A dual-mode dynamic collaborative control method and system based on a pull-tether line

Abstract

This invention discloses a dual-mode dynamic cooperative control method and system based on a tethered line, comprising: acquiring displacement electrical signals through dual potentiometers; generating a Z-axis PWM signal based on the displacement electrical signals through a PWM signal generator; generating an XY-axis PWM signal based on the angle difference between the XY joysticks on the UAV and the tethered line; splitting the Z-axis PWM signal into two paths, one transmitted to the winch ESC and the other transmitted to the UAV and merged with the XY-axis PWM signal to form a three-dimensional control signal; setting the remote control signal and the winch potentiometer signal, and the three-dimensional control signal and the winch manual signal as the first and second linkage groups, respectively; when the first linkage group is activated and the second linkage group is deactivated, the UAV enters the first control mode; when the second linkage group is activated and the first linkage group is deactivated, the UAV enters the second control mode. This invention addresses the technical problems of single control mode, insufficient human-machine collaboration, poor adaptability to environmental interference, and limited control accuracy in traditional tethered UAV control.

Description

Technical Field

[0001] This invention relates to the field of tethered unmanned aerial vehicle (UAV) control technology, specifically to a dual-mode dynamic collaborative control method and system based on pulling a tether line. Background Technology

[0002] Traditional tethered unmanned aerial vehicle (UAV) control technologies generally suffer from problems such as a single control mode, insufficient human-machine collaboration, and poor environmental adaptability, specifically manifested in the following aspects:

[0003] I. Limitations of Control Mode and Operational Coordination

[0004] Existing technologies mostly rely on remote controllers to achieve independent control of the drone and the winch for line deployment and retrieval. This requires manual adjustment of the aircraft's attitude and tether tension, resulting in high operational complexity and potential safety risks due to response delays. For example, traditional systems lack a dual-mode linkage mechanism, making it impossible to flexibly switch between conventional remote control scenarios and close-range precision tracking scenarios. This leads to insufficient control accuracy and ease of operation in specific tasks (such as emergency manual intervention and high-precision trajectory tracking).

[0005] II. Environmental and Mechanical Interference Issues

[0006] Temperature sensitivity: The elastic modulus of the tethering wire material fluctuates significantly with temperature changes (e.g., the elastic modulus of some materials decreases by 2%-5% for every 10°C increase in temperature). Traditional fixed parameter models are prone to Z-axis control signal deviations exceeding 15% under different temperature environments (-20°C to 60°C), leading to tension loss of control or decreased lifting accuracy.

[0007] Illumination and sway interference: Horizontal position detection relies on a single sensor (such as a visual or mechanical angle sensor), which is susceptible to environmental interference.

[0008] In scenes with strong light overexposure or low light noise, the deviation in extracting the pixel coordinates of the tether line increases in visual sensors.

[0009] Mechanical displacement sensors suffer from nonlinear distortion of the included angle difference data due to the high-frequency oscillation (0.5-5Hz) and elastic deformation of the tethered wire. Traditional linear fitting models are unable to compensate for this type of error.

[0010] III. Sensor Data Processing Defects

[0011] Signal noise and drift: Traditional tension detection uses a single potentiometer to collect displacement signals, which is easily affected by mechanical clearance and changes in contact resistance, resulting in a low signal-to-noise ratio; the rocker arm pulley assembly is installed at the UAV end, and the change in weight due to changes in the length of the tether line causes the potentiometer's median value to drift.

[0012] IV. Control Coupling and Dynamic Response Lag

[0013] Traditional systems do not consider the dynamic coupling effect between the tether line deployment length and the drone speed, resulting in cross-interference between XY-axis horizontal control and Z-axis vertical control; they lack a feedforward prediction mechanism, causing position response to lag behind wind disturbance changes, leading to trajectory deviation; and they do not design multi-mode protection logic based on oscillation frequency, which can easily cause attitude instability or cable entanglement risks in resonance scenarios (such as oscillation frequency exceeding 3Hz).

[0014] The aforementioned problems result in traditional tethered drones having poor adaptability and high operational thresholds in complex working conditions, making it difficult to meet the requirements of high-precision operations. Therefore, there is an urgent need for a dynamic control scheme that balances control flexibility, environmental robustness, and collaborative accuracy. Summary of the Invention

[0015] In order to overcome the shortcomings of the existing technology, the purpose of this invention is to provide a dual-mode dynamic cooperative control method and system based on a tethering line, which can solve the technical problems of single control mode, insufficient human-machine cooperation, poor adaptability to environmental interference and limited control accuracy in traditional tethered UAV control.

[0016] To solve the above problems, the technical solution adopted by the present invention is as follows:

[0017] A dual-mode dynamic collaborative control method based on a pull-tether line includes the following steps:

[0018] Displacement electrical signals are acquired by dual potentiometers configured on the rocker arm pulley block of the ground equipment;

[0019] The dynamic elastic modulus is generated by using a PWM signal generator based on the elastic coefficient model of the tether wire and the real-time collected ambient temperature data of the tether wire. The dynamic elastic modulus is then generated by combining the displacement electrical signal to generate a Z-axis PWM signal that characterizes the pulling force.

[0020] By using the angle difference between the XY joystick on the drone and the tether line, an XY-axis PWM signal representing the horizontal position is generated based on the angle difference-coordinate mapping model and dynamic filtering.

[0021] The Z-axis PWM signal is split into two paths. One path is transmitted to the winch ESC to control the winding and unwinding speed, and the other path is transmitted to the UAV and merged with the XY-axis PWM signal to form a three-dimensional control signal.

[0022] Set the remote control signal and the winch potentiometer signal as the first linkage group, and set the three-dimensional control signal and the winch manual signal as the second linkage group.

[0023] When the first linkage group is activated and the second linkage group is deactivated by the first and second remote control relays, the drone enters the first control mode; when the second linkage group is activated and the first linkage group is deactivated, the drone enters the second control mode.

[0024] Preferably, when entering the first control mode, the UAV receives a remote control signal, and at the same time the winch ESC receives a winch potentiometer signal to perform remote control control of the UAV in the XYZ three-axis direction, and to automatically rotate the winch motor forward and reverse to reel in and unload the cable by superimposing a tension fluctuation suppression factor into the Z-axis PWM signal.

[0025] When entering the second control mode, the UAV receives three-dimensional control signals, and the winch ESC receives manual signals from the winch. When the winch is manually retracting the line, the PWM signal generator outputs a Z-axis PWM signal representing descent. When the winch is manually releasing the line, the PWM signal generator outputs a Z-axis PWM signal representing ascent. Combined with the XY-axis PWM signals, a new three-dimensional control signal is generated, and the UAV's XYZ three-axis wire-controlled following is performed based on nonlinear decoupling.

[0026] Preferably, when generating the XY axis PWM signal, the following steps are included:

[0027] The drone uses a binocular vision sensor to capture the projection image of the tether line on a horizontal plane and extracts the pixel coordinates of the tether line from the projection image.

[0028] The analog signal of the joystick deflection angle is obtained by the mechanical displacement sensor of the XY joystick and converted into angle difference data;

[0029] The pixel coordinates and angle difference data are corrected based on the ambient light intensity and the oscillation frequency, respectively, and then input into the angle difference-coordinate mapping model built based on the BP neural network to output the predicted XY coordinates.

[0030] Real-time error compensation is performed on the predicted XY coordinates using the Kalman filter algorithm to generate XY axis PWM signals.

[0031] The state covariance matrix P of the filtering process is dynamically adjusted according to the real-time error.

[0032] Preferably, when correcting pixel coordinates based on ambient light intensity, the following steps are included:

[0033] Multiple ambient light intensity samples were continuously collected and divided into three intervals: low light, medium light, and high light based on the intensity. Logarithmic enhancement algorithm was used to improve pixel contrast in the low light interval, and dynamic range compression was used to suppress overexposed areas in the high light interval, resulting in the processed ambient light intensity samples.

[0034] The sample variance is calculated for the processed ambient light intensity samples. When the variance is less than or equal to the set value, the mean is directly taken as the current light intensity. When the variance is greater than the set value, median filtering is started to remove outliers and the mean is taken as the current light intensity.

[0035] The preset interval compensation coefficient is queried based on the current light intensity, and the pixel coordinates are non-linearly mapped and corrected by combining the logarithmic operation results of the current light intensity and the standard reference light intensity.

[0036] Preferably, when correcting the included angle difference data based on the oscillation frequency, the following is included:

[0037] Obtain the natural frequency curves of the tether under different tensions and establish a swing frequency-stiffness mapping model;

[0038] The current swing frequency of the mooring line is collected in real time. The current equivalent stiffness coefficient is calculated through the swing frequency-stiffness mapping model and compared with the standard stiffness coefficient to generate the stiffness correction coefficient λ.

[0039] The angle difference data is multiplied by the stiffness correction factor λ, and the swing phase compensation term is introduced for correction.

[0040] A variable step size LMS adaptive filter is used to denoise the corrected angle difference data, and the step size factor μ is dynamically adjusted according to the swing frequency.

[0041] Preferably, when outputting the predicted XY coordinates, the following are included:

[0042] The input layer assigns dynamic weights based on interval compensation coefficients to the corrected pixel coordinates and dynamic weights based on stiffness correction coefficient λ to the corrected angle difference data. The weight allocation ratio is optimized in real time through gradient descent algorithm.

[0043] The pixel coordinate features are processed by the Leaky ReLU activation function from the input layer to the first hidden layer, and the angle difference features are processed by the ELU activation function from the second hidden layer to the output layer. A Batch Normalization layer is added to the output layer to suppress gradient vanishing.

[0044] This also includes: the construction of an adversarial training sample set based on Gaussian noise, random scaling, and generative adversarial networks, and the online iterative optimization of the model based on coordinate error calculation, Bayesian optimization algorithm updates, and dynamic adjustment of the learning rate.

[0045] Preferably, when performing real-time error compensation, it includes:

[0046] When the root mean square error between the measured XY coordinates of the UAV's GPS and the filtered output XY coordinates exceeds a set value for multiple consecutive sampling periods, adaptive optimization of the Kalman gain K is initiated.

[0047] The root mean square error is divided into three levels—low error, medium error, and high error—by a fuzzy logic controller, and an adjustment coefficient corresponding to the K value is assigned. When the error level jumps from low or medium error to high error, a strong tracking factor t is triggered to perform a weighted correction on the state covariance matrix P of the Kalman filter.

[0048] Preferably, when superimposed with a tension fluctuation suppression factor, it includes:

[0049] The high-frequency fluctuation component of the tether tension was collected and decomposed into three frequency bands through wavelet transform;

[0050] Adaptive threshold noise reduction is performed on each frequency band component to retain the effective frequency band characterizing the elastic vibration of the cable;

[0051] The noise-reduced fluctuation component is converted into a reverse compensation PWM signal and linearly superimposed with the Z-axis PWM signal;

[0052] Among them, the amplitude of the reverse compensation PWM signal is positively correlated with the root mean square value of the tension fluctuation, and the phase lag is set to 90° to achieve active vibration reduction.

[0053] Preferably, when performing wired tracking of the XYZ axes of a drone, it includes:

[0054] The XY-axis PWM signal and the Z-axis PWM signal are nonlinearly decoupled by using the coupling coefficient matrix between the tether wire unfolding length and the real-time speed of the UAV. The coupling coefficient matrix is ​​dynamically updated based on the tether wire elastic coefficient model.

[0055] An inverse model of UAV dynamics is introduced to predict the UAV position deviation based on the current three-dimensional control signal and generate a feedforward compensation amount that is superimposed on the control signal. The magnitude of the compensation amount is positively correlated with the first derivative of the tether tension fluctuation.

[0056] When the tether line swing frequency exceeds the set value, the control weight of the XY axis is reduced and the attitude angle limiting protection is activated. At the same time, the Z axis control is switched to constant tension priority mode, and the tension fluctuation is controlled by the PID parameter self-tuning algorithm.

[0057] A dual-mode dynamic cooperative control system based on a pull-tether line, which, during operation, executes the above-mentioned method, including:

[0058] Tethered ground equipment includes a rocker arm pulley block, tether line, tether box winch, and PWM signal generator; the rocker arm pulley block is equipped with dual potentiometers for acquiring displacement electrical signals; the PWM signal generator is used to generate a Z-axis PWM signal based on the displacement electrical signal, and it is divided into two paths, one of which is transmitted to the winch ESC and the other of which is transmitted to the UAV.

[0059] The drone includes a flight control system, an XY joystick, and a signal merging module. The XY joystick generates XY-axis PWM signals based on the angle difference between itself and the tethering line. The signal merging module receives the Z-axis PWM signal sent by the tethered ground equipment and merges it with the XY-axis PWM signal into a three-dimensional control signal.

[0060] Switching execution unit: includes first and second remote control relays and remote controller; both the first and second remote control relays are equipped with two sets of linkage contacts; the first remote control relay is used to control the control signals received by the flight control system, and the second remote control relay is used to control the control signals received by the winch ESC; the remote controller is used to send remote control signals to the UAV and to enable the UAV to enter the first or second control mode by controlling the on or off of the remote control relays.

[0061] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0062] (1) Dual-mode collaborative control enhances operational flexibility

[0063] By dynamically switching between a first control mode (conventional remote control + automatic line reeling and deployment) and a second control mode (wire-controlled following + manual intervention), the system caters to both automated operations and high-precision manual control requirements. For example, the second control mode allows the drone to follow in all directions by pulling the tether line, achieving fully synchronized landing during the line reeling and descent phase, significantly reducing operational difficulty.

[0064] (2) Adaptive correction of environmental and mechanical characteristics to improve control accuracy

[0065] Dynamic temperature compensation: Based on the ambient temperature, the elastic modulus of the tethered line is corrected in real time, which solves the problem of excessive tension detection error in traditional fixed parameter models at different temperatures.

[0066] Illumination and sway interference suppression: By fusing data from binocular vision and mechanical displacement sensors, combined with an illumination interval enhancement algorithm (logarithmic enhancement for low light and dynamic compression for high light) and a sway frequency-stiffness mapping model, the horizontal position detection error is reduced.

[0067] (3) Multi-source data fusion and dynamic modeling enhance system robustness

[0068] A BP neural network is used to construct a nonlinear mapping model of the angle difference-coordinate, and Kalman filtering is used to dynamically adjust the state covariance matrix to achieve real-time error compensation of the predicted coordinates. Variable step size LMS adaptive filtering and wavelet transform active vibration reduction technology are introduced to effectively suppress high-frequency vibration of the mooring line and ensure signal stability in complex environments.

[0069] (4) Nonlinear decoupling and feedforward control to optimize dynamic response

[0070] The XYZ three-axis nonlinear decoupling is achieved by using the coupling coefficient matrix between the tether line unfolding length and the UAV speed. The inverse dynamic model is introduced to predict the position deviation and generate the feedforward compensation, which shortens the trajectory tracking lag time. Multi-mode protection logic triggered by oscillation frequency is designed (such as starting the constant tension priority mode when >3Hz) to avoid the risk of attitude instability caused by resonance.

[0071] In summary, this invention significantly improves the ease of operation, control accuracy, and environmental adaptability of tethered drones through innovative control modes, environmental adaptive algorithms, and dynamic collaborative mechanisms, and can be widely applied to various high-precision operation scenarios.

[0072] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0073] Figure 1 This is a flowchart illustrating the steps of the dual-mode dynamic cooperative control method according to an embodiment of the present invention;

[0074] Figure 2 This is a structural diagram of the rocker arm pulley block according to an embodiment of the present invention;

[0075] Figure 3 This is a flowchart illustrating the generation of XY-axis PWM signals according to an embodiment of the present invention;

[0076] Figure 4 This is a flowchart illustrating the pixel coordinate correction based on ambient light intensity in an embodiment of the present invention.

[0077] Figure 5 This is a flowchart illustrating the correction of the included angle difference data based on the oscillation frequency in an embodiment of the present invention;

[0078] Figure 6 This is an interaction diagram of the dual-mode dynamic collaborative control system module according to an embodiment of the present invention.

[0079] Explanation of reference numerals in the attached figures: 20. Dual-mode dynamic collaborative control system; 21. Tethered ground equipment; 211. PWM signal generator; 212. Rocker arm pulley block; 213. Dual potentiometers; 214. Winch ESC; 215. Winch motor; 22. Unmanned aerial vehicle (UAV); 221. Flight control system; 222. XY joystick; 223. Signal merging module; 224. Receiver; 231. First remote control relay; 232. Second remote control relay; 233. Remote controller. Detailed Implementation

[0080] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0081] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0082] The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the scope of this application and its application or use.

[0083] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0084] Example 1, see Figure 1 The present invention provides a step-by-step diagram of a dual-mode dynamic cooperative control method, as shown in the figure. Figure 1 The method for dual-mode dynamic collaborative control based on a pull-tether line, as shown, includes the following steps:

[0085] S101. Pulling State Detection: The relative pulling state between the UAV 22 and the tether line is detected by the rocker arm pulley assembly 212 on the ground tethering device 21; wherein, the displacement electrical signal is collected by the dual potentiometers 213 configured in the rocker arm pulley assembly 212, and the Z-axis PWM signal characterizing the pulling is generated by the PWM signal generator 211.

[0086] S102. Horizontal state detection: Generate an XY axis PWM signal representing the horizontal position by using the angle difference between the XY joystick 222 on the UAV 22 and the tether line;

[0087] S103. Signal Distribution and Transmission: The Z-axis PWM signal is divided into two paths. One path is transmitted to the winch ESC 214 to control the winding and unwinding speed. The other path is transmitted to the UAV 22 via wired or wireless means and merged with the XY-axis PWM signal to form a three-dimensional control signal.

[0088] S104. Signal synchronization logic: Set the remote control signal and the winch potentiometer signal as the first linkage group, and set the three-dimensional control signal and the winch manual signal as the second linkage group;

[0089] S105. Control mode switching: When the first linkage group is turned on and the second linkage group is turned off by the first remote control relay 231 and the second remote control relay 232, the drone 22 enters the first control mode; when the second linkage group is turned on and the first linkage group is turned off by the first remote control relay 231 and the second remote control relay 232, the drone 22 enters the second control mode.

[0090] The first remote control relay 231 is mounted on the UAV 22, and the second remote control relay 232 is mounted on the tethered ground equipment 21.

[0091] See Figure 2 The rocker arm pulley assembly structure diagram is shown below. In this embodiment of the invention, it is necessary to further explain that the rocker arm pulley assembly 212 with pull sensing is deployed on the tethered ground device 21, which solves the problem of potentiometer midpoint drift caused by weight change due to change in tether line length when the rocker arm pulley assembly 212 is installed on the UAV 22 (the horizontal XY axis control detection function is still integrated into the UAV 22).

[0092] Background Description: Traditional tethered drones 22 generally suffer from a lack of diverse control modes and operational coordination. For example, existing technologies often rely on remote controllers 233 to achieve independent control of the drone 22 and the winch for line deployment and retrieval, requiring manual adjustment of the aircraft's attitude and tether line tension. This is highly complex and prone to safety risks due to response delays. Furthermore, in specific operational scenarios (such as close-range precision following and emergency manual intervention), traditional remote control modes struggle to balance control accuracy and operational flexibility, exhibiting problems such as excessive tension fluctuations and lagging horizontal following. Therefore:

[0093] In step S105 above, when entering the first control mode, the UAV 22 receives the remote control signal, and at the same time the winch ESC 214 receives the winch potentiometer signal to perform remote control control of the UAV 22 in the XYZ three-axis direction, and to automatically rotate the winch motor 215 forward and reverse to reel in and out the wire by superimposing a tension fluctuation suppression factor into the Z-axis PWM signal.

[0094] When entering the second control mode, the UAV 22 receives three-dimensional control signals, and the winch ESC 214 receives manual winch signals. When the winch is manually retracting the line, the PWM signal generator 211 outputs a Z-axis PWM signal representing descent. When the winch is manually releasing the line, the PWM signal generator 211 outputs a Z-axis PWM signal representing ascent. Combined with the XY-axis PWM signals, a new three-dimensional control signal is generated, and the UAV 22 performs wire-controlled following of the XYZ axes based on nonlinear decoupling.

[0095] In this embodiment of the invention, it is necessary to further explain that the embodiment achieves the collaborative work of the UAV 22 and the tethered ground equipment 21 by designing a dual-mode linkage control logic: the first control mode meets the automatic cable reeling and laying requirements in conventional remote control scenarios, while the second control mode achieves high-precision tracking and emergency response capabilities through the dynamic coupling of manual operation and wire control signals, thereby improving the reliability and ease of operation of the tethered UAV 22 in diverse operation scenarios.

[0096] The second control mode achieves altitude control of the drone 22 by pulling the tether line. It can complete all-round following of the horizontal XY axis and vertical Z axis during the tethered drone 22 following process, and achieve fully synchronized landing during the line retraction and landing phase, which greatly reduces the difficulty of landing the tethered drone 22.

[0097] Background Description: Traditional tension detection solutions for tethered drones typically use a fixed elastic modulus parameter for tension-displacement conversion, neglecting the impact of ambient temperature on the physical properties of the tethering line. The elastic modulus of tethering line materials (such as high-strength fibers and composite cables) fluctuates significantly with temperature changes (e.g., for every 10°C increase in temperature, the elastic modulus of some materials may decrease by 2%-5%), leading to systematic errors in tension detection. If the static elastic modulus at standard temperatures is directly used to calculate tension, in extreme temperature environments (such as -20°C or 60°C), the Z-axis PWM signal output deviation may exceed 15%, causing a significant decrease in the drone's lift control accuracy or the risk of uncontrolled tethering line tension.

[0098] Meanwhile, traditional tension detection often relies on a single potentiometer to acquire displacement signals, which is susceptible to interference from factors such as mechanical clearance and changes in contact resistance, resulting in significant signal noise. Furthermore, installing a temperature sensor on the UAV's 22-end increases the aircraft's load and makes temperature measurement stability susceptible to airflow disturbances, while ground-based temperature acquisition can improve the accuracy and reliability of ambient temperature monitoring through fixed installation. Based on this:

[0099] In step S101 above, generating the Z-axis PWM signal representing the pulling action includes:

[0100] The displacement electrical signal collected by the dual potentiometer 213 is acquired, and the dynamic elastic modulus is generated through the tie-line elastic coefficient model. The Z-axis PWM signal is generated by combining the displacement electrical signal and the dynamic elastic modulus.

[0101] When generating the dynamic elastic modulus, the following steps are taken: constructing a model of the elastic modulus of the tether based on the elastic modulus at the standard temperature, the temperature coefficient of the tether material, the standard temperature, and the ambient temperature of the tether.

[0102] The real-time collected ambient temperature data of the tether wire is input into the tether wire elastic coefficient model to generate the difference with the standard temperature. The elastic modulus at the standard temperature is then corrected by combining the temperature coefficient of the tether wire material, and the dynamic elastic modulus is output.

[0103] The ambient temperature of the tether line is obtained by a temperature sensor installed on the tether ground device 21.

[0104] In this embodiment of the invention, it is necessary to further explain that a dynamic elastic modulus correction mechanism is proposed: the ambient temperature is collected in real time by a ground temperature sensor, and an elastic coefficient model is constructed by combining the material temperature coefficient to dynamically correct the elastic modulus parameters, so as to ensure the linearity of tension-displacement conversion under different temperature conditions; at the same time, the displacement signal is collected differentially by dual potentiometers 213 to reduce the risk of single-point failure, improve the signal-to-noise ratio of the original signal, and provide a reliable data foundation for the accurate generation of Z-axis PWM signal.

[0105] Background Description: Traditional horizontal position detection schemes for tethered UAVs often employ a single sensor (such as a purely visual or purely mechanical angle sensor) for position estimation, which suffers from poor environmental adaptability and large model linearization errors. For example, visual sensors are easily affected by changes in light intensity (such as strong light overexposure or low light noise), leading to deviations in the extraction of tether line pixel coordinates; mechanical displacement sensors, due to factors such as tether line swaying and material elastic deformation, produce angular difference data drift. Furthermore, traditional models often use linear fitting methods to establish the mapping relationship between sensor data and position, making it difficult to handle nonlinear factors such as light interference and mechanical hysteresis, resulting in insufficient accuracy of the horizontal control signal (XY axis PWM) output, especially under complex conditions (such as strong wind disturbances or rapid attitude adjustments), easily leading to following delays or overshoot. Based on this:

[0106] See Figure 3 The flowchart for generating XY-axis PWM signals includes the following steps in step S102: Generating the XY-axis PWM signal representing the horizontal position includes:

[0107] The binocular vision sensor mounted on the UAV 22 acquires the projection image of the tether line on the horizontal plane, and extracts the pixel coordinates of the tether line from the projection image;

[0108] The analog signal of the joystick deflection angle is obtained by the mechanical displacement sensor of the XY joystick 222 and converted into angle difference data;

[0109] After correcting the pixel coordinates and angle difference data, they are used as the model input.

[0110] Input the model input into the angle difference-coordinate mapping model built on the BP neural network, and output the predicted XY coordinates;

[0111] Real-time error compensation is performed on the predicted XY coordinates using the Kalman filter algorithm to generate XY axis PWM signals.

[0112] The state covariance matrix P of the filtering process is dynamically adjusted according to the real-time error.

[0113] In this embodiment of the invention, it is necessary to further explain that the proposed method of multi-source data fusion and dynamic modeling is as follows: by complementing the heterogeneous data of binocular vision and mechanical displacement sensors, and combining the ambient light and oscillation frequency correction mechanism, the reliability of the original data is improved; a nonlinear mapping model is constructed using a BP neural network to achieve high-precision conversion of the angle difference to the coordinates; and a Kalman filter is introduced to dynamically adjust the state covariance matrix to compensate for prediction errors in real time, and finally generate a more robust XY axis PWM signal to meet the horizontal position control requirements of the tethered UAV 22 in complex environments.

[0114] In one possible embodiment, correcting the pixel coordinates and angle difference data includes:

[0115] The ambient light intensity is collected in real time by a light sensor module integrated on the fuselage of the UAV 22.

[0116] The swaying frequency of the tether line is collected by a miniature triaxial accelerometer installed at the bottom anchor point of the tether line.

[0117] Pixel coordinates are corrected based on ambient light intensity, and the angle difference data is corrected based on the oscillation frequency.

[0118] The light sensor module employs a silicon-based photodiode array with a measurement range of 0-100,000 Lux and a sampling frequency ≥10Hz, and is connected via I... 2 The C-bus communicates with the flight control system 221; the miniature triaxial accelerometer has a built-in MEMS vibration detection unit with a sampling frequency ≥100Hz, and can output X / Y / Z axis acceleration components in real time. It converts the time-domain vibration signal into frequency-domain data through Fourier transform and extracts the main oscillation frequency as the tether line oscillation frequency.

[0119] In this embodiment of the invention, a multi-source environmental parameter fusion correction strategy is proposed: A high-precision illumination sensor module (silicon-based photodiode array) mounted on the UAV 22 monitors illumination intensity in real time, and dynamically corrects pixel coordinates using an interval compensation algorithm; simultaneously, a miniature triaxial accelerometer is deployed at the bottom anchor point of the tethering line, and a MEMS vibration detection unit collects the oscillation frequency; the angle difference data is optimized using a stiffness mapping model and adaptive filtering. Regarding sensor selection, the illumination module covers a range of 0-100000 Lux and supports I... 2High-speed C-band communication, an accelerometer with a sampling frequency ≥100Hz and Fourier transform frequency domain analysis capability ensure the real-time and accuracy of environmental parameter acquisition, providing high-quality input data for subsequent BP neural network coordinate mapping.

[0120] Background Description: In traditional tethered UAV 22 visual positioning systems, changes in illumination intensity significantly reduce the contrast and edge sharpness of the tethered line projection image. For example, increased image noise in low-light environments leads to pixel coordinate extraction errors, while overexposed areas in high-light environments cause the tethered line outline to be lost. Traditional solutions often employ fixed image enhancement algorithms (such as global contrast stretching) without addressing the dynamic range of illumination (0-100000 Lux), resulting in poor correction performance under different lighting conditions. Furthermore, illumination sensor sampling data is susceptible to transient interference (such as cloud cover or ground reflection), causing jumps in values. Traditional mean filtering struggles to effectively remove outliers, further exacerbating coordinate mapping errors. Additionally, the mapping relationship between pixel coordinates and actual physical coordinates is non-linear; traditional linear interpolation cannot compensate for the complex coupling between illumination intensity and imaging distortion, leading to reduced horizontal position control accuracy. Based on this:

[0121] See Figure 4 The flowchart for correcting pixel coordinates, in one possible embodiment, includes correcting pixel coordinates based on ambient light intensity, comprising:

[0122] Light intensity range division: Ten ambient light intensity samples were continuously collected by the light sensor module. The light intensity samples were divided into three ranges based on intensity: low light (0-1000 Lux), medium light (1001-50000 Lux), and high light (50001-100000 Lux). Logarithmic enhancement algorithm was used to improve pixel contrast in the low light range, and dynamic range compression was used to suppress overexposed areas in the high light range, resulting in the processed ambient light intensity samples.

[0123] Real-time noise suppression: Calculate the sample variance for the processed ambient light intensity samples. When the variance is less than or equal to 500 Lux... 2 When the variance is greater than 500 Lux, the mean value is directly taken as the current light intensity. 2 When the light intensity is high, median filtering is activated to remove outliers and the mean value is taken as the current light intensity.

[0124] Coordinate mapping correction: Based on the current light intensity, the preset interval compensation coefficient is queried, and the pixel coordinates are non-linearly mapped and corrected by combining the logarithmic operation results of the current light intensity and the standard reference light intensity.

[0125] In this embodiment of the invention, it is necessary to further explain that the proposed illumination adaptive correction mechanism is as follows: by dividing the illumination range into three levels and configuring corresponding processing strategies (logarithmic enhancement for low light and dynamic compression for high light), image quality optimization under full illumination range is achieved; a median filtering algorithm triggered by variance threshold is adopted to improve the stability of illumination sampling data; and a nonlinear mapping model combining interval compensation coefficient and logarithmic operation is combined to accurately correct pixel coordinate offset caused by illumination changes, providing high signal-to-noise ratio input data for subsequent BP neural network coordinate prediction, thereby improving the environmental robustness of tethered UAV 22 horizontal position detection.

[0126] Background Description: In the traditional mechanical displacement sensing system of the tethered UAV 22, the change in the swaying frequency of the tether line will cause dynamic fluctuations in the equivalent stiffness coefficient. However, the traditional solution does not establish a correlation model between the swaying frequency and mechanical characteristics, and directly uses a fixed stiffness coefficient to calculate the angle difference data. As a result, when the tension changes (such as tension fluctuation of ±30% during the ascent and descent of the UAV 22) or when there are external disturbances (such as the swaying frequency change of 0.5-5Hz caused by gusts), the stiffness coefficient error can reach 15%-20%, causing nonlinear distortion of the angle difference data.

[0127] Meanwhile, the tethered wire is prone to resonance in the 0.5-2Hz range, and the output signal of traditional mechanical displacement sensors exhibits significant phase lag (the lag can reach 10°-15°). Furthermore, the filtering parameters are not dynamically adjusted for different oscillation frequencies. Fixed-step LMS filtering has slow convergence speed in the high-frequency range (convergence time >200ms at >5Hz) and insufficient filtering accuracy in the low-frequency range (noise suppression ratio <20dB at <1Hz), further exacerbating the measurement error of the angle difference data. Based on this:

[0128] See Figure 5 A flowchart for correcting the included angle difference data, in one possible embodiment, includes the following steps when correcting the included angle difference data based on the oscillation frequency:

[0129] Construction of the oscillation frequency-stiffness mapping model: Obtain the natural frequency curves of the mooring line under different tensions, and establish a nonlinear mapping model between the oscillation frequency and the equivalent stiffness coefficient, where the equivalent stiffness coefficient K = ω. 2 ·m / L (ω is the oscillation frequency, m is the mass per unit length, and L is the unfolded length of the tether).

[0130] Dynamic correction coefficient generation: The current swing frequency of the mooring line is collected in real time, the current equivalent stiffness coefficient is calculated through the swing frequency-stiffness mapping model, and compared with the standard stiffness coefficient to generate the stiffness correction coefficient λ=K / K0 (K0 is the stiffness coefficient under the standard state at 25℃).

[0131] Mechanical displacement compensation: The included angle difference data is multiplied by the stiffness correction coefficient λ, and the swing phase compensation term is introduced for correction; when the swing frequency is in the resonance range of 0.5-2Hz, a phase lag compensation based on the arctangent function is superimposed, and the compensation amplitude is positively correlated with the swing amplitude.

[0132] Adaptive filtering optimization: A variable step size LMS adaptive filter is used to denoise the corrected angle difference data. The step size factor μ is dynamically adjusted according to the swing frequency. Specifically, μ is set to 0.01-0.05 in the high frequency band (>5Hz) to accelerate convergence, μ is set to 0.005-0.01 in the mid frequency band (1-5Hz) to balance convergence speed and filtering accuracy, and μ is set to 0.001-0.005 in the low frequency band (<1Hz) to improve filtering accuracy.

[0133] In this embodiment of the invention, it is necessary to further explain that the proposed oscillation frequency coupling correction mechanism is as follows: by establishing an oscillation frequency-stiffness mapping model, the equivalent stiffness coefficient is calculated in real time, and a dynamic correction coefficient λ is generated to compensate for stiffness changes; an arctangent function phase compensation term is introduced for the resonance interval to offset the mechanical hysteresis effect; and a variable step size LMS adaptive filter (μ=0.01-0.05 in the high frequency band, μ=0.005-0.01 in the mid frequency band, and μ=0.001-0.005 in the low frequency band) is adopted to achieve adaptive suppression of noise across the entire frequency band, thereby effectively controlling the measurement error of the included angle difference data and providing high-precision input for the generation of XY axis PWM signals.

[0134] Background Description: Traditional angle difference-coordinate mapping models employ a fixed weight allocation strategy when fusing multi-source data. This strategy cannot dynamically adjust feature contributions based on the reliability of sensor data. For example, when drastic changes in illumination cause a surge in pixel coordinate noise, fixed weights introduce redundant features, reducing model robustness. Furthermore, single activation functions (such as pure ReLU or Sigmoid) struggle to balance the linear distribution of pixel coordinates with the nonlinear characteristics of angle difference data, resulting in insufficient feature representation capabilities in hidden layers. Deep networks are also prone to gradient vanishing (gradient descent amplitude decay > 60%), impacting the real-time performance of coordinate mapping.

[0135] Furthermore, traditional training sample sets often rely on laboratory environment data collection, lacking data on extreme operating conditions (such as a sudden increase in mooring stiffness due to -20℃ low temperatures, or high-frequency swaying caused by strong winds of 15m / s), thus limiting the model's generalization ability. Moreover, after offline training, the parameters are fixed, making it unable to cope with sensor drift during long-term flight (such as error accumulation due to temperature drift in MEMS devices), and the deviation between predicted coordinates and GPS measurements gradually increases over time. Based on this:

[0136] In one possible embodiment, outputting the predicted XY coordinates includes:

[0137] Adaptive weight allocation is performed through the input layer of the angle difference-coordinate mapping model: dynamic weights based on interval compensation coefficients are assigned to the corrected pixel coordinates, and dynamic weights based on stiffness correction coefficient λ are assigned to the corrected angle difference data. The weight allocation ratio is optimized in real time through the gradient descent algorithm to reduce feature redundancy under extreme environmental interference.

[0138] The hidden layers of the angle difference-coordinate mapping model are processed using a combination of activation functions: the Leaky ReLU activation function is used to process pixel coordinate features from the input layer to the first hidden layer, and the ELU activation function is used to process angle difference features from the second hidden layer to the output layer. A Batch Normalization layer is added to the output layer to suppress gradient vanishing and improve the real-time convergence speed of coordinate mapping.

[0139] This also includes the steps of constructing an adversarial training sample set and online iterative optimization of the model;

[0140] The steps for constructing the adversarial training sample set are as follows: Gaussian noise (mean 0, variance 0.02-0.1) is added to the original pixel coordinate samples to simulate camera shake, and the angle difference samples are randomly scaled (scaling factor 0.8-1.2) to simulate the elastic deformation of the mooring line. Virtual samples under extreme conditions are generated by generative adversarial network (GAN) so that the training set covers the temperature range of -20℃ to 60℃ and the wind speed scene of 10-15m / s.

[0141] Online model iteration optimization steps: Every preset flight cycle (10-30 seconds), the error between the actual GPS XY coordinates of the UAV and the predicted XY coordinates output by the angle difference-coordinate mapping model is calculated. The connection weights and bias terms of the network are updated by Bayesian optimization algorithm. The learning rate is dynamically adjusted during the iteration process. When the error is less than 0.5 meters, the learning rate is reduced to 1 / 5 of the initial value to stabilize the model output.

[0142] In this embodiment of the invention, it is necessary to further explain that a dynamic modeling and online optimization mechanism is proposed: the weight allocation is optimized in real time through gradient descent (pixel coordinate weights are associated with illumination compensation coefficients, and angle difference weights are associated with stiffness correction coefficients λ), so as to achieve adaptive feature selection; the Leaky ReLU-ELU hybrid activation function and BatchNormalization are used to suppress gradient vanishing and improve feature extraction efficiency; a GAN virtual sample set is constructed to cover a wide temperature range (-20℃ to 60℃) and strong wind scenarios, and the network parameters are updated iteratively every 10-30 seconds through Bayesian optimization (the learning rate is reduced to 1 / 5 when the error is <0.5m), so as to stably control the prediction error of XY coordinates and provide high-precision initial input for Kalman filtering.

[0143] Background Description: In traditional tethered UAV navigation systems, Kalman filtering often uses a fixed gain matrix, which cannot dynamically adjust the state estimation weights based on real-time errors. When the deviation between the measured GPS coordinates and the predicted values ​​changes abruptly (e.g., electromagnetic interference causing the RMSE to briefly exceed 0.3 meters), the fixed gain leads to filtering lag, preventing the state covariance matrix P from converging quickly and causing position tracking errors to accumulate. Furthermore, traditional error grading relies heavily on empirical thresholds and lacks the ability to partition the continuous domain using fuzzy logic, making it difficult to quantify the nonlinear impact of different error levels (low / medium / high) on gain adjustment, resulting in insufficient system robustness. Therefore:

[0144] In one possible embodiment, real-time error compensation includes:

[0145] Introducing a prediction error feedback mechanism: When the root mean square error (RMSE) between the UAV's measured GPS XY coordinates and the filtered output XY coordinates exceeds 0.3 meters for three consecutive sampling periods, adaptive optimization of the Kalman gain K is initiated.

[0146] The root mean square error (RMSE) is divided into three levels—low error (<0.1m), medium error (0.1-0.3m), and high error (>0.3m)—using a fuzzy logic controller, and corresponding adjustment coefficients of 0.9, 1.0, and 1.2 are assigned to the corresponding K values.

[0147] When the error level jumps from low or medium error to high error, a strong tracking factor t (with a value of 1.5-3.0) is triggered to perform a weighted correction on the state covariance matrix P of the Kalman filter, thereby enhancing the tracking ability for sudden errors.

[0148] In this embodiment of the invention, it is necessary to further explain that the error levels are dynamically divided by the fuzzy logic controller and matched with Kalman gain coefficients (0.9 for low error, 1.0 for medium error, and 1.2 for high error) to achieve continuous domain adaptive adjustment of the gain; a strong tracking factor (1.5-3.0) is introduced to weight and correct the state covariance matrix (P) to enhance the ability to quickly track sudden errors.

[0149] Background Description: In traditional winch cable reel control systems, the Z-axis PWM signal generation does not consider high-frequency fluctuations in tether cable tension (elastic vibrations above 10Hz), and tension feedback adjustment is achieved only through simple PID closed-loop control. Because the tension signal is not frequency-domain decomposed, it is impossible to distinguish between the cable's inherent elastic vibration (10-50Hz) and external wind disturbances (<10Hz), resulting in a phase difference between the motor output torque and the actual tension requirement (lag can reach 45°-60°). In strong winds of 15m / s, the tension fluctuation amplitude reaches ±20%, causing cable slack or over-tension, resulting in 22-degree attitude oscillations in the drone (pitch angle fluctuation ±3°).

[0150] Meanwhile, traditional filtering methods (such as moving average filtering) have limited effectiveness in suppressing multi-band mixed noise, cannot effectively extract key vibration components in the 10-50Hz range, and suffer from insufficient reverse compensation accuracy. Furthermore, the compensation signal and the original PWM signal are often superimposed in phase, achieving only passive vibration reduction and failing to offset the elastic vibration energy with a 90° phase difference. This results in the cable reel control accuracy failing to meet the 22-point hovering requirement of the UAV. Based on this:

[0151] In one possible embodiment, when superimposing a tension fluctuation suppression factor, the following is included:

[0152] The high-frequency fluctuation components (above 10Hz) of the tether tension are collected and decomposed into 3 frequency bands through wavelet transform;

[0153] Adaptive threshold noise reduction is performed on each frequency band component to retain the effective frequency band characterizing the elastic vibration of the cable;

[0154] The noise-reduced fluctuation component is converted into a reverse compensation PWM signal and linearly superimposed with the Z-axis PWM signal to offset the impact of high-frequency tension oscillation on the accuracy of wire take-up and release.

[0155] Among them, the amplitude of the reverse compensation PWM signal is positively correlated with the root mean square value of the tension fluctuation, and the phase lag is set to 90° to achieve active vibration reduction.

[0156] In this embodiment of the invention, an active vibration reduction control mechanism is proposed: tension fluctuations are decomposed into three frequency bands using wavelet transform, and effective vibration components are extracted after adaptive threshold noise reduction. A reverse compensation PWM signal with a 90° phase lag is generated and linearly superimposed with the Z-axis control signal to achieve active vibration reduction. The amplitude of the compensation signal is dynamically correlated with the root mean square value of the tension fluctuation, which can improve the suppression rate of high-frequency vibrations (above 10Hz), thereby effectively controlling the amplitude of tension fluctuations during wire take-up and release, and significantly improving the vertical position control accuracy of the UAV.

[0157] Background Description: In traditional tethered UAV 22 three-axis remote control following systems, linear decoupling control strategies are often adopted, which do not consider the dynamic coupling effect between the tether line deployment length and the real-time speed of the UAV 22. This results in cross-interference between the XY axis horizontal control and the Z axis vertical control. Especially when the tether line undergoes elastic deformation, the control signals interfere with each other, causing the UAV 22 to deviate from its trajectory.

[0158] Meanwhile, traditional feedback control relies on the hysteresis adjustment of position error and lacks a feedforward prediction mechanism. When the UAV 22 encounters sudden wind disturbances, the position response lags, leading to overshoot of the tether line tension. Furthermore, the system lacks multi-mode switching logic. When the tether line oscillation frequency exceeds 3Hz (such as resonance caused by strong winds), it maintains conventional control parameters, causing attitude instability and even the risk of cable entanglement. Based on this:

[0159] In one possible embodiment, performing 22XYZ three-axis wire-controlled following of a drone includes:

[0160] Dynamic coupling compensation: The XY-axis PWM signal and the Z-axis PWM signal are nonlinearly decoupled by the coupling coefficient matrix between the tether wire unfolding length and the real-time speed of the UAV 22. The coupling coefficient matrix is ​​dynamically updated based on the tether wire elastic coefficient model.

[0161] Predictive feedforward control: Introducing the inverse dynamic model of UAV 22, the position deviation of UAV 22 is predicted 0.1-0.3 seconds later based on the current three-dimensional control signal, and a feedforward compensation amount is generated and superimposed on the control signal. The amplitude of the compensation amount is positively correlated with the first derivative of the tether tension fluctuation.

[0162] Multi-mode switching logic: When the tether line swing frequency exceeds 3Hz, the emergency following mode is automatically activated: At this time, the XY axis control weight is reduced by 20% and the attitude angle limiting protection is activated. At the same time, the Z axis control is switched to constant tension priority mode, and the tension fluctuation is controlled within ±5% of the rated value through the PID parameter self-tuning algorithm.

[0163] In this embodiment of the invention, it is necessary to further explain that a dynamic collaborative control mechanism is proposed: the coupling coefficient matrix is ​​updated in real time through the mooring linear elastic coefficient model to achieve nonlinear decoupling of the XY-Z axes; a dynamic inverse model is introduced to predict the position deviation of 0.1-0.3 seconds, generating a feedforward compensation quantity associated with the first derivative of tension fluctuation; a multi-mode switching logic based on the oscillation frequency is designed (emergency mode is activated when the oscillation frequency is >3Hz), the control weight is dynamically adjusted (XY axis is reduced by 20%) and constant tension priority control is started (fluctuation ±5% of the rated value), so that the three-axis following error is reduced and the trajectory tracking accuracy under complex working conditions is significantly improved.

[0164] Example 2, see Figure 6 The invention provides a dual-mode dynamic collaborative control system module interaction diagram, as shown in the following figure. Figure 6 The dual-mode dynamic collaborative control system 20 based on a tethering line is shown, including: tethering ground equipment 21, unmanned aerial vehicle 22, and switching execution unit.

[0165] The tethered ground equipment 21 includes a rocker arm pulley assembly 212, a tether line, a tether box winch, and a PWM signal generator 211. The rocker arm pulley assembly 212 is equipped with dual potentiometers 213 to convert mechanical displacement into electrical displacement signals. The PWM signal generator 211 generates a Z-axis PWM signal representing the pulling action based on the electrical displacement signal, and divides it into two paths. One path is transmitted to the winch ESC 214 to control the speed of line take-up and release, and the other path is transmitted to the UAV 22 via wired or wireless means.

[0166] The UAV 22 includes a flight control system 221, an XY joystick 222, and a signal merging module 223; the XY joystick 222 is used to generate XY-axis PWM signals representing the horizontal position based on the angle difference between it and the tether line; the signal merging module 223 is used to receive the Z-axis PWM signal sent by the tethered ground equipment 21 and merge it with the XY-axis PWM signal into a three-dimensional control signal;

[0167] The switching execution unit includes a first remote control relay 231, a second remote control relay 232, and a remote controller 233; both the first remote control relay 231 and the second remote control relay 232 are equipped with two sets of linkage contacts; the first remote control relay 231 is used to control the control signals received by the flight control system 221, and the second remote control relay 232 is used to control the control signals received by the winch ESC 214; the remote controller 233 is used to send remote control signals to the UAV 22 and to enable the UAV 22 to enter the first control mode or the second control mode by controlling the opening or closing of the remote control relay.

[0168] In one possible embodiment, the tethered box winch further includes: a winch potentiometer and a winch motor 215;

[0169] When entering the first control mode, the first remote control relay 231 controls the flight control system 221 to receive remote control signals, while the second remote control relay 232 controls the winch ESC 214 to receive winch potentiometer signals. The UAV 22 performs remote control control in the XYZ three-axis direction, and the winch motor 215 performs automatic forward and reverse winding and unwinding.

[0170] When entering the second control mode, the first remote control relay 231 controls the flight control system 221 to receive three-dimensional control signals, while the second remote control relay 232 controls the winch ESC 214 to receive winch manual signals. When the winch is manually retracting the line, the PWM signal generator 211 outputs a Z-axis PWM signal representing descent. When the winch is manually releasing the line, the PWM signal generator 211 outputs a Z-axis PWM signal representing ascent. Combined with the XY-axis PWM signals, a new three-dimensional control signal is generated, and the UAV 22 performs wire-controlled following on the XYZ axes.

[0171] In one possible embodiment, the drone 22 further includes a receiver 224; the receiver 224 is used to receive remote control signals of the drone 22 emitted by the remote controller 233 and control signals of the remote control relay, so as to control the flight of the drone 22 and the on / off state of the remote control relay.

[0172] In one possible embodiment, the tethered ground device 21 is also equipped with a temperature sensor for collecting the ambient temperature of the tethered line.

[0173] In one possible embodiment, the drone 22 is also equipped with a binocular vision sensor for acquiring a projection image of the tether line on a horizontal plane and extracting the pixel coordinates of the tether line from the projection image.

[0174] In one possible embodiment, the XY joystick 222 is also equipped with a mechanical displacement sensor for acquiring analog signals of the joystick deflection angle and converting them into angle difference data.

[0175] In one possible embodiment, the drone 22 is also equipped with a light sensor module for real-time acquisition of ambient light intensity. Preferably, the light sensor module employs a silicon-based photodiode array with a measurement range covering 0-100,000 Lux, a sampling frequency ≥10Hz, and is connected via I... 2 The C-bus communicates with the flight control system 221.

[0176] In one possible embodiment, a miniature triaxial accelerometer is configured at the bottom anchor point of the tether line to collect the tether line's swing frequency. Preferably, the miniature triaxial accelerometer has a built-in MEMS vibration detection unit with a sampling frequency ≥100Hz, and can output X / Y / Z axis acceleration components in real time. The time-domain vibration signal is converted into frequency-domain data through Fourier transform, and the dominant swing frequency is extracted as the tether line's swing frequency.

[0177] In one possible embodiment, the mechanical displacement sensor is configured with a variable step size LMS adaptive filter for noise reduction of the corrected angle difference data.

[0178] In one possible embodiment, the XY joystick 222 is also equipped with a fuzzy logic controller for classifying the root mean square error (RMSE) into three levels: low error (<0.1m), medium error (0.1-0.3m), and high error (>0.3m), and assigning corresponding adjustment coefficients of K value, which are 0.9, 1.0, and 1.2, respectively.

[0179] In conclusion, 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 dual-mode dynamic collaborative control method based on a pull-and-hold line, characterized in that, Includes the following steps: Displacement electrical signals are acquired by dual potentiometers configured on the rocker arm pulley block of the ground equipment; The dynamic elastic modulus is generated by using a PWM signal generator based on the elastic coefficient model of the tether wire and the real-time collected ambient temperature data of the tether wire. The dynamic elastic modulus is then generated by combining the displacement electrical signal to generate a Z-axis PWM signal that characterizes the pulling force. By using the angle difference between the XY joystick on the drone and the tether line, an XY-axis PWM signal representing the horizontal position is generated based on the angle difference-coordinate mapping model and dynamic filtering. The Z-axis PWM signal is split into two paths. One path is transmitted to the winch ESC to control the winding and unwinding speed, and the other path is transmitted to the UAV and merged with the XY-axis PWM signal to form a three-dimensional control signal. Set the remote control signal and the winch potentiometer signal as the first linkage group, and set the three-dimensional control signal and the winch manual signal as the second linkage group. When the first linkage group is activated and the second linkage group is deactivated by the first and second remote control relays, the drone enters the first control mode; when the second linkage group is activated and the first linkage group is deactivated, the drone enters the second control mode. When entering the first control mode, the drone receives the remote control signal, and at the same time the winch ESC receives the winch potentiometer signal to perform remote control control of the drone in the XYZ three-axis direction, and to automatically rotate the winch motor forward and reverse to reel in and unload the cable by superimposing a tension fluctuation suppression factor into the Z-axis PWM signal. When entering the second control mode, the UAV receives three-dimensional control signals, and the winch ESC receives manual signals from the winch. When the winch is manually retracting the line, the PWM signal generator outputs a Z-axis PWM signal representing descent. When the winch is manually releasing the line, the PWM signal generator outputs a Z-axis PWM signal representing ascent. Combined with the XY-axis PWM signals, a new three-dimensional control signal is generated, and the UAV's XYZ three-axis wire-controlled following is performed based on nonlinear decoupling. When generating XY axis PWM signals, the following is included: The drone uses a binocular vision sensor to capture the projection image of the tether line on a horizontal plane and extracts the pixel coordinates of the tether line from the projection image. The analog signal of the joystick deflection angle is obtained by the mechanical displacement sensor of the XY joystick and converted into angle difference data; The pixel coordinates and angle difference data are corrected based on the ambient light intensity and the oscillation frequency, respectively, and then input into the angle difference-coordinate mapping model built based on the BP neural network to output the predicted XY coordinates. Real-time error compensation is performed on the predicted XY coordinates using the Kalman filter algorithm to generate XY axis PWM signals. The state covariance matrix P of the filtering process is dynamically adjusted according to the real-time error.

2. The method according to claim 1, characterized in that, When correcting pixel coordinates based on ambient light intensity, the following is included: Multiple ambient light intensity samples were continuously collected and divided into three intervals: low light, medium light, and high light based on the intensity. Logarithmic enhancement algorithm was used to improve pixel contrast in the low light interval, and dynamic range compression was used to suppress overexposed areas in the high light interval, resulting in the processed ambient light intensity samples. The sample variance is calculated for the processed ambient light intensity samples. When the variance is less than or equal to the set value, the mean is directly taken as the current light intensity. When the variance is greater than the set value, median filtering is started to remove outliers and the mean is taken as the current light intensity. The preset interval compensation coefficient is queried based on the current light intensity, and the pixel coordinates are non-linearly mapped and corrected by combining the logarithmic operation results of the current light intensity and the standard reference light intensity.

3. The method according to claim 1, characterized in that, When correcting the included angle difference data based on the oscillation frequency, the following is included: Obtain the natural frequency curves of the tether under different tensions and establish a swing frequency-stiffness mapping model; The current swing frequency of the mooring line is collected in real time. The current equivalent stiffness coefficient is calculated through the swing frequency-stiffness mapping model and compared with the standard stiffness coefficient to generate the stiffness correction coefficient λ. The angle difference data is multiplied by the stiffness correction factor λ, and the swing phase compensation term is introduced for correction. A variable step size LMS adaptive filter is used to denoise the corrected angle difference data, and the step size factor μ is dynamically adjusted according to the swing frequency.

4. The method according to claim 1, characterized in that, When outputting the predicted XY coordinates, the following should be included: The input layer assigns dynamic weights based on interval compensation coefficients to the corrected pixel coordinates and dynamic weights based on stiffness correction coefficient λ to the corrected angle difference data. The weight allocation ratio is optimized in real time through gradient descent algorithm. The pixel coordinate features are processed by the Leaky ReLU activation function from the input layer to the first hidden layer, and the angle difference features are processed by the ELU activation function from the second hidden layer to the output layer. A Batch Normalization layer is added to the output layer to suppress gradient vanishing. This includes: the construction of an adversarial training sample set based on Gaussian noise, random scaling, and generative adversarial networks; and the online iterative optimization of the model based on coordinate error calculation, Bayesian optimization algorithm updates, and dynamic adjustment of the learning rate.

5. The method according to claim 1, characterized in that, Real-time error compensation includes: When the root mean square error between the measured XY coordinates of the UAV's GPS and the filtered output XY coordinates exceeds a set value for multiple consecutive sampling periods, adaptive optimization of the Kalman gain K is initiated. The root mean square error is divided into three levels—low error, medium error, and high error—by a fuzzy logic controller, and an adjustment coefficient corresponding to the K value is assigned. When the error level jumps from low or medium error to high error, a strong tracking factor t is triggered to perform a weighted correction on the state covariance matrix P of the Kalman filter.

6. The method according to claim 1, characterized in that, When superimposed with tension fluctuation suppression factors, the following are included: The high-frequency fluctuation component of the tether tension was collected and decomposed into three frequency bands through wavelet transform; Adaptive threshold noise reduction is performed on each frequency band component to retain the effective frequency band characterizing the elastic vibration of the cable; The noise-reduced fluctuation component is converted into a reverse compensation PWM signal and linearly superimposed with the Z-axis PWM signal; Among them, the amplitude of the reverse compensation PWM signal is positively correlated with the root mean square value of the tension fluctuation, and the phase lag is set to 90° to achieve active vibration reduction.

7. The method according to claim 1, characterized in that, When performing XYZ three-axis wired tracking of a drone, the following is included: The XY-axis PWM signal and the Z-axis PWM signal are nonlinearly decoupled by using the coupling coefficient matrix between the tether wire unfolding length and the real-time speed of the UAV. The coupling coefficient matrix is ​​dynamically updated based on the tether wire elastic coefficient model. An inverse model of UAV dynamics is introduced to predict the UAV position deviation based on the current three-dimensional control signal and generate a feedforward compensation amount that is superimposed on the control signal. The magnitude of the compensation amount is positively correlated with the first derivative of the tether tension fluctuation. When the tether line swing frequency exceeds the set value, the control weight of the XY axis is reduced and the attitude angle limiting protection is activated. At the same time, the Z axis control is switched to constant tension priority mode, and the tension fluctuation is controlled by the PID parameter self-tuning algorithm.

8. A dual-mode dynamic cooperative control system based on a pull-tether line, characterized in that, During runtime, the method of claim 1 is executed, comprising: Tethered ground equipment includes a rocker arm pulley block, tether line, tether box winch, and PWM signal generator; the rocker arm pulley block is equipped with dual potentiometers for acquiring displacement electrical signals; the PWM signal generator is used to generate a Z-axis PWM signal based on the displacement electrical signal, and it is divided into two paths, one of which is transmitted to the winch ESC and the other of which is transmitted to the UAV. The drone includes a flight control system, an XY joystick, and a signal merging module. The XY joystick generates XY-axis PWM signals based on the angle difference between itself and the tethering line. The signal merging module receives the Z-axis PWM signal sent by the tethered ground equipment and merges it with the XY-axis PWM signal into a three-dimensional control signal. Switching execution unit: includes first and second remote control relays and remote controller; both the first and second remote control relays are equipped with two sets of linkage contacts; the first remote control relay is used to control the control signals received by the flight control system, and the second remote control relay is used to control the control signals received by the winch ESC; the remote controller is used to send remote control signals to the UAV and to enable the UAV to enter the first or second control mode by controlling the on or off of the remote control relays. The tethered winch also includes: a winch potentiometer and a winch motor; When entering the first control mode, the first remote control relay controls the flight control system to receive remote control signals, while the second remote control relay controls the winch ESC to receive winch potentiometer signals. The UAV performs remote control control in the XYZ three-axis direction, and the winch motor performs automatic forward and reverse rotation for cable winding and unwinding. When entering the second control mode, the first remote control relay controls the flight control system to receive three-dimensional control signals, while the second remote control relay controls the winch ESC to receive manual winch signals. When the winch is manually retracting the line, the PWM signal generator outputs a Z-axis PWM signal representing descent. When the winch is manually releasing the line, the PWM signal generator outputs a Z-axis PWM signal representing ascent. Combined with the XY-axis PWM signals, a new three-dimensional control signal is generated, and the UAV performs wire-controlled following on the XYZ axes. The drone is also equipped with a binocular vision sensor to capture the projection image of the tether line on the horizontal plane and extract the pixel coordinates of the tether line from the projection image; The XY joystick is also equipped with a mechanical displacement sensor to collect analog signals of the joystick deflection angle and convert them into angle difference data; The drone is also equipped with a light sensor module to collect ambient light intensity in real time; A miniature triaxial accelerometer is installed at the bottom anchor point of the tether line to collect the tether line swing frequency.

Images