A UAV landing control method for dynamic apron

Abstract

The present invention discloses a method for controlling the landing of a UAV for a dynamic helipad, which comprises the following steps: obtaining the IMU data of the UAV in real time; obtaining the relative rotation matrix and translation vector between the UAV and the helipad; solving the Euler angle of the UAV according to the relative rotation matrix between the UAV and the helipad, and obtaining the conversion rotation matrix; obtaining the helipad coordinates according to the imaging device coordinates, the conversion rotation matrix and the translation vector, and determining the horizontal offset of the helipad relative to the UAV according to the helipad coordinates; obtaining the horizontal oscillation frequency and horizontal oscillation amplitude of the helipad according to the horizontal offset of the helipad relative to the UAV, and executing the corresponding UAV landing strategy according to the horizontal oscillation frequency and / or horizontal oscillation amplitude of the helipad. The present invention does not need to add additional sensors, and can obtain the tilt shaking and horizontal oscillation data between the helipad and the UAV only through the existing sensor data, and further realizes the posture estimation of the oscillating helipad and the precise landing control of the UAV according to the data.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and more specifically, to a method for controlling the landing of an unmanned aerial vehicle on a dynamic helipad. Background Art

[0002] With the increasing application of drones, the landing sites of drones are not only static aprons, but also vehicle-mounted mobile nests, power inspection nests on the top of towers, etc. The nests at these locations are not static and will produce tilt shaking and horizontal vibrations, which brings new challenges to the landing of drones.

[0003] As for the apron with the above-mentioned tilt shaking and horizontal vibration, it is generally necessary to install sensors on the apron to record the fluctuation state of the apron, and then calculate the posture compensation data of the UAV and compensate it to the landing control of the UAV. However, this method requires the installation of additional sensors, resulting in high application costs. At the same time, in the case of a small vibration of the apron, the response frequency and response speed of the UAV cannot meet the requirements, and the effective synchronization of the posture compensation data cannot be achieved, resulting in the UAV being unable to land accurately on the apron. Summary of the invention

[0004] The purpose of the present invention is to provide a method for controlling the landing of a drone on a dynamic apron. The method does not require additional sensors and can obtain the tilt shaking and horizontal oscillation data between the apron and the drone only through the existing sensor data, and further realizes the posture estimation of the oscillating apron and the precise landing control of the drone based on the data.

[0005] In order to achieve these purposes and other advantages according to the present invention, a method for controlling a drone landing on a dynamic helipad is provided, which comprises the following steps:

[0006] Acquire the IMU data of the drone in real time, and correct the image acquired by the drone imaging device according to the IMU data;

[0007] Get the relative rotation matrix and translation vector between the drone and the landing pad;

[0008] Calculate the Euler angle of the drone according to the relative rotation matrix between the drone and the landing pad, and obtain the conversion rotation matrix;

[0009] Obtain the apron coordinates according to the imaging device coordinates, the conversion rotation matrix and the translation vector, and determine the horizontal offset of the apron relative to the UAV according to the apron coordinates;

[0010] The horizontal oscillation frequency and the horizontal oscillation amplitude of the apron are obtained according to the horizontal offset of the apron relative to the UAV, and the corresponding UAV landing strategy is executed according to the horizontal oscillation frequency and / or the horizontal oscillation amplitude of the apron.

[0011] Preferably, the corrected image is consistent with the image obtained by the imaging device from the perspective of facing the ground directly.

[0012] Preferably, the image is set on the helipad and can be recognized by the drone to assist the drone in landing on the helipad.

[0013] Preferably, the image is a two-dimensional code.

[0014] Preferably, the corner positions in the obtained two-dimensional code image are corrected to obtain the corrected corners, and the corrected corners are consistent with the two-dimensional code corner positions obtained by the imaging device from the perspective of facing the ground directly.

[0015] Preferably, obtaining the horizontal oscillation frequency and horizontal oscillation amplitude of the helipad according to the horizontal offset of the helipad relative to the drone includes the following steps:

[0016] Draw a horizontal offset curve graph according to the horizontal offset of the helipad relative to the drone;

[0017] Use a sliding window to intercept a part of the curve, perform time-domain transformation on the intercepted part of the curve, and then map it to the frequency domain to obtain the horizontal oscillation frequency curve of the helipad, and obtain the horizontal oscillation frequency of the helipad and the horizontal oscillation amplitude within a predetermined frequency domain range from it.

[0018] Preferably, executing the corresponding drone landing strategy according to the horizontal oscillation frequency and / or horizontal oscillation amplitude of the helipad includes the following steps:

[0019] Perform time-domain conversion on the horizontal oscillation frequency curve of part / all of the helipad to obtain a time-domain curve, obtain the horizontal offset of the helipad relative to the drone from the time-domain curve, and guide the drone to move to the center of the horizontal oscillation range according to this horizontal offset.

[0020] Preferably, perform time-domain conversion on other horizontal oscillation frequency curves outside the predetermined frequency domain range.

[0021] Preferably, executing the corresponding drone landing strategy according to the horizontal oscillation frequency and / or horizontal oscillation amplitude of the helipad includes the following steps:

[0022] Compare the horizontal oscillation amplitude of the helipad within the predetermined frequency domain range with the amplitude threshold;

[0023] If E < E0, control the drone to land on the helipad in segments;

[0024] If E ≥ E0, cancel the landing.

[0025] Preferably, controlling the drone to land on the helipad in segments includes the following steps:

[0026] First, determine the horizontal oscillation period of the apron according to the horizontal oscillation frequency of the apron, and then determine the time t1 when the apron returns to the horizontal oscillation center position according to the horizontal oscillation period of the apron;

[0027] The time t2 when the drone lands on the helipad is calculated based on the landing speed and height of the drone, and t1=t2 is ensured.

[0028] The present invention has at least the following beneficial effects:

[0029] The present invention does not require additional sensors. The tilt shaking and horizontal oscillation data between the apron and the drone can be obtained only through the data of existing sensors (i.e., the IMU carried by the drone). The posture estimation of the vibrating apron and the precise landing control of the drone can be further realized based on the data. This can greatly reduce costs and expand the possibility of automatic and precise landing of drones in multiple scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a schematic diagram of the relative position relationship between the drone and the apron when the drone is flying.

[0031] Figure 2 The figure is a schematic diagram of the steps of the method for controlling the landing of a UAV on a dynamic helipad in the present invention.

[0032] Figure 3 This is a diagram showing the effect of correcting the image acquired by the drone imaging device in the present invention.

[0033] Figure 4 It is a horizontal offset curve diagram drawn according to the horizontal offset x and y of the apron relative to the UAV in the present invention.

[0034] Figure 5 It is a schematic diagram of intercepting part of the horizontal offset curve through the sliding window in the present invention.

[0035] Figure 6 It is a schematic diagram of performing time domain transformation on the intercepted horizontal offset curve in the present invention.

[0036] Figure 7 It is the horizontal oscillation frequency curve of the apron obtained in the present invention.

[0037] Figure 8 In the present invention, the horizontal oscillation frequency curve of part / all of the apron is converted into time domain to obtain the time domain curve. DETAILED DESCRIPTION

[0038] The present invention is further described in detail below in conjunction with the accompanying drawings so that those skilled in the art can implement the invention with reference to the description.

[0039] In the description of the present invention, the terms "lateral", "longitudinal", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" etc. to indicate directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings and are only for the convenience of describing the present invention and simplifying the description. They do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as a limitation on the present invention.

[0040] Embodiment 1:

[0041] like Figure 1 As shown in the figure, when the drone is flying, the roll angle and pitch angle of the drone are of the same nature, collectively referred to as the inclination angle, where α represents the inclination angle of the drone relative to the horizontal line, and R U represents the rotation matrix generated by the inclination angle α, β represents the inclination angle of the apron relative to the horizontal line, R A represents the rotation matrix generated by the inclination angle β, γ represents the relative inclination angle between the UAV and the apron, R represents the relative rotation matrix between the UAV and the apron, and γ=α+β, R=R U* R A .

[0042] On this basis, this embodiment provides a method for controlling the landing of a UAV on a dynamic helipad, such as Figure 2 As shown, it includes the following steps:

[0043] S1. Acquire the IMU (Inertial Measurement Unit) data of the drone in real time, and correct the image acquired by the drone imaging device (such as a camera) according to the IMU data, so that the corrected image is consistent with the image acquired by the imaging device at a perspective of facing the earth; the image is set on the apron and can be recognized by the drone to assist the drone to land on the apron. For example, in this embodiment, the image is a QR code;

[0044] For example, Figure 3 As shown, the corner point position in the acquired two-dimensional code image can be corrected to obtain a corrected corner point P, and the corrected corner point P is consistent with the corner point position of the two-dimensional code acquired by the imaging device at a viewing angle facing the earth;

[0045] S2, based on the PnP solution algorithm, obtain the relative rotation matrix R and translation vector T between the UAV and the apron;

[0046] Since the imaging device camera and the drone are rigidly connected, after correcting the image according to the drone's IMU data in step S1, the influence of the tilt angle α can be eliminated, and only the influence of the tilt angle β needs to be considered. Therefore, the relative rotation matrix between the drone and the apron is R = R A , by obtaining R A The rotation matrix R can be obtained, which can greatly simplify the calculation process;

[0047] S3. Calculate the Euler angles (pitch, roll, yaw) of the drone based on the relative rotation matrix R between the drone and the apron. The pitch and roll angles at this time are the rotation angles of the apron relative to the ground.

[0048] And according to the (0, 0, yaw) angle combination, the conversion rotation matrix R1 is obtained, and the conversion rotation matrix R1 is used as the relative rotation matrix between the apron and the drone;

[0049] S4, according to the imaging device coordinates P U (0,0,0), transform the rotation matrix R1 and the translation vector T to obtain the apron coordinates Pw(x,y,z), and determine the horizontal offset of the apron relative to the drone based on the apron coordinates Pw(x,y,z);

[0050] Since the imaging device coordinates P U (0,0,0) and the apron coordinates Pw(x,y,z) have the following relationship in the geodetic coordinate system:

[0051] P U (0,0,0)=R 1 *P W (x,y,z)+T;

[0052] Where P U , R 1 , T are known, therefore, according to P W (x,y,z)=R 1- 1 *(P U (0,0,0)-T) to obtain the apron coordinates Pw(x,y,z), where x and y represent the horizontal offset of the apron relative to the drone, and z represents the relative height between the drone and the apron;

[0053] Since the horizontal offset of the apron relative to the UAV obtained here is the relative offset that has eliminated the influence of the inclination angles α and β, it simplifies the flight control parameters of the UAV, which is more conducive to the control of the UAV;

[0054] S5. Obtain the horizontal oscillation frequency T0 and the horizontal oscillation amplitude E of the apron according to the horizontal offset x and y of the apron relative to the UAV, and execute the corresponding UAV landing strategy according to the horizontal oscillation frequency T0 and / or the horizontal oscillation amplitude E of the apron.

[0055] Specifically, step S5 includes the following steps:

[0056] S51, such as Figure 4 As shown, a horizontal offset curve graph is drawn according to the horizontal offset x and y of the apron relative to the UAV, wherein the horizontal axis represents time and the vertical axis represents the horizontal offset;

[0057] S52, such as Figure 5 As shown, a sliding window is used to intercept part of the curve. Figure 6 As shown, the intercepted part of the curve is transformed through the time domain (Time Domain) and then mapped to the frequency domain (Frequency Domain) to obtain the horizontal oscillation frequency curve of the apron (such as Figure 7 As shown, where the horizontal axis represents the oscillation frequency and the vertical axis represents the amplitude), and the horizontal oscillation frequency T0 of the apron and the horizontal oscillation amplitude E within a predetermined frequency domain are obtained therefrom;

[0058] For example, the apron in this embodiment is placed on a high-voltage tower. The apron and the tower oscillate synchronously horizontally, and the horizontal oscillation frequency of the high-voltage tower is within 4 Hz. Therefore, the frequency domain data of 0-4 Hz is intercepted using a window (i.e. Figure 7 The frequency domain range in the box is obtained from the amplitude E of the apron. It should be noted that, since the oscillation frequency domains of different apron carriers are different, the horizontal oscillation frequency domain of the carrier can be detected in advance (such as the oscillation frequency domain of the high-voltage tower is 0-4Hz), which is used as the horizontal oscillation frequency domain of the apron, and then the horizontal oscillation amplitude E of the apron is obtained from the oscillation frequency domain;

[0059] S53, such as Figure 8 As shown, the horizontal oscillation frequency curve of part / all of the apron is converted into time domain (in this embodiment, it is preferred to convert the horizontal oscillation frequency curve outside the predetermined frequency domain range, i.e. Figure 7 Other horizontal oscillation frequency curves outside the box are converted into time domain) to obtain a time domain curve, and the horizontal offset of the apron relative to the UAV (i.e. Figure 8 The vertical axis in the figure) and guides the UAV to move to the center of the horizontal oscillation range according to the horizontal offset;

[0060] S54, comparing the horizontal oscillation amplitude E of the apron in a predetermined frequency domain range with an amplitude threshold E0, wherein the amplitude threshold E0 is related to the size of the apron, and the larger the apron, the larger E0;

[0061] If E < E0, the drone is controlled to land on the landing pad in segments. Specifically, first determine the horizontal oscillation period of the landing pad according to the horizontal oscillation frequency T0 of the landing pad, then determine the moment t1 when the landing pad returns to the horizontal oscillation center position according to the horizontal oscillation period of the landing pad, calculate the moment t2 when the drone lands on the landing pad according to the landing speed and height of the drone, and ensure that t1 = t2;

[0062] If E ≥ E0, the landing is cancelled.

[0063] In summary, the present invention does not require additional sensors, and can obtain the inclination shaking and horizontal oscillation data between the landing pad and the drone only through the data of existing sensors (i.e., the IMU carried by the drone). Furthermore, based on this data, the pose estimation of the oscillating landing pad and the precise landing control of the drone are realized, thereby expanding the possibility of automatic and precise landing of the drone in multi-scene applications.

[0064] It should be noted that although the implementation scheme of the present invention has been disclosed as above, it is not limited to the applications listed in the specification and the implementation manner. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily achieved. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to the specific details and the drawings shown and described herein.

Claims

1. A method for controlling the landing of a UAV on a dynamic helipad, characterized in that: It includes the following steps: Obtain the IMU data of the drone in real time, and correct the images obtained by the imaging device of the drone according to the IMU data; Obtain the relative rotation matrix and translation vector between the drone and the landing pad; Solve the Euler angles of the drone according to the relative rotation matrix between the drone and the landing pad, and obtain the conversion rotation matrix; Obtain the landing pad coordinates according to the imaging device coordinates, the conversion rotation matrix, and the translation vector, and determine the horizontal offset of the landing pad relative to the drone according to the landing pad coordinates; Obtain the horizontal oscillation frequency and horizontal oscillation amplitude of the landing pad according to the horizontal offset of the landing pad relative to the drone, and execute the corresponding drone landing strategy according to the horizontal oscillation frequency and / or horizontal oscillation amplitude of the landing pad; Among them, obtaining the horizontal oscillation frequency and horizontal oscillation amplitude of the landing pad according to the horizontal offset of the landing pad relative to the drone includes the following steps: Draw a horizontal offset curve according to the horizontal offset of the landing pad relative to the drone; Use a sliding window to intercept a part of the curve, perform time-domain transformation on the intercepted part of the curve, and then map it to the frequency domain to obtain the horizontal oscillation frequency curve of the landing pad, and obtain the horizontal oscillation frequency of the landing pad and the horizontal oscillation amplitude within a predetermined frequency range from it.

2. The UAV landing control method according to claim 1, characterized in that: The corrected image is consistent with the image obtained by the imaging device in the perspective of facing the ground.

3. The UAV landing control method according to claim 1, characterized in that: The image is set on the landing pad and can be recognized by the drone, and is used to assist the drone in landing on the landing pad.

4. The UAV landing control method according to claim 2, characterized in that: The image is a two-dimensional code.

5. The UAV landing control method according to claim 4, characterized in that: Correct the corner positions in the obtained two-dimensional code image to obtain the corrected corners, and the corrected corners are consistent with the two-dimensional code corner positions obtained by the imaging device in the perspective of facing the ground.

6. The UAV landing control method according to claim 1, characterized in that: Executing the corresponding drone landing strategy according to the horizontal oscillation frequency and / or horizontal oscillation amplitude of the landing pad includes the following steps: Perform time-domain conversion on part / all of the horizontal oscillation frequency curves of the landing pad to obtain a time-domain curve, obtain the horizontal offset of the landing pad relative to the drone from the time-domain curve, and guide the drone to move to the center of the horizontal oscillation range according to the horizontal offset.

7. The UAV landing control method according to claim 6, characterized in that: Perform time-domain conversion on the other horizontal oscillation frequency curves outside the predetermined frequency range.

8. The UAV landing control method according to claim 1, characterized in that: Executing the corresponding drone landing strategy according to the horizontal oscillation frequency and / or horizontal oscillation amplitude of the landing pad includes the following steps: Compare the horizontal oscillation amplitude of the landing pad within the predetermined frequency range with the amplitude threshold; If E < E0, control the drone to land on the landing pad in segments; If E ≥ E0, cancel the landing.

9. The method for controlling the landing of a drone as claimed in claim 8, characterized in that: Controlling the drone to land on the landing pad in segments includes the following steps: First, determine the horizontal oscillation period of the landing pad according to the horizontal oscillation frequency of the landing pad, and then determine the moment t1 when the landing pad returns to the center position of the horizontal oscillation according to the horizontal oscillation period of the landing pad; Calculate the moment t2 when the drone lands on the landing pad according to the landing speed and height of the drone, and ensure that t1 = t2.

Images