Unmanned aerial vehicle autonomous landing system and method based on UWB and visual fusion

Abstract

The invention discloses an unmanned aerial vehicle autonomous landing system based on UWB and visual fusion, and the system comprises a ground unmanned vehicle and an unmanned aerial vehicle, and the ground unmanned vehicle is in communication connection with the unmanned aerial vehicle. The ground unmanned vehicle is further provided with a ground station, a landing mark and a main control module used for controlling movement output of the ground unmanned vehicle, the main control module is connected with a wireless remote control module, the main control module is further provided with a monitoring unit, and the ground station obtains flight information of the unmanned aerial vehicle for path planning; the unmanned aerial vehicle is further provided with an information collection module, a motion sensing module, a controller module and a power module, and the information collection module is used for collecting position information of the landing mark. Compared with the prior art, the unmanned aerial vehicle autonomous landing system and the unmanned aerial vehicle autonomous landing method based on UWB and vision fusion have the advantages that the unmanned aerial vehicle autonomous landing system and the unmanned aerial vehicle autonomous landing method are used by a combined navigation system formed by UWB and IMU for a scene of cooperative operation of an unmanned aerial vehicle and an unmanned vehicle.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle (UAV) application technology, and in particular to an autonomous landing system and method for an unmanned aerial vehicle (UAV) based on UWB and vision fusion. Background Art

[0002] In recent years, drone technology and unmanned vehicle technology have played an important role in many industries. Ground unmanned vehicles can carry more equipment and have a longer endurance time.

[0003] Drones are small in size and highly efficient in performing tasks, but they have a short endurance. Therefore, combining the advantages of both and working together can significantly improve operational efficiency and the degree of autonomy of the system.

[0004] With the rapid development of UAV technology, the autonomous tracking and landing technology of quadrotor UAV for mobile platforms, which integrates target recognition, tracking and landing decision-making and flight control, has broad application prospects and important research significance.

[0005] With the increasing demand for monitoring and search and rescue missions in complex environments affected by different types of disasters, drones are very helpful for autonomous navigation or human remote operation with their aerial perspective. However, due to the limitation of drone endurance, drones need to return frequently for charging or battery replacement. Therefore, an air-to-ground system composed of drones and unmanned vehicles can provide valuable insights into complex environments, such as battlefield environment reconnaissance, emergency rescue, handling nuclear accidents and other scenarios. Multi-robot systems require drones to take off autonomously, fly in coordination, perform tasks, and return to the landing platform after the tasks are completed on the moving unmanned vehicles. In order to successfully perform its tasks, an accurate navigation system is required.

[0006] like Figures 1-1 to 1-4 As shown in the figure, when drones perform power line inspection tasks, air-ground combined bomb disposal robot systems, coordinated firefighting, and maritime reconnaissance, they are limited by payload equipment and battery technology. The time for a drone to perform a flight operation is often less than an hour. When the inspection mission is completed or the battery is insufficient and needs to return, the drone can be recovered to the mobile platform for rapid charging by autonomous landing, which can greatly improve the deployment time and application efficiency of the drone.

[0007] Therefore, the following problems need to be overcome:

[0008] 1. When UAVs and unmanned vehicles work together, real-time communication needs to be established between them. However, due to the high speed of UAVs and environmental interference, there will be communication delays in the actual process, which may lead to problems such as controller response lag;

[0009] 2. When the drone is tracking the unmanned vehicle, since the movement of the unmanned vehicle is unknown, the drone needs to estimate the movement state of the drone in real time to maintain continuous tracking and control of the unmanned vehicle;

[0010] 3. When the UAV autonomously lands on the unmanned vehicle, wind disturbance will have an adverse effect on the precise landing, or even make it impossible to land. Therefore, enhancing the flight control system's ability to resist wind disturbance is a key technical problem to be solved when designing the flight control system. Summary of the invention

[0011] The technical problem to be solved by the present invention is to overcome the above technical defects and provide a drone autonomous landing system and method based on UWB and vision fusion, which is used in a combined navigation system composed of UWB and IMU for scenarios where drones and unmanned vehicles work together.

[0012] In order to solve the above technical problems, the technical solution provided by the present invention is: an autonomous landing system for unmanned aerial vehicles based on UWB and visual fusion, comprising a ground unmanned vehicle and a unmanned aerial vehicle, wherein the ground unmanned vehicle and the unmanned aerial vehicle are connected in communication;

[0013] The ground unmanned vehicle is also provided with a ground station, a landing mark and a main control module for controlling its mobile output, the main control module is connected to a wireless remote control module, and a monitoring unit is also provided on the main control module. The ground station includes obtaining the flight information of the unmanned aerial vehicle for path planning;

[0014] The drone is also provided with an information collection module, a motion sensor module, a controller module and a power supply module. The information collection module is used to collect the position information of the landing mark.

[0015] Preferably, the motion sensing module includes an inertial measurement unit, a magnetometer, a barometer, and an optical flow sensor.

[0016] Preferably, the information acquisition module includes an airborne camera and a UWB positioning unit, and the controller module obtains feedback information from the motion sensing module and is connected to the control execution unit;

[0017] The onboard camera acquires the landing mark image and performs position calculation, and the controller module acquires the position calculation information and integrates the UWB information and the motion sensor module information to output the attitude information;

[0018] The execution unit includes a PWM signal output by a receiving controller module to control the speed and direction of the drone motor.

[0019] Preferably, the main control module is a microprocessor, the microprocessor is connected to control the monitoring unit and the positioning unit, and the positioning unit uses UWB positioning;

[0020] The monitoring unit includes voltage monitoring, buzzer alarm and indicator light;

[0021] The microprocessor is also connected to control four sets of encoding motors, which control the ground unmanned vehicle to form a four-wheel differential steering unmanned vehicle.

[0022] Preferably, the microprocessor includes a function of parsing the position information of the unmanned vehicle output by the positioning unit, and sending the output information to the ground station and the unmanned aerial vehicle.

[0023] Preferably, the landing mark is an Apriltag landing mark pattern, which includes a plurality of groups of Apriltag marks, and the plurality of groups of Apriltag marks are arranged in a nine-square grid shape.

[0024] Another aspect of the present invention discloses: a method for autonomous landing of a UAV, wherein a combined navigation system is formed by UWB and IMU, and the UAV is connected to a ground unmanned vehicle through a wireless communication module;

[0025] When the UAV receives an autonomous landing command, the onboard camera captures images, detects the Apriltag landing logo pattern, obtains the position information of the target landing platform, combines the UWB / IMU into a navigation system, performs data fusion on the output, and uses PID control to adjust the vertical speed and height of the UAV to accurately land on the unmanned vehicle.

[0026] Preferably, the UWB / IMU navigation system comprises fusing UWB and INS data through a filtering algorithm;

[0027] The position and velocity of UWB and the position, velocity and attitude of INS are used as the input of the filter to build an error model. The INS parameter error output by the Kalman filter is fed back to the system for correction and then the navigation result is output.

[0028] Preferably, the navigation system formula is:

[0029]

[0030] Where k represents the time, t represents the sampling interval from the previous time to this time, E and N represent the east and north coordinates of the object to be measured in the navigation coordinate system, respectively, and v E With v N They represent the speed of the object to be measured in the east and north directions respectively, and a E with a N They represent the acceleration of the object to be measured in the east and north directions respectively;

[0031]

[0032] The coordinates measured by UWB and the distance measured by INS With heading angle As the observation quantity of UWB / INS combined positioning system;

[0033] and They respectively represent the x-axis coordinate and y-axis coordinate measured by UWB at the last moment;

[0034] State vector: x k-1 =[E k-1 N k-1 v Ek v Nk ] T ;

[0035] Observation vector:

[0036] Substitute the state vector and observation vector into the Kalman filter algorithm to obtain the estimated value.

[0037] Preferably, the PID control includes outputting the desired Z-axis speed according to the difference between the desired height and the actual height, using a speed controller to output the desired pulling force F, and obtaining the desired speed and the desired attitude angle of the drone step by step through the cascade PID, and finally converting them into a torque output τ x , τ y , τ z , the calculated tension F and moment τ x , τ y , τ z Control distribution is performed and the output is the PWM signal required by the four motors to adjust the flight attitude of the drone.

[0038] Compared with the prior art, the advantages of the present invention are as follows: the combined navigation system is composed of UWB and IMU, and the connection between the two is established through the wireless communication module, so that the UAV can autonomously track the unmanned vehicle and can cooperate to perform tasks, and after the task is completed, the UAV can autonomously and accurately land on the unmanned vehicle and return together;

[0039] In order to solve the problem that the traditional GPS+INS inertial navigation system cannot be used, UWB positioning technology is used for navigation, and the Kalman filter algorithm is used for data estimation and fusion to reduce data noise and improve navigation accuracy and stability.

[0040] Design a landing mark composed of multiple Apriltag codes to address the problem of narrowing the camera's field of view due to altitude changes during landing, ensuring that the landing mark is within the camera's field of view during the entire tracking and landing process;

[0041] Highly adaptive PID control algorithm, the controller gain changes with the flight altitude, improving the landing success rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1-1 It is a schematic diagram of the application of the prior art.

[0043] Figure 1-2 It is a schematic diagram of the application of the prior art.

[0044] Figure 1-3 It is a schematic diagram of the application of the prior art.

[0045] Figure 1-4 It is a schematic diagram of the application of the prior art.

[0046] Figure 2-1 Schematic diagram of the functional modules of quadrotor drone and ground unmanned vehicle.

[0047] Figure 2-2 It is the block diagram of the embedded system of the ground unmanned vehicle.

[0048] Figure 2-3 It is a picture of the landing platform.

[0049] Figure 2-4 It is a block diagram of the hardware system of a quadrotor drone.

[0050] Figure 3-1 Schematic diagram of UWB / INS integrated navigation.

[0051] Figure 3-2 Control block diagram of the collaborative operation of a quadrotor drone and a ground unmanned vehicle.

[0052] Figure 3-3 This is the PID schematic diagram.

[0053] Figure 3-4 It is the control block diagram of the UAV during autonomous landing.

[0054] Figure 3-5 This is the drone landing process.

[0055] Figure 4-1 It is a flowchart of the Apriltag landing mark recognition algorithm.

[0056] Figure 4-2 This is a diagram of the Apriltag tag recognition process.

[0057] Figure 4-3 This is the navigation data fusion result diagram when the UAV autonomously tracks the unmanned vehicle.

[0058] Figure 4-4 This is a diagram of the drone tracking the unmanned vehicle and landing autonomously.

[0059] Figure 4-5 It is a two-dimensional map of the drone's autonomous tracking landing position.

[0060] Figure 4-6It is a three-dimensional map of the drone's autonomous tracking landing position. DETAILED DESCRIPTION

[0061] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0062] Combined with the attached drawings:

[0063] The system's hardware platform consists of two parts: a quad-rotor UAV and a four-wheel differential steering unmanned vehicle. Combined with corresponding sensor modules, communication modules, vision modules, etc., it realizes functions such as autonomous target positioning, tracking, and autonomous landing of the UAV in collaborative operation scenarios.

[0064] The combined navigation system is composed of UWB and IMU, and the connection between the two is established through the wireless communication module, so that the drone can autonomously track the unmanned vehicle and can perform tasks collaboratively. After the mission is completed, it can autonomously and accurately land on the unmanned vehicle and return together.

[0065] like Figure 2-2 As shown, the power module provides power support for the entire system, and the main control chip is STM32F4. The receiver receives instructions from the remote control, combines the speed information fed back by the encoder motor, distributes power, and calculates the PWM signal required by each motor.

[0066] Since the microcontroller pins can only provide voltage signals, in order to control the motor speed, a motor drive module is also needed. The PWM signal output by the main control chip is converted into the voltage and current of the motor through a high-power dual H-bridge drive circuit to control the motor and change the movement mode of the unmanned vehicle.

[0067] When the drone tracks and lands the unmanned vehicle, the STM32 parses the unmanned vehicle position information output by the UWB through the serial port protocol, and uses the wireless serial port to package the unmanned vehicle's motion information and send it to other platforms, such as PCs and drones. The monitoring module has the functions of detecting the power supply voltage, whether external modules such as UWB and wireless serial ports are working properly, the operating status of motors and drive modules, and abnormal alarms. Users can use this module to detect abnormalities in hardware functions.

[0068] The following table is the parameter data table of unmanned vehicle

[0069]

[0070] based on Figure 2-4 As shown:

[0071] The power module is a lithium battery with light weight and large capacity, which provides power for the entire system. The sensor module includes an inertial measurement unit (IMU), a magnetometer, a barometer optical flow, etc. The IMU integrates an accelerometer and a gyroscope to measure the acceleration and angular velocity of the aircraft and calculate the attitude angle and position. The magnetometer is used to measure the strength of the geomagnetic field. Combined with other sensor data, the heading angle of the aircraft can be obtained. The sensor data is collected through the serial port protocol, SPI protocol, and IIC protocol.

[0072] It also includes a low-voltage power alarm function. When the battery power is lower than the threshold, the detection module will sound an alarm.

[0073] The controller module uses the TM4C123GH6PM microcontroller, which is a 32-bit microcontroller with an ARM Cortex-M4 core. It has strong computing power and communicates with peripheral sensors through communication protocols to collect various data during the flight of the drone in real time to control the flight attitude of the drone, including: receiving sensor data, running flight control algorithms, generating motor control instructions, communicating with the ground station, sending telemetry data and receiving instructions.

[0074] The actuator module includes 4 electronic regulators and brushless motors, which control the speed and direction of each motor according to the PWM signal sent by the controller, thereby adjusting the thrust and torque of the drone in all directions to achieve the desired motion state.

[0075] The wireless data transmission is used for two-way communication between the PC and the aircraft, transmitting flight data to the ground station and receiving instructions from the host computer. The ground station is mainly composed of ANO host computer software and wireless data transmission. The ANO host computer is used to monitor the aircraft status, adjust parameters and plan flight routes. The visual module is used to obtain the landing mark image on the ground unmanned vehicle and perform position calculation.

[0076] The following table shows the hardware and parameters of the quadcopter drone:

[0077]

[0078] The AprilTag mark is composed of a square mark with a wide border of alternating black and white and an internal binary matrix. The present invention adopts a combination marker based on multiple AprilTags, and adds other AprilTag marks around an AprilTag mark. When a certain deviation occurs during the landing process, the camera can collect any AprilTag mark in the surrounding area, allowing the drone to move closer to the middle and find the AprilTag mark in the middle again, ensuring that the drone effectively tracks the unmanned vehicle and finally lands successfully.

[0079] UWB / INS integrated navigation algorithm based on Kalman filtering:

[0080] Although INS can provide real-time information on the carrier's motion status, its navigation accuracy is not high due to the influence of hardware quality and sensor cumulative errors. The UWB positioning system has the advantages of high precision and high stability, but the signal is easily interfered with, especially in obstructed areas or when the signal is lost, it can no longer complete positioning.

[0081] The UWB positioning system used in this system fuses UWB and INS data through a filtering algorithm. INS can provide continuous attitude and acceleration information to ensure the continuity of navigation. By fusing the position update of UWB, the drift error of INS can be reduced and the accuracy of long-term navigation can be improved. UWB and inertial navigation systems work independently and give navigation results respectively. In order to improve navigation accuracy, the position and velocity of UWB and the position, velocity and attitude of INS are used as inputs of the filter. The filter compares the difference between the two and constructs an error model to estimate the error of INS. In the feedback correction, the INS parameter error estimated by the Kalman filter is fed back to the inertial navigation system to correct the error, thereby outputting a navigation result that is closer to the true value, such as Figure 3-1 shown.

[0082] The inertial navigation error differential equation is established, and on this basis, the UWB / INS combined navigation error state space model with velocity error and position error as observation values ​​is obtained. The Kalman filter fuses the positioning data measured by UWB with the positioning data measured by INS to output smoother and more accurate results.

[0083] In the UWB / INS combined positioning method, by adjusting the layout of the base station, the coordinate system of the UWB positioning is made to coincide with the navigation coordinate system, that is, the x-axis is in the east direction and the y-axis is in the north direction. The state equation of the navigation system is as follows:

[0084]

[0085] Where k represents the time, t represents the sampling interval from the previous time to this time, E and N represent the east and north coordinates of the object to be measured in the navigation coordinate system, respectively, and v E With v N They represent the speed of the object to be measured in the east and north directions respectively, and a E with a N They represent the acceleration of the object to be measured in the east and north directions respectively.

[0086] Observation equation:

[0087]

[0088] In the formula, the coordinates measured by UWB and the distance measured by INS are With heading angle As the observation quantity of UWB / INS combined positioning system. and They represent the x-axis coordinate and y-axis coordinate measured by UWB at the last moment. State vector: x k-1 =[E k-1 N k-1 v Ek v Nk ] T , observation vector: Substitute the state vector and observation vector into the Kalman filter algorithm to obtain a more accurate estimate.

[0089] Introduction to Kalman filtering:

[0090] Kalman filtering is a very classic estimation method that estimates the system state in real time through noisy measurement data. The traditional Kalman filtering method is built on the premise that the system model is accurate and the statistical characteristics of the random interference signal are known. It uses a linear state space model to estimate the system state, and the algorithm is relatively simple to implement. Since Kalman filtering achieves optimal estimation based on the minimum mean square error criterion, it is widely applicable to linear time-varying discrete systems containing Gaussian white noise.

[0091] The state space model of Kalman filter:

[0092]

[0093] Where, X k is the n-dimensional state vector; Z k is the m-dimensional measurement vector; Φ k / k-1 , Γ k / k-1 and H k They are respectively called the n-order state one-step transfer matrix, the n×l-order system noise allocation matrix, and the m×n-order measurement matrix; W k-1 is the l-dimensional system noise vector, V k is an m-dimensional measurement noise vector. Assume that both process and measurement noise are zero-mean Gaussian white noise vectors and are uncorrelated, that is, they satisfy:

[0094]

[0095] The basic process of the Kalman filter algorithm is as follows:

[0096] (1) Priori

[0097] ① One-step state prediction:

[0098]

[0099] ② Calculate the prior mean square error matrix:

[0100]

[0101] (2) Posterior

[0102] ① Kalman filter gain:

[0103]

[0104] ② Corrected state estimation:

[0105]

[0106] ③ Corrected state estimation mean square error matrix:

[0107] P k =(IK k H k ) k / k-1

[0108] Before the Kalman filter algorithm starts iterating, it needs to be initialized. and the initial value of P0.

[0109]

[0110] like Figure 3-2 As shown in the figure, the flight control adopts cascade control, which includes position control, attitude control and altitude control. Control target: The UAV takes off autonomously, reaches the specified altitude, and keeps a certain distance to track the unmanned vehicle. The UWB / INS combined navigation system provides accurate movement information of the unmanned vehicle, controls the position coordination of the UAV and the unmanned vehicle, and enables the UAV to stably track the movement of the ground unmanned vehicle.

[0111] PID principle Figure 3-3 As shown:

[0112] As a linear controller, the conventional PID controller forms a deviation based on the given value r(t) and the actual output value c(t): e(t) = r(t)-c(t). The proportion P, integral I and differential D of the deviation are linearly combined to form the control quantity to control the controlled object. Its control law is:

[0113]

[0114] The transfer function is:

[0115]

[0116] Where K P is the proportionality coefficient, T i is the integration time constant, T dis the differential time constant, K i is the integration coefficient, K d is the differential coefficient.

[0117] When used in the present invention:

[0118] The height PID controller calculates the expected speed on the z-axis based on the difference between the expected height and the actual height, and outputs the expected tension F through the speed controller. Under the action of the positioning system, the UAV obtains the real-time position information of the unmanned vehicle through the wireless communication module and converts it into the expected position of the target that the UAV needs to track after maintaining a certain distance. The expected speed and expected attitude angle of the UAV are obtained step by step through the cascade PID, and finally converted into the torque output τ x , τ y , τ z . The calculated tension F and moment τ x , τ y , τ z Control distribution is carried out and converted into the PWM signals required by the four motors in the hardware system. The motor speed is controlled to adjust the flight attitude of the UAV, ensuring that the UAV can track the ground unmanned vehicle in real time and stably, completing the coordinated operation of the two.

[0119] Under the UWB / INS combined navigation system, the UAV keeps a certain distance from the unmanned vehicle and stably tracks the ground UAV at a vertical height h. When the mission is completed and the landing command is received, the onboard camera recognizes the landing mark composed of the Apriltag code, and obtains the position information of the UAV relative to the ground unmanned vehicle through algorithm detection and calculation, and integrates it with the original navigation data to control the UAV to continuously approach the target center in the horizontal direction, and maintain a certain height directly above the unmanned vehicle and the horizontal relative position with the unmanned vehicle is zero. Through the highly adaptive PID control algorithm, the UAV can land stably and quickly on the unmanned vehicle.

[0120] In one embodiment:

[0121] Consider possible emergencies, such as loss of landing target, and plan emergency response plans in advance, such as Figure 3-5 As shown in the figure, when the landing target is lost, the UAV will rise to the initial height on the spot, expand the field of view of the onboard camera, and recapture the image; when the UAV executes the landing command and the flight altitude is 20cm above the ground, it will lock the landing and control the UAV to autonomously and safely land on the ground unmanned vehicle.

[0122] During the flight of the drone, different altitudes correspond to different expected speeds, that is, the output response of the controller will be different. When the relative altitude is high, it is hoped that the drone will respond faster to achieve a faster landing, so during the landing process, the altitude controller uses an exponential function to dynamically adjust the PID control parameters:

[0123]

[0124] Where, j = p, i, d are the gain parameters of the highly adaptive PID controller, p z is the vertical height of the UAV relative to the landing plane, D and a are the parameters of the function, which are measured experimentally and used to adjust the gain.

[0125] During the landing mark detection test, the camera was installed at the center of mass of the drone and below. The image resolution captured by the camera was a 640*480 color image. The mark detection includes the first step of detecting the candidate mark and converting the RGB image captured by the camera into a grayscale image. In order to detect the candidate squares, an adaptive threshold method is used to convert the grayscale image into a binary image during the detection process to shorten the image processing time; the purpose of the second step is to obtain the internal code of the Apriltag tag. The size of the Apriltag code is 6*6 squares, each of which consists of black and white, corresponding to binary codes 0 and 1. By separating the white and black parts, the binary code of the Apriltag in the captured image is obtained, and its ID in the family can be determined. After obtaining the ID, the relative position and other information can be analyzed by calculation, such as Figure 4-1 shown.

[0126] like Figure 4-2 As shown in the figure, when part of the Apriltag code is blocked, positioning can continue based on other Apriltag codes. Each Apriltag code has a different ID. By designing their positions on the landing plane, the position of any Apriltag code center from the landing mark center can be obtained by calculation, providing a position target for the drone to approach the center of the landing platform.

[0127] like Figure 4-3 As shown in the figure, after the data is fused by the Kalman filter algorithm, the fluctuation of the position data is small, there is no obvious jump, and it presents a smoother change trend;

[0128] By formula The standard deviation (SD) of the position data before and after fusion using the Kalman filter algorithm is calculated, where X k is the average value of the sampled data, X kis the data corresponding to the sampling point, n represents the number of signal samples, and the specific values ​​are shown in the following table:

[0129]

[0130] After data fusion processing, the position standard deviation becomes significantly smaller, because the Kalman filter algorithm can estimate and update the noise variance matrix in a timely manner.

[0131] Combined with Figures 4-4 to 4-6 It can be seen that the UAV has a better effect of tracking the unmanned vehicle, and the two can maintain a certain distance in vertical height. In addition, the height of the UAV's autonomous take-off point and autonomous landing point is about 0.1m from the landing platform surface. This is because the UAV is equipped with a landing gear, which gives the UAV an initial height.

[0132] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

[0133] The scope of the present invention is defined by the appended claims and their equivalents, and therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the present invention as claimed but merely represents selected embodiments of the present invention.

[0134] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of the present invention.

[0135] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. UAV autonomous landing system based on UWB and vision fusion, characterized by: Including ground unmanned vehicles and unmanned aerial vehicles, and communication connection between ground unmanned vehicles and unmanned aerial vehicles; The ground unmanned vehicle is also provided with a ground station, a landing mark and a main control module for controlling its mobile output, the main control module is connected to a wireless remote control module, and a monitoring unit is also provided on the main control module. The ground station includes obtaining the flight information of the unmanned aerial vehicle for path planning; The drone is also provided with an information collection module, a motion sensor module, a controller module and a power supply module. The information collection module is used to collect the position information of the landing mark.

2. The UAV autonomous landing system based on UWB and vision fusion according to claim 1 is characterized in that: The motion sensing module includes an inertial measurement unit, a magnetometer, a barometer, and an optical flow sensor.

3. The UAV autonomous landing system based on UWB and vision fusion according to claim 1 or 2, characterized in that: The information acquisition module includes an airborne camera and a UWB positioning unit, and the controller module obtains feedback information from the motion sensor module and is connected to the control execution unit; The onboard camera acquires the landing mark image and performs position calculation, and the controller module acquires the position calculation information and integrates the UWB information and the motion sensor module information to output the attitude information; The execution unit includes a PWM signal output by a receiving controller module to control the speed and direction of the drone motor.

4. The UAV autonomous landing system based on UWB and vision fusion according to claim 1 is characterized in that: The main control module is a microprocessor, which is connected to the control monitoring unit and the positioning unit, and the positioning unit uses UWB positioning; The monitoring unit includes voltage monitoring, buzzer alarm and indicator light; The microprocessor is also connected to control four sets of encoding motors, which control the ground unmanned vehicle to form a four-wheel differential steering unmanned vehicle.

5. The UAV autonomous landing system based on UWB and vision fusion according to claim 4 is characterized in that: The microprocessor includes a device for analyzing the position information of the unmanned vehicle output by the positioning unit, and sending the output information to the ground station and the unmanned aerial vehicle.

6. The UAV autonomous landing system based on UWB and vision fusion according to claim 1 is characterized in that: The landing mark is an Apriltag landing mark pattern, which includes multiple groups of Apriltag marks, and the multiple groups of Apriltag marks are arranged in a nine-square grid shape.

7. A method for autonomous landing of a UAV, based on the autonomous landing system of a UAV based on UWB and visual fusion according to any one of claims 1 to 6, characterized in that: It includes a combined navigation system composed of UWB and IMU, and the UAV and the ground unmanned vehicle are connected through a wireless communication module; When the UAV receives an autonomous landing command, the onboard camera captures images, detects the Apriltag landing logo pattern, obtains the position information of the target landing platform, combines the UWB / IMU into a navigation system, performs data fusion on the output, and uses PID control to adjust the vertical speed and height of the UAV to accurately land on the unmanned vehicle.

8. The autonomous landing method for a drone according to claim 7, characterized in that: The UWB / IMU navigation system comprises fusing UWB and INS data through a filtering algorithm; The position and velocity of UWB and the position, velocity and attitude of INS are used as the input of the filter to build an error model. The INS parameter error output by the Kalman filter is fed back to the system for correction and then the navigation result is output.

9. The autonomous landing method for a drone according to claim 8, characterized in that: The navigation system formula is: Where k represents the time, t represents the sampling interval from the previous time to this time, E and N represent the east and north coordinates of the object to be measured in the navigation coordinate system, respectively, and v E With v N They represent the speed of the object to be measured in the east and north directions respectively, and a E with a N Respectively represent the acceleration of the object to be measured in the due east and due north directions; The coordinates measured by UWB and the distance measured by INS With heading angle As the observation quantity of UWB / INS combined positioning system; and They respectively represent the x-axis coordinate and y-axis coordinate measured by UWB at the last moment; State vector: x k-1 =[E k-1 N k-1 v Ek v Nk ] T ; Observation vector: Substitute the state vector and observation vector into the Kalman filter algorithm to obtain the estimated value.

10. The autonomous landing method of a UAV according to claim 7, characterized in that: The PID control includes outputting the desired Z-axis speed according to the difference between the desired height and the actual height, using the speed controller to output the desired tension F, and obtaining the desired speed and attitude angle of the drone step by step through the cascade PID, and finally converting it into a torque output τ x , τ y , τ z , the calculated tension F and moment τ x , τ y , τ z Control distribution is performed and the output is the PWM signal required by the four motors to adjust the flight attitude of the drone.

Images