A control method and system for precise landing of unmanned aerial vehicle

Abstract

The present invention discloses a control method for accurate landing of a drone, comprising the following steps: S1, establishing a communication connection between the drone and the unmanned boat, sending the current position and posture data of each to the other in real time, and the drone tracking to the top of the unmanned boat based on the position data of the unmanned boat; S2, using a depth camera to collect the positioning mark on the unmanned boat, obtaining the coordinates of the positioning mark in the earth coordinate system, and the drone tracking the positioning marks of different sizes during the descent process; S3, when the drone descends to a height below a height threshold H from the unmanned boat, the drone stops the landing and stops the landing. 1 When the control starts to enter the posture synchronization mode, that is, the posture of the UAV and the unmanned boat are controlled to be synchronized; S4, when the UAV descends to a height lower than the height threshold H from the unmanned boat 2 , the position of the unmanned boat remains unchanged, the UAV accelerates to land at the center of the positioning mark, exits the posture synchronization mode, and completes the landing operation. This improves the success rate of the UAV landing when the unmanned boat is tilted.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle landing, and more specifically, the present invention relates to a control method and system for accurate unmanned aerial vehicle landing. Background Art

[0002] In recent years, my country has comprehensively accelerated the construction of a strong maritime nation, and marine military has become an important part of my country's military development. As a new weapon system for future maritime wars, the technological development of maritime unmanned systems is of great significance to safeguarding my country's maritime rights and interests. Maritime unmanned systems generally include drones and unmanned boats. The former has the characteristics of a large search range but short endurance, while the latter has the characteristics of a long endurance but limited search range. Therefore, the advantages of the two are complementary to form a maritime aircraft-boat collaborative combat system.

[0003] Compared with fixed-wing drones and unmanned helicopters, quad-rotor drones have simpler structures and are more maneuverable. They can achieve vertical take-off and landing when the landing platform area is limited. The use of quad-rotor drones can better play the advantages of aircraft-boat coordinated operations.

[0004] Autonomous landing of unmanned aerial vehicles is a key process in coordinated combat between aircraft and boats. Compared with large ships, unmanned aerial vehicles are smaller in size, and the corresponding landing platform area is also smaller, which puts higher requirements on the precise landing control of unmanned aerial vehicles. During the landing process of unmanned aerial vehicles, different sea conditions will cause dynamic changes in the sea surface. The unmanned aerial vehicle is also affected by waves, and will experience heave and sway, resulting in changes in pitch angle, roll angle, and yaw angle. When the unmanned aerial vehicle is in the final landing state, the unmanned aerial vehicle may be in a tilted posture. Therefore, the unmanned aerial vehicle is very likely to fail to land and cannot achieve high-precision landing operations during the landing process. However, the ability of unmanned aerial vehicles to land accurately determines the key and difficulty of the successful execution of combat missions by aircraft and boats. Summary of the invention

[0005] The present invention provides a control method for precise landing of a UAV, aiming to improve the above-mentioned problem.

[0006] The present invention is implemented as follows: a control method for precise landing of a drone, the method comprising the following steps:

[0007] S1. Establish a communication connection between the UAV and the unmanned boat, send each other's current position and posture data in real time, and the UAV tracks to the top of the unmanned boat based on the position data of the unmanned boat;

[0008] S2. Use a depth camera to collect the positioning mark on the unmanned boat, obtain the coordinates of the positioning mark in the earth coordinate system, and the UAV tracks the positioning marks of different sizes during the descent;

[0009] S3: When the UAV descends to a height lower than the height threshold H from the unmanned boat 1 When , the control starts to enter the posture synchronization mode, that is, the posture of the UAV and the unmanned boat are controlled to be synchronized;

[0010] S4. When the UAV descends to a height lower than the height threshold H from the unmanned boat 2 When the position of the unmanned boat is controlled to remain unchanged, the UAV accelerates to land at the center of the positioning mark, exits the posture synchronization mode, and completes the landing operation.

[0011] Furthermore, the positioning mark is composed of a coarse positioning mark, a sub-precise positioning mark and a fine positioning mark, the sizes of the coarse positioning mark, the sub-precise positioning mark and the fine positioning mark decrease in sequence, and the centers of the three coincide.

[0012] Furthermore, the method for tracking the positioning mark of the drone during descent is as follows:

[0013] S21, obtaining the coordinates of the coarse positioning mark in the earth coordinate system, controlling the UAV to fly to directly above the coarse positioning mark, and lowering the altitude of the UAV;

[0014] S22: Check whether the height of the UAV from the unmanned boat is less than the height threshold H. 1 If the detection result is yes, then execute step S23, if the detection result is no, then execute step S21;

[0015] S23, using a depth camera to collect the positioning mark on the unmanned boat, obtaining the coordinates of the sub-precision positioning mark in the earth coordinate system, controlling the UAV to fly to the top of the sub-precision positioning mark, and lowering the altitude of the UAV;

[0016] S24, detect whether the height of the UAV from the unmanned boat is lower than the height threshold H 2 If the detection result is yes, then execute step S25, if the detection result is no, then execute step S23;

[0017] S25. Use a depth camera to collect the positioning mark on the unmanned boat, obtain the coordinates of the precise positioning mark in the earth coordinate system, and control the UAV to fly to directly above the precise positioning mark.

[0018] Furthermore, the method for maintaining the posture synchronization between the UAV and the unmanned boat is as follows:

[0019] S31, the next moment's posture data of the unmanned boat predicted by the bidirectional long short-term memory neural network, including: the rolling angle of the unmanned boat Pitch Angle Yaw angle

[0020] S32. The UAV uses the next moment's position of the unmanned boat as a reference position and controls the UAV's position to be consistent with the reference position.

[0021] Furthermore, the control method for keeping the drone posture consistent with the reference posture is as follows:

[0022] First, calculate the attitude angle offset of the drone relative to the reference posture, as follows:

[0023]

[0024] φ A ,θ A , A Respectively represent the current roll angle, pitch angle and yaw angle of the drone. Indicates the predicted values ​​of the roll angle, pitch angle and yaw angle of the unmanned boat at the next moment;

[0025] The PID algorithm is used to control the drone to move in the direction of decreasing offset until the attitude angle deviation is reduced to zero.

[0026] Furthermore, the coarse positioning mark is a circular mark, the sub-precise positioning mark is Aruco QR code I, Aruco QR code I is located inside the circular mark, and the centers of the two coincide, and the precise positioning mark is Aruco QR code II, Aruco QR code II is located inside Aruco QR code I, and the centers of the two coincide.

[0027] The present invention is implemented as follows: a control system for precise landing of an unmanned aerial vehicle, the system comprising:

[0028] Unmanned boat, the landing platform of the unmanned boat is provided with a positioning mark;

[0029] A drone, wherein a depth camera is fixed on the drone, and the depth camera is connected to a processing unit, and the processing unit is connected to a receiving unit and a storage unit;

[0030] The receiving unit is used to receive the position data and posture data sent by the unmanned boat and send them to the processing unit. The depth camera sends the collected images to the processing unit. The processing unit controls the unmanned boat to land at the positioning mark on the unmanned boat landing platform based on the control method for precise landing of the unmanned boat.

[0031] Furthermore, the system also includes:

[0032] The storage unit is connected to the receiving unit and is used to store the training data of the bidirectional long short-term memory neural network; the bidirectional long short-term memory neural network is trained based on the unmanned boat attitude data received by the receiving unit, and the trained bidirectional long short-term memory neural network is used for the unmanned boat attitude data prediction during the landing process of the unmanned boat.

[0033] Furthermore, the positioning mark is composed of a coarse positioning mark, a sub-precise positioning mark and a fine positioning mark, and the centers of the coarse positioning mark, the sub-precise positioning mark and the fine positioning mark coincide with each other.

[0034] The method for controlling the precise landing of a UAV provided by the present invention has the following beneficial technical effects: in the final stage of the landing of the UAV, the posture data of the unmanned boat is processed by a neural network prediction algorithm to achieve real-time response of the UAV to the posture changes of the unmanned boat, thereby achieving synchronous movement of the UAV and the unmanned boat; the control method for the synchronous movement of the UAV and the unmanned boat enables the UAV to land precisely even if the unmanned boat is in a tilted posture, making the landing process of the UAV more stable and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A flow chart of a control method for precise landing of a UAV provided in an embodiment of the present invention;

[0036] Figure 2 A schematic diagram of the structure of a control system for precise landing of a UAV provided in an embodiment of the present invention;

[0037] Figure 3 A schematic diagram of a positioning mark provided in an embodiment of the present invention;

[0038] 21. Drone, 22. Airborne wireless data transmission, 23. Depth camera, 24. Unmanned boat, 25. Positioning mark. DETAILED DESCRIPTION

[0039] The specific implementation modes of the present invention are further explained in detail below by describing the embodiments with reference to the accompanying drawings, so as to help those skilled in the art to have a more complete, accurate and in-depth understanding of the inventive concept and technical solution of the present invention.

[0040] Figure 1 A flow chart of a control method for precise landing of a UAV provided in an embodiment of the present invention, the method specifically comprises the following steps:

[0041] S1. Establish a communication connection between the UAV and the unmanned boat, send each other's current position and posture data in real time, and the UAV tracks to the top of the unmanned boat based on the position data of the unmanned boat;

[0042] In the embodiment of the present invention, the drone and the unmanned boat communicate with each other through airborne wireless data transmission, and set their IP addresses to the same local area network. The data sent by the drone and the unmanned boat should include GPS coordinates and their own attitude data, including roll angle. The pitch angle θ, the yaw angle ψ, and the GPS coordinates are the coordinates in the earth coordinate system.

[0043] Assume that the current coordinates of the drone in the earth coordinate system are (X A ,Y A ,Z A ), the current coordinates of the unmanned boat in the earth coordinate system are (X S ,Y S ,Z S ), the UAV is tracked to the top of the unmanned boat, and the objective function J in the tracking process is 1 The details are as follows:

[0044] J 1 =[X A Y A ] T -[X S Y S ] T

[0045] When J 1 When it is zero, the UAV has tracked to the top of the unmanned boat. At this time, the depth camera on the UAV is controlled to take pictures to detect whether there is a coarse positioning mark in the image. If the detection result is yes, the coordinates of the coarse positioning mark in the earth coordinates are obtained. If the detection result is no, the UAV is controlled to rise.

[0046] S2. Use a depth camera to collect the positioning mark on the unmanned boat, obtain the coordinates of the coarse positioning mark in the earth coordinate system, control the UAV to fly to the top of the coarse positioning mark, and reduce the height of the UAV;

[0047] In an embodiment of the present invention, the method for obtaining the coordinates of the coarse positioning marker in the earth coordinate system is specifically as follows:

[0048] Read the coordinates of the center of the coarse positioning mark in the image (u m ,v m ), convert the read image coordinates to the earth coordinate system, and obtain the coordinates of the rough positioning center in the earth coordinate system (X m ,Y m ), the conversion formula is as follows:

[0049]

[0050] Among them, φ A ,θ A , A They represent the current roll angle, pitch angle and yaw angle of the drone, respectively. u , k v Respectively represent the pixel focal length of the camera's intrinsic parameter matrix in the X(u) and Y(u) directions, (c u , c v ) represents the center coordinate of the image, (X A ,YA ,Z A ) represents the current coordinates of the drone in the earth coordinate system.

[0051] In the embodiment of the present invention, the coarse positioning mark is a circular mark. The recognition speed of the circular mark is faster than that of the Aruco two-dimensional code, and it is suitable for the coarse positioning of the UAV at high altitude. The positioning accuracy of the Aruco two-dimensional code is relatively high, and its recognition speed is relatively slow. It is suitable for the precise positioning of the UAV before landing. The specific positioning method is as follows:

[0052]

[0053] When the height between the UAV and the unmanned boat is H>H 1 When the circular mark is detected, the Canny operator is used to detect the edge, and then the circular mark can be detected based on the Hough transform method; at this time, the Aruco QR code I and Aruco QR code II are too small in the image and cannot be recognized. After the circular mark is recognized in the image, the coordinates of the center of the circular mark in the image are determined, and then the coordinates are converted to the earth coordinate system. The coordinate conversion method is the same as above, and then the drone is controlled to fly directly above the center of the circular mark. The drone control method is the same as above.

[0054] S3: Check whether the height of the UAV from the unmanned boat is less than the height threshold H. 1 If the test result is yes, then execute step S4, if the test result is no, then execute step S2;

[0055] S4, using a depth camera to collect the positioning mark on the unmanned boat, obtain the coordinates of the sub-precision positioning mark in the earth coordinate system, control the UAV to fly to the top of the sub-precision positioning mark, lower the altitude of the UAV, and start entering the posture synchronization mode, that is, start controlling the posture of the UAV to keep synchronization with that of the unmanned boat;

[0056] In the embodiment of the present invention, the secondary precise positioning mark is Aruco QR code I, and the UAV is at a height H above the unmanned boat. 2 <H<H 1 At this time, the circular mark is incomplete in the image due to the low altitude, and the Aruco QR code II is too small in the image due to the high altitude and cannot be recognized. At this time, the drone tracks the center of the Aruco QR code I. After identifying the Aruco QR code I (corresponding ID) in the image, the coordinates of the center of the Aruco QR code I in the image can be determined based on the coordinates of the four corner points of the outer frame of the Aruco QR code I, and then the coordinates are converted to the earth coordinate system. The coordinate conversion method is the same as above, and then the drone is controlled to fly directly above the center of the Aruco QR code I. The drone control method is the same as above.

[0057] In an embodiment of the present invention, the method for maintaining the posture synchronization between the UAV and the unmanned boat is specifically as follows:

[0058] The next moment's posture data of the unmanned boat predicted by the bidirectional long short-term memory neural network, including: the rolling angle of the unmanned boat Pitch Angle Yaw angle

[0059] The UAV uses the next moment's posture of the unmanned boat as a reference posture, and controls the posture of the UAV to be consistent with the reference posture; first calculate the attitude angle offset of the UAV relative to the reference posture, as follows:

[0060]

[0061] φ A ,θ A , A Respectively represent the current roll angle, pitch angle and yaw angle of the drone. It indicates the predicted value of the rolling angle, pitch angle and yaw angle of the unmanned boat at the next moment. Δθ and Δψ represent the roll angle offset, pitch angle offset, and yaw angle offset, respectively.

[0062] The PID algorithm is used to control the UAV to move in the direction of decreasing offset until the attitude angle deviation is reduced to zero, thereby synchronizing the attitudes of the UAV and the unmanned boat.

[0063] After the UAV starts communicating with the unmanned boat, the UAV will store the received unmanned boat attitude data in chronological order, use the stored unmanned boat attitude data for training the bidirectional long short-term memory neural network, and then use the trained bidirectional long short-term memory neural network to predict the attitude.

[0064] S5. Check whether the height of the UAV from the unmanned boat is lower than the height threshold H. 2 If the detection result is yes, the position of the unmanned boat is controlled to remain unchanged, and step S6 is executed; if the detection result is no, step S4 is executed;

[0065] S6. Use a depth camera to collect the positioning mark on the unmanned boat, obtain the coordinates of the precise positioning mark in the earth coordinate system, and control the UAV to fly to the top of the precise positioning mark;

[0066] The precise positioning mark is Aruco QR code Ⅱ, and the UAV is at a height of H<H from the unmanned boat. 2 When H 2The range should be within 0.1m to 0.2m. Due to the low height, the circular logo and Aruco QR code I are incomplete in the image and cannot be identified. At this time, the drone tracks the center of Aruco QR code II. After identifying Aruco QR code II in the image (the corresponding ID, Aruco QR code II and Aruco QR code I have different IDs), the coordinates of the center of Aruco QR code II in the image can be determined based on the coordinates of the four corner points of the outer frame of Aruco QR code II, and then the coordinates are converted to the earth coordinate system. The coordinate conversion method is the same as above, and then the drone is controlled to fly directly above the center of Aruco QR code II. The drone control method is the same as above.

[0067] S7. The UAV accelerates and lands at the center of the positioning mark, exits the posture synchronization mode, and completes the landing operation.

[0068] Figure 2 The structure of the control system for precise landing of a UAV provided in an embodiment of the present invention is schematically shown. For ease of explanation, only the part related to the embodiment of the present invention is shown.

[0069] The system includes:

[0070] The unmanned boat has a positioning mark set on its landing platform. The positioning mark is composed of a coarse positioning mark, a sub-precise positioning mark and a fine positioning mark. The centers of the coarse positioning mark, the sub-precise positioning mark and the fine positioning mark coincide with each other.

[0071] A drone, wherein a depth camera is fixed on the drone, and the depth camera is connected to a processing unit, and the processing unit is connected to a receiving unit and a storage unit;

[0072] The receiving unit is used to receive the position data and posture data sent by the unmanned boat and send it to the processing unit. The depth camera sends the collected images to the processing unit. The processing unit controls the drone to land at the positioning mark on the unmanned boat landing platform based on the above-mentioned control method for precise landing of the drone.

[0073] In the embodiment of the present invention, the coarse positioning mark is a circular mark, the sub-precise positioning mark is Aruco two-dimensional code I, Aruco two-dimensional code I is located in the circular mark, and the centers of the two coincide, and the precise positioning mark is Aruco two-dimensional code II, Aruco two-dimensional code II is located in Aruco two-dimensional code I, and the centers of the two coincide, such as Figure 3 shown.

[0074] In an embodiment of the present invention, the system further comprises: a storage unit connected to the receiving unit and used to store training data of the bidirectional long short-term memory neural network;

[0075] The bidirectional long short-term memory neural network is trained based on the unmanned boat attitude data received by the receiving unit. The trained bidirectional long short-term memory neural network is used to predict the unmanned boat attitude data during the landing process of the UAV.

[0076] The method for controlling the precise landing of a UAV provided by the present invention has the following beneficial technical effects: in the final stage of the landing of the UAV, the posture data of the unmanned boat is processed by a neural network prediction algorithm to achieve real-time response of the UAV to the posture changes of the unmanned boat, thereby achieving synchronous movement of the UAV and the unmanned boat; the control method for the synchronous movement of the UAV and the unmanned boat enables the UAV to land precisely even if the unmanned boat is in a tilted posture, making the landing process of the UAV more stable and efficient.

[0077] The present invention has been described exemplarily. Obviously, the specific implementation of the present invention is not limited to the above-mentioned method. As long as various non-substantial improvements are made using the method concept and technical solution of the present invention, or the concept and technical solution of the present invention are directly applied to other occasions without improvement, they are all within the protection scope of the present invention.

Claims

1. A control method for precise landing of a UAV, characterized in that: The method comprises the following steps: S1. Establish a communication connection between the UAV and the unmanned boat, send each other's current position and posture data in real time, and the UAV tracks to the top of the unmanned boat based on the position data of the unmanned boat; S2. Use a depth camera to collect the positioning mark on the unmanned boat, obtain the coordinates of the positioning mark in the earth coordinate system, and the UAV tracks the positioning marks of different sizes during the descent; S3, when the UAV descends to a height lower than the height threshold H1 from the unmanned boat, the control starts to enter the posture synchronization mode, that is, the posture of the UAV and the unmanned boat are controlled to be synchronized; S4. When the UAV descends to a height lower than the height threshold H2 from the unmanned boat, the position of the unmanned boat is controlled to remain unchanged, and the UAV accelerates to land at the center of the positioning mark, exits the posture synchronization mode, and completes the landing operation; The method for maintaining the synchronization of the posture of the UAV and the unmanned boat is as follows: S31, the next moment's posture data of the unmanned boat predicted by the bidirectional long short-term memory neural network, including: the rolling angle of the unmanned boat Pitch Angle Yaw angle S32, the UAV uses the next moment's posture of the unmanned boat as a reference posture, and controls the posture of the UAV to be consistent with the reference posture; The specific control method to keep the drone posture consistent with the reference posture is as follows: First, calculate the attitude angle offset of the drone relative to the reference posture, as follows: φ A ,θ A , A Respectively represent the current roll angle, pitch angle and yaw angle of the drone. Indicates the predicted values ​​of the roll angle, pitch angle and yaw angle of the unmanned boat at the next moment; The PID algorithm is used to control the drone to move in the direction of decreasing offset until the attitude angle deviation is reduced to zero.

2. The control method for accurate landing of a UAV as claimed in claim 1, characterized in that: The positioning mark is composed of a coarse positioning mark, a sub-precise positioning mark and a fine positioning mark. The sizes of the coarse positioning mark, the sub-precise positioning mark and the fine positioning mark decrease in sequence, and the centers of the three coincide.

3. The control method for accurate landing of a UAV as claimed in claim 2, characterized in that: The tracking method of the drone’s positioning mark during descent is as follows: S21, obtaining the coordinates of the coarse positioning mark in the earth coordinate system, controlling the UAV to fly to directly above the coarse positioning mark, and lowering the altitude of the UAV; S22, detecting whether the height of the UAV from the unmanned boat is less than the height threshold H1, if the detection result is yes, executing step S23, if the detection result is no, executing step S21; S23, using a depth camera to collect the positioning mark on the unmanned boat, obtaining the coordinates of the sub-precision positioning mark in the earth coordinate system, controlling the UAV to fly to the top of the sub-precision positioning mark, and lowering the altitude of the UAV; S24, detecting whether the height of the UAV from the unmanned boat is lower than the height threshold H2, if the detection result is yes, executing step S25, if the detection result is no, executing step S23; S25. Use a depth camera to collect the positioning mark on the unmanned boat, obtain the coordinates of the precise positioning mark in the earth coordinate system, and control the UAV to fly to the top of the precise positioning mark.

4. The control method for accurate landing of a UAV as claimed in claim 2, characterized in that: The coarse positioning mark is a circular mark, the sub-precise positioning mark is Aruco QR code I, Aruco QR code I is located inside the circular mark, and the centers of the two coincide, and the precise positioning mark is Aruco QR code II, Aruco QR code II is located inside Aruco QR code I, and the centers of the two coincide.

5. A control system for precise landing of unmanned aerial vehicles, characterized in that: The system comprises: Unmanned boat, the landing platform of the unmanned boat is provided with a positioning mark; A drone, wherein a depth camera is fixed on the drone, and the depth camera is connected to a processing unit, and the processing unit is connected to a receiving unit and a storage unit; The receiving unit is used to receive the position data and posture data sent by the unmanned boat and send it to the processing unit. The depth camera sends the collected image to the processing unit. The processing unit controls the unmanned boat to land at the positioning mark on the unmanned boat landing platform based on the control method for precise landing of the unmanned boat as described in any one of claims 1 to 4.

6. The control system for precise landing of unmanned aerial vehicle according to claim 5, characterized in that: The system further comprises: The storage unit is connected to the receiving unit and is used to store the training data of the bidirectional long short-term memory neural network; the bidirectional long short-term memory neural network is trained based on the unmanned boat attitude data received by the receiving unit, and the trained bidirectional long short-term memory neural network is used for the unmanned boat attitude data prediction during the landing process of the unmanned boat.

7. The control system for precise landing of unmanned aerial vehicle according to claim 5, characterized in that: The positioning mark is composed of a coarse positioning mark, a sub-precise positioning mark and a fine positioning mark, and the centers of the coarse positioning mark, the sub-precise positioning mark and the fine positioning mark coincide.

Images