position of this cluster is the target position: .

[0024] The recovery system includes a blocking device 100, a robot system 200, a ground rail 300, a multi-source positioning system 400, and a UAV 500; the blocking device 100 is installed at the power execution end of the robot system 200 to achieve blocking and deceleration of the UAV; the robot system 200 has the ability of spatial multi-degree-of-freedom attitude adjustment to support the active adjustment of the blocking device according to the UAV pose state; the ground rail 300 is installed below the robot system to support the robot system to move along the track, so as to match the dynamic trajectory of the UAV; the multi-source positioning system 400 is arranged on both sides of the robot system 200 and the ground rail 300 to achieve high-precision real-time positioning of the UAV within a limited range.

[0025] The robot system 200 is a six-axis robot, starting from its root are axis, axis, axis, axis, axis, axis. The above rotation axes enable the robot flange position to have the ability of spatial six-degree-of-freedom attitude adjustment. By arranging the blocking device at the execution end of the six-axis robot, the active capture of the UAV is realized; where the default axis rotation angle is zero;

[0026] During the UAV recovery process, the robot rotates the axis to make the recovery plane parallel to the flight path plane ; rotates the axis to make the blocking device adapt to the UAV roll angle; the translation of the robot on the ground rail can make the recovery plane coincide with the flight path plane;

[0027] In the recovery plane the robot system adjusts through Axis Axis The axis causes the arresting device to reach the predetermined recovery point coordinate position at the pitch angle , let Axis Axis The rotation angles of the axis are respectively , , ;

[0028] Establish a coordinate system at the robot base , Axis Axis Separate coordinate systems are established at the axis , , , Based on the adjacent link coordinate system transformation matrix , Calculate the homogeneous transformation matrix of the coordinate system relative to the coordinate system : :

[0029] ;

[0030] The coordinate system at the A5 axis relative to the base coordinate system is represented as:

[0031] ;

[0032] where and represent the position of the coordinate system in the recovery plane:

[0033] ;

[0034] ;

[0035] Thus, the rotation angles of the axis, axis, axis , , are respectively:

[0036] ;

[0037] ;

[0038] ;

[0039] ​Finally, obtain axis, axis, axis rotation angle , , , and input axis rotation angle and axis rotation angle to control the robot system.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] 1. By introducing a two-way collaborative control mechanism, the drone can not only autonomously plan its flight path according to the position information of the ground recovery system, but also the recovery system can guide the drone to achieve precise docking in real time through sensing feedback, thereby improving the recovery success rate and attitude matching accuracy.

[0042] 2. The multi-source positioning system uses lidar and vision sensors for integrated positioning, which greatly improves the positioning accuracy and anti-interference ability compared with the method that only relies on GPS.

[0043] 3. The two-way collaborative recovery method based on multi-source positioning can achieve positioning error correction and dynamic adjustment of the recovery attitude, enabling the recovery system to have a deviation correction feedback function during the recovery process of the drone, effectively solving problems such as drone trajectory deviation and attitude disturbance, and improving the adaptability of the recovery system in complex scenarios; at the same time, this method considers the ability of the drone to take off again after recovery failure. By setting a recovery fault tolerance judgment logic, it realizes the takeoff again and recovery retry of the drone after docking failure, thus ensuring the system stability and mission safety.

[0044] In summary, the present invention proposes a two-way collaborative recovery method for drones and recovery systems that integrates perception fusion, flight path correction, and attitude adjustment, which is applicable to various autonomous recovery scenarios of drones such as vehicle-mounted, ship-mounted, and fixed platforms. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a schematic diagram of a drone recovery system based on multi-source positioning,

[0046] Figure 2a is a schematic diagram of the arresting device,

[0047] Figure 2b is a schematic diagram of the structure of the arresting block and electromagnetic lock,

[0048] Figure 2c is a schematic diagram of the structure of the front limit and guiding device,

[0049] Figure 2d is a schematic diagram of the buffer and arresting cable layout,

[0050] Figure 2eSchematic diagram of the UAV interception process

[0051] Figure 3 Schematic diagram for the definition description of each axis of the robot system

[0052] Figure 4 Schematic diagram of the coordinate system in the robot control method

[0053] Figure 5 Schematic diagram of the recovery position in the robot control method

[0054] Figure 6 Flowchart of the two-way collaborative recovery method for UAVs based on multi-source positioning

[0055] Figure 7 Schematic diagram of the multi-source positioning method Detailed implementation manners

[0056] To clearly illustrate the technical features of the present invention, the present invention will be elaborated in detail below through specific implementation manners and in conjunction with its accompanying drawings.

[0057] Figure 1 Shows the state of a two-way collaborative UAV recovery system based on multi-source positioning proposed by the present invention during UAV recovery operations. The recovery system includes a blocking device 100, a robot system 200, a ground rail 300, a multi-source positioning system 400, and a UAV 500.

[0058] The blocking device 100 is installed at the robot execution end with six-degree-of-freedom attitude adjustment ability and can actively adjust according to the position and attitude information of the UAV 500; the ground rail 300 is installed below the robot system 200 and is used to support the robot system 200 to move along the track, so as to match the dynamic trajectory of the UAV 500.

[0059] Both the UAV 500 and the recovery system are equipped with independent positioning and communication devices, and the two can establish a communication connection to achieve two-way positioning and data exchange; in addition to its own positioning system, the recovery system is also equipped with a multi-source positioning system 400 for more accurate positioning of the UAV 500. The multi-source positioning system 400 is arranged on both sides of the robot 200 system and can perform higher-precision real-time positioning of the UAV 500 within a limited range.

[0060] The above has made an overall description of the recovery system proposed by the present invention, elaborating the overall layout and functions of the blocking device 100, the robot system 200, the ground rail 300, the multi-source positioning system 400, and the UAV 500. The composition and principles of the above-mentioned parts will be described in detail below.

[0061] See Figure 2a 、 2b, 2c, 2d, 2e, the arresting device 100 includes an arresting frame 101, an arresting block 102, a slide rail 103, a front limit 104, a rear limit 105, a buffer 108, and an arresting cable 108c;

[0062] The arresting frame 101 is fixedly installed at the output end of the multi-axis robot. A pair of slide rails 103 are arranged in parallel and are both fixedly installed on the arresting frame 101. The arresting block 102 is slidably installed on each slide rail 103. The front limit 104 and the rear limit 105 are fixedly installed at the head and tail ends of the slide rail 103 respectively. A number of pulleys are installed at the front part and below of the slide rail 103. The cavity 108a of the buffer 108 is fixedly installed below the slide rail 103, and a pulley is also installed on the pull rod 108b of the buffer 108. One end of the arresting cable 108c is fixedly connected to the arresting frame 101, and after sequentially passing around each pulley, the other end is fixedly connected to the arresting block 102. A card slot adapted to the wing of the fixed-wing unmanned aerial vehicle 500 is provided on one side of the arresting block 102.

[0063] Regarding the specific structure of the arresting block:

[0064] The arresting block 102 includes a horseshoe plate 102a, impact sponge 102b, a trolley 102c, and a limit pulley 102d. The limit pulley 102d is rotatably connected to the bottom of the trolley 102c and is accommodated in the slide rail 103. The horseshoe plate 102a is fixedly installed on the trolley 102c. The impact sponge 102b is fixedly installed in the horseshoe plate 102a, and the card slot is provided in the impact sponge 102b.

[0065] Further, the arresting block 102 further includes a braking lug 102e;

[0066] As a supporting part, a number of electromagnetic locks 107 are fixedly installed on the slide rail 103 at equal intervals. The braking lug 102e is located in the area directly above the electromagnetic lock 107 and is used to lock the position of the arresting block 102 when the unmanned aerial vehicle is arrested.

[0067] Regarding the specific structure of the arresting frame:

[0068] The main body of the arresting frame 101 adopts a frame structure. The front side is connected to a guiding device 106 in an arc shape, and there are two cross beams extending backward, on which the slide rail 103 is installed;

[0069] The guiding device 106 includes a roller base 106a, rollers 106b, and an I-beam 106c. The I-beam 106c has an arc structure and is fixedly connected to the arresting frame 101. The roller bases 106a are arranged in series thereon. Each roller base 106a has the same structure, and a flexible roller 106b is installed thereon to guide the UAV smoothly into the arresting frame during the UAV recovery process.

[0070] The guiding device 106 has four and is arranged symmetrically in pairs. The space between the I-beams 106c of the same pair of guiding devices 106 is the wing receiving space, and the wing receiving space gradually narrows and then accesses the area where the arresting block 102 is located. During the process of recovering the UAV 500, the guiding device 106 is used for guiding, so that the UAV can be smoothly docked with the arresting frame 101, and the wings are introduced into the arresting block 102.

[0071] The main body of the arresting frame 101 adopts a frame structure, with a hollow middle part for the UAV to pass through. A trapezoidal steel plate is installed below it to enhance the strength of the frame body. Ten mounting holes are drilled on one side of the steel plate to connect with the power output end of the robot system. Two cross beams extending backward are arranged on the left and right sides of the arresting frame 101, and slide rails 103 are installed thereon. The arresting block 102 is slidably installed on each slide rail 103, thus forming a deceleration channel during the UAV recovery process.

[0072] The front side of the arresting frame 101 is connected to an arc-shaped guiding device 106, which has four and is arranged symmetrically in pairs. The space between the same pair of guiding devices 106 is the wing receiving space, and the wing receiving space gradually narrows and then accesses the area where the arresting block 102 is located. During the process of recovering the UAV 500, the guiding device 106 is used for guiding, so that the UAV can be smoothly docked with the arresting frame 101, and the wings are introduced into the arresting block 102.

[0073] There are two buffers 108, which are symmetrically installed below the arresting frame 101. A number of pulleys are installed at the front and below of the arresting frame 101, and pulleys are also installed on the pull rods of the buffers 108. Both ends of the arresting cable are fixed to the middle steel plate of the arresting frame 101 and the front end of the arresting block 102 respectively. The middle of the arresting cable is supported by a number of fixed pulleys below the frame body of the arresting frame and bypasses the movable pulley at the end of the buffer pull rod, so as to increase the deceleration stroke during the UAV recovery and further reduce the UAV recovery overload.

[0074] Regarding the installation position of the robot system:

[0075] The bottom of the robot system 200 is movably installed on the ground rail, and a traveling mechanism is provided at the bottom of the robot system 200.

[0076] See Figure 3, the robot system has six independent rotating shafts, which are respectively from its root as axis, axis, axis, axis, axis, axis. The above rotating shafts enable the robot flange position to have the ability to adjust the spatial six-degree-of-freedom attitude. Among them, the default axis rotation angle is zero; in addition, the arresting device is arranged at the execution end of the robot to achieve the cooperative recovery of the UAV. The recovery window is a rectangular opening at the front end of the arresting device for realizing the UAV arrest.

[0077] For the robot system in this embodiment, the present invention provides a control method for attitude adjustment according to the pose state of the UAV:

[0078] See Figure 4 , during the UAV recovery process, the robot rotates the axis to make the recovery plane parallel to the track plane ; rotates the axis to make the arresting device adapt to the UAV roll angle; the robot can translate on the ground track to make the recovery plane coincide with the track plane.

[0079] See Figure 5 , in the recovery plane , the robot system adjusts the axis, axis, axis to make the arresting device reach the predetermined recovery point coordinate position at the pitch angle . Let the axis, axis, axis rotation angles be , , .

[0080] Establish a coordinate system at the robot base, axis, axis, axis respectively establish coordinate systems , , . Based on the adjacent link coordinate system transformation matrix , calculate the homogeneous transformation matrix of the coordinate system relative to the coordinate system :

[0081] ;

[0082] Coordinate system at the A5 axis Relative to the base coordinate system The transformation matrix is expressed as:

[0083] ;

[0084] Where and represent the position of the coordinate system in the recovery plane:

[0085] ;

[0086] ;

[0087] Thus, we obtain axis, axis, axis rotation angles , , respectively as:

[0088] ;

[0089] ;

[0090] ;

[0091] Finally, obtain axis, axis, axis rotation angles , , , and input axis rotation angle, axis rotation angle, to control the robot system.

[0092] Figure 6 A two-way collaborative recovery method for unmanned aerial vehicles based on multi-source positioning provided by this embodiment. The application objects of this recovery method include a recovery system and an unmanned aerial vehicle, and the two achieve two-way collaborative recovery operations through information interaction, feedback correction, and dynamic adjustment. When the unmanned aerial vehicle needs to perform a recovery operation through the recovery system, the recovery method provided by the present invention is triggered, which specifically includes the following steps:

[0093] S1: The unmanned aerial vehicle sends a recovery request and the current flight parameters of the unmanned aerial vehicle to the ground end.

[0094] In some specific embodiments, this step can be carried out as follows: after the drone sends a recovery request to the ground terminal, if the ground terminal has the recovery condition, it will return a confirmation recovery instruction to the drone, and then the drone will continuously send its flight parameters to the ground terminal and update them in real time to ensure that the ground terminal can grasp the latest flight data in real time.

[0095] It should be noted that the recovery request also includes the flight parameters at the current moment; the flight parameters include data such as the longitude, latitude, altitude, speed, heading, remaining flight duration, and flight attitude of the drone.

[0096] S2: After the ground terminal receives the recovery request from the drone, it sends the ground terminal position and recovery channel parameters to the drone.

[0097] After the ground terminal receives the recovery request sent by the drone, it will make a recovery judgment based on the drone's flight parameters and the position and status of the ground terminal. If the recovery condition requirements are met, it will return a confirmation recovery instruction to the drone and send the specific position of the ground terminal and the recovery channel parameters to the drone. Among them, the recovery channel is a rectangular airspace with a certain width, height, and length, and this airspace is only open to the target drone.

[0098] S3: The drone plans an expected recovery trajectory based on the ground terminal position, recovery channel parameters, and the current position and heading of the drone, and performs tracking control based on the planned trajectory.

[0099] The drone planning an expected recovery trajectory based on the ground terminal position, recovery channel parameters, and the current position and heading of the drone also includes: if the current flight altitude of the drone is too far from the recovery channel and the descent route exceeds the maximum maneuverability of the drone, it should perform a slow descent in a spiral. When the flight altitude of the drone drops to an acceptable range, then plan the recovery trajectory and adjust the flight heading of the drone to align with the recovery channel.

[0100] Performing tracking control based on the planned trajectory also includes: longitudinal altitude and speed control, and lateral roll and lateral offset control in the horizontal and vertical directions. The drone should maintain longitudinal and horizontal stability before entering the recovery channel. If the drone is affected by gusts or other environments and causes large attitude and heading fluctuations, it should immediately perform a go-around operation, and plan the route again and re-execute the recovery process after recovery.

[0101] S4: When the drone enters the recovery channel, it sends a docking request to the ground terminal; the ground terminal performs real-time positioning on the drone through a multi-source positioning system and sends a trajectory correction instruction to the drone according to the positioning result, so that the drone flies according to the trajectory correction instruction. At the same time, the recovery system adjusts the attitude according to the real-time positioning result to meet the docking requirements with the drone.

[0102] After the UAV enters the recovery channel, the ground end performs real-time positioning on the UAV through a multi-source positioning system, which specifically includes the following solutions:

[0103] See Figure 7 , the multi-source positioning system in this embodiment has four sensing sources, and each positioning source is equipped with a vision camera and a lidar. By respectively fusing the image data and point cloud data generated by each positioning source, accurate screening of the target UAV point cloud under a single positioning source is achieved, and then the target point clouds generated by all positioning sources are stitched and clustered to finally determine the positioning data of the target UAV.

[0104] Among them, since the data acquisition frequencies of the vision sensor and the lidar are different, it is necessary to align the timestamps of the image data and the point cloud data to ensure that the data near the same time point can be effectively time-synchronized, and then the position of each point cloud in the image, that is, the projected point cloud, can be obtained through coordinate transformation. In addition, the image target detection algorithm based on deep learning can effectively identify and screen the target UAV in the image data, so as to determine the pixel range of the target UAV in the image. By fusing the above projected point cloud and the pixel range of the target UAV, the target point cloud data generated by the target UAV can be accurately obtained.

[0105] S5: When the UAV reaches the recovery distance threshold, the ground end judges the recovery conditions; if the recovery conditions are not met, the UAV executes a reflight and repeats the recovery process; if the recovery conditions are met, the UAV continues to maintain its heading until the recovery is completed.

[0106] The recovery distance threshold represents the necessary distance that the UAV needs to maintain from the recovery system when it can execute a reflight operation. If the UAV is closer to the recovery system than this threshold, dangerous situations such as rubbing or collision will occur during the reflight operation.

[0107] The judgment of the recovery conditions may include the following aspects: whether the UAV is in a stable attitude state, whether the communication and positioning signals are good, whether the docking error between the recovery system and the UAV is within an acceptable range, and whether there are only obstacles in the recovery channel.

[0108] Regarding the method for calculating the position of the projected point cloud in the image:

[0109] Let the vision camera coordinate system be and the lidar coordinate system be , and the image coordinate system be .

[0110] A point detected by the lidar in the environment has a coordinate of in the lidar coordinate system, and its coordinate in the vision camera coordinate system is , and there is , where and are the rotation matrix and translation matrix from the lidar coordinate system to the visual camera coordinate system respectively, and The specific values of depend on the relative installation positions of the visual camera and the lidar.

[0111] Then the position of point in the image is , where represents the horizontal direction in the image, represents the vertical direction in the image, is the internal parameter matrix of the visual camera and is determined by its own structure.

[0112] That is, the coordinates of the projected point cloud in the image coordinate system are:

[0113] ;

[0114] Regarding the calculation method of fusing the projected point cloud and the pixel range of the target UAV:

[0115] After target detection based on the image information of the visual camera, the pixel range of the target UAV in the image is obtained, denoted as , where , are the horizontal and vertical coordinate positions of the target in the image, , are the horizontal coordinate region and vertical coordinate region occupied by the target respectively.

[0116] The projected point cloud coordinates are screened, and the screening conditions are shown in the following formula:

[0117] ;

[0118] That is, only the point cloud within the target pixel range is retained, which reduces the number of point clouds and thus improves the operation speed. On the other hand, the screening of the target point cloud helps to improve the positioning accuracy.

[0119] Regarding the calculation method of point cloud fusion and positioning:

[0120] For ease of description, the lidar arranged at the positioning source is denoted as lidar .

[0121] Denote as the representation of the target point cloud data generated by lidar using a four-dimensional coordinate vector. Among them, is the lidar The coordinate vector of the point cloud in this radar coordinate system.

[0122] Let:

[0123] ;

[0124] be the transformation relationship between the lidar coordinate system and the lidar coordinate system, where and are the rotation relationship and the translation relationship respectively. The coordinate vectors of the point clouds of each lidar in the lidar coordinate system can be expressed as:

[0125] ;

[0126] Thus, the spliced point cloud data in the lidar coordinate system is obtained.

[0127] For each point after point cloud splicing is completed, its neighborhood is defined as:

[0128] ;

[0129] If , then the point is a core point. Expand all density-reachable points into a cluster to obtain several clustering clusters . Let the cluster with the largest number of points be the target UAV cluster, that is . Then the average position of this cluster is the target position:

[0130] .

[0131] There are many specific implementation ways of the present invention. The above description is only the preferred implementation manner of the present invention. It should be pointed out that for those of ordinary skill in the art of this technology, without departing from the principle of the present invention, several improvements can still be made, and these improvements should also be regarded as the protection scope of the present invention.

Claims

1. A two-way collaborative recovery method for unmanned aerial vehicles based on multi-source positioning, characterized in that, It includes the following steps: S1: The drone sends a recovery request and the current flight parameters of the drone to the ground terminal; the flight parameters include the longitude, latitude, altitude, speed, heading, remaining endurance, and flight attitude of the drone. S2: After receiving the recovery request from the drone, the ground terminal sends the ground terminal position and recovery channel parameters to the drone. S3: The drone plans an expected recovery flight path based on the ground terminal position, recovery channel parameters, and the current position and heading of the drone, and flies towards the recovery channel based on the planned flight path. S4: When the drone enters the recovery channel, it sends a docking request to the ground terminal; the ground terminal real-time locates the drone through a multi-source positioning system, and sends a flight path correction instruction to the drone according to the positioning result, so that the drone flies according to the flight path correction instruction. At the same time, the recovery system adjusts the attitude according to the real-time positioning result to meet the docking requirements with the drone. S5: When the drone reaches the recovery distance threshold, the ground terminal judges the recovery conditions; if the recovery conditions are not met, the drone performs a re-flight and repeats the recovery process; if the recovery conditions are met, the drone continues to maintain its heading until the recovery is completed.

2. The two-way collaborative recovery method of an unmanned aerial vehicle based on multi-source positioning according to claim 1, wherein The multi-source positioning system includes a plurality of sensing sources arranged around the recovery system, and the sensing sources include visual cameras and lidar sensors. After the drone enters the recovery channel, the ground terminal real-time locates the drone through a multi-source positioning system, which specifically includes the following steps: Synchronize the point cloud data and image data obtained by each sensing source in time. Perform coordinate transformation on the point cloud data to obtain the projected point cloud in the image coordinate system. Denote a point detected by the lidar in the environment The coordinate in the lidar coordinate system is , and its coordinate in the visual camera coordinate system is , and there is , where and are the rotation matrix and translation matrix from the lidar coordinate system to the visual camera coordinate system respectively. The coordinate of the projected point cloud in the image coordinate system can be expressed as: , where represents the horizontal direction in the image, represents the vertical direction in the image, and K represents the transformation matrix; Based on the deep learning model, the drones in the image data are target-recognized, and the pixel range of the image occupied by the target drones is screened out and denoted as , where , are the horizontal and vertical coordinate positions of the target in the image, , are the horizontal coordinate region and the vertical coordinate region occupied by the target respectively. The projected point cloud coordinates are screened, and the screening conditions are shown in the following formula: ； Fuse and calculate the processed different perception source data, denoted as the target point cloud data generated by the lidar using a four-dimensional coordinate vector representation. Let the following equation be established: , namely, the lidar coordinate system to the lidar conversion relationship between coordinate systems, where and are the rotation relationship and the translation relationship respectively. The coordinate vectors of each lidar point cloud in the lidar coordinate system can be expressed as: , where Nx is the number of point clouds generated by the x-th radar; Obtain lidar The point cloud data after stitching in the coordinate system , for each point after completing the point cloud stitching , define its neighborhood: , where is the set of point clouds after splicing, is the neighborhood radius; If , is the quantity threshold, then the point is the core point, and all density-reachable points are expanded into a cluster to obtain several clustering clusters . Let the cluster with the largest number of points be the target UAV cluster, that is , then the average position of this cluster is the target position: .

3. The two-way collaborative recovery method for unmanned aerial vehicles based on multi-source positioning according to claim 1, characterized in that The recovery system includes a blocking device (100), a robot system (200), a ground rail (300), a multi-source positioning system (400), and a drone (500); the blocking device (100) is installed at the power execution end of the robot system (200) to achieve blocking and deceleration of the drone; the robot system (200) has the ability of spatial multi-degree-of-freedom attitude adjustment to support the blocking device to actively adjust according to the pose state of the drone; the ground rail (300) is installed below the robot system to support the robot system to move along the rail to match the dynamic trajectory of the drone; the multi-source positioning system (400) is arranged on both sides of the robot system (200) and the ground rail (300) to achieve high-precision real-time positioning of the drone within a limited range.

4. A two-way collaborative recovery method for unmanned aerial vehicles based on multi-source positioning according to claim 1, characterized in that, The robot system (200) is a six-axis robot. Starting from its root, there are respectively axis, axis, axis, axis, axis, axis. The above-mentioned rotating shafts enable the robot flange position to have the ability to adjust the spatial six-degree-of-freedom attitude. By arranging the arresting device at the execution end of the six-axis robot, the active capture of the UAV is realized; where the default axis rotation angle is zero; During the UAV recovery process, the robot rotates the axis to make the recovery plane parallel to the flight path plane ; rotates the axis to make the arresting device adapt to the UAV roll angle; the translation of the robot on the ground track can make the recovery plane coincide with the flight path plane; In the recovery plane the robot system adjusts axis, axis, axis to make the arresting device reach the predetermined recovery point coordinate position at the pitch angle , let axis, axis, axis rotation angles be , , ; Establish a coordinate system at the robot base , axis, axis, respectively establish coordinate systems at the , , . Based on the adjacent link coordinate system transformation matrix , calculate the homogeneous transformation matrix of the coordinate system relative to the coordinate system : : ； Coordinate system at the A5 axis Relative to the base coordinate system The transformation matrix is expressed as: ； Among them and represents the coordinate system Position in the recovery plane: ； ； Thus, shaft, shaft, shaft rotation angle , , are respectively: ； ； ； Finally, obtain axis, axis, axis rotation angle , , , and input axis rotation angle and axis rotation angle to control the robot system.

Images