A method for calibrating a drone

Abstract

The application discloses a kind of unmanned aerial vehicle calibration methods, belong to unmanned aerial vehicle ground control system technical field.The method is controlled by driving tool control front wheel rotation, during rotation process, angle encoder is used to collect the real rotation angle of each set collection point, the landing gear control unit of unmanned aerial vehicle obtains the measured angle electric signal digital quantity of each collection point, the real rotation angle and measured angle electric signal digital quantity are fitted by host computer, and the fitting function model reflecting the mapping relationship of both is obtained, and the fitting function is written into the memory of landing gear control unit, realize unmanned aerial vehicle front wheel zero automatic calibration calibration, landing gear control unit can realize the calibration of measured angle based on the function model.The application solves the problem that the existing front wheel turning calibration calibration method based on two-point calibration has precision dependent on artificial, cannot overcome nonlinearity, low efficiency.

Description

Technical Field

[0001] This invention relates to the field of UAV ground control system technology, specifically to a UAV calibration method, belonging to the zero-position / angle calibration and calibration technology of UAV front wheel turning system. Background Technology

[0002] The nose wheel steering system is a key subsystem for ensuring directional control during the ground taxiing phase of an aircraft. Its performance directly affects the straightness of the taxiing, correction efficiency, and takeoff and landing safety. For modern aircraft using fly-by-wire or electric servo motors, it is necessary to accurately calibrate the "electrical zero position" and "mechanical zero position" through electrical means. ("Mechanical zero position" refers to the reference position where the nose wheel is completely parallel to the longitudinal axis of the fuselage and facing forward, with no nose wheel deflection angle. "Electrical zero position" refers to the position where the steering angle sensor outputs 0°, ideally coinciding with the "mechanical zero position," but in practice, calibration is required to eliminate deviations.)

[0003] Unlike manned aircraft pilots who can directly perceive the aircraft's attitude and direction of the roll, UAV operators rely entirely on ground station telemetry data and images for judgment, resulting in a certain "perception delay" and "lack of tactile feedback." If there is a deviation in the zero-point or full-range calibration of the front wheel steering system, it will directly lead to an incorrect mapping between the control surface deflection commands issued by the ground station and the actual wheel hub steering angle. Unlike manned aircraft pilots, UAV operators cannot quickly detect and correct this deviation intuitively, which can easily lead to risks such as roll deviation, overcorrection, or even loss of control. Therefore, achieving high-precision, high-reliability automatic calibration of the full-range steering angle, especially accurate zero-point calibration, is a key technological prerequisite for improving the safety and autonomy of UAV ground operation.

[0004] Currently, the industry commonly uses a linear fitting method based on two-point calibration for zero-position calibration. For example, the front wheel is manually placed at the mechanical zero position and another specific angular position (such as 4° or 8°), and the angle sensor data and controller mechanical position information at the two points are read to calculate the linear relationship parameters. Although this method achieves zero-position calibration to a certain extent and is easy to implement, its inherent limitations are also quite obvious: First, its calibration accuracy heavily depends on the absolute accuracy of the two calibration points; any sensor noise, manual alignment error, or mechanical clearance at any single point will lead to systematic deviations. Second, this method is based on the ideal assumption of a perfectly linear system and cannot identify or compensate for nonlinear characteristics that may exist in actual mechanisms. Third, the entire process is highly dependent on manual operation, resulting in low efficiency, and the consistency between different operators or different operations is difficult to guarantee. Finally, existing methods lack an immediate and automatic verification step after calibration, and the correctness of the results cannot be confirmed immediately.

[0005] Therefore, given the higher requirements placed on ground control precision, maintenance efficiency, and autonomous support capabilities by aircraft such as drones, there is an urgent need for a front wheel turning zero-position calibration method that can overcome the shortcomings of existing technologies, achieve automation, high precision, and self-verification capabilities, so as to fundamentally improve the safety margin and system intelligence level of ground taxiing control. Summary of the Invention

[0006] The technical problem to be solved:

[0007] To avoid the shortcomings of existing technologies, this invention provides a UAV calibration method that achieves full automation of the calibration process. By using multi-point data fitting, it improves accuracy and robustness, ensuring the accuracy and reliability of the calibration results. This solves the problems of existing front wheel turning calibration methods based on two-point calibration, which rely on manual precision, cannot overcome nonlinearity, and are inefficient. This method meets the urgent need of UAVs for high-precision and high-reliability ground control.

[0008] The technical solution of this invention is characterized by the following: the method employs a calibration device in conjunction with the entire UAV, the calibration device comprising a drive fixture integrating an angle encoder, a host computer, and a data acquisition module; the calibration method includes the following steps:

[0009] The front wheel of the drone is rotated by a drive fixture to adjust the front wheel positioning to the mechanical zero position; when the mechanical zero position is reached, the reading of the angle encoder is set to zero, and the digital value of the current measured angle electrical signal of the front wheel is marked as the electrical zero position.

[0010] The host computer sets the front wheel rotation angle range and rotation angle acquisition points. The drive fixture controls the front wheel to rotate within the preset rotation angle range. The angle encoder acquires the rotation angle at the output end of the drive fixture according to the set acquisition points. The rotation angle at the output end of the drive fixture is taken as the actual rotation angle of the front wheel and transmitted to the host computer through the data acquisition module. The landing gear control unit of the UAV acquires the digital quantity of the electrical signal of the measurement angle corresponding to each acquisition point during the rotation of the front wheel and imports it into the host computer.

[0011] The host computer performs data fitting on the digital quantities of the measured angle electrical signals and the actual rotation angles corresponding to all acquisition points to obtain a fitting function that reflects the mapping relationship between the digital quantities of the measured angle electrical signals and the actual rotation angles.

[0012] Based on the current digital measurement angle electrical signal, and using a fitting function, the physical steering angle of the front wheel is obtained, and the physical steering angle is used as the calibration angle.

[0013] A further technical solution of the present invention is: the fitting function is written into the memory of the landing gear control unit. When the landing gear control unit is running, the landing gear control unit calls the fitting function to analyze the digital quantity of the measured angle electrical signal to obtain the calibration angle.

[0014] A further technical solution of the present invention is: the host computer performs data fitting through built-in data processing software. When fitting, if the data is linear within the full range, a linear function or low-order polynomial is used for fitting; if the data is nonlinear with a certain regular change within the full range, the full range is divided into multiple intervals, and then linear or low-order polynomial fitting is performed.

[0015] A further technical solution of the present invention is that the data processing software built into the host computer is MATLAB data processing software.

[0016] A further technical solution of the present invention is as follows: Before calibration, the drive fixture and the front wheel steering mechanism of the UAV are connected, and the front wheel is placed on a pad with a friction coefficient of 0.25~0.65, and the front wheel steering system of the landing gear control unit is placed in a reduced sway state; during the calibration process, it is ensured that the front wheel does not leave the pad and the front wheel steering system remains in a reduced sway state.

[0017] A further technical solution of the present invention is: the calibration method further includes a calibration verification method to verify whether the error of the calibration angle exceeds a preset error value, and to analyze the repeatability of the calibration angle error; the calibration verification method includes the following steps:

[0018] Multiple verification angle points are set on the drone, and the drone's flight control system controls the front wheel to rotate to each verification angle point in sequence;

[0019] Perform the following operations for each verification angle point:

[0020] a. The angle encoder acquires the actual angle of the front wheel when it rotates to the verification angle point, and the landing gear control unit acquires the calibration angle when the front wheel rotates to the verification angle point; the host computer acquires the actual angle and the calibration angle, calculates the absolute error between the actual angle and the calibration angle at the verification angle point, and then compares the absolute error with the preset error value to determine whether the verification angle point is qualified.

[0021] b. Perform multiple positioning operations at the same verification angle point. The host computer calculates the absolute error between the actual angle and the calibration angle at each positioning point, then calculates the standard deviation, and compares the standard deviation with the preset error value to determine whether the standard deviation exceeds the limit.

[0022] If all verification angle points are deemed to be qualified and the standard deviation is within the limit, the calibration is deemed qualified; if any angle point fails to pass verification, the host computer will issue an out-of-tolerance alarm, check the calibration process or the status of the on-board equipment, recalibrate and re-verify; if the standard deviation exceeds the limit, the host computer will issue an out-of-tolerance alarm, check the status of the on-board equipment, and re-verify.

[0023] A further technical solution of the present invention is: when determining whether a verification angle point is qualified, if the absolute error is less than a preset error value, the verification angle point is deemed qualified; if the absolute error is greater than the preset error value, the verification is deemed unqualified.

[0024] A further technical solution of the present invention is: when determining whether the standard deviation exceeds the limit, if the standard deviation is less than the preset error value, it is determined that the standard deviation does not exceed the limit and the calibration angle has repeatability; if the standard deviation is greater than the preset error value, it is determined that the standard deviation exceeds the limit and the calibration angle does not have repeatability.

[0025] A further technical solution of the present invention is: when setting multiple verification angle points, the principle is that the density of verification points in a small angle range is greater than the density of verification points in a large angle range.

[0026] A further technical solution of the present invention is: setting a verification angle point at 0.5° intervals within ±4° on both sides of the mechanical zero position of the front wheel rotation, and setting a verification angle point at 1° intervals within the range exceeding ±4°.

[0027] The beneficial effects of this invention are as follows: In this UAV calibration method, a calibration device drives the UAV's front wheel to rotate. An angle encoder integrated into the calibration device measures a set of true angle data at each acquisition point during the front wheel's rotation. The host computer of the calibration device acquires this set of true angle data through its data acquisition module. Simultaneously, the UAV's landing gear control unit acquires a set of digital electrical signals corresponding to each acquisition point via the front wheel turning angle sensor. The host computer fits the set of true angle data and the set of digital electrical signals at each acquisition point to generate a fitting function model. This function model is then written into the memory of the landing gear control unit. When the landing gear control unit is operating, it can correct the angle information acquired by the front wheel turning angle sensor based on this function model. After calibrating the UAV's front wheel zero position using this invention, when the flight control system issues a turning angle command, the landing gear control unit of the UAV's front wheel turning system controls the front wheel to turn according to the command. Simultaneously, the front wheel turning angle sensor acquires the electrical signal of the turning angle, which is parsed by the landing gear control unit to obtain the turning angle, i.e., the calibration angle value. This angle value more accurately approximates the turning angle command value issued by the flight control system.

[0028] Compared with existing methods for zero-position calibration of UAV front wheels, this method has the following advantages:

[0029] 1. This invention abandons the outdated method of relying on manual positioning and collecting only a few calibration points. Instead, it uses a calibration device to drive the front wheels to rotate, enabling continuous or stepped rotational motion while simultaneously and densely collecting data points across the entire measurement range. This allows for a complete one-time mapping of the input-output characteristics of the front wheel angle sensor, and the fitting of a function model to establish the correspondence between input and output for calibration. This process is efficient and consistent, completely eliminating human error, and is particularly suitable for rapid field support.

[0030] 2. This method incorporates verification and diagnostic steps. By comparing the deviations of the calibration model's predicted values ​​(calibrated angles calculated from information collected by the angle sensor) with the external true values ​​(angles measured by the angle encoder) at multiple points, the correctness of a single calibration process is determined. Furthermore, by analyzing the consistency of repeated positioning data, potential problems such as additional deformation caused by abnormal tire pressure, wear or loosening of the steering actuator's transmission chain, and excessive load on the front wheel can be effectively diagnosed. The introduction of the verification step improves the accuracy of calibration and provides a new technical means for predictive maintenance of UAVs. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart of a UAV calibration method according to the present invention;

[0033] Figure 2 This is a schematic diagram illustrating the working principle of a UAV calibration method according to the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] This embodiment provides a method for calibrating unmanned aerial vehicles (UAVs), applicable to the automatic zero-position calibration of the nose wheel of an unmanned aerial vehicle (UAV). In this method, an external calibration device works in conjunction with the entire UAV to automatically calibrate the nose wheel's zero position. The external calibration device includes a drive fixture with an integrated angle encoder, a host computer, and a data acquisition module. The drive fixture drives the UAV's nose wheel to rotate, the angle encoder measures the actual steering angle of the nose wheel, and the data acquisition module collects the measurement data from the angle encoder and transmits it to the host computer. The host computer controls the operation of the drive fixture and the angle encoder, and performs subsequent data processing and display. The UAV's flight control system, nose wheel steering control system, and landing gear control unit participate in the calibration process.

[0036] like Figure 1 and Figure 2 As shown, the UAV calibration method of the present invention is as follows:

[0037] S0. Preparations before calibration: Ensure a reliable transmission connection between the UAV's front wheel steering mechanism and the drive fixture.

[0038] Specifically, the entire drone is towed into the hangar, with the front wheel steering mechanism positioned next to the drive fixture. Simultaneously, the front wheel is placed on a mat with a friction coefficient of 0.25–0.65 laid on the ground. The power output of the drive fixture is then connected to the power input of the front wheel steering mechanism, enabling the drive fixture to control the rotation of the front wheel. The drone's front wheel steering system is set to a "reduced yaw" state. During subsequent calibration, the front wheel remains on the mat and rotates on it under the drive of the drive fixture. Placing the front wheel on the mat reduces friction during rotation. In another embodiment, a low-friction medium (with a friction coefficient of 0.25–0.65) can be coated on the ground, and the front wheel can be placed on the coated medium to reduce rotational friction. It should be noted that when selecting a low-friction mat or other low-friction medium, it must meet safety requirements such as non-flammability to ensure the safety of the test process.

[0039] In this embodiment, the drive fixture includes a servo motor. The motor's output shaft is coaxially connected to the drive fixture's power output end (drive shaft) via a coupling. The drive shaft is connected to the power input end (e.g., a towing joint) of the UAV's front wheel steering mechanism. The motor in the drive fixture controls the rotation of the front wheel. An angle encoder is integrated into the servo motor in the drive fixture. The angle encoder measures the rotation angle of the drive shaft, which is the actual rotation angle of the front wheel. The host computer, as the control processing core, controls the operation of the drive fixture and the angle encoder, and performs subsequent data processing and display.

[0040] S1. Adjust the front wheel alignment to the mechanical zero position and mark it as the electrical zero position.

[0041] Specifically, the front wheel of the drone is rotated using a drive fixture to adjust its positioning to the mechanical zero position. When the mechanical zero position is reached, the reading of the angle encoder is set to zero, and the digital value of the current measured angle electrical signal of the front wheel is marked as the electrical zero position.

[0042] In this embodiment, during mechanical zero-position adjustment, the front wheel is adjusted to its inherent mechanical zero position by manually observing the mechanical positioning scale built into the UAV landing gear system. When the front wheel is positioned at the mechanical zero position, the reading of the angle encoder is set to zero, and the measured angle signal detected by the UAV landing gear control unit at the front wheel mechanical zero position is manually recorded. This angle signal is the digital quantity of the electrical signal measured by the front wheel turning angle sensor of the landing gear control unit, and the digital quantity of the electrical signal at this time is marked as the electrical zero position.

[0043] S2. Acquisition of electrical signal data of actual front wheel steering angle and measured angle.

[0044] Specifically, the host computer sets the front wheel rotation angle range and rotation angle acquisition points. The drive fixture controls the front wheel to rotate within the preset rotation angle range to perform continuous or step rotational motion. The angle encoder acquires the rotation angle at the output end of the drive fixture according to the set acquisition points, and uses the rotation angle at the output end of the drive fixture as the actual rotation angle of the front wheel, which is then transmitted to the host computer through the data acquisition module. The landing gear control unit of the UAV acquires the digital quantity of the measured angle electrical signal corresponding to each acquisition point during the rotation of the front wheel, and imports it into the host computer.

[0045] In this embodiment, the control software of the host computer sets the front wheel rotation range and multiple front wheel angle acquisition points. The set front wheel rotation range covers the commonly used steering angles of the front wheels and is as close as possible to the front wheel safety mechanical limit. The host computer sends instructions to the drive fixture, which controls the front wheels to rotate at a low speed and uniformly within the set front wheel rotation range. This low-speed and uniform rotation avoids impact on the steering mechanism. During the front wheel turning process, the angle encoder of the calibration device sequentially collects the actual front wheel angle data at each acquisition point. The host computer obtains a set of actual angle data collected by the angle encoder through the data acquisition module. Simultaneously, the landing gear control unit collects the digital quantity of the measured angle electrical signal at each acquisition point. The set of digital quantity data of the measured angle electrical signal output by the landing gear control unit is manually imported into the host computer. This set of digital quantity data of the measured angle electrical signal output by the landing gear control unit is a digital quantity of the measured angle signal based on the electrical zero position.

[0046] Data acquisition is performed synchronously at a frequency of 100Hz. The true angle value at each acquisition point is measured by a high-precision angle encoder of the calibration device. , i For the first iEach data collection point yields a set of true steering angle data. The landing gear control unit measures the digital electrical signal of the angle at each collection point using the front wheel steering angle sensor. , i For the first i Each acquisition point yields a set of digital data for the measured angle electrical signal. For each acquisition point, a corresponding data pair is obtained. , Approximately 2,000 valid data pairs can be collected in about 20 seconds of single-pass rotation.

[0047] S3. Perform data fitting processing on the collected data to obtain the fitting function.

[0048] Specifically, the host computer performs data fitting on the digital quantities of the measured angle electrical signals and the actual rotation angles corresponding to all acquisition points to obtain a fitting function that reflects the mapping relationship between the digital quantities of the measured angle electrical signals and the actual rotation angles.

[0049] In this embodiment, the host computer has built-in data processing software, such as MATLAB data processing software. Through this software, the actual angle collected by the angle encoder and the digital quantity of the measured angle electrical signal of the landing gear control unit are fitted, thereby realizing the automatic zero-position calibration of the front wheel of the unmanned aircraft.

[0050] Before data fitting, the data is first smoothed and filtered (e.g., by moving average filtering) to suppress high-frequency noise. Then, the host computer uses its built-in data processing software to perform a full-point fitting of the acquired set of real rotation angle data and a set of measured angle electrical signal digital data, obtaining the fitting function, expressed as follows: = In the formula, This is the actual angle value. To measure the digital quantity of the angular electrical signal, that is, the digital quantity of the electrical signal measured by the front wheel steering angle sensor, this fitting function model reflects the mapping relationship between the digital quantity of the electrical signal and the actual steering angle.

[0051] The MATLAB data processing software built into the host computer is a commercial scientific computing and simulation software developed by MathWorks, Inc. The fitting method used is an existing method within the software. During data fitting, different complexity models are selected based on the distribution characteristics of the data points. If the data has good linearity, meaning it exhibits linearity across the entire range, a linear function or low-order polynomial is used for fitting. If the data exhibits nonlinearity with a certain regularity across the entire range, the entire range is divided into multiple intervals, and linear or low-order polynomial fitting is performed to obtain higher accuracy.

[0052] S4. Obtain the calibration angle by fitting the function.

[0053] Specifically, based on the current digital value of the measured angle electrical signal, and based on the fitting function, the physical steering angle of the front wheel is obtained, and the physical steering angle is used as the calibration angle.

[0054] In this embodiment, the fitting function is written into the memory of the landing gear control unit. Specifically, the fitting function model is burned into the non-volatile memory of the UAV's landing gear control unit via a bus. The landing gear control unit software adds processing for calibrating the function model, so that the function model can be called to correct the measured angle during landing gear operation. This completes the automatic zero-position calibration of the front wheel.

[0055] During the operation of the landing gear control unit, the landing gear control unit, based on the mapping relationship provided by the function model, resolves the digital quantity of the original measured angle electrical signal collected by its angle sensor into the corresponding physical rotation angle. This physical rotation angle is the calibrated angle after calibration, thereby obtaining the front wheel physical rotation angle with higher reliability.

[0056] The raw digital measurement angle signal acquired by the angle sensor is conditioned and converted from analog to digital by the landing gear control unit to obtain the analyzed angle value. This process requires calibration using the aforementioned function model. For AC sensors (such as rotary transformers), signal conditioning and analog-to-digital conversion mainly include signal demodulation, amplitude scaling, bias adjustment, and filtering; for DC sensors (such as potentiometers), they mainly include amplitude scaling, bias adjustment, and filtering. These are existing technologies and will not be elaborated upon here.

[0057] S5. Calibration and verification.

[0058] After the S4 fitting function model is written and solidified, an automatic verification process is performed. Verification determines whether the automatic front wheel zero-position calibration was successful. The specific method used by the flight control system is as follows:

[0059] Multiple verification angle points are set in the UAV's flight control system. When setting these verification angle points, the principle is that the density of verification points within a smaller angle range is greater than that within a larger angle range. In this embodiment, a relatively dense interval of 0.5° is set within ±4° on both sides of the mechanical zero position of the front wheel rotation; a relatively loose interval, such as 1°, is set in the larger angle range exceeding ±4°.

[0060] After the verification angle points are set, the drone's flight control system issues rotation commands according to the set verification angle points, controlling the front wheel to rotate to each verification angle point, and performs the following operations for each verification angle point:

[0061] a. An angle encoder acquires the actual angle at which the front wheel rotates to the verification angle point. , represents the true angle of the verification angle point, and n represents the nth verification angle point. The landing gear control unit obtains the calibration angle when the front wheel turns to this verification angle point , represents the calibration angle, and n represents the nth verification angle point.

[0062] The upper computer obtains the true angle of the verification angle point through the data acquisition module, and manually imports the calibration angle of the verification point output by the landing gear control unit into the upper computer. The upper computer calculates the absolute error △ between the true angle and the calibration angle of this verification angle point, △ = | - |, and compares the absolute error with the preset error value to determine whether this verification angle point is verified qualified. The preset error value is the measurement angle error threshold set according to the design requirements. Within this threshold range, it is considered that the measurement angle collected by the angle sensor has accurately approached the true angle, and the error is acceptable. When determining whether the verification angle point is verified qualified, if the absolute error is less than the preset error value, it is determined that this verification angle point is verified qualified; if the absolute error is greater than the preset error value, the verification is unqualified.

[0063] b. Perform multiple front wheel alignment operations at the same verification angle point. The upper computer obtains multiple true angles and calibration angles at this point, calculates the absolute error between the true angle and the calibration angle for each positioning to this point, then calculates the standard deviation, and compares the standard deviation with the preset error value to determine whether the standard deviation exceeds the limit. The purpose of this step of verification is to verify repeatability. If the standard deviation is less than the preset error value, it is determined that the standard deviation does not exceed the limit, and the calibration angle has repeatability; if the standard deviation is greater than the preset error value, it is determined that the standard deviation exceeds the limit, and the calibration angle does not have repeatability.

[0064] The verification evaluation criterion is: If it is determined that all verification angle points are verified qualified and the standard deviation does not exceed the limit, it is determined that the automatic calibration is qualified this time, the calibration is successful, and the system automatically generates a calibration report. If any angle point is verified unqualified, the upper computer issues an out-of-tolerance alarm: "The model accuracy is out of tolerance. Please check the calibration process or the status of the on-board equipment." At this time, it is necessary to check the calibration process or the status of the on-board equipment and re-calibrate and then verify. If the standard deviation exceeds the limit, the upper computer issues an out-of-tolerance alarm: "The structural connection repeatability is poor. Please check the status of the on-board equipment." At this time, check the status of the on-board equipment and re-verify.

[0065] By adopting a calibration method for an unmanned aerial vehicle of the present invention to perform automatic zero-position calibration of the front wheel of an unmanned aircraft, it can be realized that the measurement angle of the landing gear control unit accurately approaches the actual true angle. In theory, the turning instruction issued by the flight control system and the measurement angle (calibrated angle) of the landing gear control unit will be close to consistent, and the error is within the allowable range, improving the accuracy of the zero-position calibration of the front wheel of the unmanned aerial vehicle.

[0066] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for calibrating an unmanned aerial vehicle (UAV), characterized in that, This method uses a calibration device in conjunction with the entire UAV. The calibration device includes a drive fixture with an integrated angle encoder, a host computer, and a data acquisition module. The calibration method includes the following steps: The front wheel of the drone is rotated by a drive fixture to adjust the front wheel positioning to the mechanical zero position; when the mechanical zero position is reached, the reading of the angle encoder is set to zero, and the digital value of the current measured angle electrical signal of the front wheel is marked as the electrical zero position. The host computer sets the front wheel rotation angle range and rotation angle acquisition points. The drive fixture controls the front wheel to rotate within the preset rotation angle range. The angle encoder acquires the rotation angle at the output end of the drive fixture according to the set acquisition points. The rotation angle at the output end of the drive fixture is taken as the actual rotation angle of the front wheel and transmitted to the host computer through the data acquisition module. The landing gear control unit of the UAV acquires the digital quantity of the electrical signal of the measurement angle corresponding to each acquisition point during the rotation of the front wheel and imports it into the host computer. The host computer performs data fitting on the digital quantities of the measured angle electrical signals and the actual rotation angles corresponding to all acquisition points to obtain a fitting function that reflects the mapping relationship between the digital quantities of the measured angle electrical signals and the actual rotation angles. Based on the current digital measurement angle electrical signal, and using a fitting function, the physical steering angle of the front wheel is obtained, and the physical steering angle is used as the calibration angle.

2. The UAV calibration method according to claim 1, characterized in that, The fitting function is written into the memory of the landing gear control unit. When the landing gear control unit is running, the landing gear control unit calls the fitting function to analyze the digital quantity of the measured angle electrical signal in order to obtain the calibration angle.

3. The UAV calibration method according to claim 1, characterized in that, The host computer uses built-in data processing software to fit the data. When the data is linear across the entire range, a linear function or low-order polynomial is used for fitting. When the data is nonlinear across the entire range, the entire range is divided into multiple intervals, and then linear or low-order polynomial fitting is performed.

4. The UAV calibration method according to claim 3, characterized in that, The host computer has MATLAB built-in data processing software.

5. The UAV calibration method according to claim 1, characterized in that, Before calibration, connect the drive fixture and the front wheel steering mechanism of the drone, and place the front wheel on a pad with a friction coefficient of 0.25~0.65, and put the drone's front wheel turning system in a reduced-sway state; during calibration, ensure that the front wheel does not leave the pad and the front wheel turning system remains in a reduced-sway state.

6. The UAV calibration method according to claim 1, characterized in that, The calibration method further includes a calibration verification method to verify whether the error of the calibration angle exceeds a preset error value, and to analyze the repeatability of the calibration angle error; the calibration verification method includes the following steps: Multiple verification angle points are set on the drone, and the drone's flight control system controls the front wheel to rotate to each verification angle point in sequence; Perform the following operations for each verification angle point: a. The angle encoder acquires the actual angle of the front wheel when it rotates to the verification angle point, and the landing gear control unit acquires the calibration angle when the front wheel rotates to the verification angle point; the host computer acquires the actual angle and the calibration angle, calculates the absolute error between the actual angle and the calibration angle at the verification angle point, and then compares the absolute error with the preset error value to determine whether the verification angle point is qualified. b. Perform multiple positioning operations at the same verification angle point. The host computer calculates the absolute error between the actual angle and the calibration angle at each positioning point, then calculates the standard deviation, and compares the standard deviation with the preset error value to determine whether the standard deviation exceeds the limit. If all verification angle points are deemed to be qualified and the standard deviation is within the limit, the calibration is deemed qualified; if any angle point fails to pass verification, the host computer will issue an out-of-tolerance alarm, check the calibration process or the status of the on-board equipment, recalibrate and re-verify; if the standard deviation exceeds the limit, the host computer will issue an out-of-tolerance alarm, check the status of the on-board equipment, and re-verify.

7. The UAV calibration method according to claim 6, characterized in that, When determining whether a verification angle point is qualified, if the absolute error is less than the preset error value, the verification angle point is deemed qualified; if the absolute error is greater than the preset error value, the verification is deemed unqualified.

8. The UAV calibration method according to claim 6, characterized in that, When determining whether the standard deviation exceeds the limit, if the standard deviation is less than the preset error value, the standard deviation is determined to be within the limit and the calibration angle is repeatable; if the standard deviation is greater than the preset error value, the standard deviation is determined to exceed the limit and the calibration angle is not repeatable.

9. The UAV calibration method according to claim 6, characterized in that, When setting multiple verification angle points, the principle is that the density of verification points in the smaller angle range is greater than the density of verification points in the larger angle range.

10. A UAV calibration method according to claim 9, characterized in that, Set a verification angle point at 0.5° intervals within ±4° on both sides of the mechanical zero position of the front wheel rotation, and set a verification angle point at 1° intervals within the range exceeding ±4°.

Images