A method for detecting and recovering control of a drone in flight
By using inertial measurement unit data conversion and a phased control recovery algorithm, the inaccuracy problem of multi-rotor UAV launch detection and recovery was solved, achieving accurate detection and stable recovery of arbitrary launch attitudes, thus improving the flight reliability and safety of UAVs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG UNIV
- Filing Date
- 2026-04-03
- Publication Date
- 2026-07-21
AI Technical Summary
Existing methods for detecting the launch of multi-rotor UAVs rely on manual signal intervention, which cannot adapt to launch situations involving violent rotation or low speed. Launch recovery depends on inaccurate position and speed information observation, leading to unsafe control or failure.
Data detected by the inertial measurement unit is converted into the center of mass acceleration through a throw-and-fly transformation algorithm with external parameter calibration. The state of the UAV is calculated by combining the least squares estimation algorithm and the differential flatness assumption. The UAV is then controlled in stages to restore it to horizontal hovering, and intervention is carried out by a cascade feedback controller and a safety lever.
It enables accurate detection and stable recovery of arbitrary launch attitudes, reduces reliance on observation information, and improves the flight reliability and safety of UAVs under extreme conditions.
Smart Images
Figure CN121957095B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) control technology, and in particular relates to a method for detecting and controlling the recovery of UAVs after they are thrown into flight. Background Technology
[0002] Multirotor drones are vertical takeoff and landing (VTOL) aircraft that generate lift by driving rotors with multiple independent motors. They possess hovering, low-speed maneuvering, and stationary flight capabilities, and are widely used in aerial photography, inspection, rescue, and logistics. Multirotor drone drop detection and recovery technologies are commonly used in emergency rescue and rapid response, operations in confined spaces or complex terrain, industrial inspection, and portable operations. However, existing drop detection and recovery technologies still have the following problems:
[0003] (1) Limited throw detection: Most existing throw detection methods rely on manual signal intervention to determine the throw recovery time, which is essentially a human-driven throw detection. Alternatively, they rely on simple judgment of acceleration, speed, or even position and direction conditions to determine the throw, which greatly limits the throw action and cannot adapt to arbitrary throw situations, such as violently rotating throws or horizontal or downward throws. For example, the open-source flight control system PX4 detects the throw by detecting whether the current speed of the UAV is greater than a set threshold, which requires a large speed for the throw and cannot adapt to situations with low speed.
[0004] (2) Too much reliance on the state for throw-and-recovery: Most existing throw-and-recovery methods rely on the observation of the UAV's position and speed information at the beginning of the throw-and-recovery. However, the violent rotation of the throw-and-recovery often leads to inaccurate observation information, resulting in either restricted throw-and-recovery actions or unsafe or failed control recovery. Summary of the Invention
[0005] The purpose of this application is to address the problems existing in the prior art by providing a method for detecting and controlling the flight of a drone.
[0006] According to a first aspect of the embodiments of this application, a method for detecting and controlling the flight of a drone is provided, comprising:
[0007] The flight data of the UAV detected in real time by the inertial measurement unit is acquired, and the flight data is synchronously identified and converted by the external parameter calibration flight transformation algorithm to obtain the center of mass acceleration.
[0008] The drone is determined to be in a launch state based on the acceleration of the center of mass. When the acceleration of the center of mass is continuously close to zero and exceeds a time threshold, it is determined to be in a launch state.
[0009] If the UAV is in a jettison state, it is controlled in three stages: attitude recovery, velocity convergence, and position convergence. During the staged control process, the UAV state is calculated based on the differential flatness assumption, and angular velocity and thrust commands are generated according to the UAV state and the staged control target. The UAV is controlled according to the commands to restore the UAV state to a horizontal hovering state.
[0010] Furthermore, the process of calibrating the external parameters is as follows:
[0011] Establish a dataset of UAV launches with several launch attitudes, where each flight data includes acceleration and angular velocity;
[0012] Using a least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms, combined with studentized residuals, the external parameters between the UAV inertial measurement unit and the centroid are fitted.
[0013] Furthermore, the data in the UAV throwing dataset consists of flight data of the UAV after it is thrown in various throwing postures and before it lands, which is measured by the inertial measurement unit.
[0014] Furthermore, using a least-squares estimation algorithm with compensation for centrifugal and tangential acceleration terms, combined with studentized residuals, the extrinsic parameters between the UAV inertial measurement unit and the center of mass are fitted, specifically as follows:
[0015] Using a least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms, the extrinsic parameters between the UAV inertial measurement unit and the center of mass are initially fitted. Outliers in the initial fitting results are filtered out by studentized residuals. The least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms is used again for fitting to obtain the extrinsic parameters between the UAV inertial measurement unit and the center of mass.
[0016] Furthermore, phased control is implemented in three stages: attitude recovery, velocity convergence, and position convergence, including:
[0017] Based on the UAV flight data detected in real time by the inertial measurement unit, the UAV state is estimated, and with the horizontal attitude as the control target, angular velocity and thrust commands are generated. The UAV is controlled to recover its attitude according to the commands until the error between the UAV attitude and the horizontal attitude is within the predetermined attitude error range.
[0018] The desired speed is set to 0. The UAV state is calculated based on the differential flatness assumption. Angular velocity and thrust commands are generated according to the UAV state and the phased control target. The UAV speed is controlled to converge according to the commands until the error between the UAV speed and the desired speed is within the predetermined speed error range.
[0019] Using the current position as the desired position, the UAV state is calculated based on the differential flatness assumption. Angular velocity and thrust commands are generated according to the UAV state and the phased control target. The UAV position is controlled to converge according to the commands until the UAV is in a hovering state.
[0020] Furthermore, the phased control adopts a cascaded feedback controller, and the position error feedback gain and velocity error feedback gain are set to zero in the attitude recovery phase, and the position error feedback gain is set to zero in the velocity convergence phase.
[0021] Furthermore, if the safety lever integrated on the drone is detected to be in an disabled state, the launch detection or control recovery will not be initiated, or the drone will be immediately controlled to stop the launch detection or control recovery.
[0022] According to a second aspect of the embodiments of this application, a drone flight detection and control recovery device is provided, comprising:
[0023] The flight data conversion module is used to acquire UAV flight data detected in real time by the inertial measurement unit, and synchronously identify and convert the flight data using the external parameter calibration flight transformation algorithm to obtain the center of mass acceleration;
[0024] The launch detection module is used to determine whether the drone is in a launch state based on the center of mass acceleration. When the center of mass acceleration is continuously close to zero and exceeds a time threshold, it is determined to be in a launch state.
[0025] The control recovery module is used to perform phased control in three stages: attitude recovery, velocity convergence, and position convergence if the UAV is in a jettison state. During the phased control process, the UAV state is calculated based on the differential flatness assumption, and angular velocity and thrust commands are generated according to the UAV state and the phased control target. The UAV is controlled according to the commands to restore the UAV state to a horizontal hovering state.
[0026] According to a third aspect of the embodiments of this application, an unmanned aerial vehicle (UAV) system is provided, including a UAV body, on which a flight control system and an onboard computer are provided. The flight control system includes an inertial measurement unit, and the onboard computer integrates the UAV drop-off detection and control recovery method described in the first aspect.
[0027] Furthermore, the system also includes a remote controller for controlling the flight of the drone. The drone is equipped with a safety lever, which is enabled during the launch detection and control recovery process. If the safety lever is disabled, the launch detection or control recovery will not begin, or the drone will be immediately controlled to stop the launch detection or control recovery.
[0028] The technical solutions provided by the embodiments of this application may include the following beneficial effects:
[0029] This invention converts inertial measurement unit (IMU) detection data into center-of-mass acceleration, enabling it to adapt to various throwing attitudes for flight detection. Employing a phased recovery logic, it effectively handles various intense flight situations, relying on minimal information in the initial stage and gradually increasing the amount of observational information as the stages progress, minimizing dependence on position and velocity observations. Furthermore, it continuously adjusts the recovery strategy based on the dynamic environment, ensuring safe and stable flight of the UAV under extreme conditions. Combined with a supporting control system, this forms a complete multi-rotor UAV flight detection and recovery technology system, enabling extreme flight and stable recovery of UAVs in complex application scenarios, providing strong technical support for the rapid deployment of multi-rotor UAVs. Thus, whether in emergency rescue or other missions requiring efficient response, this technology significantly improves the flight capability and reliability of UAVs.
[0030] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description
[0031] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.
[0032] Figure 1 This is a flowchart illustrating a method for detecting and controlling the repatriation of a drone according to an exemplary embodiment.
[0033] Figure 2 This is a flowchart illustrating a throw-out detection and recovery process according to an exemplary embodiment.
[0034] Figure 3 This is a schematic diagram illustrating phased control according to an exemplary embodiment.
[0035] Figure 4 This is a block diagram illustrating a drone drop detection and control recovery device according to an exemplary embodiment.
[0036] Figure 5 This is a block diagram illustrating an unmanned aerial vehicle (UAV) system according to an exemplary embodiment.
[0037] Figure 6 This is a schematic diagram of an electronic device according to an exemplary embodiment. Detailed Implementation
[0038] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.
[0039] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.
[0040] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."
[0041] like Figure 1 and Figure 2 This application provides a method for detecting and controlling the recapture of a drone after it has been thrown out of control, which may include:
[0042] S1: Acquire the UAV flight data detected in real time by the inertial measurement unit, and use the external parameter calibration flight transformation algorithm to synchronously identify and transform the flight data to obtain the center of mass acceleration;
[0043] Specifically, the UAV flight data detected by the inertial measurement unit (IMU) is acceleration and angular velocity data in the IMU coordinate system, which includes timestamp information. The flight data is then transformed using an external parameter calibration-based launch transformation module to obtain acceleration data in the body's center-of-mass coordinate system.
[0044] Specifically, the external parameter calibration process of the launch-and-fly transformation module is as follows:
[0045] S11: Establish a dataset of drones after they are launched in various throwing postures, where each flight data includes acceleration and angular velocity;
[0046] Specifically, the data in the UAV throwing dataset must be flight data of the UAV after it is thrown in various throwing postures and before it lands, measured by the inertial measurement unit, and can be filtered according to the timestamps corresponding to the flight data.
[0047] S12: Using a least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms, the extrinsic parameters between the UAV inertial measurement unit and the center of mass are fitted to account for the effects of violent rotation; studentized residuals are introduced to filter out outliers caused by various errors; the least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms is used again for fitting to obtain the extrinsic parameters between the UAV inertial measurement unit and the center of mass, thereby converting the flight data into acceleration data of the body's center of mass.
[0048] Specifically, the inertial measurement unit (IMU) measures the acceleration at its location. and angular velocity The data assumes no rotation between the inertial measurement unit coordinate system and the center-of-mass coordinate system, only translation. Then the acceleration value at the center of mass The relationship between the acceleration and angular velocity of the inertial measurement unit coordinate system is as follows:
[0049]
[0050] In this formula, Represents angular acceleration, by Differentiation is obtained. Indicates tangential acceleration. This indicates centrifugal acceleration.
[0051] During the launch phase, theoretically Since the formula only has one unknown variable, r, it can be estimated using the least squares method. Furthermore, using studentized residuals to estimate the overall distribution of the data, thereby removing outliers, can effectively improve the estimation accuracy of r and achieve better results.
[0052] S2: Determine whether the drone is in a jettison state based on the centroid acceleration. When the centroid acceleration continuously approaches zero and exceeds the time threshold, it is determined to be in a jettison state.
[0053] Specifically, theoretically, the acceleration at the aircraft's center of mass is zero during the launch. However, in reality, slight shifts in the center of mass may occur due to sensor measurement errors, aircraft deformation, or loose components, or external forces (such as air resistance) may cause the calculated acceleration data to be non-zero. Therefore, a state where the acceleration at the center of mass remains close to zero for an extended period exceeding a certain time threshold is defined as the launch state. In one embodiment, if the acceleration modulus is less than 3 m / s², the launch state is defined as the acceleration at the center of mass remaining close to zero for an extended period exceeding a certain time threshold. 2 If so, it can be determined that the acceleration is close to zero; the time threshold can be set to 5ms~80ms.
[0054] During a launch, the aircraft is in a state of weightlessness. Ignoring air resistance, if we consider the aircraft as a point mass, the acceleration measured by the inertial measurement unit at that point should be zero. The core principle of launch detection is to detect whether the acceleration data is close to zero. Based on this principle, launch detection can theoretically adapt to any launch maneuver.
[0055] In reality, an airplane is not a point mass and may rotate when it is launched, but the conditions above are still met at the center of mass of the airplane. Therefore, the launch can still be determined by detecting the acceleration data at the center of mass.
[0056] Furthermore, in reality, the aircraft's inertial measurement unit (IMU) is often not located at the aircraft's center of mass. Therefore, to perform a throw-off test, it is necessary to calculate the acceleration data at the aircraft's center of mass. This invention proposes a throw-off test method that converts flight data measured by an IMU located at any position on the aircraft into data at the center of mass, thereby enabling throw-off detection under any circumstances. In the case of a violently rotating throw-off, the angular velocity and angular acceleration are not zero, meaning the centrifugal acceleration and tangential acceleration terms are not zero. Therefore, least-squares estimation with compensation for the centrifugal and tangential acceleration terms is required.
[0057] S3: If the UAV is in a jettison state, it is controlled in three stages: attitude recovery, velocity convergence, and position convergence. During the staged control process, the UAV state is calculated based on the differential flatness assumption, and angular velocity and thrust commands are generated according to the UAV state and the staged control target. The UAV is controlled according to the commands to restore the UAV state to a horizontal hovering state.
[0058] In step S3, the calculation of the UAV state using the differential flatness assumption specifically involves: calculating the magnitude of the thrust relative to the Z-axis direction of the UAV based on the target position, velocity, acceleration, and UAV mass parameters.
[0059] Furthermore, the differential flatness assumption makes the jettison recovery control method widely applicable. Based on this assumption, it is possible to simplify the modeling of the UAV's dynamic behavior, focusing on the calculation of thrust and the UAV's Z-axis direction. This means that by simply inputting information such as the target position, velocity, and acceleration, the complete state of the UAV can be calculated using the differential flatness property.
[0060] Specifically, in step S3, when performing phased control in three stages—attitude recovery, velocity convergence, and position convergence—the error feedback gain of the UAV's position / velocity is set to zero. This design decision effectively achieves the effect of phased control without the need for complex parameter adjustments or meticulous debugging. Specifically, the controller in this invention is designed as a cascade feedback controller. In one embodiment, a cascade PID (Proportion Integration Differential) controller is used. Specifically, the position error is fed back and superimposed on the original velocity to form a new desired velocity, while the velocity error is also fed back and superimposed on the desired acceleration. The acceleration is then used as input to the differential flattening to calculate the desired complete state of the UAV (attitude, thrust, angular velocity, etc.). Because it is cascade control, setting the error feedback gain of both the position loop and the velocity loop to zero during the attitude recovery phase is equivalent to abandoning position and velocity control, which allows for simple but effective implementation of only attitude control in staged control. During the velocity convergence phase, setting the position error feedback gain to zero allows for control of only attitude and velocity. During the velocity convergence phase, all error feedback gains are not set to zero, allowing for simultaneous control of position, velocity, and attitude, thus achieving staged control.
[0061] Specifically, in step S3, the UAV state is initialized step by step in three stages: attitude recovery, velocity convergence, and position convergence, as follows: Figure 3 As shown, the phased control process is as follows:
[0062] S31: Based on the UAV flight data detected in real time by the inertial measurement unit, estimate the UAV state, and with the horizontal attitude as the control target, generate angular velocity and thrust commands. Control the UAV to perform attitude recovery according to the commands until the error between the UAV attitude and the horizontal attitude is within the predetermined attitude error range. The predetermined attitude error range can be set according to actual needs, and can be 5°~20°.
[0063] S32: Set the desired speed to 0, calculate the UAV state based on the differential flatness assumption, and generate angular velocity and thrust commands according to the UAV state and the phased control target. Control the UAV speed to converge according to the commands until the error between the UAV speed and the desired speed is within a predetermined speed error range. The predetermined speed error range can be set according to actual needs, and can be set to 0.1 m / s or 0.2 m / s.
[0064] S33: Using the current position as the desired position, calculate the UAV state based on the differential flatness assumption, and generate angular velocity and thrust commands according to the UAV state and the phased control target. Control the UAV position convergence according to the commands until the UAV is in a hovering state.
[0065] Specifically, such as Figure 3As shown, attitude control accepts the desired attitude as horizontal and feeds back the actual attitude. The outer control loop is inactive because the control gain is set to zero. Speed control accepts the desired speed as zero and feeds back the actual speed. At this time, the attitude loop control is still active, while the position loop is closed. During position control, the desired hovering position (i.e., the current position) is given, and the actual position of the aircraft is fed back. The attitude loop, speed loop, and position loop all operate.
[0066] The phased control process involves initial positioning, i.e., initializing position and velocity, which can be done using any existing UAV positioning initialization algorithm, and this application does not impose any restrictions on it.
[0067] Furthermore, the method also includes a safety intervention mechanism: a safety lever is integrated into the drone's physical controller (such as a remote controller). During the jettison detection and control recovery process, the lever is in an enabled state, allowing the recovery process to proceed. When the lever is switched to the disabled state, the drone is immediately controlled to stop propellers and terminate the recovery. Specifically, in the event of an unexpected situation or if the user wishes to terminate the recovery, the user only needs to switch the lever to the disabled state, and the system will immediately stop propellers to protect the safety of those around them.
[0068] Corresponding to the aforementioned embodiments of the UAV launch detection and control recovery method, this application also provides embodiments of a UAV launch detection and control recovery device.
[0069] Figure 4 This is a block diagram illustrating a drone flight detection and control recovery device according to an exemplary embodiment. (Refer to...) Figure 4 The device may include:
[0070] The flight data conversion module 21 is used to acquire the UAV flight data detected in real time by the inertial measurement unit, and to synchronously identify and convert the flight data using the external parameter calibration flight transformation algorithm to obtain the center of mass acceleration.
[0071] The launch detection module 22 is used to determine whether the UAV is in a launch state based on the center of mass acceleration. When the center of mass acceleration is continuously close to zero and exceeds a time threshold, it is determined to be in a launch state.
[0072] The control recovery module 23 is used to perform phased control in three stages: attitude recovery, velocity convergence, and position convergence if the UAV is in a jettison state. During the phased control process, the UAV state is calculated based on the differential flatness assumption, and angular velocity and thrust commands are generated according to the UAV state and the phased control target. The UAV is controlled according to the commands to restore the UAV state to a horizontal hovering state.
[0073] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.
[0074] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.
[0075] Correspondingly, such as Figure 5 The present invention also provides an unmanned aerial vehicle (UAV) system, including a remote controller, a flight control system, and an onboard computer; the remote controller integrates a safety lever, the flight control system has an inertial measurement unit, and the onboard computer integrates a flight data conversion module 21, a flight detection module 22, a control recovery module 23, and a positioning module. The flight data conversion module 21, the flight detection module 22, and the control recovery module 23 are used to implement the above-mentioned UAV flight detection and control recovery methods.
[0076] Inertial Measurement Unit: Real-time acquisition of data from the drone's flight.
[0077] Positioning module: locates the drone's position based on the drone's status initialization signal and coordinates with the drone's speed convergence; it is mainly used to control the phased control process of the recovery module 23 and can integrate any existing drone positioning initialization algorithm;
[0078] Safety lever: When the safety lever is enabled throughout the entire process from launch detection to recovery, the recovery process is allowed; if the safety lever is disabled, launch detection or control recovery will not begin, or the drone will be immediately controlled to stop launch detection or control recovery, and the drone will be controlled to stop propellers to ensure safety; wherein, the lever signal is sent to the onboard computer through the flight control system.
[0079] Accordingly, this application also provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the above-described method for detecting and controlling the unmanned aerial vehicle (UAV) flight.
[0080] Accordingly, this application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; and, when the one or more programs are executed by the one or more processors, causing the one or more processors to implement the UAV drop-off detection and control recovery method described above. Figure 6The diagram shown is a hardware structure diagram of any data processing-capable device within a UAV flight detection and control recovery system provided in an embodiment of the present invention. (Except for...) Figure 6 In addition to the processor, memory, and network interface shown, any data processing device in the embodiment may also include other hardware depending on the actual function of the data processing device, which will not be described in detail here.
[0081] Accordingly, this application also provides a computer-readable storage medium storing computer instructions thereon, which, when executed by a processor, implement the UAV flight detection and control recovery method described above. The computer-readable storage medium can be an internal storage unit of any data-processing device as described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, smart media card (SMC), SD card, flash card, etc., equipped on the device. Furthermore, the computer-readable storage medium can include both internal storage units of any data-processing device and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the data-processing device, and can also be used to temporarily store data that has been output or will be output.
[0082] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.
Claims
1. A method for detecting and controlling the recovery of a drone after it has been thrown out of control, characterized in that, include: The flight data of the UAV detected in real time by the inertial measurement unit is acquired, and the flight data is synchronously identified and converted by the external parameter calibration flight transformation algorithm to obtain the center of mass acceleration. The drone is determined to be in a launch state based on the acceleration of the center of mass. When the acceleration of the center of mass is continuously close to zero and exceeds a time threshold, it is determined to be in a launch state. If the UAV is in a jettison state, it is controlled in three stages: attitude recovery, velocity convergence, and position convergence. During the staged control process, the UAV state is calculated based on the differential flatness assumption, and angular velocity and thrust commands are generated according to the UAV state and the staged control target. The UAV is controlled according to the commands to restore the UAV state to a horizontal hovering state. The process of calibrating the external parameters is as follows: Establish a dataset of UAV launches with several launch attitudes, where each flight data includes acceleration and angular velocity; Using a least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms, combined with studentized residuals, the extrinsic parameters between the UAV inertial measurement unit and the center of mass are fitted. Specifically, the least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms is used to initially fit the extrinsic parameters between the UAV inertial measurement unit and the center of mass. Outliers in the initial fitting results are filtered out by studentized residuals. The least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms is used again to fit the extrinsic parameters between the UAV inertial measurement unit and the center of mass.
2. The method according to claim 1, characterized in that, The data in the UAV launch dataset consists of flight data of the UAV after it is launched in various launching postures and before it lands, measured by an inertial measurement unit.
3. The method according to claim 1, characterized in that, The control is implemented in three stages: attitude recovery, velocity convergence, and position convergence, including: Based on the UAV flight data detected in real time by the inertial measurement unit, the UAV state is estimated, and with the horizontal attitude as the control target, angular velocity and thrust commands are generated. The UAV is controlled to recover its attitude according to the commands until the error between the UAV attitude and the horizontal attitude is within the predetermined attitude error range. The desired speed is set to 0. The UAV state is calculated based on the differential flatness assumption. Angular velocity and thrust commands are generated according to the UAV state and the phased control target. The UAV speed is controlled to converge according to the commands until the error between the UAV speed and the desired speed is within the predetermined speed error range. Using the current position as the desired position, the UAV state is calculated based on the differential flatness assumption. Angular velocity and thrust commands are generated according to the UAV state and the phased control target. The UAV position is controlled to converge according to the commands until the UAV is in a hovering state.
4. The method according to claim 1, characterized in that, The phased control employs a cascaded feedback controller, with the position error feedback gain and velocity error feedback gain set to zero during the attitude recovery phase and the position error feedback gain set to zero during the velocity convergence phase.
5. The method according to claim 1, characterized in that, If the safety lever integrated on the drone is detected to be disabled, the launch detection or control recovery will not be initiated; or, the drone will be immediately controlled to stop the launch detection or control recovery.
6. A device for detecting and controlling the flight of a drone, characterized in that, include: The flight data conversion module is used to acquire UAV flight data detected in real time by the inertial measurement unit, and synchronously identify and convert the flight data using the external parameter calibration flight transformation algorithm to obtain the center of mass acceleration; The launch detection module is used to determine whether the drone is in a launch state based on the center of mass acceleration. When the center of mass acceleration is continuously close to zero and exceeds a time threshold, it is determined to be in a launch state. The control recovery module is used to perform phased control in three stages: attitude recovery, velocity convergence, and position convergence if the UAV is in a jettison state. During the phased control process, the UAV state is calculated based on the differential flatness assumption, and angular velocity and thrust commands are generated according to the UAV state and the phased control target. The UAV is controlled according to the commands to restore the UAV state to a horizontal hovering state. The process of calibrating the external parameters is as follows: Establish a dataset of UAV launches with several launch attitudes, where each flight data includes acceleration and angular velocity; Using a least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms, combined with studentized residuals, the extrinsic parameters between the UAV inertial measurement unit and the center of mass are fitted. Specifically, the least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms is used to initially fit the extrinsic parameters between the UAV inertial measurement unit and the center of mass. Outliers in the initial fitting results are filtered out by studentized residuals. The least squares estimation algorithm with compensation for centrifugal and tangential acceleration terms is used again to fit the extrinsic parameters between the UAV inertial measurement unit and the center of mass.
7. An unmanned aerial vehicle (UAV) system, characterized in that, The device includes a drone body, on which a flight control system and an onboard computer are installed. The flight control system includes an inertial measurement unit, and the onboard computer integrates the drone flight detection and control recovery method according to any one of claims 1-5.
8. The system according to claim 7, characterized in that, The system also includes a remote controller for controlling the flight of the drone. The drone is equipped with a safety lever. During the flight detection and control recovery process, the safety lever is in an enabled state. If the safety lever is in an disabled state, the flight detection or control recovery will not start, or the drone will be immediately controlled to stop the flight detection or control recovery.