CN-121791891-B - Unmanned aerial vehicle double-loop beam tracking system and method based on intelligent super surface

CN121791891BCN 121791891 BCN121791891 BCN 121791891BCN-121791891-B

Abstract

The invention discloses an unmanned aerial vehicle double-loop beam tracking system and method based on an intelligent super-surface, belonging to the field of wireless communication and electromagnetic beam control, wherein a binocular depth camera at the ground end is used as a main sensor to collect unmanned aerial vehicle images in real time, target detection, association tracking and track estimation are carried out on an image sequence of the unmanned aerial vehicle, and azimuth angle and pitch angle taking the center of the super-surface as a reference are output to generate an intelligent super-surface discrete coding matrix so as to finish beam state updating; the double-loop coupling control module extracts amplitude and trend characteristics from the returned voltage sequence, and adopts an improved Kalman filter to correct the angular state of the unmanned aerial vehicle so as to form beam pointing closed loop updating. The scheme is suitable for application scenes such as low-altitude communication, monitoring and relay.

Inventors

JI XIAO
YE ZHENG
Gao Longtao
GENG XIONG
HAO WENYU
ZHOU WEI
GUO YUEHAN

Assignees

华中科技大学

Dates

Publication Date: 20260512
Application Date: 20260304

Claims (8)

1. An unmanned aerial vehicle double-loop beam tracking method based on intelligent super-surface is characterized by comprising the following steps: s1, receiving a current frame unmanned aerial vehicle image acquired by a binocular camera at a current sampling time n+1 Target detection algorithm is adopted for unmanned aerial vehicle images of current frames Performing target detection to obtain a detection result, and adopting ByteTrack algorithm to combine the detection result with the previous unmanned aerial vehicle image Generating array discrete control quantity of RIS according to the unmanned aerial vehicle pixel coordinates so as to update the beam state of the RIS; s2, calculating the previous frame of unmanned aerial vehicle image In unmanned aerial vehicle pixel speed Receiving voltage signals corresponding to power signals fed back by unmanned aerial vehicle after updating beam state of RIS According to the pixel speed And voltage signal Adopting an improved ByteTrack algorithm to combine the detection result with the previous unmanned aerial vehicle image Generating corrected array discrete control quantity of RIS according to the corrected unmanned aerial vehicle pixel coordinates so as to update the beam state of RIS again; s3, updating n to n+1, returning to S1, and until the termination condition is reached.
2. The method of claim 1, wherein in step S2, the detection result is compared with a previous frame of unmanned aerial vehicle image by using a modified ByteTrack algorithm according to the pixel speed and voltage signals Performing data association to obtain corrected unmanned aerial vehicle pixel coordinates, including: S201, adopting an improved Kalman filter according to the pixel speed And voltage signal Predicting previous frame unmanned aerial vehicle image Correction result of position of middle unmanned aerial vehicle ; S202, the unmanned aerial vehicle image of the previous frame is processed Correction result of position of middle unmanned aerial vehicle Performing data association with the detection result to obtain corrected unmanned aerial vehicle pixel coordinates; Wherein, the The calculation formula of (2) is as follows: ; for improved Kalman filter output The correction result of the position of the unmanned aerial vehicle, Last two-frame unmanned aerial vehicle image output for improved Kalman filter The correction result of the position of the unmanned aerial vehicle, Is that Is provided with the unmanned aerial vehicle pixel coordinates, For improved kalman gain of the kalman filter, , In order to measure the noise of the light, , In order to set the coefficient to be the preset value, Is that Covariance of correction result of unmanned plane position according to iterative formula The method is obtained by iterative calculation, , As an initial value of the process noise, , , , , In order to smooth the coefficient of the coefficient, Unmanned aerial vehicle image for last two frames Is used for the voltage signal of the (a), Is that 、 Is used for the sampling time interval of (a).
3. The method of claim 1 or 2, wherein the last frame of drone image In unmanned aerial vehicle pixel speed The calculation formula of (2) is as follows: Wherein, the , , wherein, 、 Unmanned aerial vehicle images of the last frame respectively Unmanned aerial vehicle pixel in (a) The coordinates of the two points of the coordinate system, 、 Unmanned aerial vehicle images respectively of two frames Unmanned aerial vehicle pixel of (a) The coordinates of the two points of the coordinate system, Is that 、 Is used for the sampling time interval of (a).
4. The method of claim 1, wherein in step S1, the generating an array discrete control quantity of RIS from the unmanned aerial vehicle pixel coordinates comprises: S101, pixel coordinates of the unmanned aerial vehicle are set according to camera internal parameters Depth and depth Conversion to three-dimensional points in camera coordinate system ; S102, calibrating offset pair three-dimensional points based on camera and RIS center Coordinate correction is carried out to obtain three-dimensional points taking RIS as reference ; Wherein, the , , = , Calibrating and biasing the binocular camera and the RIS center; S103, according to Calculating azimuth angle of unmanned aerial vehicle relative to RIS And pitch angle ; S104, according to azimuth angle of unmanned aerial vehicle relative to RIS And pitch angle The array discrete control quantity of RIS is calculated.
5. The method of claim 4, wherein in step S104, the azimuth angle is determined And pitch angle Inputting the data into a pre-trained neural network to obtain the array discrete control quantity of the RIS.
6. The method of claim 5, wherein in step S104, the azimuth angle of the drone with respect to the RIS is based on And pitch angle Before calculating the array discrete control quantity of the RIS, the method further comprises the following steps: Will azimuth angle And pitch angle As the current azimuth angle And pitch angle Respectively corresponding to the historical azimuth angles and the historical pitch angles one by one to fuse the current azimuth angles And pitch angle Smoothing is performed.
7. The method of claim 5, wherein the neural network is a CNN-fransformer network.
8. Unmanned aerial vehicle dicyclo wave beam tracking system based on intelligence super surface, its characterized in that includes: The RIS is used for receiving and reflecting the communication signals sent by the base station; The ground end control device comprises a binocular depth camera, a double-loop coupling control module and a phase/voltage issuing module; The binocular depth camera and the RIS are positioned on the same plane and are used for acquiring images of the unmanned aerial vehicle in real time; The dual loop coupling control module for performing the method of any of claims 1-4 to update the beam state of the RIS; The phase/voltage issuing module is used for converting the array discrete control quantity into a voltage control signal of the RIS and issuing the voltage control signal to a voltage control board of the RIS so as to enable the RIS to update the beam state; the unmanned aerial vehicle terminal power feedback unit is used for receiving a power signal received by the unmanned aerial vehicle after the RIS updates the beam state, converting the power signal into a voltage signal, performing analog-to-digital conversion on the voltage signal to obtain a digital voltage signal, and sending the digital voltage signal to the control module.

Description

Unmanned aerial vehicle double-loop beam tracking system and method based on intelligent super surface Technical Field The invention belongs to the field of wireless communication and electromagnetic wave beam control, and particularly relates to an unmanned aerial vehicle double-loop wave beam tracking system and method based on an intelligent super surface. Background In low-altitude mobile platform communications, monitoring and relay applications, link stability and pointing consistency typically depend on beam alignment. A smart supersurface (RIS, reconfigurable Intelligent Surface) as a reconfigurable electromagnetic steering structure can change the reflection/transmission phase profile under external control to form a beam in a specific direction. In practical deployment, the beam state update of the intelligent subsurface is accomplished by the control and bias voltage application process, and the update period is affected by the hardware and control links. Meanwhile, the unmanned aerial vehicle target continuously moves in time, frame rate and processing delay exist for visual observation of the unmanned aerial vehicle, and a single information source can be shielded or lost in a short time. Disclosure of Invention Aiming at the defects or improvement demands of the prior art, the invention provides an unmanned aerial vehicle double-loop beam tracking system and method based on an intelligent super surface, wherein a closed loop relationship is established among intelligent super surface beam updating, visual angle estimation and energy domain feedback so as to support continuous pointing updating of an unmanned aerial vehicle target. To achieve the above object, according to a first aspect of the present invention, there is provided an unmanned aerial vehicle dual-loop beam tracking method based on intelligent super-surface, including: s1, receiving a current frame unmanned aerial vehicle image acquired by a binocular camera at a current sampling time n+1 Target detection algorithm is adopted for unmanned aerial vehicle images of current framesPerforming target detection to obtain a detection result, and adopting ByteTrack algorithm to combine the detection result with the previous unmanned aerial vehicle imageGenerating array discrete control quantity of RIS according to the unmanned aerial vehicle pixel coordinates so as to update the beam state of the RIS; s2, calculating the previous frame of unmanned aerial vehicle image In unmanned aerial vehicle pixel speedReceiving voltage signals corresponding to power signals fed back by unmanned aerial vehicle after updating beam state of RISAccording to the pixel speedAnd voltage signalAdopting an improved ByteTrack algorithm to combine the detection result with the previous unmanned aerial vehicle imageGenerating corrected array discrete control quantity of RIS according to the corrected unmanned aerial vehicle pixel coordinates so as to update the beam state of RIS again; s3, updating n to n+1, returning to S1, and until the termination condition is reached. According to a second aspect of the present invention, there is provided an unmanned aerial vehicle dual-loop beam tracking system based on intelligent super-surface, comprising: The RIS is used for receiving and reflecting the communication signals sent by the base station; The ground end control device comprises a binocular depth camera, a double-loop coupling control module and a phase/voltage issuing module; The binocular depth camera and the RIS are positioned on the same plane and are used for acquiring images of the unmanned aerial vehicle in real time; The dual loop coupling control module is configured to perform the method of the first aspect to update the beam state of the RIS; The phase/voltage issuing module is used for converting the array discrete control quantity into a voltage control signal of the RIS and issuing the voltage control signal to a voltage control board of the RIS so as to enable the RIS to update the beam state; the unmanned aerial vehicle terminal power feedback unit is used for receiving a power signal received by the unmanned aerial vehicle after the RIS updates the beam state, converting the power signal into a voltage signal, performing analog-to-digital conversion on the voltage signal to obtain a digital voltage signal, and sending the digital voltage signal to the control module. In general, the above technical solutions conceived by the present invention, compared with the prior art, enable the following beneficial effects to be obtained: The invention provides an intelligent super-surface-based unmanned aerial vehicle double-loop beam tracking system and method, which are characterized in that a binocular depth camera at the ground end is used as a main sensor to acquire unmanned aerial vehicle images in real time, target detection, associated tracking and track estimation are carried out on an unmanned aerial vehicle image sequence, an azimuth angle phi