Unmanned aerial vehicle take-off and landing control method and system

Abstract

The invention discloses an unmanned aerial vehicle take-off and landing control method and system, and belongs to the technical field of unmanned aerial vehicle autonomous take-off and landing control, and the system comprises a vibration suppression mode library construction module, a vibration suppression mode library activation module, a comparison analysis module, a comparison judgment module and an unmanned aerial vehicle landing execution module. The method comprises the following steps: constructing a vibration suppression mode library and configuring a condition trigger of the vibration suppression mode library, receiving and analyzing vertical acceleration data output by an inertial measurement unit in real time based on vibration suppression modes in the vibration suppression mode library, and comparing the vertical acceleration data with invalid vibration characteristic modes in the vibration suppression mode library in real time; the unmanned aerial vehicle is determined as vertical vibration of the mobile platform when a preset matching degree threshold value is met, and the unmanned aerial vehicle is instructed as a corresponding instruction for not executing the vertical vibration of the mobile platform. And the take-off and landing stability of the unmanned aerial vehicle in a complex environment is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of autonomous take-off and landing control of unmanned aerial vehicles (UAVs), and in particular to a method and system for controlling the take-off and landing of UAVs. Background Art

[0002] As a flexible and efficient aerial work platform, drones have shown broad application prospects in industrial environments such as automated logistics and smart warehousing, especially when performing tasks such as precise cargo transfer and inspection.

[0003] In actual applications, the mobile platform used in conjunction with the drone and providing landing support may encounter uneven surfaces, joints, or obstacles during its movement, causing the mobile platform itself to generate transient, high-frequency mechanical vibrations. For example, in a warehouse, when a ground transfer vehicle runs over a joint in a floor panel or a metal cover, the vehicle body and the landing platform on top of it will experience short-lasting but large-amplitude vertical vibrations and momentary attitude tilt. The drone's onboard inertial measurement unit (IMU) accurately senses and records these vertical vibrations as sharp changes in acceleration. If the drone's control system responds entirely to the IMU's data, attempting to offset these vibrations by rapidly adjusting the rotor speed, it will cause the drone to experience a series of unnecessary and violent up and down jerks.

[0004] In addition, the visual markers that the drone relies on for positioning will vibrate when the mobile platform rolls over the seams of the bottom plates or metal covers, resulting in difficulty in recognition, intermittent loss of information or position jumps. The confidence level is significantly reduced, and it cannot provide the drone with stable and reliable positioning and attitude references.

[0005] Therefore, there is an urgent need to improve the solutions to the above problems. Summary of the Invention

[0006] The purpose of this invention is to provide a method and system for controlling the takeoff and landing of a UAV, which can identify and suppress high-frequency, ineffective vibrations of a mobile platform and maintain a stable landing configuration when visual positioning information is unreliable, thereby improving the takeoff and landing stability of the UAV on complex mobile platforms.

[0007] In order to achieve the above object, the technical solutions of the present invention are:

[0008] As one aspect of the present application, a method for controlling the take-off and landing of a drone is provided, comprising:

[0009] S1. Provide drones and mobile platforms equipped with inertial measurement units and vision systems;

[0010] S2. Build a vibration suppression pattern library and configure a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, wherein the invalid vibration characteristic patterns indicate acceleration change characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform in inertial measurement unit data detected by the inertial measurement unit;

[0011] S3. When it is determined that the distance between the UAV and the mobile platform is less than a preset distance threshold, and the visual positioning information about the mobile platform obtained by the visual system is less than a preset confidence threshold, a vibration suppression mode is enabled, where the vibration suppression mode indicates that a vibration suppression mode library is applied when a conditional trigger is triggered in the current operating state of the UAV and the mobile platform;

[0012] S4. In the vibration suppression mode, receiving and analyzing vertical acceleration data output by the inertial measurement unit in real time, and performing real-time comparison between the vertical acceleration data currently output by the inertial measurement unit and invalid vibration characteristic modes in the vibration suppression pattern library;

[0013] S5. When it is determined that the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern meet a preset matching threshold, it is determined that the vertical acceleration data output by the current inertial measurement unit indicates vertical vibration of the mobile platform;

[0014] S6. When it is determined that the mobile platform is vibrating vertically, the UAV is instructed not to execute the vertical attitude or altitude adjustment instruction generated corresponding to the vertical vibration of the mobile platform, and the UAV maintains the preset landing configuration.

[0015] Compared with the prior art, the present application provides a method for controlling the take-off and landing of a UAV. By constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, the present application enables a vibration suppression mode when it is determined that the distance between the UAV and the mobile platform is less than a preset distance threshold and the visual positioning information about the mobile platform obtained by the visual system is less than a preset confidence threshold. In this mode, the method receives and analyzes vertical acceleration data output by the inertial measurement unit in real time, performs a real-time comparison based on the vertical acceleration data output by the current inertial measurement unit with invalid vibration characteristic patterns in the vibration suppression pattern library. When a preset matching threshold is met, it is determined to be vertical vibration of the mobile platform. The UAV is instructed not to execute the vertical attitude or altitude adjustment instructions corresponding to the vertical vibration of the mobile platform, and the UAV maintains a preset landing configuration. This method effectively improves the take-off and landing stability of the UAV in complex environments by identifying and suppressing invalid vibrations while maintaining a stable landing configuration when visual information is unreliable.

[0016] Furthermore, the steps of constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, wherein the invalid vibration characteristic patterns indicate that the inertial measurement unit data obtained by the detection of the mobile platform has acceleration change characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform, include:

[0017] S21. Before receiving a landing request message from the drone to the mobile platform and when the mobile platform passes through the seam, compare the vertical acceleration data output by the inertial measurement unit with the vertical motion image data output by the vision system, and obtain a comparison result;

[0018] S22. Determine, based on the comparison result, whether the inertial measurement unit data obtained by the current inertial measurement unit detection contains an acceleration change feature that is inconsistent with the overall low-speed motion trend of the mobile platform, and if the displacement direction of the mobile platform indicated by the acceleration change feature is inconsistent with the vertical motion image data output by the vision system, mark the acceleration change feature as an invalid vibration characteristic pattern;

[0019] S23. The identified one or more invalid vibration characteristic patterns are stored to form a vibration suppression pattern library.

[0020] Furthermore, the steps of constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, wherein the invalid vibration characteristic pattern indicates that acceleration change characteristics inconsistent with the overall low-speed motion trend of the mobile platform appear in the inertial measurement unit data detected by the inertial measurement unit, specifically include:

[0021] S221. Periodically obtain vertical acceleration data output by the inertial measurement unit of the drone and vertical motion image data output by the visual system of the drone;

[0022] S222. Perform frequency analysis on the vertical acceleration data output by the inertial measurement unit to obtain frequency components and energy distribution of the vertical acceleration data output by the inertial measurement unit; perform motion trend analysis on the vertical motion image data output by the visual system to obtain a low-frequency motion trend of the mobile platform;

[0023] S223. Determine, based on the frequency components and energy distribution of the inertial measurement unit data and the low-frequency motion trend of the mobile platform, whether an extreme value in the energy distribution of the vertical acceleration data exceeds a preset energy threshold and the vertical motion image data of the visual system does not show corresponding vertical motion;

[0024] S224: If it exists, extract the feature of the extreme value in the vertical acceleration data as the currently identified acceleration change feature, and mark the acceleration change feature as an invalid vibration feature mode.

[0025] Furthermore, the confidence threshold is set based on the completeness of the visual markings obtained by the visual system;

[0026] The visual marking completeness indicator is one or more of the percentage of the number of visual feature points detected by the visual system to the total number of feature points on the mobile platform and the number of frames in which the visual system continuously recognizes a set number of feature points on the mobile platform.

[0027] Furthermore, when the vertical vibration of the mobile platform is determined to occur, the step of instructing the drone not to execute the vertical attitude or altitude adjustment command generated in response to the vertical vibration of the mobile platform, and the drone maintaining a preset landing configuration includes:

[0028] S61. When it is determined that the mobile platform is vibrating vertically, reduce the weight of the vertical acceleration data output by the current inertial measurement unit in the attitude and altitude control solution of the UAV;

[0029] S62: Instruct the UAV not to execute the rapid vertical attitude or altitude adjustment instruction corresponding to the vertical vibration of the mobile platform, and the UAV maintains the preset landing configuration.

[0030] Furthermore, after the step of instructing the drone not to execute the rapid vertical attitude or altitude adjustment command generated in response to the vertical vibration of the mobile platform when the vertical vibration is determined to be occurring, and the drone maintaining a preset landing configuration, the method further includes:

[0031] S7. When the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the estimated position and attitude of the current UAV are calibrated using the restored visual positioning information output by the visual system.

[0032] Furthermore, when the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the step of calibrating the estimated position and attitude of the current UAV using the restored visual positioning information output by the visual system includes:

[0033] S71. When the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the restored visual positioning information is acquired;

[0034] S72. Calculating a deviation between the restored visual positioning information and the inferred relative position and attitude of the UAV relative to the mobile platform based on the restored visual positioning information and the inferred relative position and attitude;

[0035] S73. Determine calibration parameters based on the deviation and the confidence level of the restored visual positioning information;

[0036] S74. Within a preset time window, gradually adjust the estimated position and attitude of the UAV according to the calibration parameters.

[0037] Furthermore, the step of determining calibration parameters according to the deviation and the confidence of the restored visual positioning information includes:

[0038] S731. Determine an initial adjustment weight according to the magnitude of the deviation;

[0039] S732: Modify the initial adjustment weights according to the confidence level of the restored visual positioning information to obtain final calibration parameters.

[0040] Furthermore, the step of gradually adjusting the estimated position and attitude of the UAV according to the calibration parameters within the preset time window includes:

[0041] S741. Generate an intermediate target point within the preset time window based on the currently calculated relative position and attitude of the UAV relative to the mobile platform, the recovered visual positioning information, and the calibration parameters;

[0042] S742: The UAV is caused to follow the intermediate target points in sequence, so that the relative position and attitude of the UAV and the mobile platform smoothly transition from the estimated relative position and attitude to a position and attitude consistent with the restored visual positioning information.

[0043] As a second aspect of this application, a drone take-off and landing control system is provided for enabling the drone to stably follow a target platform and achieve precise landing. The system is applied to drones and mobile platforms equipped with an inertial measurement unit and a vision system, and includes:

[0044] a vibration suppression pattern library construction module, the vibration suppression pattern library construction module being used to construct a vibration suppression pattern library and configure a conditional trigger for applying the vibration suppression pattern library, the vibration suppression pattern library including a plurality of invalid vibration characteristic patterns, the invalid vibration characteristic patterns being indicated by acceleration change characteristics appearing in inertial measurement unit data obtained by an inertial measurement unit of the mobile platform that are inconsistent with an overall low-speed motion trend of the mobile platform;

[0045] a vibration suppression pattern library activation module, the vibration suppression pattern library activation module being configured to activate a vibration suppression mode when it is determined that the distance between the UAV and the mobile platform is less than a preset distance threshold and the visual positioning information about the mobile platform obtained by the visual system is less than a preset confidence threshold, wherein the vibration suppression mode indication is the application of the vibration suppression pattern library when a conditional trigger is triggered by the UAV and the mobile platform in the current operating state;

[0046] a comparison and analysis module, the comparison and analysis module being configured to receive and analyze vertical acceleration data output by the inertial measurement unit in real time in the vibration suppression mode, and to perform real-time comparison between the vertical acceleration data currently output by the inertial measurement unit and invalid vibration characteristic modes in the vibration suppression pattern library;

[0047] a comparison and judgment module, configured to determine that the vertical acceleration data output by the current inertial measurement unit indicates vertical vibration of the mobile platform when the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern meet a preset matching threshold;

[0048] The drone landing execution module is used to instruct the drone not to execute the rapid vertical attitude or altitude adjustment instructions corresponding to the vertical vibration of the mobile platform when it is determined that the mobile platform is vibrating vertically, and the drone maintains a preset landing shape.

[0049] The present application discloses a drone takeoff and landing control system, comprising a vibration suppression pattern library construction module, a vibration suppression pattern library activation module, a comparison and analysis module, a comparison and judgment module, and a drone landing execution module. By constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, a vibration suppression mode is activated when it is determined that the distance between the drone and the mobile platform is less than a preset distance threshold and the visual positioning information about the mobile platform obtained by the vision system is less than a preset confidence threshold. In this mode, vertical acceleration data output by an inertial measurement unit is received and analyzed in real time. The vertical acceleration data output by the current inertial measurement unit is compared in real time with invalid vibration characteristic patterns in the vibration suppression pattern library. When a preset matching threshold is met, the vertical vibration of the mobile platform is determined to be present. The drone is then instructed not to execute the vertical attitude or altitude adjustment command generated in response to the vertical vibration of the mobile platform, and the drone maintains a preset landing configuration. This method effectively improves the takeoff and landing stability of the drone in complex environments by identifying and suppressing invalid vibrations while maintaining a stable landing configuration when visual information is unreliable.

[0050] For better understanding and implementation, the present invention is described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 1 is a flow chart of a method for controlling the take-off and landing of a UAV in this embodiment;

[0052] Figure 2 is a flow chart indicating step S2 in a method for controlling the take-off and landing of a UAV in this embodiment;

[0053] Figure 3 is a flow chart indicating step S6 in a method for controlling the take-off and landing of a UAV in this embodiment;

[0054] Figure 4 1 is a flow chart of a method for controlling the take-off and landing of a UAV in this embodiment, including step S7;

[0055] Figure 5 is a flow chart indicating step S7 in a method for controlling the take-off and landing of a UAV in this embodiment;

[0056] Figure 6 This is a system structure block diagram of a UAV take-off and landing control system in this embodiment.

[0057] Description of the accompanying drawings: 101. Vibration suppression pattern library construction module; 102. Vibration suppression pattern library activation module; 103. Comparison and analysis module; 104. Comparison and judgment module; 105. UAV landing execution module. DETAILED DESCRIPTION

[0058] In order to better illustrate the present invention, the present invention is described in further detail below with reference to the accompanying drawings.

[0059] It should be clear that in order to make the purpose, technical solutions and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. The components of the embodiments of the present disclosure generally described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present disclosure provided in the drawings is not intended to limit the scope of the present disclosure for protection, but merely represents selected embodiments of the present disclosure. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present disclosure.

[0060] The following is a specific example for description. In this example:

[0061] First, as Figure 1 As shown, a method for controlling the take-off and landing of a UAV is provided, comprising:

[0062] S1. Provide drones and mobile platforms equipped with inertial measurement units and vision systems;

[0063] S2. Build a vibration suppression pattern library and configure a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, wherein the invalid vibration characteristic patterns indicate acceleration change characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform in inertial measurement unit data detected by the inertial measurement unit;

[0064] S3. When it is determined that the distance between the UAV and the mobile platform is less than a preset distance threshold, and the visual positioning information about the mobile platform obtained by the visual system is less than a preset confidence threshold, a vibration suppression mode is enabled, where the vibration suppression mode indicates that a vibration suppression mode library is applied when a conditional trigger is triggered in the current operating state of the UAV and the mobile platform;

[0065] S4. In the vibration suppression mode, receiving and analyzing vertical acceleration data output by the inertial measurement unit in real time, and performing real-time comparison between the vertical acceleration data currently output by the inertial measurement unit and invalid vibration characteristic modes in the vibration suppression pattern library;

[0066] S5. When it is determined that the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern meet a preset matching threshold, it is determined that the vertical acceleration data output by the current inertial measurement unit indicates vertical vibration of the mobile platform;

[0067] S6. When it is determined that the mobile platform is vibrating vertically, the UAV is instructed not to execute the vertical attitude or altitude adjustment instruction generated corresponding to the vertical vibration of the mobile platform, and the UAV maintains the preset landing configuration.

[0068] First, this application provides a drone take-off and landing control method, which aims to solve the control problems caused by platform vibration and visual information uncertainty when the drone takes off and lands on a mobile platform. By constructing a vibration suppression pattern library and dynamically adjusting the control strategy in combination with the visual information confidence, the stability of take-off and landing is improved.

[0069] In this method, an inertial measurement unit (IMU) is a sensor capable of measuring an object's acceleration and angular velocity. It can be implemented using a micro-electromechanical system (MEMS) inertial sensor, a fiber optic gyroscope, or a laser gyroscope. Its primary purpose is to obtain information about the motion state of a mobile platform, including changes in its vertical acceleration. A visual system is a system that uses optical sensors to acquire image information and process it to achieve positioning and recognition. Its primary purpose is to obtain visual positioning information about the drone relative to the mobile platform.

[0070] By combining a pre-built vibration suppression pattern library with a conditional triggering mechanism based on the distance between the UAV and the mobile platform and the confidence of the visual system, it is possible to effectively identify and suppress the interference of the vertical vibration of the mobile platform on the UAV control instructions during the critical landing phase when the UAV approaches the mobile platform and the visual positioning information is unreliable, thereby achieving the effect of improving the UAV's take-off and landing stability on complex mobile platforms.

[0071] First, a mobile platform equipped with an inertial measurement unit (IMU) and a drone equipped with a vision system are provided, laying the foundation for subsequent data acquisition and control. Before the drone begins landing, the system pre-builds a vibration suppression pattern library and stores multiple invalid vibration signature patterns. These patterns are pre-identified acceleration variation characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform and represent vertical vibrations that may be generated by the mobile platform during operation. Furthermore, the system configures conditional triggers based on the vibration suppression pattern library, enabling the vibration suppression function to be activated under specific conditions. When the distance between the drone and the mobile platform falls below a preset distance threshold and the visual positioning information of the mobile platform obtained by the vision system falls below a preset confidence threshold, the current environment is determined to pose a challenge to the drone's precise landing. At this point, vibration suppression mode is activated. In vibration suppression mode, the drone receives and analyzes vertical acceleration data output by the mobile platform's IMU in real time. This real-time data is compared with the invalid vibration signature patterns in the vibration suppression pattern library. The system determines whether the vertical acceleration data output by the IMU meets a preset matching threshold with the invalid vibration signature patterns. Once this matching threshold is met, the system determines that the vertical acceleration data output by the IMU indicates vertical vibration of the mobile platform. Once this is determined to be vertical vibration, suppression measures are immediately taken, instructing the drone not to execute the rapid vertical attitude or altitude adjustments generated by this vertical vibration. This means that even if the IMU reports a dramatic change in vertical acceleration, the drone will not blindly make compensatory adjustments.

[0072] Through the above technical solutions, the present application can effectively solve the challenges faced by drones when taking off and landing on mobile platforms. First, by constructing a vibration suppression pattern library and combining it with a conditional trigger mechanism, the present solution can intelligently identify and suppress invalid inertial measurement unit data caused by vertical vibrations of the mobile platform, preventing the drone from making unnecessary rapid vertical attitude or altitude adjustments to these interference signals. Secondly, during the critical landing phase where the confidence level of visual positioning information is low, the present solution can switch to vibration suppression mode, effectively dealing with the problem of intermittent loss or jumps of visual information, and ensuring that the drone can maintain a stable landing trajectory when visual information is unreliable. Ultimately, the drone can follow the mobile platform more accurately and achieve a smooth landing.

[0073] In this embodiment, if Figure 2 As shown, the steps of constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, wherein the invalid vibration characteristic pattern indicates that acceleration change characteristics inconsistent with the overall low-speed motion trend of the mobile platform appear in the inertial measurement unit data detected by the inertial measurement unit, include:

[0074] S21. Before receiving a landing request message from the drone to the mobile platform and when the mobile platform passes through the seam, compare the vertical acceleration data output by the inertial measurement unit with the vertical motion image data output by the vision system, and obtain a comparison result;

[0075] S22. Determine, based on the comparison result, whether the inertial measurement unit data obtained by the current inertial measurement unit detection contains an acceleration change feature that is inconsistent with the overall low-speed motion trend of the mobile platform, and if the displacement direction of the mobile platform indicated by the acceleration change feature is inconsistent with the vertical motion image data output by the vision system, mark the acceleration change feature as an invalid vibration characteristic pattern;

[0076] S23. The identified one or more invalid vibration characteristic patterns are stored to form a vibration suppression pattern library.

[0077] Before the drone sends a landing request to the mobile platform and while the platform passes through a joint, the vertical acceleration data output by the inertial measurement unit (IMU) is compared with the vertical motion image data output by the vision system to obtain a comparison result. This comparison process exploits the complementary nature of the two sensor types: the IMU is sensitive to instantaneous acceleration changes, while the vision system provides relatively stable image information reflecting the overall low-speed motion of the platform. Based on this comparison result, it is possible to determine whether there are acceleration variation signatures in the IMU data that are inconsistent with the overall low-speed motion trend of the mobile platform. Furthermore, the displacement direction indicated by this signature is inconsistent with the vertical motion image data output by the vision system. This inconsistency is a key criterion for identifying invalid vibration signature patterns, indicating that the vertical vibration changes detected by the IMU are not caused by the actual displacement of the entire platform, but rather by local vibration or turbulence. Once such signatures are identified, they are marked as invalid vibration signature patterns and stored to form a vibration suppression pattern library.

[0078] By constructing a vibration suppression pattern library in this manner, the subsequent drone can activate vibration suppression mode during landing if it determines that the distance between the drone and the mobile platform is below a preset distance threshold and the visual positioning information is below a preset confidence threshold. In vibration suppression mode, the drone can receive and analyze the vertical acceleration data output by the inertial measurement unit in real time, and compare it with the invalid vibration characteristic patterns in the pre-built vibration suppression pattern library in real time. When the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern meet the preset matching threshold, the system can accurately determine that the vertical acceleration data output by the current inertial measurement unit indicates vertical vibration of the mobile platform. Based on this judgment, the drone is instructed not to execute the rapid vertical attitude or altitude adjustment instructions generated in response to the vertical vibration of the mobile platform, and the drone maintains the preset landing configuration.

[0079] This solution uses data from two different sensor types, an inertial measurement unit (IMU) and a visual system, to cross-compare at specific moments when the mobile platform passes through joints, potentially generating vertical vibration. This allows for precise identification of high-frequency acceleration changes in the IMU data that are caused by localized vibration rather than overall motion. This multi-source data comparison approach avoids potential misjudgments based on single-sensor data and improves the accuracy of identifying invalid vibration signature patterns. These identified invalid vibration signature patterns are stored and constructed into a vibration suppression pattern library, enabling the drone to rely on reliable pattern library data during subsequent landings.

[0080] In this embodiment, the step of determining, based on the comparison result, that the inertial measurement unit data obtained by the detection of the mobile platform contains an acceleration change feature that is inconsistent with the overall low-speed motion trend of the mobile platform, and the displacement direction of the mobile platform indicated by the acceleration change feature is inconsistent with the vertical motion image data output by the vision system, and marking the acceleration change feature as an invalid vibration characteristic pattern specifically includes:

[0081] S221. Periodically obtain vertical acceleration data output by the inertial measurement unit of the drone and vertical motion image data output by the visual system of the drone;

[0082] S222. Perform frequency analysis on the vertical acceleration data output by the inertial measurement unit to obtain frequency components and energy distribution of the vertical acceleration data output by the inertial measurement unit; perform motion trend analysis on the vertical motion image data output by the visual system to obtain a low-frequency motion trend of the mobile platform;

[0083] S223. Determine, based on the frequency components and energy distribution of the inertial measurement unit data and the low-frequency motion trend of the mobile platform, whether an extreme value in the energy distribution of the vertical acceleration data exceeds a preset energy threshold and the vertical motion image data of the visual system does not show corresponding vertical motion;

[0084] S224: If it exists, extract the feature of the extreme value in the vertical acceleration data as the currently identified acceleration change feature, and mark the acceleration change feature as an invalid vibration feature mode.

[0085] By periodically acquiring vertical acceleration data from the mobile platform's inertial measurement unit (IMU) and vertical motion image data from the drone's vision system, the system ensures real-time and synchronized data acquisition, laying the foundation for subsequent refined analysis. Furthermore, frequency analysis of the IMU's vertical acceleration data accurately captures its frequency components and energy distribution, effectively separating vertical vibration from low-frequency motion. Motion trend analysis of the vertical motion image data from the vision system reveals the mobile platform's true low-frequency motion trends, providing a critical external reference for determining whether high-frequency components in the IMU data represent invalid vibrations. By comprehensively comparing the high-frequency energy in the IMU data with the corresponding vertical motion in the vision system data, the system can accurately determine whether high-frequency energy in the vertical acceleration data exceeds a preset energy threshold, while the vision system does not indicate corresponding vertical motion. This dual-source data cross-validation mechanism enables the system to effectively distinguish actual low-speed motion of the mobile platform from transient vertical vibrations caused by factors such as uneven ground. If such a situation is determined to exist, the high-frequency feature is extracted and marked as an invalid vibration feature pattern, thereby preventing the drone from making incorrect posture or altitude adjustments to these vibrations that should not be followed, ensuring the stability of the drone's takeoff and landing on the mobile platform.

[0086] In a specific example, vertical acceleration data from the mobile platform's inertial measurement unit (IMU) can be periodically acquired at a frequency of 100 Hz, while vertical motion image data from the drone's visual system can be acquired at a frame rate of 30 frames per second. A fast Fourier transform (FFT) can be used to perform frequency analysis on the IMU's vertical acceleration data to obtain its frequency components and corresponding energy distribution within the 0-50 Hz range. Simultaneously, the low-frequency motion trends of the mobile platform within the 0-5 Hz range can be determined based on the vertical motion image data from the visual system. The energy distribution of components with frequencies above 10 Hz in the IMU data is compared with a preset energy threshold, which can be set to an empirical value determined through experimental calibration or statistical analysis of historical data. Furthermore, a determination is made as to whether the visual system's vertical motion image data exhibits significant vertical motion with a frequency above 10 Hz within the corresponding time period. If the inertial measurement unit data shows that the frequency energy exceeds the threshold, but the visual system does not show the corresponding vertical motion, the features in the inertial measurement unit vertical acceleration data with a frequency higher than 10 Hz and an energy exceeding the threshold are extracted as the currently identified acceleration change features and marked as invalid vibration feature patterns so that the subsequent vibration suppression module can recognize and ignore these features.

[0087] In some of the aforementioned embodiments, it is mentioned that when the distance between the UAV and the mobile platform is lower than a preset distance threshold and the visual positioning information obtained by the visual system is lower than a preset confidence threshold, the vibration suppression mode is enabled. The preset confidence threshold can be a fixed empirical value, such as 0.7, so that whether to enable the vibration suppression mode can be determined based on the reliability of the visual positioning information.

[0088] In this embodiment, the confidence threshold is preferably set based on the completeness of the visual mark obtained by the visual system;

[0089] The visual marking completeness indicator is one or more of the percentage of the number of visual feature points detected by the visual system to the total number of feature points on the mobile platform and the number of frames in which the visual system continuously recognizes a set number of feature points on the mobile platform.

[0090] Among them, the confidence threshold refers to the standard for judging the reliability of visual positioning information. Its purpose is to provide the system with a basis for judging whether the visual data is usable;

[0091] Among them, visual marker integrity is a quantitative indicator that indicates the clarity and completeness of the visual system's recognition of visual markers. Its purpose is to reflect the quality of visual positioning information.

[0092] The solution dynamically adjusts the confidence threshold, allowing drones to more intelligently evaluate the reliability of visual positioning information.

[0093] In some specific examples, the vision system on a drone can be equipped with an image processor capable of real-time analysis of mobile platform images captured by an onboard camera. During image processing, feature points on the visual marker can be first identified using a feature extraction algorithm, such as the Scale-Invariant Feature Transform (SIFT) or Speeded Up Robust Features (SURF) algorithm. The ideal total number of feature points that the mobile platform's visual marker should contain can be determined in advance. One measure of visual marker completeness can be calculated as the percentage of the currently detected number of visual feature points relative to the total number of mobile platform feature points. For example, if the mobile platform has 100 total feature points and 85 are currently detected, the percentage is 85%. The vision system can continuously track these feature points and record the number of frames in which a set number of feature points on the mobile platform (e.g., at least 80% of the feature points) are continuously recognized. When the visual marker completeness (e.g., the feature point percentage) is high and the number of consecutive recognition frames is long, the confidence threshold can be dynamically set to a higher value, such as 0.8. Conversely, when the visual marker integrity is low or the number of consecutive recognition frames is short, the confidence threshold can be set to a lower value, such as 0.6. This allows the drone to make adaptive adjustments based on the actual condition of the visual markers when determining whether the visual positioning information is reliable, thereby more accurately deciding whether to enable vibration suppression mode.

[0094] In this embodiment, if Figure 3 As shown, when it is determined that the mobile platform is vibrating vertically, the UAV is instructed not to execute the vertical attitude or altitude adjustment instruction generated corresponding to the vertical vibration of the mobile platform, and the UAV maintains a preset landing configuration, including:

[0095] S61. When it is determined that the mobile platform is vibrating vertically, reduce the weight of the vertical acceleration data output by the current inertial measurement unit in the attitude and altitude control solution of the UAV;

[0096] S62: Instruct the UAV not to execute the rapid vertical attitude or altitude adjustment instruction generated corresponding to the vertical vibration of the mobile platform, and the UAV maintains the preset landing configuration.

[0097] By fine-tuning the inertial measurement unit data and optimizing the drone control strategy, stable landing can be achieved under vertical vibration of the mobile platform.

[0098] Specifically, when the system determines that the mobile platform is experiencing vertical vibration, this is usually based on the result of comparing the vertical acceleration data output by the inertial measurement unit with pre-constructed invalid vibration characteristic patterns, indicating that the current inertial measurement unit data contains a large amount of noise components that should not be followed by the drone. In this case, the weight of the vertical acceleration data output by the current inertial measurement unit in the drone's attitude and altitude control solution is reduced. This means that the influence of the inertial measurement unit data on the drone's control system decision-making is weakened, effectively filtering out the instantaneous, drastic, and unhelpful acceleration change information caused by vertical vibration.

[0099] In a specific example, when the drone control system determines through comparative analysis that the mobile platform is experiencing vertical vibration, for example, when the vertical acceleration data output by the inertial measurement unit matches the invalid vibration characteristic pattern in the vibration suppression pattern library to a preset threshold, the control system can dynamically adjust the parameters of its state estimation and control loop. Specifically, an adaptive Kalman filter algorithm can be used to reduce the weight of the vertical acceleration data currently output by the inertial measurement unit in the attitude and altitude control solution of the drone. In this algorithm, when vertical vibration is detected, the gain associated with the inertial measurement unit vertical acceleration measurement noise covariance matrix in the Kalman filter can be reduced in real time, or the terms associated with the vertical position and velocity in the state estimation covariance matrix can be increased, thereby reducing the contribution of the inertial measurement unit data to the drone state estimation.

[0100] Through the above technical solution, the present application can effectively solve the problem that it is difficult for drones to land stably under vertical vibration of the mobile platform. Specifically, by reducing the weight of the vertical acceleration data output by the inertial measurement unit in the drone attitude and altitude control solution, the interference of the noise caused by vertical vibration on the drone control system can be effectively suppressed, and the drone can be prevented from making unnecessary drastic attitude or altitude adjustments due to incorrect response to invalid vibration. At the same time, by instructing the drone not to execute the rapid adjustment instructions generated by vertical vibration and maintain the preset landing shape, it is ensured that the drone can still maintain a smooth and controllable descent trajectory when it is disturbed by vibration, thereby significantly improving the stability of the drone's landing in complex dynamic environments and reducing the risk of landing failure.

[0101] In this embodiment, if Figure 4 As shown, after the step of instructing the drone not to execute the rapid vertical attitude or altitude adjustment command corresponding to the vertical vibration of the mobile platform when the vertical vibration is determined to be caused by the mobile platform, and the drone maintains a preset landing configuration, the method further includes:

[0102] S7. When the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the estimated position and attitude of the current UAV are calibrated using the restored visual positioning information output by the visual system.

[0103] By further introducing a position and attitude calibration mechanism based on positioning information recovered by the drone's visual system, the system addresses the cumulative error caused by the drone's long-term reliance on inertial measurement unit (IMU) inference during a drone's landing configuration, while avoiding vibration-induced attitude instability, the IMU's position and attitude estimates inevitably drift and accumulate errors over time. When external environmental conditions improve, the drone's visual system is able to reliably identify the mobile platform, and the confidence level of the visual positioning information it outputs reaches a preset confidence threshold. At this point, the system compares this recovered, high-confidence visual positioning information with the drone's current position and attitude inferred by the IMU. This comparison accurately calculates the deviation between the inferred and true values. The system then corrects the drone's inferred position and attitude based on this deviation.

[0104] Specifically, step S7 is further explained. In this embodiment, if Figure 5 As shown, when the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the step of calibrating the estimated position and attitude of the current UAV using the restored visual positioning information output by the visual system includes:

[0105] S71. When the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the restored visual positioning information is acquired;

[0106] S72. Calculating a deviation between the restored visual positioning information and the inferred relative position and attitude of the UAV relative to the mobile platform based on the restored visual positioning information and the inferred relative position and attitude;

[0107] S73. Determine calibration parameters based on the deviation and the confidence level of the restored visual positioning information;

[0108] S74. Within a preset time window, gradually adjust the estimated position and attitude of the UAV according to the calibration parameters.

[0109] Through a refined calibration strategy, the calibration problems that may occur after the visual positioning information of the drone is restored are solved, thereby improving flight stability.

[0110] Specifically, when the vision system obtains visual positioning information about the mobile platform and recovers it to a preset confidence threshold, the system first acquires this recovered visual positioning information. This step ensures that the calibration process is triggered only when the visual information is sufficiently reliable, avoiding the introduction of new errors due to the use of low-quality data. The system then calculates the deviation between this recovered visual positioning information and the drone's current estimated relative position and attitude with respect to the mobile platform. This deviation quantifies the potential accumulated positioning error of the drone during the period when visual information was unavailable. Furthermore, the core of the solution lies in intelligently determining calibration parameters based on this deviation and the confidence level of the recovered visual positioning information. This means that the intensity and method of calibration are no longer fixed, but are dynamically adjusted based on the error level and the reliability of the visual information. Furthermore, to ensure a smooth calibration process, the solution gradually adjusts the drone's estimated position and attitude within a preset time window according to the determined calibration parameters. This gradual adjustment method avoids sudden jumps in the drone's position and attitude, allowing the drone to smoothly transition from its estimated state to the accurate state indicated by the recovered visual information.

[0111] In a specific example, when the drone's vision system regains visual positioning information about the mobile platform and the confidence level of this information reaches a preset confidence threshold—for example, when the vision system successfully identifies and stably tracks more than a certain number of visual feature points on the mobile platform, or when its positioning error for multiple consecutive frames is below a preset value—the system acquires this recovered visual positioning information. The drone's control system then calculates the deviation between the currently estimated relative position and attitude of the drone relative to the mobile platform and the recovered visual positioning information. This deviation can be a vector representing six degrees of freedom (position deviation in the X, Y, and Z directions, and attitude deviation in roll, pitch, and yaw angles) obtained by directly comparing the two sets of pose data. Based on this deviation, the system dynamically determines calibration parameters based on the calculated deviation and the confidence level of the recovered visual positioning information. For example, a nonlinear function or lookup table mechanism can be designed that takes the absolute value of the deviation and the confidence level of the visual positioning information as input and outputs a calibration gain between 0 and 1. When the deviation is large and the confidence level is high, the calibration gain can be set relatively large to achieve faster calibration. When the confidence level is low, the calibration gain is limited to a smaller value even if the deviation is large to avoid overcalibration due to unreliable information. Finally, to ensure the smoothness of the calibration process, the system gradually adjusts the drone's estimated position and attitude based on the determined calibration parameters within a preset time window. For example, the calibration amount can be broken down into multiple small steps, and a calibration amount of one step can be applied every control cycle within the preset time window, or a smooth trajectory can be generated along which the drone's position and attitude gradually transition from the current estimated value to the target value indicated by the restored visual information. This method can effectively avoid sudden displacement or attitude changes of the drone during the calibration process, thereby ensuring flight stability.

[0112] In this embodiment, the step of determining the calibration parameters according to the deviation and the confidence of the restored visual positioning information includes:

[0113] S731. Determine an initial adjustment weight according to the magnitude of the deviation;

[0114] S732: Modify the initial adjustment weights according to the confidence level of the restored visual positioning information to obtain final calibration parameters.

[0115] The magnitude of the deviation refers to the quantitative representation of the difference between the estimated position and attitude of the drone and the recovered visual positioning information. Its purpose is to quantify the degree of deviation between the current state of the drone and the target state.

[0116] The initial adjustment weight refers to the calibration intensity factor initially set based on the magnitude of the deviation, which aims to reflect that the larger the deviation, the larger the adjustment required in theory;

[0117] To illustrate, when a drone needs to calibrate its estimated position and attitude relative to a mobile platform, it can first calculate the difference between the current estimated drone position and attitude and the recovered visual positioning information and perform a weighted average to obtain a deviation value. An initial adjustment weight can then be determined based on this deviation value. For example, a piecewise function can be pre-set, where the initial adjustment weight is set to a smaller value when the deviation is small, a medium value when the deviation is medium, and a larger value when the deviation is large. To adjust this initial adjustment weight, the system can then obtain the confidence level of the recovered visual positioning information. For example, the visual system can output a confidence score based on the number of identified visual feature points, the uniformity of their distribution in the image, and the stability of feature point matching between consecutive frames. This confidence score can be a number between 0 and 1, with 1 indicating high confidence and 0 indicating low confidence. This confidence score can then be used to adjust the initial adjustment weight. Specifically, the initial adjustment weight can be multiplied by the confidence score, or a weighted average model can be used. For example, if the initial adjustment weight is 0.8 and the confidence of the visual positioning information is 0.5, then the final calibration parameter can be calculated as 0.8 multiplied by 0.5. In this way, when the reliability of the visual positioning information is not high, even if the deviation is large, the final calibration parameter will be appropriately reduced to avoid excessive adjustment. Conversely, if the confidence is higher, such as 0.9, then the final calibration parameter will be 0.8 multiplied by 0.9, so that the calibration strength is closer to the initial adjustment weight to make full use of reliable visual information. In this way, the process of determining the final calibration parameters can dynamically adapt to the quality of the visual positioning information to ensure the smoothness of the calibration.

[0118] In this embodiment, the step of gradually adjusting the estimated position and attitude of the drone according to the calibration parameters within the preset time window includes:

[0119] S741. Generate an intermediate target point within the preset time window based on the currently calculated relative position and attitude of the UAV relative to the mobile platform, the recovered visual positioning information, and the calibration parameters;

[0120] S742: The UAV is caused to follow the intermediate target points in sequence, so that the relative position and attitude of the UAV and the mobile platform smoothly transition from the estimated relative position and attitude to a position and attitude consistent with the restored visual positioning information.

[0121] Among them, the intermediate target points refer to a series of discrete or continuous transition points generated by calculation between the current estimated position and attitude of the drone and the position and attitude indicated by the restored visual positioning information. These points constitute the trajectory nodes that the drone needs to reach in sequence during the calibration process. The purpose is to decompose a one-time large-scale position and attitude adjustment into multiple small, controllable adjustment steps, thereby avoiding sudden changes in the drone's movement.

[0122] By generating a series of intermediate target points within a preset time window based on the current estimated relative position and attitude of the drone relative to the mobile platform, the recovered visual positioning information, and calibration parameters, the drone's position and attitude calibration process is broken down into multiple controllable, small adjustment stages. It is precisely because of the introduction of these intermediate target points, and having the drone follow them sequentially, that the relative position and attitude of the drone and the mobile platform can smoothly transition from the estimated relative position and attitude to a position and attitude consistent with the recovered visual positioning information.

[0123] In one specific example, when the visual positioning information about the mobile platform obtained by the vision system returns to a preset confidence threshold, the drone control system initiates the position and attitude calibration process. Specifically, within a preset time window, for example, set to 2 seconds, the control system generates a series of intermediate target points based on the drone's current relative position and attitude relative to the mobile platform, calculated through inertial navigation, the restored visual positioning information at that moment, and previously determined calibration parameters. For example, assuming that N intermediate target points need to be generated within the preset time window, the position and attitude of the kth intermediate target point can be calculated as: P_k = P_imu + (P_vis - P_imu) * (k / N) * α. Where P_imu represents the estimated position and attitude, P_vis represents the visual positioning information, and k increases from 1 to N. As k increases, the intermediate target point gradually approaches the position and attitude indicated by the visual positioning information from the estimated position and attitude, with the speed and degree of approach controlled by the calibration parameter α. The drone flight control system then sequentially follows these generated intermediate target points. This means that the drone does not jump directly to the final position and attitude indicated by the visual positioning information. Instead, it adjusts its current position and attitude to the next intermediate target point within each time step. For example, the flight controller calculates the attitude and thrust commands required to move from the current position to the next intermediate target point and executes them smoothly. In this way, the relative position and attitude of the drone and the mobile platform can gradually transition from the calculated relative position and attitude to a position and attitude consistent with the recovered visual positioning information in a controlled and smooth manner, thus avoiding the body shaking or attitude instability caused by sudden calibration.

[0124] Second, if Figure 6 As shown, a UAV take-off and landing control system 100 is provided, which is used to achieve stable tracking of the UAV on the target platform and achieve precise landing. The system is applied to UAVs and mobile platforms equipped with an inertial measurement unit and a vision system. The system includes:

[0125] a vibration suppression pattern library construction module 101, configured to construct a vibration suppression pattern library and configure a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, each of which indicates acceleration variation characteristics in inertial measurement unit data obtained by an inertial measurement unit of the mobile platform that are inconsistent with the overall low-speed motion trend of the mobile platform;

[0126] a vibration suppression pattern library activation module 102, configured to activate a vibration suppression mode when it is determined that the distance between the UAV and the mobile platform is less than a preset distance threshold and the visual positioning information about the mobile platform obtained by the visual system is less than a preset confidence threshold, wherein the vibration suppression mode indicates that the vibration suppression pattern library is applied when a conditional trigger is triggered by the UAV and the mobile platform in the current operating state;

[0127] A comparison and analysis module 103 is configured to receive and analyze vertical acceleration data output by the inertial measurement unit in real time in the vibration suppression mode, and perform real-time comparison between the vertical acceleration data currently output by the inertial measurement unit and invalid vibration characteristic modes in the vibration suppression mode library;

[0128] A comparison and judgment module 104 is configured to determine that the vertical acceleration data output by the current inertial measurement unit indicates vertical vibration of the mobile platform when the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern meet a preset matching threshold;

[0129] The drone landing execution module 105 is used to instruct the drone not to execute the rapid vertical attitude or altitude adjustment instructions corresponding to the vertical vibration of the mobile platform when it is determined that the mobile platform is vibrating vertically, and the drone maintains a preset landing shape.

[0130] In this embodiment, a drone takeoff and landing control system 100 includes a vibration suppression pattern library construction module 101, a vibration suppression pattern library activation module 102, a comparison and analysis module 103, a comparison and judgment module 104, and a drone landing execution module 105. By constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, when it is determined that the distance between the drone and the mobile platform is less than a preset distance threshold and the visual positioning information about the mobile platform obtained by the vision system is less than a preset confidence threshold, a vibration suppression mode is activated. In this mode, vertical acceleration data output by an inertial measurement unit is received and analyzed in real time. The vertical acceleration data output by the current inertial measurement unit is compared in real time with invalid vibration characteristic patterns in the vibration suppression pattern library. When a preset matching threshold is met, the vertical vibration of the mobile platform is determined to be present. The drone is instructed not to execute the vertical attitude or altitude adjustment command generated in response to the vertical vibration of the mobile platform, and the drone maintains a preset landing configuration. This method effectively improves the takeoff and landing stability of the drone in complex environments by identifying and suppressing invalid vibrations while maintaining a stable landing configuration when visual information is unreliable.

[0131] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present disclosure, which are used to illustrate the technical solutions of the present disclosure rather than to limit them. The protection scope of the present disclosure is not limited thereto. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, ordinary technicians in this field should understand that any technician familiar with this technical field can still modify the technical solutions recorded in the aforementioned embodiments within the technical scope disclosed in the present disclosure, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure, and should all be covered by the protection scope of the present disclosure.

Claims

1. A method for controlling the take-off and landing of an unmanned aerial vehicle, characterized in that: include: S1. Provide drones and mobile platforms equipped with inertial measurement units and vision systems; S2. Build a vibration suppression pattern library and configure a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, wherein the invalid vibration characteristic patterns indicate acceleration change characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform in inertial measurement unit data detected by the inertial measurement unit; S3. When it is determined that the distance between the UAV and the mobile platform is less than a preset distance threshold, and the visual positioning information about the mobile platform obtained by the visual system is less than a preset confidence threshold, a vibration suppression mode is enabled, where the vibration suppression mode indicates that a vibration suppression mode library is applied when a conditional trigger is triggered in the current operating state of the UAV and the mobile platform; S4. In the vibration suppression mode, the vertical acceleration data output by the inertial measurement unit is received and analyzed in real time, and the vertical acceleration data output by the current inertial measurement unit is compared with the invalid vibration characteristic patterns in the vibration suppression pattern library in real time; S5. When it is determined that the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern meet a preset matching threshold, the vertical acceleration data output by the current inertial measurement unit is determined to indicate vertical vibration of the mobile platform; S6. When it is determined to be vertical vibration of the mobile platform, the UAV is instructed not to execute the vertical attitude or altitude adjustment instructions corresponding to the vertical vibration of the mobile platform, and the UAV maintains the preset landing configuration.

2. The method for controlling the take-off and landing of a UAV according to claim 1, wherein: The steps of constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes multiple invalid vibration characteristic patterns, wherein the invalid vibration characteristic pattern indicates that the inertial measurement unit data obtained by the inertial measurement unit of the mobile platform detected and obtained have acceleration change characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform, include: S21, before receiving a landing request message sent by the drone to the mobile platform and when the mobile platform passes through a joint, comparing the vertical acceleration data output by the inertial measurement unit with the vertical motion image data output by the vision system, and obtaining a comparison result; S22, based on the comparison result, determining whether the inertial measurement unit data obtained by the current inertial measurement unit detected and obtained have acceleration change characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform, and if the displacement direction of the mobile platform indicated by the acceleration change characteristics is inconsistent with the vertical motion image data output by the vision system, marking the acceleration change characteristics as invalid vibration characteristic patterns; and S23, storing the identified one or more invalid vibration characteristic patterns to form the vibration suppression pattern library.

3. The method for controlling the take-off and landing of a UAV according to claim 2, wherein: The steps of constructing a vibration suppression pattern library and configuring a conditional trigger for applying the vibration suppression pattern library, wherein the vibration suppression pattern library includes a plurality of invalid vibration characteristic patterns, wherein the invalid vibration characteristic pattern indicates that the acceleration change characteristics inconsistent with the overall low-speed motion trend of the mobile platform appear in the inertial measurement unit data detected by the inertial measurement unit, specifically include: S221, periodically obtaining the vertical acceleration data output by the inertial measurement unit of the drone and the vertical motion image data output by the visual system of the drone; S222, performing frequency analysis on the vertical acceleration data output by the inertial measurement unit to obtain the vertical acceleration data output by the inertial measurement unit; S223, based on the frequency components and energy distribution of the inertial measurement unit data and the low-frequency motion trend of the mobile platform, determine whether there is an extreme value in the energy distribution in the vertical acceleration data that exceeds a preset energy threshold and the vertical motion image data of the visual system does not show the corresponding vertical motion; S224, if so, extract the feature of the extreme value in the vertical acceleration data as the currently identified acceleration change feature, and mark the acceleration change feature as an invalid vibration feature pattern.

4. The method for controlling the take-off and landing of an unmanned aerial vehicle according to claim 1, wherein: The confidence threshold is set based on the completeness of the visual markers obtained by the visual system; the visual marker completeness is indicated as one or more of the percentage of the number of visual feature points detected by the visual system to the total number of feature points on the mobile platform and the number of frames in which the visual system continuously recognizes a set number of feature points on the mobile platform.

5. The method for controlling the take-off and landing of a UAV according to claim 1, wherein: The step of instructing the drone not to execute the vertical attitude or altitude adjustment instruction corresponding to the vertical vibration of the mobile platform when it is determined to be the vertical vibration of the mobile platform, and the drone maintains a preset landing shape, includes: S61, when it is determined to be the vertical vibration of the mobile platform, reducing the weight of the vertical acceleration data output by the current inertial measurement unit in the attitude and altitude control solution of the drone; S62, instructing the drone not to execute the rapid vertical attitude or altitude adjustment instruction corresponding to the vertical vibration of the mobile platform, and the drone maintains the preset landing shape.

6. The method for controlling the take-off and landing of a UAV according to claim 1, wherein: After the step in which, when it is determined that the vertical vibration of the mobile platform occurs, the UAV is instructed not to execute the rapid vertical attitude or altitude adjustment instruction corresponding to the vertical vibration of the mobile platform, and the UAV maintains a preset landing shape, the method further includes: S7, when the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the visual positioning information output by the restored visual system is used to calibrate the estimated position and attitude of the current UAV.

7. The method for controlling the take-off and landing of a UAV according to claim 6, characterized in that: When the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the step of calibrating the estimated position and attitude of the current UAV using the restored visual positioning information output by the visual system includes: S71. When the visual positioning information about the mobile platform obtained by the visual system is restored to a preset confidence threshold, the restored visual positioning information is obtained; S72. Based on the restored visual positioning information and the estimated relative position and attitude of the UAV relative to the mobile platform, the deviation between the restored visual positioning information and the estimated relative position and attitude is calculated; S73. Based on the deviation and the confidence of the restored visual positioning information, the calibration parameters are determined; S74. Within a preset time window, the estimated position and attitude of the UAV is gradually adjusted according to the calibration parameters.

8. The method for controlling the take-off and landing of a UAV according to claim 7, characterized in that: The step of determining the calibration parameters based on the deviation and the confidence of the restored visual positioning information includes: S731, determining the initial adjustment weight based on the amplitude of the deviation; S732, correcting the initial adjustment weight based on the confidence of the restored visual positioning information to obtain the final calibration parameters.

9. The method for controlling the take-off and landing of a UAV according to claim 7, characterized in that: The step of gradually adjusting the estimated position and attitude of the UAV within a preset time window based on the calibration parameters includes: S741, generating an intermediate target point within the preset time window based on the currently estimated relative position and attitude of the UAV relative to the mobile platform, the restored visual positioning information and the calibration parameters; S742, causing the UAV to follow the intermediate target points in sequence, so that the relative position and attitude of the UAV and the mobile platform smoothly transition from the estimated relative position and attitude to a position and attitude consistent with the restored visual positioning information.

10. A UAV take-off and landing control system for achieving stable tracking of a target platform and precise landing by the UAV, and the system is applied to UAVs and mobile platforms equipped with an inertial measurement unit and a visual system, characterized in that: The system includes: a vibration suppression pattern library construction module, the vibration suppression pattern library construction module is used to construct a vibration suppression pattern library and configure a conditional trigger for applying the vibration suppression pattern library, the vibration suppression pattern library includes multiple invalid vibration characteristic patterns, the invalid vibration characteristic pattern indication is that the acceleration change characteristics that are inconsistent with the overall low-speed motion trend of the mobile platform appear in the inertial measurement unit data obtained by the inertial measurement unit of the mobile platform; a vibration suppression pattern library activation module, the vibration suppression pattern library activation module is used to enable the vibration suppression mode when it is determined that the distance between the drone and the mobile platform is lower than a preset distance threshold and the visual positioning information about the mobile platform obtained by the visual system is lower than a preset confidence threshold, the vibration suppression mode indication is that the drone and the mobile platform trigger the conditional trigger in the current operating state and apply the vibration suppression a comparison and analysis module, wherein the comparison and analysis module is used to receive and analyze the vertical acceleration data output by the inertial measurement unit in real time under the vibration suppression mode, and perform real-time comparison based on the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern in the vibration suppression pattern library; a comparison and judgment module, wherein the comparison and judgment module is used to judge that when the vertical acceleration data output by the current inertial measurement unit and the invalid vibration characteristic pattern meet a preset matching threshold, the vertical acceleration data output by the current inertial measurement unit is judged to indicate vertical vibration of the mobile platform; a UAV landing execution module, wherein when the UAV landing execution module is judged to be the vertical vibration of the mobile platform, the UAV is instructed not to execute the rapid vertical attitude or height adjustment instruction corresponding to the vertical vibration of the mobile platform, and the UAV maintains the preset landing shape.

Images