CN107727945B - Large parabolic antenna surface accuracy testing system based on UAV - Google Patents

Abstract

The invention discloses a large parabolic antenna surface accuracy testing system based on a drone, which consists of a detection part and a data processing part connected in sequence; the detection part includes a drone equipped with a camera, an identification mark, and a laser calibration reference Subsystem, signal receiving and transmitting module I, the data processing part includes signal receiving and transmitting module II, and computer; this system is based on the camera probe mounted on the drone to capture the identification mark on the surface of the antenna, which can quickly and quasi-real-time at different observation angles of the antenna Obtaining the current antenna panel accuracy and quickly correcting the surface shape can quickly improve the antenna performance, overcoming the problems of pitch angle consideration, light pollution and other problems in the current mainstream antenna surface measurement.

Description

Large parabolic antenna surface accuracy testing system based on UAV

Technical field

The invention relates to the fields of large-scale parabolic antenna measurement and radio astronomy research, and is suitable for real-time evaluation of antenna performance when using large-scale parabolic antennas for radio astronomical observations to obtain current antenna performance.

Background technique

With the development of science and technology, large parabolic antennas have been used in many aspects such as ground communication antennas, space-borne deployable antennas, radio astronomical telescopes, etc. At the same time, higher requirements have been put forward for the electrical performance indicators of reflective antennas, such as high gain, narrow Beam, high efficiency, etc., which will inevitably increase the difficulty of the electromagnetic design and structural design of the parabolic antenna. Reflector antennas usually work in higher frequency bands. In order to pursue higher electrical performance indicators, higher requirements are placed on the accuracy of the antenna structure; therefore, it is necessary to quantitatively describe the impact of various error information on the electrical performance of the antenna structure; Among them, the error information mainly includes installation, manufacturing errors and system errors, and is mainly the error information of the reflective panel; installation and manufacturing errors are random errors, which are a type of rapidly changing errors. The impact of such errors on electrical performance can be measured through probability method for estimation; Ruze was the first to give the relationship between random error and gain loss, Vu expanded its formula and studied the relationship between uneven random error and aperture field, Rahma-t samii in order to carry out structural Parametric research provides a mathematical model of the impact of random errors on the average power pattern; Ruze, Vu, Rahma-t samii and others only studied the impact of random errors on the reflective surface (panel manufacturing accuracy) on electrical properties. ; The influence of systematic errors is not considered; the systematic errors are inherent in the antenna structure; their sources are caused by loads such as self-weight, temperature, inertia, vibration, etc. acting on the antenna structure. This type of error changes slowly and can be passed through the structure. Analyze and determine structural deformation information.

In 2005, Bahadori K gave a relationship model of the impact of system errors on electrical performance and analyzed the impact of system errors on gain and side lobe levels; however, Bahadori K did not give the impact of random errors on the reflective surface and the back support of the reflective surface. ; But in the actual working conditions of the antenna, various error information (system error and random error) of the antenna exist at the same time. It is in line with the actual situation to analyze the comprehensive impact of various error information on electrical performance. For reflective panels, the random error depends on the manufacturing accuracy of the panel and will not change with the working state and environment; the systematic error is affected by the working state of the antenna, the environment and its back frame support structure. Based on this, a mathematical model of the impact of the reflective surface on the electrical performance when random errors and systematic errors exist simultaneously is established. Structural analysis is used to give the systematic error information of the reflecting surface, and the influence of error ε on the gain and side lobe level of the parabolic antenna is analyzed and calculated through numerical methods.

At present, there are the following methods for measuring the accuracy of antenna panels at home and abroad: theodolite method, microwave holography method and digital photography method. The theodolite method has been eliminated due to its low accuracy.

(1) Microwave holography method:

The microwave holography method has become increasingly mature through long-term exploration by Wang Jinqing and others from the Shanghai Observatory. The method is: since there is a two-dimensional Fourier transform relationship between the aperture field and the far field of the parabolic antenna, this relationship can be used to measure the high-frequency phase diagram of the large aperture antenna, thereby determining the accuracy of the antenna panel.

The main advantages of microwave holography:

1. Does not depend on external light;

2. Quickly obtain panel distribution.

Main disadvantages:

1. It relies heavily on satellite frequency and position. For example, when measuring the 40-meter radio telescope of Yunnan Observatory in 2012, we used the same satellite as the Shanghai Observatory as the far-field signal source. However, this star is relatively small at the geographical latitude of Kunming, Yunnan. In the Shanghai area, it is nearly 12° higher, which causes the test adjustment angle of the 40-meter antenna to be too high, and the antenna performance decreases at low pitch positions (below 35°);

2. High-frequency electromagnetic waves are greatly affected by weather, and measurements in rainy weather are affected;

3. The measuring pitch plane is fixed and cannot take into account multiple pitch planes.

(2) Digital photography method

The basic principle of photogrammetry is very similar to the dual theodolite system. If you use a camera to take pictures of the measured target at two locations, you can get photos of the measured target at two different angles. These two different angle photos are A stereoscopic image pair is formed. If the measured target point is photographed from multiple camera stations, multiple stereo image pairs of the measured target can be obtained to form a multi-eye three-dimensional model.

If the target is photographed from multiple camera stations, multiple stereoscopic image pairs of the measured object can be obtained to form a multi-view stereoscopic model. Assuming that the object-direction point is only intersected by i camera stations (i rays), there are i collinear equations:

Among them, x _s , y _s and z _s , a _i b _i and c _i (i=1, 2, 3) are the translation amounts of the external orientation elements of the image and the elements of the rotation matrix respectively; x ₀ , y ₀ , f, Δx, and Δy are the internal parameters of the image, which have been calibrated in advance and can be regarded as known values; x and y are the image point coordinates corresponding to the object coordinates X, Y, and Z. According to the principle of least squares, the spatial coordinates (X, Y, Z) of the object square point can be obtained by solving the collinear equations of multiple rays (beams) simultaneously (beam method adjustment). According to the obtained spatial coordinates, the 3D model of the antenna can be reconstructed and obtained.

The accuracy measurement of the antenna panel of the 13.7-meter millimeter-wave radio telescope at the Qinghai Delingha Observatory of the Purple Mountain Observatory adopted the digital photography method. In the measurement in 2009, the accuracy error of the front panel was: 0.768mm. After the measurement and adjustment, the panel accuracy was obtained. The error is: 0.083 mm.

The Sino-German submillimeter wave radio telescope kosma also adopted this method. In the end, the surface accuracy of the inner ring panel reached 14 μm (r.m.s.), and the overall panel surface accuracy was better than 24 μm (r.m.s.), achieving submillimeter wave 235GHz observation. surface accuracy requirements.

The main advantages of digital photography:

1. The measurement accuracy is very high and is not affected by microwave frequency;

2. Taking into account various pitch angles and not affected by humidity and temperature;

3. Quickly calculate the surface shape accuracy.

Main disadvantages:

1. Work during the day is affected by sunlight;

2. When the antenna is tilted at a high pitch, it requires a guide rail or a crane, which is troublesome to operate;

3. Reflective materials are easily contaminated.

Different wind speeds and temperatures will cause different antenna surface accuracy even at the same pitch angle, such as the newly built 65-meter radio telescope of the Shanghai Observatory, the 110-meter radio telescope planned to be built by the Xinjiang Observatory, and the new Jingdong radio telescope planned to be built by the Yunnan Observatory. All take into account active surface shape control.

Based on the identification mark on the antenna surface captured by the camera probe mounted on the drone, the current antenna panel accuracy can be obtained quickly and quasi-real-time at different observation angles of the antenna, and the surface shape can be quickly corrected to quickly improve the antenna performance.

Contents of the invention

The purpose of the present invention is to provide a large parabolic antenna surface accuracy testing system based on an unmanned aerial vehicle, which is mainly suitable for the field of radio astronomy observation of large parabolic antennas. It is a device and solution for quickly and safely obtaining antenna surface parameters. It overcomes the problems of considering the pitch angle and light pollution in the current mainstream antenna surface measurement.

The present invention is achieved through the following technical solutions:

A UAV-based large-scale parabolic antenna surface accuracy testing system consists of a detection part and a data processing part connected in sequence.

The detection part consists of the following:

1. A multi-axis UAV aircraft and its equipped camera constitute the shooting subsystem. The multi-axis aircraft can be a four-axis, six-axis or eight-axis aircraft, and then the camera is hung. A memory card is integrated in the camera to record the captured pictures. Backup storage;

2. The UAV carries a signal receiving and transmitting module I. This module is connected to the memory card of the camera and the flight control circuit of the UAV. It is used to modulate the image captured by the camera into a wireless signal and transmit it to the data processing part on the ground. At the same time Receive the stop shooting instruction from the ground data processing part. The module uses wireless frequencies such as wifi, 3G or 4G to transmit and receive;

3. The identification mark is pasted on the surface of the main reflecting surface of the antenna. It is made of highly photosensitive material and is easy to be recognized by the camera;

At the same time, each identification mark has two different identification features, which can be numbers printed on the mark made of photosensitive material, or can be composed of different combinations of shapes;

4. The laser calibration reference subsystem includes a laser transmitter and a laser absorber. The relative positions of the laser transmitter and the absorber can be accurately determined. The system is fixed at a conspicuous place on the antenna surface and is suitable for drone shooting at any angle. To this laser beam, the distance from the laser emitted by the laser emitter to the laser absorber can be accurately measured, which provides a benchmark for the entire measurement.

The data processing part consists of the following parts:

1. Signal receiving and transmitting module II, which receives image signals from the drone and transmits flight control commands. This module uses wireless frequencies such as wifi, 3G or 4G to receive and transmit;

2. Computer. The computer includes conventional graphics processing software. After identifying the image sent by the drone, it finds the location of the antenna where the image is located, and performs virtual three-dimensional positioning to achieve a three-dimensional restoration of the entire antenna surface. When the drone After completing the shooting of all identification marks and completely restoring the antenna surface shape, send stop shooting information to the drone through the signal receiving and transmitting module II;

According to the length of the captured laser beam, the relative position between each identification mark is accurately calibrated. This restores the precise three-dimensional structure of the antenna surface, obtains the error distribution of the current parabolic surface, and achieves the purpose of measuring the accuracy of the antenna surface. .

Effects of the present invention:

By completing the real-time measurement of the surface shape of a large parabolic antenna, we can quickly learn the relevant performance parameters of the antenna in the current pitch angle, wind speed, temperature and even sunlight environment. This has a crucial impact on the antenna performance parameters in actual radio astronomy observation data. It can improve the quality of observation data; at the same time, if combined with the active surface control system, a quasi-real-time closed-loop control system can be formed, which can truly optimize the performance of the antenna at any pitch angle;

For this reason, the computer vision technology proposed by this invention using drones to quickly shoot provides a new idea for the accuracy measurement of large domestic parabolic antenna panels. At the same time, the project also hopes to promote the results to future antenna measurement systems and improve data quality.

Description of the drawings

Figure 1 is a schematic structural diagram of the device of the present invention;

Figure 2 is a schematic diagram of the antenna surface error;

Figure 3 shows the antenna panel number distribution;

Figure 4 is a schematic diagram of the marked picture;

Figure 5 is a schematic diagram of the labeled image.

Detailed ways

As shown in Figure 1-5, the UAV-based large parabolic antenna surface accuracy testing system consists of two parts: the detection part and the data processing part, which are connected in sequence:

The detection part consists of a multi-axis UAV equipped with a camera, an identification mark affixed to the surface of the antenna, a laser calibration reference subsystem and a signal receiving and transmitting module I;

A. The multi-copter UAV and its mounted camera constitute the shooting subsystem. The multi-copter is a six-copter, and then the camera is suspended. A memory card is integrated in the camera to back up and store the captured pictures;

At the same time, the UAV carries a signal receiving and transmitting module I, which is connected to the camera and the flight control circuit of the UAV. This module uses wireless frequencies such as wifi, 3G or 4G to transmit and receive; it is used to modulate the images captured by the camera. The wireless signal is transmitted to the ground data processing part, and at the same time, the ground data processing part receives the stop shooting command. After shooting all the marks, the ground data processing part sends a stop flight command, and the drone stops flying;

B. The position identification mark is pasted on the surface of the main reflective surface of the antenna. It is composed of highly photosensitive materials and is easy to be recognized by the camera. Each mark has two different identification features, which can be made of photosensitive materials printed on the mark. The number represents the number of the antenna panel (Figure 4);

It can also be composed of different shape combinations; for example: the identification mark is designed to be circular, and a color is assigned to the panel target point of each loop of the antenna; a color is assigned to the panel of each sector of the antenna, and in a certain The target point on the panel at the intersection of the ring channel and a certain sector is composed of semicircles of two colors (Figure 5);

C. The laser calibration reference subsystem includes a laser transmitter and a laser absorber;

The laser calibration reference subsystem is fixed at a conspicuous place on the antenna surface. It is suitable for drones to capture the laser beam at any angle. The distance between the laser emitted by the laser transmitter and the laser absorber can be accurately measured. This is The entire measurement provides a baseline.

The data processing part includes signal receiving and transmitting module II and computer.

The signal receiving and transmitting module II uses wireless frequencies such as wifi, 3G or 4G to receive and transmit. It mainly receives image signals from the drone and transmits flight control commands, and sends them to the drone after the drone has photographed all the signs;

The computer uses conventional data processing software to identify each image sent by the drone, find the location of the antenna where the image is located, and perform virtual three-dimensional positioning to achieve a three-dimensional restoration of the entire antenna surface. When the drone completes After all marked areas are photographed and the antenna surface shape can be completely restored, a stop shooting message is sent to the drone through the data transmitting and receiving module;

According to the length of the photographed laser beam, the relative position between each position mark is accurately calibrated. This restores the precise three-dimensional structure of the antenna surface, obtains the error distribution of the current parabolic surface, and achieves the purpose of measuring the accuracy of the antenna surface. .

During the test, it is assumed that a parabolic reflector antenna with a radius a and a focal length f is used. Since the reflector is located in the far area of the feed source, the electromagnetic waves emitted by the feed source reach the aperture surface through the reflector. Usually the surface error is not very large. It can be considered that the impact of the surface error on the amplitude of the electromagnetic field on the aperture surface can be ignored, and it will only cause the phase error of the electromagnetic field on the aperture surface. The expression forms of surface error include axial error, radial error and normal error. Here, axial error is used to represent surface error information. The dotted line in Figure 2 is the position of the reflective surface after deformation. According to the geometric relationship in the figure, the optical path difference is:

The optical path difference caused by the presence of Δz when the electromagnetic wave is reflected by the reflector is:

Δδ＝Δz(1+cosξ)＝2Δzcos ² (ξ/2) (1)

The phase error caused by Δz for different electromagnetic waves (λ) is:

Δz includes systematic errors and random errors:

Δz＝Δz _r +Δz _s (3)

So

Random errors can be determined by the manufacturing accuracy of the panel, usually represented by manufacturing tolerances, and need to be processed through probabilistic methods; systematic errors require finite element analysis of the antenna structure to determine its deformation information Δz;

Assume that the ideal design shape of the reflecting surface antenna is: z ₀ = f ₀ (ρ', φ'); under the influence of various loads (wind, gravity, etc.), the actual form of the reflecting surface is:

z(ρ',φ')＝z ₀ +Δz _s =f ₀ (ρ',φ')+Δz _s (ρ',φ') (6)

According to the structural stiffness equation:

[K]{δ}＝{p} (7)

Where [K] is the structural stiffness matrix, {δ} is the node displacement vector, and {p} is the load vector.

Solving the equation, we can get the node displacement vector of the reflecting surface:

{δ _i }={Δρ _i ', Δφ _i ', Δz _i '} (8)

The node coordinate information after the antenna surface deformation is:

{ρ _i ',φ _i ',z _0i }+{δ _i }＝{ρ _i '+Δρ _i ',φ _i '+Δφ _i ',z _0i +Δz _0i } (9)

Through interpolation or fitting methods, the shape z(ρ',φ') of the deformed reflective surface can be given; using equation (5), the axial error Δz _s (ρ',φ') of the deformed reflective surface can be determined ');

According to the calculation formula of the antenna pattern:

E(θ,φ)＝∫∫E ₀ (ρ',φ')e ^{jkρ'sinθcos(φ-φ')} ρ'dρ'dφ' (10)

According to the aperture field method, the far-field pattern function of the deformed reflector antenna is:

in: is the phase error on the aperture surface A, and E ₀ (ρ',φ') is the field distribution function on the aperture surface A; since random errors need to be processed through the probability method; therefore, the aperture surface is meshed using As shown in Figure 3, it is divided into N rings along the radial direction, and each ring is divided into K _n grids. It is assumed that there is a random phase error in each grid/> are the same and equal to the phase error at the center point in the grid area;

The discretization of the equation is expressed as (random error effect):

The far-field potential generated by a single panel can be expressed as (system error effect):

The antenna radiation power pattern written in discrete form is:

The average radiated power pattern is:

Assume that the random errors between the grids are independent of each other, and the random errors on each ring satisfy the normal distribution with a mean of 0 and a standard deviation of ε _n ε _m , then we get:

The relationship between the root mean square and standard deviation of panel manufacturing errors is:

((εrms) _n is the surface manufacturing error of the nth ring) (17)

Simplifying the formula of the average radiated power of the antenna, the error model that affects the average radiated power pattern of the antenna when the surface error contains systematic and random errors is as follows:

Through the mathematical model of the impact of surface errors on the electrical performance of the reflector antenna, the impact of random errors and systematic errors on the electrical performance when they exist alone is analyzed.

In actual measurement, labels are first pasted on the surface of the parabolic antenna. Generally speaking, a 40-meter diameter parabolic antenna requires about 1,400 labels, and they are evenly pasted on the surface of the antenna;

Then set the position of the laser transceiver, and fix the transceiver at a conspicuous place of the antenna. Generally, it is fixed on a support pole with different antenna feed sources. The distance is preferably 3-5 meters. The ruler distance is obtained and then transmitted to the processing computer;

Again, when the weather permits, use the drone to take pictures of all the signs and the length of the laser ruler. Pay attention to the reflection of the antenna and sunlight, and try not to be polluted by sunlight;

While shooting, the captured images are transmitted to the computer through the signal receiving and transmitting module I carried by the drone;

The image processing software of AICON's photogrammetry system is used in the computer to splice the images captured by the drone, and at the same time complete the comparison with the length of the laser ruler. After the splicing is completed, a stop shooting command is sent to the drone, and the drone is Can be recovered.

Specific operations: 1. Paste the identification mark; 2. Coordinate positioning, use the identification mark pasted on the flange on the antenna center body to determine the reference coordinates; 3. Use the professional photogrammetry system to collect data and virtual three-dimensional positioning of each point of the target to achieve Three-dimensional reconstruction of the entire antenna surface.

Claims (1)

1. A large-scale parabolic antenna surface accuracy testing system based on a drone, which is characterized by: consisting of a detection part and a data processing part connected in sequence; The detection part includes a drone equipped with a camera, an identification mark, a laser calibration reference subsystem, and a signal receiving and transmitting module I. The identification mark and laser calibration reference subsystem are set on the surface of a large parabolic antenna. The camera communicates with the signal receiver through a memory card. The transmitting module I is connected; the flight control circuit of the drone is connected to the signal receiving and transmitting module I; The data processing part includes the signal receiving and transmitting module II and the computer. The computer is connected to the signal receiving and transmitting module I through the signal receiving and transmitting module II; The laser calibration reference subsystem includes a laser transmitter and a laser absorber; the identification marks have different identification characteristics, and each pair is different; The UAV obtains the position information of different identification marks on the antenna surface through the camera. At the same time, the laser beam emitted by the laser calibration reference subsystem is used as a reference signal to obtain the current surface shape of the antenna through three-dimensional reconstruction and infer the surface shape accuracy of the paraboloid. , to achieve surface accuracy measurement.