Cross-medium unmanned aerial vehicle and tracking control method thereof

Abstract

The cross-medium unmanned aerial vehicle and the tracking control method thereof provided in the application solve the position tracking error and identify the disturbance estimation value by synchronously collecting real-time state, expected trajectory and real-time height data of the unmanned aerial vehicle, build a three-dimensional ILOS guidance model relying on real-time height to distinguish the medium flight state of the unmanned aerial vehicle, fuse the flight state, position error and disturbance estimation to synchronously output complete three-dimensional kinematics reference instructions containing longitudinal velocity, vertical velocity, yaw angle and pitch angle, solve the air / water propulsion control data adapted to the medium, build three-dimensional guidance logic, synchronously compensate and correct the two types of flow field disturbances of lateral and vertical directions, eliminate the double trajectory deviation problems in the underwater, low-altitude water-adhesion and water-air transition working conditions, accelerate the tracking error convergence speed, greatly reduce the steady-state tracking deviation, still maintain high-precision trajectory tracking in the complex three-dimensional flow field environment composed of wind, wave and ocean current, and fully meet the use requirements of water-air three-dimensional high-precision cross-domain operation.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) technology, and in particular to cross-media UAVs and their tracking and control methods. Background Technology

[0002] Transmedia unmanned aerial vehicles (UAVs) can perform integrated multi-domain operations including aerial patrol, surface maneuvering, and underwater navigation, and are widely used in scenarios such as marine environmental monitoring, underwater target detection, and near-shore three-dimensional inspection. In actual operating environments, UAVs are continuously subjected to three-dimensional spatial coupling interference throughout their operation: during aerial flight, there is time-varying turbulent wind disturbance; during surface navigation and underwater navigation, ocean currents and waves bring lateral and vertical complex water flow disturbances, which simultaneously cause the UAV to drift laterally and deviate from its altitude.

[0003] Traditional two-dimensional planar line-of-sight guidance algorithms can only compensate for lateral disturbances in the plane and cannot correct for vertical water flow and airflow disturbances. Furthermore, they do not distinguish the relationship between the trajectory and the combined motion of the water flow. Under conditions such as underwater submersion, low-altitude water-skimming flight, and water-air cross-medium transition, the UAV will experience continuous vertical and lateral trajectory deviations relative to the preset path. The tracking error convergence speed is slow and the steady-state deviation is large. When facing complex three-dimensional flow fields, the trajectory tracking accuracy drops significantly, making it difficult to meet the requirements of high-precision cross-domain operations. Summary of the Invention

[0004] The technical problem this application aims to solve is that traditional two-dimensional planar line-of-sight guidance algorithms can only compensate for lateral disturbances in the planar plane, and cannot correct for vertical water flow and airflow disturbances. Furthermore, they do not distinguish the relationship between the trajectory and the combined motion of the water flow. Under conditions such as underwater submersion, low-altitude water-skimming flight, and water-air cross-medium transition, the UAV will experience continuous vertical and lateral trajectory deviations relative to the preset path. The tracking error convergence speed is slow, the steady-state deviation is large, and the trajectory tracking accuracy drops significantly when facing complex three-dimensional flow fields, making it difficult to meet the requirements of high-precision cross-domain operations.

[0005] To address the aforementioned issues, this application provides a cross-media unmanned aerial vehicle and its tracking and control method.

[0006] In a first aspect, the present invention discloses a cross-media unmanned aerial vehicle (UAV) tracking and control method, which includes, Acquire real-time status data, desired trajectory data, and real-time altitude data of the UAV. Calculate position tracking error and disturbance estimate using real-time status data and desired trajectory data. Real-time UAV status data includes position data and attitude data. The flight status of the UAV is determined based on real-time altitude data, a three-dimensional ILOS guidance model is constructed, and the kinematic reference commands are calculated by inputting the UAV flight status, position tracking error and disturbance estimate. The kinematic reference commands include the desired longitudinal velocity, desired vertical velocity, desired yaw angle and desired pitch angle. By combining kinematic reference commands, real-time UAV status data, and disturbance estimates, UAV control data is calculated. The UAV drives its aerial propulsion mechanism or underwater navigation mechanism based on the UAV control data, which includes aerial propulsion data and underwater propulsion data.

[0007] Preferably, the real-time status data, desired trajectory data, and real-time altitude data of the UAV are acquired, and the position tracking error and disturbance estimate are calculated using the real-time status data and desired trajectory data. Specifically, this includes the following steps: Obtain desired trajectory data from the host system, collect UAV flight attitude data, position data, and altitude data in real time, and obtain real-time status data and real-time altitude data; The desired trajectory data undergoes coordinate transformation, converting the position and height data to the same coordinate system as the desired trajectory data. The three-dimensional position tracking error of the UAV is calculated by combining real-time status data and expected trajectory data; A Gaussian RBF neural network observer model is constructed. The attitude data and position data are input, and the combined disturbance force and disturbance moment received by the UAV are calculated to obtain the disturbance estimate.

[0008] Preferably, a Gaussian RBF neural network observer model is constructed, and attitude data and position data are input to calculate the combined disturbance force and disturbance moment received by the UAV, thereby obtaining the disturbance estimate. This specifically includes the following steps: Construct a Gaussian RBF neural network observer model, input real-time UAV status data, and update the weight matrix in the model in real time based on the gradient descent adaptive law; By comparing the standard dynamic model with real-time state data, the composite disturbances formed by the superposition of wind, waves, ocean currents, model mismatch, and actuator deviations are identified online through the dynamic residual equation, and the composite disturbance force and disturbance torque are output.

[0009] Preferably, the UAV's flight status is determined based on real-time altitude data, and kinematic reference commands are calculated by combining the UAV's flight status, position tracking error, and disturbance estimation values. The kinematic reference commands include desired longitudinal velocity, desired vertical velocity, desired yaw angle, and desired pitch angle, specifically including the following steps: Based on the comparison between real-time altitude data and preset altitude threshold, the medium in which the drone is located is determined, the medium state is classified, and the drone flight status is output. Construct a trajectory-water flow composite coordinate system, map the tracking position error to the composite coordinate system, and obtain the tracking position error after coordinate transformation; The composite pitch angle and composite yaw angle are calculated based on the disturbance estimate to adapt to the influence of water flow. An ILOS guidance model is constructed. The tracking position error after coordinate transformation, the composite pitch angle, the composite yaw angle, and the preset forward look distance are substituted into the ILOS guidance model to calculate the desired longitudinal velocity, desired vertical velocity, desired yaw angle, and desired pitch angle, thereby obtaining the kinematic reference command.

[0010] Preferably, the UAV control data is calculated by combining kinematic reference commands, real-time UAV status data, and disturbance estimates. The UAV then drives the aerial propulsion mechanism or underwater navigation mechanism based on the UAV control data, specifically including the following steps: Calculate the current real-time three-axis speed of the drone based on real-time drone status data; The flight mode is distinguished based on the drone's flight status, including aerial mode and underwater mode; The tracking error is calculated based on kinematic reference commands, real-time UAV status data, and real-time three-axis velocity of the UAV in various modes. Specifically, the air mode calculates the speed tracking error, and the underwater mode calculates the attitude angle tracking error. The disturbance estimate is fed forward to compensate the sliding mode control law. The super-helical sliding mode algorithm is used to solve the total resultant force and total resultant torque of the body required to offset the tracking error, and output virtual control quantities, including three-axis virtual resultant force and three-axis virtual resultant torque. The corresponding control efficiency matrix is ​​retrieved based on the flight mode. The matrix is ​​inverted to establish the mapping relationship between the virtual control quantity and the actuator thrust. The required output thrust of each air rotor or underwater thruster is calculated. Finally, the rotor speed command or underwater thrust speed command is output and sent to the UAV.

[0011] Secondly, the present invention discloses a cross-medium unmanned aerial vehicle (UAV) applicable to the above-mentioned cross-medium UAV tracking and control method, which includes an air propulsion mechanism, an underwater navigation mechanism, a fuselage, and a controller. The controller is installed in the fuselage, and the fuselage is connected to the air propulsion mechanism and the underwater navigation mechanism respectively. The controller is electrically connected to the air propulsion mechanism and the underwater navigation mechanism respectively.

[0012] Preferably, the aerial propulsion mechanism includes a support rod, a rotor, and a drive motor. One end of the support rod is connected to the fuselage, and the drive motor and rotor are mounted on the other end of the support rod and connected to the rotor.

[0013] Preferably, the underwater navigation mechanism includes a first thruster and a second thruster, with the first thruster mounted on the fuselage and the second thruster mounted at the tail of the fuselage.

[0014] Preferably, the body includes a buoyancy shell, a wave-breaking block, and a flow-guiding wedge, with the buoyancy shell connecting the wave-breaking block and the flow-guiding wedge respectively.

[0015] Preferably, it includes a quick-release mechanism and an energy storage component. The quick-release mechanism is connected to the body, the energy storage component is connected to the quick-release mechanism, and the energy storage component is electrically connected to the controller.

[0016] The technical solution provided in this application has the following advantages compared with the prior art: The cross-medium UAV and its tracking control method provided in this application mention that, by simultaneously collecting real-time UAV status, desired trajectory, and real-time altitude data, the position tracking error is first calculated and the disturbance estimate is identified. Based on the real-time altitude, the UAV's medium flight state is determined and a three-dimensional ILOS guidance model is built. The flight state, position error, and disturbance estimate are integrated to synchronously output a complete three-dimensional kinematic reference command including longitudinal velocity, vertical velocity, yaw angle, and pitch angle. Then, the reference command, the real-time airframe status, and the disturbance estimate are combined to solve for the air / underwater propulsion control data adapted to the medium. This scheme constructs a three-dimensional guidance logic that couples trajectory and flow field, and simultaneously compensates and corrects for lateral and vertical flow field disturbances. It eliminates the problem of dual trajectory deviation in underwater, low-altitude water-skimming, and water-air transition conditions, accelerates the convergence speed of tracking error, and significantly reduces steady-state tracking deviation. It can still maintain high-precision trajectory tracking in complex three-dimensional flow field environments composed of wind, waves, and ocean currents, fully meeting the requirements for high-precision cross-domain operations in water and air.

[0017] The cross-medium UAV is described as having independent aerial propulsion and underwater navigation mechanisms. The fuselage houses controllers electrically linked to both propulsion mechanisms. The hardware architecture can receive aerial and underwater propulsion data from the control system and drive the corresponding actuators according to the media. The entire hardware can match multi-media operating modes based on real-time altitude division, fully accepting three-dimensional motion control commands output by 3D ILOS guidance. This provides the hardware execution foundation for three-dimensional flow field disturbance compensation and lateral and vertical synchronous trajectory correction, ensuring that the three-dimensional kinematic commands and media-specific propulsion control commands output by the control system can be executed. This enables the UAV to stably respond to high-precision trajectory tracking control commands in all operating conditions, including underwater submersion, low-altitude water-skimming flight, and water-air cross-media transitions. From a hardware perspective, it supports cross-domain operation capabilities with no significant deviation and low steady-state error in complex three-dimensional flow fields. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0019] 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, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1A flowchart illustrating a cross-media UAV tracking and control method provided in this application; Figure 2 A flowchart illustrating step S1 of a cross-media UAV tracking and control method provided in this application; Figure 3 A flowchart illustrating step S2 of a cross-media UAV tracking and control method provided in this application; Figure 4 A flowchart illustrating step S3 of a cross-media UAV tracking and control method provided in this application; Figure 5 This application provides a schematic diagram of the motion state and force generated by the actuator of a cross-medium unmanned aerial vehicle. Figure 6 A schematic diagram of a three-dimensional improved line-of-sight guidance law for a cross-medium unmanned aerial vehicle provided in this application; Figure 7 A block diagram of a three-dimensional ILOS-STC control architecture based on an RBFNN observer for a cross-media unmanned aerial vehicle (UAV) is provided in this application. Figure 8 A structural schematic diagram of a cross-media unmanned aerial vehicle provided in this application Figure 1 ; Figure 9 A structural schematic diagram of a cross-media unmanned aerial vehicle provided in this application Figure 2 ; Figure 10 This application provides a schematic diagram of the connection structure between a quick-release mechanism and an energy storage device for a cross-medium unmanned aerial vehicle.

[0021] Explanation of reference numerals in the attached figures: 1. Cross-media unmanned aerial vehicles (UAVs); 11. Aerial propulsion mechanism; 111. Support rod; 112. Rotor; 113. Drive motor; 12. Underwater navigation mechanism; 121. First thruster; 122. Second thruster; 13. Body; 131. Buoyancy shell; 132. Wave-breaking block; 133. Flow-guiding wedge; 14. Controller; 15. Quick-release mechanism; 16. Energy storage components; 17. Positioning and navigation module; 18. Communication module. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] Firstly, see Figures 1-4 This invention discloses a cross-media unmanned aerial vehicle (UAV) tracking and control method, which includes: Step S1: Acquire the real-time status data, expected trajectory data, and real-time altitude data of the UAV. Calculate the position tracking error and disturbance estimate using the real-time status data and expected trajectory data. The real-time status data of the UAV includes position data and attitude data. Step S2: Determine the UAV's flight status based on real-time altitude data, construct a three-dimensional ILOS guidance model, input the UAV's flight status, position tracking error, and disturbance estimation value to calculate kinematic reference commands. The kinematic reference commands include the desired longitudinal velocity, desired vertical velocity, desired yaw angle, and desired pitch angle. Step S3: Combine kinematic reference commands, real-time UAV status data, and disturbance estimates to calculate UAV control data. The UAV drives the aerial propulsion mechanism or underwater navigation mechanism based on the UAV control data. The UAV control data includes aerial propulsion data and underwater propulsion data.

[0024] Specifically, in step S1, real-time UAV status data including position and attitude data, expected trajectory data issued by the mission plan, and real-time altitude data output by airborne sensors are collected. The position tracking error is obtained by performing difference calculation between the real-time status data and the expected trajectory data. At the same time, relying on the RBFNN observer combined with the real-time status data of the aircraft, the combined disturbances caused by external interference such as wind, waves, and ocean currents and model mismatch are identified online, and the disturbance estimate is output. The purpose of this step is to complete the basic data collection of the whole scene, the tracking error calculation, and the real-time estimation of unknown disturbances, providing all the original input data for the subsequent three-dimensional guidance model and controller, realizing the early identification of three-dimensional flow field interference, and supporting lateral and vertical dual disturbance compensation from the data source.

[0025] Specifically, in step S2, the acquired real-time altitude data is substituted into the segmented judgment logic to distinguish between three types of UAV flight states: fully airborne, partially submerged in water, and fully underwater. A three-dimensional ILOS guidance model adapted to the three-dimensional flow field is built. The determined UAV flight state, the position tracking error output in step S1, and the disturbance estimation value are input into the guidance model for calculation to obtain the kinematic reference command including the desired longitudinal velocity, desired vertical velocity, desired yaw angle, and desired pitch angle. The purpose of this step is to abandon the limitations of traditional two-dimensional guidance, construct a three-dimensional solution logic coupled with trajectory and water flow, simultaneously compensate for lateral and vertical water flow and airflow disturbances, eliminate the problem of dual trajectory offset under cross-medium conditions, output high-precision three-dimensional motion target command, accelerate the convergence of tracking error, and reduce steady-state deviation.

[0026] Specifically, in step S3, the output kinematic reference commands, the real-time UAV status data collected in step S1, and the disturbance estimation values ​​are input into the controller for calculation. Based on the UAV flight state determined in step S2, the aerial propulsion data and underwater propulsion data adapted to different media are solved as UAV control data. The controller sends out the corresponding control data to drive the aerial propulsion mechanism or underwater navigation mechanism to complete the motion response. The function of this step is to convert the three-dimensional motion reference commands into hardware-executable propulsion drive commands, match the output of the corresponding power system according to the medium state, realize the coordinated switching and precise drive of the air and water dual power systems, implement the three-dimensional anti-disturbance tracking control effect, and ensure that the UAV operates stably and with high precision along the preset trajectory under all working conditions of underwater, low altitude, and air-water transition.

[0027] It is understandable that by synchronously collecting real-time UAV status, desired trajectory, and real-time altitude data, the position tracking error is first calculated and the disturbance estimate is identified. Based on the real-time altitude, the UAV's medium flight state is determined and a three-dimensional ILOS guidance model is built. The flight state, position error, and disturbance estimate are integrated to synchronously output a complete three-dimensional kinematic reference command including longitudinal velocity, vertical velocity, yaw angle, and pitch angle. Then, the reference command, the real-time airframe status, and the disturbance estimate are combined to solve for the air / underwater propulsion control data adapted to the medium, constructing a three-dimensional guidance logic that couples trajectory and flow field. Simultaneously, compensation and correction are achieved for lateral and vertical flow field disturbances, eliminating the dual trajectory deviation problem under underwater, low-altitude water-skimming, and water-air transition conditions. This accelerates the convergence speed of tracking error, significantly reduces steady-state tracking deviation, and can maintain high-precision trajectory tracking even in complex three-dimensional flow field environments composed of wind, waves, and ocean currents, fully meeting the requirements for high-precision cross-domain operations in water and air.

[0028] Step S1 specifically includes the following steps: Step S11: Obtain the desired trajectory data from the host system, and collect the UAV flight attitude data, position data, and altitude data in real time to obtain real-time status data and real-time altitude data; Step S12: Perform coordinate transformation on the desired trajectory data, converting the position and height data to the same coordinate system as the desired trajectory data; Step S13: Calculate the UAV's three-dimensional position tracking error by combining real-time status data and expected trajectory data; Step S14: Construct a Gaussian RBF neural network observer model, input attitude data and position data, calculate the composite disturbance force and disturbance torque received by the UAV, and obtain the disturbance estimate.

[0029] Specifically, firstly, the system receives the desired trajectory data from the host system. Then, through the onboard positioning and navigation module and inertial sensing unit, it collects the UAV's position, attitude, and altitude data in real time. The position and attitude data are integrated to form real-time status data, while the altitude information is extracted separately as real-time altitude data. This completes the acquisition of the mission target path information and the UAV's own real-time motion and altitude information, providing all the original measured data for subsequent coordinate unification, error calculation, and disturbance observation. A coordinate transformation matrix is ​​used to perform coordinate system transformation on the desired trajectory data. Simultaneously, the same transformation matrix is ​​used to map the UAV's measured position and altitude data to the unified coordinate system of the desired trajectory, achieving alignment between the carrier's state and the target path coordinate reference. The general formula for coordinate transformation is: In the formula Coordinates in a globally unified coordinate system For coordinate rotation matrix, The original measurement coordinates of the machine body. To compensate for coordinate translation, this eliminates calculation biases caused by inconsistencies in coordinate systems, ensuring consistent data benchmarks for position error calculation and guidance calculation, and preventing additional trajectory deviations caused by benchmark misalignment. Under a unified coordinate system, the 3D position tracking error is calculated by subtracting the desired trajectory's 3D coordinates from the UAV's real-time 3D position coordinates; the 3D position tracking error calculation formula is as follows: In the formula , , The three-dimensional coordinates of the desired trajectory, , , For the real-time position coordinates of the drone, , , These are the longitudinal, lateral, and vertical position tracking errors, respectively. This step quantifies the three-dimensional offset of the UAV relative to the preset path, intuitively reflecting the lateral and vertical trajectory deviations, and providing core error input for the three-dimensional ILOS guidance model.

[0030] A Gaussian RBF neural network observer model is constructed, using attitude and position data in a unified coordinate system as network input. Through Gaussian function mapping and adaptive weight iteration, external disturbances and model mismatch are identified online, outputting a perturbation estimate consisting of the composite perturbation force and perturbation moment. The output formula of the hidden layer neurons in the RBF neural network is as follows: The formula for calculating disturbance estimation is as follows: Adaptive weight update formula: In the above formula, The input state vector is (position, orientation). Let i be the center point of the i-th neuron. For the width of the Gaussian plane, Here, W represents the estimated composite perturbation value, and W is the network weight matrix. The learning rate matrix is ​​positive definite. To mitigate velocity observation errors, the system estimates the three-dimensional composite disturbances caused by wind, waves, ocean currents, and model uncertainties in real time, and outputs disturbance estimates for feedforward compensation in guidance and control systems, thereby reducing trajectory tracking deviations caused by the three-dimensional flow field at their source.

[0031] Step S14 specifically includes the following steps: Step S141: Construct a Gaussian RBF neural network observer model, input real-time UAV state data, and update the weight matrix in the model in real time based on the gradient descent adaptive law; Step S142: Compare the standard dynamic model with the real-time state data, identify the composite disturbance formed by the superposition of wind, waves, ocean currents, model mismatch, and actuator deviation through the dynamic residual equation online, and output the composite disturbance force and disturbance torque.

[0032] Specifically, a Gaussian RBF neural network observer model is built, using real-time UAV state data including position and attitude as the network input vector. The hidden layer of the network uses the Gaussian function to complete the state mapping. The Gaussian function formula is as follows: ,in The input state vector is (position, orientation). Let i be the center point of the i-th neuron. The network weight matrix is ​​updated in real time based on the adaptive gradient descent law, with a Gaussian width. Weight update formula In the formula The learning rate matrix is ​​positive definite. To mitigate velocity observation errors, the network fitting accuracy is dynamically corrected throughout the process, enabling the neural network to adaptively adapt to the time-varying motion state of the UAV and continuously optimize its fitting capabilities. This provides a precise network fitting foundation for subsequent complex disturbance identification, ensuring that disturbance observation possesses adaptive learning capabilities. The current real-time state data is substituted into the UAV's standard nominal dynamic model to calculate the model's theoretical output value. The difference between the theoretical dynamic output and the actual motion state of the aircraft is then used to construct the dynamic residual equation. , For dynamic residuals, For actual acceleration, The standard model's theoretical acceleration is combined with the previously trained RBF network's output perturbation estimation. The system performs online calculations of the combined disturbances caused by wind, waves, ocean currents, model parameter mismatch, and actuator output deviations, ultimately outputting an estimated disturbance value consisting of the combined disturbance force and the disturbance moment. The purpose of this step is to mine all unknown composite disturbances in the dynamic residual system, quantify the magnitude of the comprehensive disturbance in three-dimensional space in real time, and output the disturbance estimate for subsequent three-dimensional ILOS guidance and sliding mode controller feedforward compensation, so as to offset the lateral and vertical tracking offsets caused by the three-dimensional flow field and reduce the steady-state tracking error.

[0033] Step S2 specifically includes the following steps: Step S21: Based on the comparison between real-time altitude data and preset altitude threshold, determine the medium in which the UAV is located, classify the medium state, and output the UAV flight status; Step S22: Construct a trajectory-water flow composite coordinate system, map the tracking position error to the composite coordinate system, and obtain the tracking position error after coordinate transformation; Step S23: Calculate the composite pitch angle and composite yaw angle to adapt to the influence of water flow based on the disturbance estimate; Step S24: Construct the ILOS guidance model, substitute the tracking position error after coordinate transformation, composite pitch angle, composite yaw angle, and preset forward look distance into the ILOS guidance model, calculate the desired longitudinal velocity, desired vertical velocity, desired yaw angle, and desired pitch angle, and obtain the kinematic reference command.

[0034] Specifically, the real-time altitude data is read and compared with two preset altitude thresholds (underwater and aerial) within the program. The results are then substituted into the segmentation judgment function. , It is a cross-medium unmanned aerial vehicle that is completely in an underwater medium. This is the transitional state of a cross-medium UAV partially submerged in water. This refers to a cross-medium UAV operating entirely in an aerial state, where h is the real-time altitude, and h1 and h2 are the preset underwater and aerial altitude thresholds, respectively. To identify the medium state, i.e., the UAV's flight state, by distinguishing different operating conditions of the UAV in water and air, corresponding medium dynamic logic is assigned for subsequent guidance and control, avoiding model mismatch errors caused by single-medium algorithms. A trajectory-flow composite coordinate system is established, and a coordinate transformation matrix is ​​used to unify the original three-dimensional position tracking error under a unified coordinate system. Mapping to a composite coordinate system, coordinate transformation formula: In the formula, The rotation transformation matrix is ​​the trajectory-flow composite coordinate system. To address the tracking position error after coordinate transformation, flow field disturbances are coupled into the error calculation benchmark. This cancels out lateral and vertical foundational disturbances caused by ocean currents and airflows at the coordinate system level, overcoming the limitation of traditional two-dimensional guidance systems that can only handle planar errors. The output disturbance estimate (flow field velocity estimate) is imported and combined with the carrier's motion velocity to solve for the trajectory-current composite pitch angle and composite yaw angle. The calculation formula is as follows: In the formula, For compound pitch angle, For a compound yaw angle, , The lateral and vertical water flow velocities are extracted from the disturbance estimates, and U is the composite forward velocity of the carrier. , These are the lateral and vertical error components. This step incorporates water flow interference into the line-of-sight angle calculation, obtaining a target line-of-sight angle adapted to the three-dimensional flow field and eliminating angular offset caused by water flow. A three-dimensional ILOS improved line-of-sight guidance model is constructed. The transformed tracking error enc, composite pitch angle θc, composite yaw angle ψc, preset lateral forward-looking distance Δy, and longitudinal forward-looking distance Δz are substituted into the guidance formula to solve for the four-dimensional kinematic reference command. In the formula, For the desired longitudinal velocity, For the desired vertical velocity, For the desired yaw angle, For the desired pitch angle, , , To adjust the convergence gain parameter, a three-dimensional kinematic reference command is output, simultaneously suppressing lateral and vertical trajectory offsets, accelerating the convergence speed of tracking errors, reducing steady-state deviation, and adapting to the high-precision trajectory tracking requirements under all working conditions, including underwater, water-skimming flight, and cross-medium transitions. Additionally, in the formula, This refers to the longitudinal tracking error in the composite coordinate system. The lateral tracking error in the composite coordinate system; The three components are vertical tracking errors in the composite coordinate system, and are all error components obtained after mapping the original three-dimensional position error through the trajectory-flow coordinate transformation matrix.

[0035] Step S3 specifically includes the following steps: Step S31: Calculate the current real-time three-axis speed of the drone based on the real-time drone status data; Step S32: Differentiate the flight mode based on the drone's flight status, including aerial mode and underwater mode; Step S33: Calculate the tracking error based on the kinematic reference command, real-time UAV status data, and real-time three-axis velocity of the UAV in each mode. Specifically, calculate the speed tracking error in the aerial mode and the attitude angle tracking error in the underwater mode. Step S34: Feedforward the disturbance estimate to the sliding mode control law, use the super-helical sliding mode algorithm to solve the total resultant force and total resultant torque of the body required to offset the tracking error, and output the virtual control quantity, including the three-axis virtual resultant force and the three-axis virtual resultant torque; Step S35: Retrieve the corresponding control efficiency matrix according to the flight mode, invert the matrix to establish the mapping relationship between virtual control quantity and actuator thrust, calculate the required output thrust of each air rotor or underwater thruster, and finally output the rotor speed command or underwater thruster thrust speed command to the UAV.

[0036] Specifically, the real-time position and attitude data of the collected UAV are extracted, and the real-time three-axis velocities in the body coordinate system are calculated using differential operations. The velocity calculation formula is as follows: In the formula This is the real-time 3D position vector of the UAV. The system obtains the real-time three-axis velocity of the aircraft, acquiring its real-time motion rate. This serves as the foundational input for calculating tracking errors under different media modes, enabling a quantitative comparison between actual motion and target commands. It reads the output media status identifier and classifies the aircraft into two operating modes: aerial mode when the media identifier indicates the aircraft is completely out of the water, and underwater mode when the aircraft is submerged / partially submerged. This completes the control channel partition selection, distinguishing between the two independent dynamic control logics for the aerial quadcopter and the underwater six-thruster system. This provides a basis for mode determination for differentiated error calculation and control allocation. The system calculates the corresponding tracking error for each mode. In aerial mode, the desired longitudinal and vertical velocities are used as references, combined with the real-time three-axis velocity to solve for the velocity tracking error. The formula for calculating the speed tracking error is: In underwater mode, the desired pitch and yaw angles are used as references, and the attitude angle tracking error is solved by combining the real-time attitude angles of the aircraft. The formula for calculating attitude angle tracking error is: In the formula For the desired velocity vector, For real-time three-axis speed, Let be the desired attitude angle vector. For the real-time attitude angle vector, differentiated error indices are constructed to address the differences in air and hydrodynamic characteristics, accurately quantifying the deviation between the body motion and the target command under different media. The composite disturbance estimate d^ is fed forward and superimposed onto the super-helical sliding mode control law. Substituting the tracking error corresponding to each mode, the total resultant force and total resultant torque of the body used to eliminate tracking error are solved through the super-helical sliding mode algorithm. The core formula for solving the virtual control quantity is as follows: In the formula For sliding surface functions, , For sliding mode adjustable gain matrix, The virtual control variables consist of three-axis virtual resultant force and three-axis virtual resultant torque. They rely on disturbance feedforward compensation to offset the combined interference caused by wind, waves, ocean currents, and model mismatch, and utilize super-spiral sliding mode to suppress chattering. The output is a virtual force / torque command that can uniformly represent the body's motion requirements, thus improving the system's robustness against disturbances. The sliding mode surface function is used to mitigate velocity tracking error. With attitude angle tracking error Calculations show that the sliding surface for the aerial mode adopts the velocity error sliding surface. The underwater sliding surface adopts the attitude angle error sliding surface. ; , The gain matrix is ​​a sliding mode adjustable matrix, calibrated through simulation iterations. This is the error proportional gain, used for rapid convergence to track deviations; The integral gain is used to eliminate steady-state residual deviations; its values ​​are all greater than 0 and can be fine-tuned on-site according to sea conditions and aircraft size. The preset control efficiency matrix M is retrieved based on the determined flight mode, and the allocation matrix is ​​obtained by inverting the matrix. Establish the mapping relationship between virtual control quantities and the thrust of each actuator. In the formula, F represents the thrust vector of all rotors / underwater thrusters, combined with the thrust-speed conversion formula for thrusters. Calculate the target rotational speed of each actuator, generate air rotor speed command or underwater thruster speed command, and send it to the UAV drive unit. In the formula, For thrust coefficient, Given fluid density, n (rotation speed), and D (blade diameter), the unified virtual resultant torque is decomposed into rotation speed control commands that can be executed by each actuator, adapting to both air and water power layouts, and enabling precise execution of control commands.

[0037] The specific operation of the cross-media UAV tracking and control method is as follows: Currently, tracking control methods introduce improved line-of-sight (ILOS) guidance laws in the three-dimensional spatial domain, requiring the establishment of a multi-parameter collaborative guidance mechanism that integrates velocity, yaw angle, and pitch angle. Specifically, a three-dimensional disturbance observer is constructed using RBFNN to achieve dynamic estimation of composite disturbances; subsequently, an improved three-dimensional ILOS guidance algorithm without sideslip compensation is developed; and a super-helical sliding mode controller based on the RBFNN observer is designed to enhance the system's robustness to model uncertainties. For cross-medium motion characteristics, a water-air propulsion system switching rule based on altitude information is constructed to improve the robustness of cross-medium UAVs in continuous cross-medium processes.

[0038] Based on relative velocity, a control system resistant to three-dimensional spatial disturbances is designed to achieve the desired trajectories of a cross-medium UAV in underwater, aerial, and water-air cross-medium environments, under the influence of thruster, model uncertainties, complex spatial wind, wave, ocean current interference, and unknown external disturbances. Robust tracking. Specifically, this goal can be broken down into three sub-problems: (1) Kinematics score system: given the desired trajectory , and By designing a guidance law resistant to space interference, a relative desired longitudinal velocity can be generated. Desired vertical velocity Desired yaw angle and desired pitch angle Position tracking error , and It converges to a small neighborhood near 0.

[0039] (2) Dynamic subsystem: given the desired relative velocity , and the perspective of expectations , and By designing robust control laws that resist interference, control force can be generated. The relative velocity tracking error , and angle tracking error , and It converges to a small neighborhood near 0.

[0040] (3) Water-air cross-medium system: In order to verify the adaptability of the designed controller to cross-medium trajectory tracking, it is necessary to establish a dynamic model of the cross-medium process and to design the fusion and switching of the underwater navigation control subsystem and the air flight control subsystem.

[0041] Among them, the physical operating conditions of cross-medium UAVs under different media, such as Figure 5 As shown: (1) Aerial state F1, F2, F3, F4: The four rotors in the air generate an upward thrust (Z-axis). xF1, xF2, xF3, xF4: The distances (in the X direction) of the four rotors from the center of gravity, and are assumed to be completely equal; yF1, yF2, yF3, yF4: The distances (in the Y direction) of the four rotors from the center of gravity, and are assumed to be completely equal; zF1, zF2, zF3, zF4: The distances (in the Z direction) of the four rotors from the center of gravity, and are assumed to be completely equal; Dynamic equations: ; In the formula, This indicates the thrust coefficient of the motor; This indicates the reverse torque coefficient of the motor; Indicates the motor's rotational speed; This indicates the total lift of the motor; This indicates the torque in the three axial directions.

[0042] (2) Underwater state T1, T2: Pointing up and down (Z-axis), arranged forward and backward (controlling pitch and buoyancy); T3, T4: Points up and down (Z-axis), arranged left and right (controls roll); T5, T6: Pointing towards the rear of the aircraft (negative X-axis), located behind the aircraft (controls forward and backward movement + yaw); l1, l2: X-direction distances (front and rear lever arms) from T1, T2 to the center of gravity. l3, l4: Y-direction distances (left and right lever arms) from T3, T4 to the center of gravity. l: The distance in the X direction from T5, T6 to the center of gravity (backwards, so X = L); l5, l6: Y-direction distances (left and right lever arms) from T5, T6 to the center of gravity.

[0043] ; The dynamic motion model of the UAV is constructed, and according to the medium, it is divided into aerial motion state and underwater motion state, such as... Figure 5 As shown, where, The total thrust along the X-axis of the body coordinate system, i.e., the resultant force in the forward and backward directions. This is the total thrust along the Y-axis of the body coordinate system, also known as the lateral resultant force.

[0044] (1) Aerial motion state When this cross-medium drone is in flight, it is propelled by four waterproof brushless motors in the body coordinate system. Down, thruster to The virtual control quantity can be expressed as: ; In the formula, This represents the force generated by each motor rotor; This represents the control efficiency matrix for four motors. The underwater control allocation matrix can be solved (inverted) based on the above formula. .

[0045] (2) Underwater motion state When this cross-medium UAV is navigating underwater, its underwater motion attitude is controlled by six underwater actuators within the body coordinate system. Down, thruster to The virtual control quantity can be expressed as: ; In the formula, Let be the control efficiency matrix of the thruster. The control allocation of the thruster can be solved (inverted) based on the above formula. The thrust of each propeller can be expressed as, ,in Thrust coefficient (dimensionless, determined by propeller type and advance ratio). Indicates fluid density, Indicates the thruster rotation speed. This indicates the diameter of the propeller.

[0046] Then, the improved line-of-sight guidance law, which is resistant to ocean current interference in space, transforms the tracking error into a trajectory-current composite coordinate system. This automatically compensates for spatial disturbances in water flow. The resulting lateral and vertical interference. For example... Figure 6 As shown, the trajectory-water flow combined pitch angle is defined. and yaw angle for: , ;Calculated using RBF observer The estimated value is as follows: , ;in, , The tracking error is transformed to a trajectory-flow composite coordinate system. middle: , At the same time, it is noted that , ; beg The time derivative is obtained. , , , In the formula, Let be the combined velocity of the desired trajectory and the water flow, where Desired flight speed for designing a three-dimensional ILOS guidance law. Desired yaw angle and desired pitch angle for: , , In the formula, —Adjustable parameters, and ; —Adjustable parameters, and Lateral forward viewing distance; —Adjustable parameters, and Longitudinal forward sight distance.

[0047] Finally, the design of the dynamic controller consists of two steps: first, designing an RBFNN observer to estimate disturbances; and second, designing the velocity tracking subsystem and the attitude tracking subsystem. The superspiral controller (RBF-STC) based on the RBFNN observer can improve the system's ability to estimate and compensate for unknown dynamic disturbances.

[0048] (1) RBFNN observer design To estimate and compensate for unknown dynamic disturbances, an RBF neural network observer is designed to estimate the dynamic disturbances experienced by the aircraft in real time, and the estimated disturbance force and torque are directly compensated in the subsequent controller design. Considering that the hidden layer contains... An adaptive RBF neural network observer with 10 neurons: ; In the formula, , The observed values ​​are relative velocities, and ; , This is an estimate of the dynamic disturbance, and ; , It is a positive definite matrix, and ; For the estimated weights of the neural network, ; The output of the hidden layer neurons, Specifically, for The One neuron, In the formula, ——No. The center point vector of each hidden layer neuron; —The width of the Gaussian function. Weights are adjusted using gradient descent. In the formula, The learning rate of the neural network. ,and It is a positive definite matrix.

[0049] (2) Design of speed and attitude tracking controller like Figure 7 As shown, the desired trajectories of a cross-medium UAV for underwater, aerial, and water-air cross-medium environments are defined. The system transforms the absolute path coordinates into real-time deviation information of the current cross-medium UAV relative to the path through trajectory coordinate system transformation. Based on the positional deviation, the desired line-of-sight angle is calculated. The RBFNN is used to estimate unknown disturbances (wind, waves, ocean currents, etc.) in real time and feeds them back to the kinematic control law for compensation. The desired line-of-sight angle is then converted into the desired velocity / angular velocity command. The kinematic part is responsible for the conversion from the reference path to the desired velocity command, while the dynamic part is responsible for ultimately implementing the velocity command into actuator thrust.

[0050] Specifically: After coordinate system transformation, the trajectory coordinates are input into the improved line-of-sight guidance law to calculate the desired line-of-sight angle, which is then sent to the kinematic control law to generate the desired velocity / angular velocity command, completing the kinematic solution. Subsequently, the command is split to the velocity tracking controller (airborne mode) and attitude tracking controller (underwater mode), and after processing, the corresponding control distribution module maps the force and torque commands to the airborne actuators (F1-F4) and the underwater actuators (T1-T6), thereby driving the aircraft / vehicle's movement. At the same time, the RBFNN state observer uses the actual system state to estimate unknown disturbances such as wind, waves, and ocean currents in real time, and feeds them back to the kinematic control law for compensation, forming a complete closed-loop control.

[0051] The criteria for determining whether a cross-medium UAV is underwater or in the air are handled in the following way: set up sea ​​level ( To determine the sea level height under real-time sea state conditions that vary over time, the following piecewise function is designed: ; From the above formula, we know that when The cross-medium drone is completely submerged underwater, experiencing full buoyancy and hydrodynamic forces. When Transmedia drones fly entirely in the air, unaffected by buoyancy or hydrodynamics. When a cross-medium UAV is partially submerged in water, the buoyancy and hydrodynamic forces it experiences vary with the immersion depth. This piecewise function can be used to further control the activation status of its underwater and aerial propulsion systems. For example... The underwater thruster can be controlled to start at any time. It can control the start of the rotor in the air.

[0052] Secondly, see Figures 8-10 The present invention discloses a cross-medium unmanned aerial vehicle (UAV) 1, which is applicable to the above-mentioned cross-medium UAV 1 tracking and control method. It includes an air propulsion mechanism 11, an underwater navigation mechanism 12, a body 13, and a controller 14. The controller 14 is installed in the body 13. The body 13 is connected to the air propulsion mechanism 11 and the underwater navigation mechanism 12 respectively. The controller 14 is electrically connected to the air propulsion mechanism 11 and the underwater navigation mechanism 12 respectively.

[0053] Specifically, the fuselage 13 is made of lightweight materials and serves as the load-bearing base for the entire aircraft. It is mechanically fixed to the air propulsion mechanism 11 and the underwater navigation mechanism 12. The controller 14, which is built into the fuselage 13, is electrically connected to the air propulsion mechanism 11 and the underwater navigation mechanism 12 through circuits. The fuselage 13 provides buoyancy support for the entire aircraft and reduces underwater travel resistance, while also bearing all functional components. The air propulsion mechanism 11 serves as the power source for air flight. It receives air propulsion speed commands from the controller 14 to generate lift and achieve air flight attitude and position control for the UAV. The underwater navigation mechanism 12 serves as the power source for underwater and surface operations. It receives underwater propulsion thrust speed commands from the controller 14 to complete submersion, surfacing, turning, and forward and backward navigation actions. The controller 14 can run the aforementioned cross-medium UAV 1 tracking and control method, receive real-time status and altitude data collected by the sensors of the fuselage 13, execute a complete set of guidance and control calculations, and output matching drive commands according to the medium to drive the air propulsion mechanism 11 or the underwater navigation mechanism 12 to achieve high-precision trajectory tracking of the UAV in air, underwater, and water-air transition conditions.

[0054] It is understandable that the cross-medium UAV 1 is equipped with an independent aerial propulsion mechanism 11 and an underwater navigation mechanism 12. The fuselage has a built-in controller 14 that is electrically linked to the two types of propulsion mechanisms. The hardware architecture can receive aerial propulsion data and underwater propulsion data output by the method and drive the corresponding actuators according to the medium. The hardware of the whole machine can match the multi-medium working mode based on real-time altitude division, fully accept the three-dimensional motion control commands output by the three-dimensional ILOS guidance, and provide the hardware execution basis for three-dimensional flow field disturbance compensation and lateral and vertical synchronous trajectory correction. It ensures that the three-dimensional kinematic commands and medium-specific propulsion control commands output by the control method can be implemented and executed. This enables the UAV to stably respond to high-precision trajectory tracking control commands in all working conditions of underwater submersion, low-altitude water-skimming flight, and water-air cross-medium transition. From the hardware level, it supports the cross-domain operation capability with no obvious deviation and low steady-state error under complex three-dimensional flow fields.

[0055] The aerial propulsion mechanism 11 includes a support rod 111, a rotor 112, and a drive motor 113. One end of the support rod 111 is connected to the fuselage 13, and the drive motor 113 and the rotor 112 are mounted on the other end of the support rod 111 and connected to the rotor 112.

[0056] Specifically, one end of the support rod 111 is mechanically fixed to the body 13, and the other end of the support rod 111 is equipped with a drive motor 113. The output shaft of the drive motor 113 is directly connected to the rotor 112. When the whole machine is running, the controller 14 outputs an air propulsion speed electrical signal to the drive motor 113. After receiving the command, the drive motor 113 rotates and drives the coaxial rotor 112 to rotate synchronously. The high-speed rotation of the rotor 112 generates a vertical upward thrust. By adjusting the speed difference of each drive motor 113 by the controller 14, the output thrust and output torque of the quadcopter 112 can be changed, thereby realizing the attitude adjustment and position movement of the UAV such as take-off, roll, pitch, and yaw, providing full power support for the UAV in all air flight conditions.

[0057] In this embodiment, the air propulsion mechanism 11 is equipped with four rotors 112 and four drive motors 113, extending from the body 13 in four directions.

[0058] The underwater navigation mechanism 12 includes a first thruster 121 and a second thruster 122. The first thruster 121 is mounted on the fuselage 13, and the second thruster 122 is mounted at the tail of the fuselage 13.

[0059] Specifically, the first thruster 121 is fixedly mounted inside the duct of the fuselage 13, and the second thruster 122 is fixedly arranged below the tail of the fuselage 13. Both are electrically connected to the controller 14 inside the fuselage 13 via wiring. After the controller 14 issues an underwater thrust speed command, the first thruster 121 operates to generate vertical thrust. The speed and steering of multiple first thrusters 121 are coordinated to complete the underwater diving, surfacing, pitching and rolling attitude adjustment of the UAV. The second thruster 122 at the tail can output forward and backward thrust. The synchronous operation of the two second thrusters 122 enables the fuselage 13 to move forward and backward. Differential operation generates yaw torque to complete underwater turning. The two work together to provide a complete set of power for the UAV's surface navigation and underwater submersion, realizing underwater multi-degree-of-freedom maneuvering.

[0060] In this embodiment, four first thrusters 121 and two second thrusters 122 are provided.

[0061] The body 13 includes a buoyancy shell 131, a wave-breaking block 132, and a flow-guiding wedge block 133. The buoyancy shell 131 is connected to the wave-breaking block 132 and the flow-guiding wedge block 133 respectively. Specifically, the fuselage 13 uses a buoyancy shell 131 as its main load-bearing structure. A wave-breaking block 132 is fixedly installed at the lower jaw of the buoyancy shell 131, and a flow-guiding wedge block 133 is fixedly connected at the stern. The three are integrated into a single assembly. The buoyancy shell 131 has a streamlined structure and its buoyancy is slightly greater than the weight of the entire aircraft. When sailing on the water surface, the buoyancy allows most of the fuselage 13 to be submerged in water, ensuring floating stability. At the same time, it provides a sealed and waterproof installation space for internal components such as control modules, batteries, and thrusters. The wave-breaking block 132 at the bow divides and guides the water flow at the front when traveling on the water surface, reducing wave-making resistance. The flow-guiding wedge block 133 at the stern regulates the water flow direction at the tail of the fuselage 13, reducing the pressure difference drag caused by tail eddies. The two work together to significantly reduce the fluid resistance of the UAV during surface and underwater navigation, reduce power consumption, and extend underwater endurance.

[0062] The cross-media drone 1 includes a quick-release mechanism 15 and an energy storage device 16. The quick-release mechanism 15 is connected to the body 13, the energy storage device 16 is connected to the quick-release mechanism 15, and the energy storage device 16 is electrically connected to the controller 14.

[0063] Specifically, the quick-release mechanism 15 is fixedly installed on the underside of the fuselage 13. The energy storage component 16 is detachably mounted on the quick-release mechanism 15. The energy storage component 16 is electrically connected to the controller 14 inside the fuselage 13 via a power supply line. The quick-release mechanism 15 is provided with a sliding groove and a movable latch. The protrusion on the outer wall of the energy storage component 16 matches and slides with the sliding groove. The energy storage component 16 can be locked or quickly disassembled by moving the latch without the need for additional disassembly tools. The energy storage component 16 serves as the power supply for the entire machine, continuously supplying power to the controller 14, the aerial propulsion mechanism 11, the underwater navigation mechanism 12, and the airborne sensing and navigation sensors, ensuring the stable operation of the entire tracking and control method, various power actuators, and sensing equipment. The quick-release mechanism 15 greatly simplifies the replacement process of the energy storage component 16, improving the maintenance efficiency of field operations and the continuous operation capability of the equipment.

[0064] The cross-media unmanned aerial vehicle 1 includes a positioning and navigation module 17 and a communication module 18, which are mounted on one side surface of the body 13.

[0065] Specifically, the positioning and navigation module 17 and the communication module 18 are both fixedly installed on the outer surface of the back of the fuselage 13. They are electrically connected to the built-in controller 14 of the fuselage 13 through internal signal lines. The positioning and navigation module 17 collects real-time status data such as the three-dimensional position, altitude, and attitude of the UAV and transmits it to the controller 14. This provides basic measured data for the tracking control method to perform coordinate transformation, position tracking error calculation, medium state determination, and disturbance observation. The communication module 18 is responsible for bidirectional data interaction with the ground host system. On the one hand, it receives the desired trajectory command issued by the host system. On the other hand, it transmits the real-time position and operating status of the UAV back to the ground terminal. The two work together to ensure that the controller 14 obtains complete mission and fuselage 13 status information, realizing remote mission issuance and real-time monitoring of the overall status.

[0066] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0067] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0068] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0069] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0070] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0071] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. The illustrative expressions of the above terms in this specification should not be construed as necessarily referring to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0072] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Since these modifications and variations fall within the scope of the claims and their equivalents, this invention also intends to include these modifications and variations.

[0073] The above description describes specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A cross-media unmanned aerial vehicle (UAV) tracking and control method, characterized in that, include, Acquire real-time status data, desired trajectory data, and real-time altitude data of the UAV. Calculate position tracking error and disturbance estimate using real-time status data and desired trajectory data. Real-time UAV status data includes position data and attitude data. The flight status of the UAV is determined based on real-time altitude data, a three-dimensional ILOS guidance model is constructed, and the kinematic reference commands are calculated by inputting the UAV flight status, position tracking error and disturbance estimate. The kinematic reference commands include the desired longitudinal velocity, desired vertical velocity, desired yaw angle and desired pitch angle. By combining kinematic reference commands, real-time UAV status data, and disturbance estimates, UAV control data is calculated. The UAV drives its aerial propulsion mechanism or underwater navigation mechanism based on the UAV control data, which includes aerial propulsion data and underwater propulsion data.

2. The method according to claim 1, characterized in that, Acquire real-time status data, desired trajectory data, and real-time altitude data of the UAV. Calculate the position tracking error and disturbance estimate using the real-time status data and desired trajectory data. The specific steps include: Obtain desired trajectory data from the host system, collect UAV flight attitude data, position data, and altitude data in real time, and obtain real-time status data and real-time altitude data; The desired trajectory data undergoes coordinate transformation, converting the position and height data to the same coordinate system as the desired trajectory data. The three-dimensional position tracking error of the UAV is calculated by combining real-time status data and expected trajectory data; A Gaussian RBF neural network observer model is constructed. The attitude data and position data are input, and the combined disturbance force and disturbance moment received by the UAV are calculated to obtain the disturbance estimate.

3. The method according to claim 1, characterized in that, Constructing a Gaussian RBF neural network observer model, inputting attitude and position data, calculating the combined perturbation force and perturbation moment received by the UAV, and obtaining the perturbation estimate, specifically includes the following steps: Construct a Gaussian RBF neural network observer model, input real-time UAV status data, and update the weight matrix in the model in real time based on the gradient descent adaptive law; By comparing the standard dynamic model with real-time state data, the composite disturbances formed by the superposition of wind, waves, ocean currents, model mismatch, and actuator deviations are identified online through the dynamic residual equation, and the composite disturbance force and disturbance torque are output.

4. The method according to claim 1, characterized in that, The UAV's flight status is determined based on real-time altitude data. Kinematic reference commands are calculated by combining the UAV's flight status, position tracking error, and disturbance estimation values. These kinematic reference commands include desired longitudinal velocity, desired vertical velocity, desired yaw angle, and desired pitch angle. The specific steps include: Based on the comparison between real-time altitude data and preset altitude threshold, the medium in which the drone is located is determined, the medium state is classified, and the drone flight status is output. Construct a trajectory-water flow composite coordinate system, map the tracking position error to the composite coordinate system, and obtain the tracking position error after coordinate transformation; The composite pitch angle and composite yaw angle are calculated based on the disturbance estimate to adapt to the influence of water flow. An ILOS guidance model is constructed. The tracking position error after coordinate transformation, the composite pitch angle, the composite yaw angle, and the preset forward look distance are substituted into the ILOS guidance model to calculate the desired longitudinal velocity, desired vertical velocity, desired yaw angle, and desired pitch angle, thereby obtaining the kinematic reference command.

5. The method according to claim 1, characterized in that, By combining kinematic reference commands, real-time UAV status data, and disturbance estimates, UAV control data is calculated. The UAV then drives its aerial propulsion mechanism or underwater navigation mechanism based on this control data, specifically including the following steps: Calculate the current real-time three-axis speed of the drone based on real-time drone status data; The flight mode is distinguished based on the drone's flight status, including aerial mode and underwater mode; The tracking error is calculated based on kinematic reference commands, real-time UAV status data, and real-time three-axis velocity of the UAV in various modes. Specifically, the air mode calculates the speed tracking error, and the underwater mode calculates the attitude angle tracking error. The disturbance estimate is fed forward to compensate the sliding mode control law. The super-helical sliding mode algorithm is used to solve the total resultant force and total resultant torque of the body required to offset the tracking error, and output virtual control quantities, including three-axis virtual resultant force and three-axis virtual resultant torque. The corresponding control efficiency matrix is ​​retrieved based on the flight mode. The matrix is ​​inverted to establish the mapping relationship between the virtual control quantity and the actuator thrust. The required output thrust of each air rotor or underwater thruster is calculated. Finally, the rotor speed command or underwater thrust speed command is output and sent to the UAV.

6. A cross-medium unmanned aerial vehicle (UAV), employing the cross-medium UAV tracking and control method according to any one of claims 1-5, characterized in that, It includes an air propulsion mechanism, an underwater navigation mechanism, a fuselage, and a controller. The controller is installed inside the fuselage, and the fuselage is connected to both the air propulsion mechanism and the underwater navigation mechanism. The controller is electrically connected to both the air propulsion mechanism and the underwater navigation mechanism.

7. The cross-media UAV according to claim 6, characterized in that, The aerial propulsion mechanism includes a support rod, a rotor, and a drive motor. One end of the support rod is connected to the fuselage, and the drive motor and rotor are mounted on the other end of the support rod and connected to the rotor.

8. The cross-media UAV according to claim 6, characterized in that, The underwater navigation mechanism includes a first thruster and a second thruster. The first thruster is mounted on the fuselage, and the second thruster is mounted at the tail of the fuselage.

9. The cross-media UAV according to claim 6, characterized in that, The body includes a buoyancy shell, wave-breaking blocks, and flow-guiding wedges, with the buoyancy shell connecting the wave-breaking blocks and the flow-guiding wedges respectively.

10. The cross-media UAV according to claim 6, characterized in that, It includes a quick-release mechanism and an energy storage component. The quick-release mechanism is connected to the main body, the energy storage component is connected to the quick-release mechanism, and the energy storage component is electrically connected to the controller.

Images