Calibration method, apparatus and system, device, medium, and product

Abstract

A calibration method, apparatus and system, a device, a medium, and a product, relating to the technical fields of optics and robots. The calibration method comprises: mounting an image acquisition device at a distal end of a first robot, and mounting a calibration device at a distal end of a second robot; obtaining motion trajectories of the first robot and the second robot, the motion trajectories comprising N predetermined measurement points; controlling the first robot and the second robot to move to a first predetermined pose relative to a first predetermined measurement point, and using the image acquisition device to acquire a first calibration image of the calibration device; according to the first calibration image, obtaining, at the first predetermined measurement point, a first relative pose between the calibration device and the image acquisition device; controlling, according to the first relative pose, the first robot and / or the second robot to make adjustments until the first relative pose between the calibration device and the image acquisition device meets an alignment condition at the first predetermined measurement point; recording a corrected pose of the first robot and the second robot at the first predetermined measurement point; and controlling the first robot and the second robot to move to a second predetermined pose relative to a second predetermined measurement point until N corrected poses of the N predetermined measurement points in the motion trajectories are obtained.

Description

Calibration methods, apparatus, systems, equipment, media and products Technical Field

[0001] This disclosure relates to the field of calibration and testing technology, and more specifically, to a calibration method, calibration apparatus, calibration system, electronic device, computer-readable storage medium, and computer program product. Background Technology

[0002] Precisely aligning two or more robots in free space without any external reference is a challenging task. It typically relies on expensive equipment such as laser trackers, but the results are inconsistent. A particularly difficult aspect when using laser trackers is ensuring simultaneous alignment of the tilt axis (attitude) while achieving precise alignment in the lateral plane (position). Therefore, laser trackers are more often used for calibrating / aligning robot bases, where the position and attitude accuracy of the end effector axis often depends on assumptions. Summary of the Invention

[0003] This disclosure provides a calibration method, comprising: installing an image acquisition device at the end effector of a first robot and a calibration device at the end effector of a second robot; obtaining motion trajectories of the first and second robots, the motion trajectories corresponding to N predetermined detection points, where N is an integer greater than 1, and the N predetermined detection points include a first predetermined detection point and a second predetermined detection point; controlling the first and second robots to move to a first predetermined pose relative to the first predetermined detection point, and using the image acquisition device to acquire a first calibration image of the calibration device; obtaining a first relative pose between the calibration device and the image acquisition device at the first predetermined detection point based on the first calibration image; controlling the first and / or second robots to adjust according to the first relative pose until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point; recording a corrected pose between the first and second robots relative to the first predetermined detection point; and controlling the first and second robots to move to a second predetermined pose relative to the second predetermined detection point until N corrected poses of the N predetermined detection points are obtained.

[0004] This disclosure provides a calibration method, comprising: obtaining motion trajectories of a first robot and a second robot, the motion trajectories including N predetermined detection points, where N is an integer greater than 1, and the N predetermined detection points including a first predetermined detection point and a second predetermined detection point; controlling the first robot with an image acquisition device mounted on its end effector and the second robot with a calibration device mounted on its end effector to move to a first predetermined pose relative to the first predetermined detection point, and using the image acquisition device to acquire a first calibration image of the calibration device; obtaining a first relative pose between the calibration device and the image acquisition device at the first predetermined detection point based on the first calibration image; controlling the first robot and / or the second robot to adjust according to the first relative pose until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point; recording a corrected pose between the first robot and the second robot relative to the first predetermined detection point; and controlling the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point until N corrected poses of the N predetermined detection points are obtained.

[0005] This disclosure provides a calibration device, comprising: an acquisition unit, configured to acquire motion trajectories of a first robot and a second robot, the motion trajectories including N predetermined detection points, where N is an integer greater than 1, and the N predetermined detection points including a first predetermined detection point and a second predetermined detection point; a control unit, configured to control the first robot with an image acquisition device mounted on its end effector and the second robot with a calibration device mounted on its end effector to move to a first predetermined pose relative to the first predetermined detection point, and to acquire a first calibration image of the calibration device using the image acquisition device; a processing unit, configured to obtain a first relative pose between the calibration device and the image acquisition device at the first predetermined detection point based on the first calibration image; the control unit is further configured to control the first robot and / or the second robot to adjust according to the first relative pose until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point; the processing unit is further configured to record a corrected pose between the first robot and the second robot relative to the first predetermined detection point; the control unit is further configured to control the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point until N corrected poses of the N predetermined detection points are obtained.

[0006] This disclosure provides a calibration system, including: a first robot and a second robot; an image acquisition device installed at the end of the first robot, and a calibration device installed at the end of the second robot; a controller, the controller being configured to obtain the motion trajectories of the first robot and the second robot, the motion trajectories including N predetermined detection points, where N is an integer greater than 1, and the N predetermined detection points including a first predetermined detection point and a second predetermined detection point; the controller being further configured to control the first robot with the image acquisition device installed at its end and the second robot with the calibration device installed at its end to move to a first predetermined pose relative to the first predetermined detection point, and to acquire a first calibration image of the calibration device using the image acquisition device; based on the first calibration image, to obtain a first relative pose between the calibration device and the image acquisition device at the first predetermined detection point; based on the first relative pose, to control the first robot and / or the second robot to adjust until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point; to record a corrected pose between the first robot and the second robot relative to the first predetermined detection point; and to control the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point until N corrected poses of the N predetermined detection points are obtained.

[0007] This disclosure provides an electronic device, including: one or more processors; and a memory configured to store one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the method described in any embodiment of this disclosure.

[0008] This disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when run on a computer, causes the computer to execute the methods described in any embodiment of this disclosure.

[0009] This disclosure provides a computer program product, including a computer program that, when executed by a computer, implements the methods described in this disclosure. Attached Figure Description

[0010] Figure 1 is a flowchart of a calibration method according to an exemplary embodiment of the present disclosure.

[0011] Figure 2 is a schematic diagram of the structure of an exemplary dual-robot system of this disclosure.

[0012] Figure 3 is a schematic diagram of the structure of another exemplary dual-robot system of this disclosure.

[0013] Figure 4 is a schematic diagram of the structure of a calibration system according to an exemplary embodiment of the present disclosure.

[0014] Figure 5 is a schematic diagram of the structure of a calibration system according to another exemplary embodiment of the present disclosure.

[0015] Figures 6 to 8 are schematic diagrams showing the alignment of an image acquisition device and a calibration device according to an exemplary embodiment of the present disclosure.

[0016] Figure 9 is an exploded view of an image acquisition device according to an exemplary embodiment of the present disclosure.

[0017] Figure 10 is an exploded view of a calibration device according to an exemplary embodiment of the present disclosure.

[0018] Figure 11 is a schematic diagram of the architecture of an radome testing system according to an exemplary embodiment of the present disclosure.

[0019] Figures 12 and 13 are schematic diagrams illustrating the use of a radome testing system to test a radome according to an exemplary embodiment of this disclosure.

[0020] Figures 14 to 17 are schematic diagrams illustrating the alignment of two exemplary test antennas according to this disclosure.

[0021] Figure 18 is an exploded view of a test antenna according to an exemplary embodiment of the present disclosure.

[0022] Figure 19 is a structural schematic diagram of a single-speaker mode of an exemplary embodiment of the present disclosure.

[0023] Figure 20 is a schematic diagram of the dual-speaker mode of an exemplary embodiment of the present disclosure.

[0024] Figure 21 is a schematic diagram of the structure of a multi-speaker mode of an exemplary embodiment of the present disclosure.

[0025] Figure 22 is a waveform diagram showing the alignment effect before and after calibration in an exemplary embodiment of the present disclosure.

[0026] Figure 23 is a schematic diagram of an elevation / azimuth radar turntable according to an exemplary embodiment of the present disclosure.

[0027] Figure 24 is a schematic diagram of the transmission efficiency test results of an radome according to an exemplary embodiment of the present disclosure.

[0028] Figure 25 is a schematic diagram of the architecture of an antenna radome testing system according to another exemplary embodiment of this disclosure.

[0029] Figure 26 is a flowchart of a calibration method according to another exemplary embodiment of this disclosure.

[0030] Figure 27 is a schematic diagram of the structure of an exemplary multi-robot system of this disclosure.

[0031] Figure 28 is a schematic diagram of the structure of another exemplary multi-robot system of this disclosure.

[0032] Figure 29 is a schematic diagram of the structure of an exemplary calibration device of this disclosure. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of this disclosure more apparent, exemplary embodiments according to this disclosure will now be described in detail with reference to the accompanying drawings. In the drawings, the same reference numerals denote the same components throughout. It should be understood that the embodiments described herein are merely illustrative and should not be construed as limiting the scope of this disclosure.

[0034] In this disclosure, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0035] In this disclosure, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; the term "multiple" refers to two or more unless otherwise expressly defined. The terms "install," "connect," "link," and "fix" should be interpreted broadly. For example, "connect" can be a fixed connection, a detachable connection, or an integral connection; "link" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this disclosure according to the specific circumstances.

[0036] In the description of this specification, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to a specific feature, structure, material, or characteristic described in connection with that embodiment or example, which is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer 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.

[0037] The above description is merely a preferred embodiment of this disclosure and is not intended to limit this disclosure. Various modifications and variations can be made to this disclosure by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

[0038] As shown in Figure 1, the calibration method provided in this embodiment includes the following steps. It should be noted that the method provided in this embodiment can be applied to dual-robot calibration or multi-robot (three or more) robot calibration. The first robot and the second robot described below can be two robots in a dual-robot system, or any two robots in a multi-robot system.

[0039] This disclosure does not limit the specific mechanical structure of the robot; it can be any device capable of moving and performing tasks in three-dimensional space. In some embodiments, it can be a device with environmental perception and autonomous movement capabilities. The robot can include, but is not limited to, industrial robotic arms, mobile robots, and biomimetic robots. The robot may include a base, a multi-degree-of-freedom robotic arm mounted on the base, and a controller for controlling the trajectory of the robotic arm's end effector (such as the tool described below).

[0040] In S110, an image acquisition device is installed at the end of the first robot, and a calibration device is installed at the end of the second robot.

[0041] In this embodiment of the disclosure, the robot's end effector (including a first robot end effector, a second robot end effector, and in other embodiments, in a multi-(three or more) robot system, a third robot end effector, etc.) refers to the mechanical interface of the robot's last joint. It can be a flange, a fixed, physical component belonging to the robot body, and serves as a base for connecting tools. Its function is to mount tools, such as image acquisition equipment, calibration equipment, test antennas, metal plates, welding torches, grippers, spray guns, sanders, etc.

[0042] In some embodiments, the first robot and the second robot work collaboratively. Both the first robot and the second robot include an Axis 6 / Final axis, which refers to the last rotary joint of the robot's kinematic chain and its driven flange, directly connected to the tool or load (such as a radar horn antenna), controlling the spatial attitude orientation of the tool or load (especially the roll around the Z-axis), and enabling precise angular adjustments of the tool around its own axis.

[0043] In this embodiment, the Tool Centre Point (TCP) is a virtual point defined on the tool mounted on the robot. It is not a physical component, but rather the origin of a defined and measured coordinate system. It represents the tool's working point or point of action. This embodiment controls the robot's motion to ensure that the TCP of the tool mounted on the robot's end effector reaches a designated position and maintains the correct posture.

[0044] For example, when an image acquisition device is mounted on the end effector of a first robot, this device serves as a sensor tool that the robot grasps. The TCP (Center Point of the Image Acquisition Device) of this device (referred to as the first tool center point or first TCP) can be located near the optical center (lens principal point) of the device, from which light rays follow the perspective model of the device. Knowing the pixel coordinates of a point in space within the image acquired by the device, a line of sight originating from the TCP can be constructed. In other embodiments, for ease of measurement, the TCP of the image acquisition device can be defined at a corner of the device's housing or at the center of its mounting surface; however, this requires additional conversion during visual computation. When the first robot moves the image acquisition device to a specific viewing angle (e.g., perpendicular to the surface of the object being measured), it controls the position and orientation (or simply pose) of the TCP of the image acquisition device.

[0045] The image acquisition device in this disclosure is a device capable of converting optical information in physical space into digital or analog signals that can be understood by a processor or controller, and may include a camera, a webcam, a scanning device, etc. In the following embodiments, a camera is used as an example.

[0046] The calibration device in this embodiment is used to provide a spatial reference for an image acquisition device. Its surface or interior contains one or more identifiable feature patterns with known geometric dimensions and distributions, including checkerboard patterns, dot arrays, Charuco plates, etc. Charuco plates will be used as examples below.

[0047] For example, when a calibration device is installed at the end effector of the second robot, the calibration device is an observed passive object; it is not a tool itself. For ease of calibration and measurement, a reference coordinate system similar to a TCP (called the second tool center point or second TCP) is defined for it, called the calibration plate coordinate system or workpiece coordinate system. This TCP can be defined at a fixed feature point on the calibration plate, or at a feature point on its own pattern, serving as the origin of the entire plate coordinate system. It is the spatial reference point during calibration and measurement. For example, for a checkerboard pattern, the TCP can be defined at the first interior corner point in the upper left corner (world coordinates (0,0,0)). For a dot array, the TCP can be defined at the center dot or a corner point. The Z-axis is perpendicular to the calibration plate plane and points outwards, while the X and Y axes are along the row and column directions of the calibration plate.

[0048] When the calibration device uses a Charuco board, a unique 3D world coordinate is assigned to each corner point on the board when generating the Charuco board pattern. The first interior corner point is defined as (0,0,0). The coordinates of subsequent corner points are incremented by the grid rows and columns based on this. Using this point as the TCP (Coordinate Connectivity) of the calibration device ensures seamless integration between the TCP definition and the visual inspection output, requiring no additional coordinate conversion and avoiding the introduction of human error. In the Charuco board's coordinate system, the origin is the first interior corner point. The X-axis runs along the first row of the board (horizontally) and points to the second interior corner point. The Y-axis runs along the first column of the board (vertically) and points to the first interior corner point of the next row. The Z-axis is perpendicular to the board surface and points outward (towards the direction of the image acquisition device), defining the "front" of the board.

[0049] For example, when a first test device is mounted on the end effector of a first robot, if the first test device is a test antenna, such as a horn antenna, the TCP (Cyclic Coordinate System) of the first test device can be defined at the geometric center of the radiating aperture plane of the first test device, and the Z-axis of the TCP coordinate system coincides with the normal direction of the aperture plane, pointing towards the principal radiation direction of the antenna. The position of the TCP can be equivalently adjusted or calibrated according to the phase center of the test antenna. In some embodiments, the TCP of the first test device can be defined at the phase center of the test antenna. The phase center is the virtual point where the spherical wavefront of the electromagnetic wave radiated by the test antenna is emitted when it is equivalent to a spherical wave in the far field. For a well-designed horn antenna, its phase center is located inside the horn or near the aperture plane.

[0050] For example, when a second test device is mounted on the end effector of a second robot, if the second test device is a test antenna, such as a horn antenna, the TCP (Cyclic Coordinate System) of the second test device can be defined at the geometric center of the radiating aperture plane of the second test device, and the Z-axis of the TCP coordinate system coincides with the normal direction of the aperture plane, pointing towards the principal radiation direction of the antenna. The position of the TCP can be equivalently adjusted or calibrated according to the phase center of the test antenna. In some embodiments, the TCP of the second test device can be defined at the phase center of the test antenna. The phase center is the virtual point where the spherical wavefront of the electromagnetic wave radiated by the test antenna is emitted when it is equivalent to a spherical wave in the far field. For a well-designed horn antenna, its phase center is located inside the horn or near the aperture plane.

[0051] The calibration method provided in this disclosure can achieve TCP alignment.

[0052] In S120, the motion trajectories of the first robot and the second robot are obtained. The motion trajectories correspond to N predetermined detection points, where N is an integer greater than 1. The N predetermined detection points include the first predetermined detection point and the second predetermined detection point.

[0053] In this embodiment, N predetermined detection points are pre-set according to the task or operation to be completed. For example, for an antenna radome to be tested, these N predetermined detection points are distributed on the surface of the radome. After determining these N predetermined detection points, their normal directions relative to the antenna radome, and the detection sequence, the motion trajectories of the first robot and the second robot can be planned accordingly. This ensures that during detection, when the first robot and the second robot are performing a task at a predetermined detection point, the first testing device held at the end of the first robot and the second testing device held at the end of the second robot are located on opposite sides of the predetermined detection point, with the distance between them satisfying a pre-set condition, and their postures also satisfying pre-set conditions. For example, the distance between the first testing device and the second testing device remains constant, and the line connecting their axes is perpendicular to the normal direction of any predetermined detection point on the antenna radome. To achieve this ideal alignment condition as much as possible, the first robot and the second robot are first calibrated using a camera and calibration equipment. The motion trajectories of the first robot and the second robot acquired during the calibration process are pre-set according to the detection task.

[0054] In S130, the first robot and the second robot are controlled to move to a first predetermined pose relative to the first predetermined detection point, and the first calibration image of the calibration device is acquired using the image acquisition device.

[0055] In the motion trajectories of the first and second robots, predetermined poses are pre-set for each predetermined detection point. These predetermined poses are the ideal positions and postures that the first and second robots should move to during the planning process. When the first and second robots are in their respective predetermined poses, a first calibration image is obtained by capturing a calibration device mounted on the end effector of the second robot using a camera mounted on the end effector of the first robot. The motion trajectory includes the first predetermined poses.

[0056] In S140, based on the first calibration image, a first relative pose between the calibration device and the image acquisition device is obtained at the first predetermined detection point.

[0057] Taking the Charuco plate as an example of calibration equipment, each corner point on the Charuco plate has a precisely known 3D coordinate (X, Y, Z) in the plate coordinate system (world coordinate system). board ,Y board(Z = 0) (because the corner points are all on a plane). Simultaneously, the first calibration image captured by the camera can detect the 2D pixel coordinates (u, v) corresponding to each corner point. Based on this, the angular deviation between the Z-axis of the calibration device and the Z-axis of the image acquisition device, the angular deviation between the X-axis of the calibration device and the X-axis of the image acquisition device, the angular deviation between the Y-axis of the calibration device and the Y-axis of the image acquisition device, the positional deviation between the origin of the calibration device and the origin of the image acquisition device, and the distance deviation between the calibration device and the image acquisition device can be calculated, thereby obtaining the first relative pose between the calibration device and the image acquisition device.

[0058] In S150, based on the first relative pose, the first robot and / or the second robot are controlled to adjust until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point.

[0059] Taking the test radome as an example, the system pre-stores an ideal pose, meaning that under ideal conditions, the calibration device and the image acquisition device are aligned, and an ideal calibration image is acquired. By comparing the first relative pose with the ideal relative pose, it is determined whether the deviation between the two meets the alignment conditions. If the alignment conditions are not met, the first robot and / or the second robot can be controlled to adjust their poses, and then the calibration device is photographed again to obtain a new first calibration image. Based on the new first calibration image, a new first relative pose between the calibration device and the image acquisition device is calculated. The new first relative pose is compared with the ideal relative pose. If the alignment conditions are still not met, the poses of the first robot and / or the second robot are adjusted until the alignment conditions are met.

[0060] In a dual-robot alignment system, the ideal pose is defined as follows: when the two robots are perfectly aligned, the relative pose between the camera and the calibration board is known (e.g., they are perfectly aligned, and the measured distance is a fixed value). This pose is the ideal pose. The Charuco board is photographed by the camera, and the current pose, relative pose, or actual pose is calculated. Position deviation = Current pose (X,Y,Z) - Ideal pose (X,Y,Z). Attitude deviation = Current pose (roll, pitch, yaw) - Ideal pose (roll, pitch, yaw). The calculated 6-dimensional pose deviation is sent to the robot controller, instructing it to fine-tune for precise alignment. By repeating this process multiple times throughout the workspace, the system can build an error mapping table, thereby achieving high-precision alignment across the entire workspace.

[0061] In S160, the corrected poses of the first robot and the second robot relative to the first predetermined detection point are recorded.

[0062] The corrected pose recorded here refers to the pose adjusted by the first robot and / or the second robot relative to the first predetermined pose at the first predetermined detection point, so that the first relative pose between the calibration device and the image acquisition device meets the alignment conditions. For example, if the alignment conditions are met only after multiple adjustments to the poses of the first robot and / or the second robot, then the corrected pose refers to the deviation of the first robot and / or the second robot relative to the first predetermined pose under the alignment conditions, including positional deviation and / or posture deviation.

[0063] In S170, the first robot and the second robot are controlled to move in a second predetermined pose relative to the second predetermined detection point until N corrected poses of N predetermined detection points are obtained.

[0064] For example, the motion trajectory includes the second predetermined pose. When the first robot and the second robot each have the second predetermined pose, a second calibration image is obtained by capturing a calibration device mounted on the end effector of the second robot using a camera mounted on the end effector of the first robot. The second relative pose between the calibration device and the image acquisition device can be calculated based on the second calibration image. By comparing the second relative pose with the ideal relative pose, it is determined whether the deviation between the two meets the alignment conditions. If the alignment conditions are not met, the first robot and / or the second robot can be controlled to adjust their poses, and then the calibration device can be captured again by the camera to obtain a new second calibration image. A new second relative pose between the calibration device and the image acquisition device is then calculated based on the new second calibration image. The new second relative pose is compared with the ideal relative pose. If the alignment conditions are still not met, the poses of the first robot and / or the second robot are continued to be adjusted until the alignment conditions are met. The corrected pose of the second predetermined detection point is recorded, which refers to the pose adjusted by the first robot and / or the second robot relative to the second predetermined pose of the second predetermined detection point, so that the second relative pose between the calibration device and the image acquisition device meets the alignment conditions. By analogy, N corrected poses for N predetermined detection points can be obtained.

[0065] The calibration method provided in this disclosure improves the pose accuracy of the robots by installing an image acquisition device at the end of a first robot and a calibration device at the end of a second robot, and then having the first robot acquire corresponding calibration images at various predetermined detection points.

[0066] For example, the motion trajectory is on a curved surface; wherein, the method further includes: determining a first normal line on the curved surface at the first predetermined detection point; when the angle between the axis (Z-axis) of the image acquisition device and the axis (Z-axis) of the calibration device and the first normal line is both less than a first angle threshold, the angle between the first horizontal axis (e.g., X-axis or Y-axis) of the image acquisition device and the first horizontal axis (e.g., X-axis or Y-axis) of the calibration device is less than a second angle threshold, and the deviation between the origin of the image acquisition device and the origin of the calibration device in the first horizontal axis and second horizontal axis plane is less than a predetermined threshold, then it is determined that at the first predetermined detection point, the first relative pose between the calibration device and the image acquisition device meets the alignment condition.

[0067] In this embodiment, the values ​​of the first angle threshold, the second angle threshold, and the predetermined threshold can be set according to the required alignment accuracy, and this disclosure does not limit their values. The determination of alignment conditions for other predetermined detection points can refer to the determination of the first predetermined detection point, and will not be elaborated here. Through the calibration method provided in this embodiment, the alignment of the first tool center point and the second tool center point is achieved.

[0068] Exemplarily, the method provided in this disclosure includes: obtaining a first transformation matrix T1 from the image acquisition device to its first tool center point and optical correction parameters of the image acquisition device; and obtaining a second transformation matrix T2 from the calibration device to its second tool center point. The first relative pose is obtained based on the first calibration image, the first transformation matrix T1, the optical correction parameters, and the second transformation matrix T2.

[0069] For example, M valid Charuco corner points can be extracted from the first calibration image, i.e., the 2D pixel coordinates p of the M corner points in the first calibration image are known. p i xe l M is a positive integer greater than 1. Optical correction parameters can include the camera intrinsic matrix K and the distortion coefficients dist. For any corner point on the calibration equipment, the plate coordinates P of the corner point are transformed using the second transformation matrix T2. board Transform to the calibration device's TCP coordinate system: P B_TCP =T2×P board Using the first relative pose X (i.e., Transform it to the camera TCP coordinate system: P C_TCP =X×P B_TCP Using the known inverse matrix T1 of the first transformation matrix T1 -1 Transform the TCP coordinates of the calibration equipment to the camera's TCP coordinate system: P cam =T1-1 ×P C_TCP Using the camera intrinsic parameter matrix K, the 3D camera coordinates are projected onto the 2D pixel plane: p proj =K×P cam P here cam For normalized camera coordinates, then based on p... proj and p pixel Construct a reprojection error function to calculate the first relative pose X (i.e., ).

[0070] For example, obtaining the first transformation matrix from the image acquisition device to its first tool center point and the optical correction parameters of the image acquisition device includes: controlling the second robot to move to a first position to keep the calibration device fixed; controlling the first robot to move to multiple different first preset poses, and controlling the image acquisition device to acquire the first image of the calibration device in different first preset poses respectively; and obtaining the first transformation matrix and the optical correction parameters based on the first images in different first preset poses.

[0071] For example, obtaining a second transformation matrix from the calibration device to its second tool center point includes: controlling the first robot to move to a second position to keep the image acquisition device fixed; controlling the second robot to move to multiple different second preset poses, and controlling the image acquisition device to acquire second images of the calibration device in different second preset poses respectively; and obtaining the second transformation matrix based on the second images in different second preset poses.

[0072] The calibration method provided in this disclosure can achieve precise six-degree-of-freedom alignment of two robots. The aligned dual-robot or multi-robot system can then be used to align two test antennas. This method is used to compensate for deviations caused by the motion system or system deformation caused by prolonged motion, either after the test antennas are installed or at maintenance cycle points. Before calibration begins, the radome / antenna cover is removed, and the RF (radio frequency) antennas (assuming both the first and second test devices are RF antennas) are removed from the two robotic arms (i.e., the robotic arms of the first robot and the second robot) and optical alignment devices (including image acquisition devices and calibration devices) are installed respectively. Exemplarily, a calibration plate / calibration device is installed on the inner robotic arm (assuming the second robot is installed inside the radome), and a camera is installed on the outer robotic arm (assuming the first robot is installed outside the radome).

[0073] For example, the calibration steps include:

[0074] The first step is to move the internal robotic arm to a fixed position (referred to as the first position for distinction), that is, fix the calibration plate, control the external robotic arm to take pictures in multiple positions (i.e. multiple different first preset poses), realize eye-on-hand calibration, accurately calculate the TCP transformation matrix from the camera to the external robotic arm (i.e. the first transformation matrix T1) and obtain the camera optical correction parameters.

[0075] In this embodiment, camera optical correction parameters refer to the camera's intrinsic parameters and distortion coefficients. These parameters are used to correct optical distortion caused by the camera lens and establish an accurate mapping between image pixel coordinates and 3D (three-dimensional) world coordinates. Intrinsic parameters, such as focal length, principal point, and tilt coefficient, describe the camera's optical characteristics and imaging geometry. Distortion coefficients, such as radial and tangential distortion, are used to correct distortions in the image caused by the lens shape. These parameters are obtained during camera calibration to improve the accuracy of visual measurements and ensure that feature points in the image accurately correspond to their real-world locations.

[0076] For example, a calibration plate (such as a Charuco plate or checkerboard) is fixed in a certain position; an external robotic arm (carrying a camera) is controlled to move to multiple different poses (i.e., different positions and postures, referred to as the first preset poses), and images of the calibration plate (i.e., "first images") are captured from different angles. For example, at least three first images of non-coplanar poses are acquired; for example, 10-15 or more first images can be acquired to cover the entire field of view, and the calibration plate is presented in various tilt and rotation states in the first images to ensure the robustness of the calibration. For each first image, an image processing algorithm (e.g., the PnP algorithm) is used to calculate the transformation from the calibration plate to the camera coordinate system. Simultaneously, the transformation from the external robot's TCP to the base coordinate system is recorded. The first transformation matrix from the camera coordinate system to the external robot's TCP coordinate system is determined according to the hand-eye calibration equation.

[0077] For example, feature points (such as corner points) on the calibration board are detected. For the Charuco board, checkerboard corner points and ArUco-marked corner points are detected simultaneously. The image coordinates (pixel coordinates) and corresponding world coordinates (based on the known physical dimensions of the calibration board) of these feature points are recorded. Using the feature point data from all the first images, optimization calculations are performed using a camera calibration algorithm. Camera optical correction parameters are estimated by minimizing the reprojection error (i.e., the difference between the projected points and the actual detected points).

[0078] The second step is for the external robot (i.e., the first robot) to move to a fixed position (for distinction, this is called the second position; the first and second positions can be the same or different), i.e., the camera is fixed, and the internal robotic arm (i.e., the second robot) is controlled to move to multiple positions (i.e., the second preset pose) so that the first robot can take multiple second images, thereby achieving eye-to-hand calibration and accurately obtaining the TCP transformation matrix from the calibration plate to the internal robotic arm (i.e., the second transformation matrix T2).

[0079] Specifically, second images are obtained under multiple different poses. For each pose, the transformation from the calibration board to the camera coordinate system is calculated using the PnP algorithm, while the transformation from the internal robot TCP to the base coordinate system is recorded. Then, the second transformation matrix is ​​obtained according to the calibration equation. The base coordinate system refers to the fixed reference coordinate system of each robot, which can be located at the geometric center of the robot base.

[0080] The radar antenna and test antenna can be precisely aligned with the mounting bracket through mechanical planes and positioning features, while the optical system (including the camera and calibration plate) and the mounting bracket (including the first mounting bracket and the second mounting bracket) are aligned through the first and second steps mentioned above.

[0081] It should be noted that there is no order between the first and second steps mentioned above; you can perform the first step first or the second step first.

[0082] The third step involves controlling the inner and outer robots to move to the corresponding predetermined detection points based on the existing motion trajectory. Using a camera and calibration board, the relative pose of the calibration board in the camera coordinate system is calculated (e.g., the first relative pose, the second relative pose, etc. mentioned above). The inner (i.e., the second robot) or outer robot is controlled to make small movements through the inverse matrix of the corresponding relative pose. Through multiple rounds of small adjustments, the alignment result is made less than the threshold, that is, the origin deviation between the image acquisition device and the calibration device is less than the predetermined threshold, and the axis angle deviation (including the angle deviation between the axis of the image acquisition device and the calibration device and the normal of the corresponding predetermined detection point, and the angle between the first horizontal axis of the image acquisition device and the calibration device) is less than the first angle threshold and the second angle threshold. The current corrected position of the inner and outer robots (i.e., the corresponding corrected pose) is recorded, and then they move to the next measurement point / predetermined detection point, such as the second predetermined detection point mentioned above.

[0083] For example, the outer surface of the radome / antenna is a low-curvature convex surface. When planning the predetermined detection points, the normal (Z-axis direction) is determined at each predetermined detection point through plane fitting. During the calibration process, when the inner and outer robots carrying the camera and calibration plate move to the corrected detection position, such that the angle between the axis of the camera and calibration plate and the normal of the curved surface of the radome / antenna is less than 0.1 degrees, the X-axis angle between the camera and calibration plate is less than 0.2 degrees, and the deviation of the origin of the camera and calibration plate in the XY plane is less than 0.5 mm, it is considered that the camera and calibration plate are precisely aligned.

[0084] Specifically, for each predetermined detection point, the actual pose of the internal robot TCP relative to the external robot TCP (including the aforementioned first relative pose and second relative pose, etc.) is calculated. The actual pose is compared with the ideal pose to obtain the current translational and rotational deviations. For example, a PID (Proportional-Integral-Derivative) control strategy can be used to adjust the pose of the first or second robot, but this disclosure is not limited to this.

[0085] For example, the motion trajectories of the first and second robots can be adjusted according to all the corrected positions / corrected poses to achieve test antenna measurement alignment at each predetermined detection point.

[0086] For example, during the testing process, when the inner and outer robots carry the two test antennas to the corrected detection position, such that the angle between the axis of the two test antennas and the normal of the curved surface of the radome / antenna radome is less than 0.1 degrees, the X-axis angle between the two test antennas is less than 0.2 degrees, and the deviation of the origin of the two test antennas in the XY plane is less than 0.5 mm, it is considered that the two test antennas are precisely aligned.

[0087] In this embodiment, the first and second preset poses are pre-set to ensure the systematic and comprehensive nature of data acquisition, avoid deviations caused by subjective position selection, and ensure that the relative distance between the calibration board and the camera is relatively short and that there are multiple rotation angles to observe the calibration board from different perspectives, covering as many rotational degrees of freedom (pitch, yaw, roll) as possible. For example, 20 first preset poses are set in the first step, and 20 second preset poses are set in the second step, providing sufficient data to ensure calibration accuracy.

[0088] For example, the calibration device includes a checkerboard-coded hybrid marking calibration pattern, meaning the calibration device can employ a Charuco plate.

[0089] In this embodiment of the disclosure, when the calibration device uses a Charuco plate, hand-eye calibration or standardization can be performed based on the Charuco plate. Hand-eye calibration based on the Charuco plate is used to accurately determine the spatial transformation relationship between the camera coordinate system and the robot end effector (tool) coordinate system (TCP). The Charuco plate provides sub-pixel level corner detection accuracy, and combined with the robustness of ArUco markers and the high-precision corner detection of the checkerboard pattern, effective calibration can still be performed even if some markers are obscured. The entire process can be fully automated, and the calibration results are stable and reliable, making it suitable for industrial applications. The calibration method provided in this embodiment of the disclosure can be implemented using an optical system equipped with a Charuco plate.

[0090] For example, as shown in Figures 4 and 5, the calibration device 7 includes a calibration plate pattern and a disk. The calibration plate pattern is drawn on a disk, which is mounted on the end effector axis or end effector 21 of a 6-axis collaborative robot (second robot 2). An image acquisition device (e.g., a camera) 8 is mounted on the end effector axis or end effector 11 of another collaborative robot (first robot 1) to image the calibration plate pattern. Exemplarily, the calibration device 7 and the image acquisition device 8 constitute an optical target system equipped with a Charuco calibration plate. Charuco is a calibration plate that incorporates checkerboard and ArUco markings for high-precision camera calibration and pose estimation.

[0091] For example, the first robot end effector includes a first mounting bracket, and the second robot end effector includes a second mounting bracket; wherein, mounting an image acquisition device on the first robot end effector and mounting a calibration device on the second robot end effector includes: removing the antenna cover between the first robot and the second robot; mounting the image acquisition device on the first mounting bracket via a first fixing adapter, and mounting the calibration device on the second mounting bracket via a second fixing adapter.

[0092] For example, the camera and calibration board are respectively fixed to the mounting bracket of the corresponding horn antenna via a first fixing adapter and a second fixing adapter to ensure alignment between the optical equipment and the horn antenna. This embodiment of the present disclosure ensures reference uniformity between the RF measurement system and the optical alignment system through a precise mechanical interface, thereby achieving high-precision, repeatable alignment. One end of the first and second fixing adapters has a mounting interface identical to that of the horn antenna (i.e., an interface that mates with the positioning post), while the other end is used to mount the camera or calibration board. The pose of the optical center of the first and second fixing adapters is precisely calibrated and set to match the pose of the phase center of the horn antenna. When the first and second fixing adapters are in place, the position and orientation of the camera or calibration board in space have a known and fixed transformation relationship with the position and orientation of the horn antenna during installation, making the entire alignment and testing process efficient and reliable.

[0093] In this embodiment of the disclosure, when testing the radome, a first testing device can be mounted on a first mounting bracket on a first robot, and a second testing device can be mounted on a second mounting bracket on a second robot. During calibration, an image acquisition device is mounted on the end effector of the first robot via the same first mounting bracket, and a calibration device is mounted on the end effector of the second robot via the same second mounting bracket. Therefore, after TCP alignment of the first and second robots is achieved through the image acquisition device and the calibration device, automatic alignment of the first and second testing devices can be automatically achieved during testing.

[0094] This disclosure provides a novel method for precise alignment of the center point (TCP) of a dual-robot tool, optimizing the alignment accuracy of a pair of robots or collaborative robots working together. It requires high accuracy in both translational (position) and rotational (attitude) coordinate systems. Simultaneously considering three translational degrees of freedom and three rotational degrees of freedom, it achieves high alignment accuracy in all six degrees of freedom (6DoF). The relative pose of the camera and the calibration board is determined based on calibration images (including a first calibration image and a second calibration image) captured by the camera. This pose is compared with the ideal or expected pose to obtain the alignment error (including translational and / or rotational deviations). This error is then used to obtain the corrected pose / correction amount for each predetermined position within the working range of the two robots (referring to all key points the robots need to reach during calibration, corresponding to predetermined detection points). The corrected poses are then used to correct the predetermined poses in the motion trajectory (including the aforementioned first and second predetermined poses) to determine the optimal spatial pose for each robot axis, encompassing optimization of translation and rotation.

[0095] The calibration method provided in this disclosure solves the alignment problem in six-dimensional space. It optimizes the alignment of a pair of robots or collaborative robots that work together with high precision in a rotational and translational coordinate system. It considers three lateral axes and three rotational axes. The three lateral axes refer to forward / backward, left / right, and up / down movement (X, Y, Z) in three-dimensional space. The three rotational axes refer to rotation (yaw, pitch, roll) around these three directions X, Y, and Z. The end effectors of the two robots are aligned not only in position but also in attitude, requiring precise control in a total of six dimensions. Any minute deviation will lead to measurement errors. In the testing system, when both the first and second robots are equipped with radar horn antennas as the first and second testing devices, the above calibration ensures that the beam is perfectly perpendicular to the predetermined detection point on the radome.

[0096] For example, the robot TCP is the origin of a coordinate system relative to the robot's end effector (sixth axis or flange). When no physical tools are installed on the robot, the center point of the flange at the robot's end effector (i.e., the rotation center of the sixth axis) can be defined as a default TCP; this point is the geometric endpoint of the robot arm itself. When a tool is installed on the robot's end effector, the effective working point of this tool is defined as the new TCP. When the robot controls the new TCP to move in space, the distance from the robot's TCP to the new TCP is automatically calculated and compensated.

[0097] For example, obtaining the motion trajectories of the first robot and the second robot includes: obtaining a three-dimensional model of the radome; planning a scanning path for the radome based on the three-dimensional model, and determining N predetermined detection points on the scanning path and the normal direction of each predetermined detection point; generating the arrival pose of the first tool center point of the first robot and the second tool center point of the second robot based on the normal direction and the set measurement distance between the first test device and the second test device, so as to generate the motion trajectory.

[0098] In this embodiment of the disclosure, a testing system is provided for testing the performance of an radome. This testing system is implemented using a first testing device mounted on a first robot and a second testing device mounted on a second robot. In this scenario, the motion trajectories of the first and second robots can be planned and generated based on the 3D model of the radome. This ensures that when the first and second robots reach a predetermined pose according to the motion trajectory, N predetermined detection points on the pre-set scanning path of the radome can be accurately detected. For example, a constant measurement distance is maintained between the first and second testing devices, and the line connecting the axes of the first and second testing devices is parallel to the normal direction of each predetermined detection point.

[0099] For example, an antenna radome / radome test bench is provided, on which a first robot and a second robot operate. Through calibration, the attitude and normal of the first TCP of the first robot and the second TCP of the second robot are perfectly matched, thereby ensuring the precise alignment of the polarized radar signal beams of the first test device and the second test device when testing the antenna radome.

[0100] For example, during calibration, the camera captures images of the calibration board pattern. Any translational or rotational deviations are identified by the camera as alignment errors, and a correction is calculated accordingly. Subsequently, two collaborative robots move to each predetermined position (corresponding to predetermined detection points) within their working range and record image data for all positions (including the first and second calibration images). This image data is used to determine the optimal spatial pose (i.e., the optimal rotation and translation state) of each robot axis, thereby achieving optimal alignment of the first testing device (e.g., a radar horn antenna) and the second testing device (also a radar horn antenna) in the radome testing scenario.

[0101] The calibration method provided in this disclosure, on the one hand, enables a high-quality optical camera to capture extremely small perpendicularity (normal) deviations through a special feature pattern (such as a Charuco calibration plate) on the alignment disk / calibration device, resulting in high-precision optical alignment performance. The system can simultaneously calibrate the robot's end effector axis in both translational (lateral) and rotational (attitude) dimensions, providing multi-dimensional calibration capabilities. Simultaneously, the use of a small-diameter alignment disk (e.g., 100 mm) significantly reduces the risk of collisions during the movement of two or more robots, resulting in higher safety.

[0102] As shown in Figure 2, the dual-robot system includes a first robot 1 and a second robot 2. The first robot 1 includes an end effector 11, a first robotic arm 12, and a first base 13, which is attached to the motion mechanism, i.e., the linear track 4. The second robot 2 includes an end effector 21, a second robotic arm 22, and a second base 23, which is attached to a second turntable 3, which has a turntable center 31.

[0103] Figure 3 illustrates another dual-robot system, which differs from the embodiment in Figure 2 in that the first base 13 of the first robot 1 is mounted on the first turntable 6 via an L-shaped bracket 5. The second base 23 of the second robot 2 is mounted on the turntable center 31 of the second turntable 3.

[0104] In Figures 2 and 3, x and z represent one axis in the ideal xyz direction of the robot, while x' and z' represent axes with actual small-angle deflections. When the robot's fixed point is far away or unstable, it introduces errors into the robot's base coordinates. When there are two robots installed at distant locations in the system, or when any robot is attached to an additional motion mechanism (turntable or linear track), the two robots need to cooperate during operation. In this case, if there is a calibration deviation (or insufficient calibration accuracy) in the robot's base coordinate calibration, or if the base flange deforms during robot movement, it will introduce deviations into the robot's end-effector center (TCP) motion or positioning.

[0105] For example, in a test radome / radome scenario, the system uses two robots. The smaller robot (the robot inside the radome, i.e., the second robot 2) is positioned at the center 31 of the turntable 3 on the second turntable 3, while the larger robot (the robot outside the radome, i.e., the first robot 1) is fixed to the first turntable / first turntable 6 via an L-shaped bracket 5. The entire system must simultaneously control both turntables (i.e., the first turntable 6 and the second turntable 3). The two robots, with a total of 14 axes, ensure that the two measuring / test antennas in the first and second test devices reach their respective detection positions to detect the corresponding predetermined detection points. Because the external robot (the first robot 1) bracket (i.e., the L-shaped bracket 5) is fixed to the first turntable 6, and the bracket arm is relatively long (the lever arm from fulcrum to fulcrum is approximately 1.5 meters), there is a deviation in the path points of the external robot.

[0106] The aforementioned 14 axes include 6 joint axes of the first robot (base rotation, shoulder swing, elbow extension, wrist rotation, etc.), 6 joint axes of the second robot, one rotation axis of the first turntable for driving the first robot and its L-shaped support fixed on it to rotate as a whole, and one rotation axis of the second turntable for driving the second robot placed at the center of its turntable to rotate.

[0107] Figures 2 and 3 illustrate common situations that can cause the robot base to twist. Both Figures 2 and 3 show instability. In Figure 2, the robot is fixed to a turntable or linear track, causing deflection. In Figure 3, the external robotic arm (i.e., the first robotic arm) is fixed to the turntable and L-shaped support.

[0108] The embodiments disclosed herein compensate for the TCP bias introduced by the above two types of problems using optical methods.

[0109] For example, as shown in Figures 4 and 5, based on Figures 2 and 3 respectively, during system calibration, an optical calibration disk (i.e., calibration device 7) is installed at the end 22 of the inner robot (second robot 2), and an image acquisition device (e.g., a camera) 8 is installed at the end 11 of the outer robot (first robot 1). By calibrating and compensating for the deviation of the path points of the outer robot, the transmission main lobe direction / polarization direction axis of the two microwave antennas (test antennas) in the first and second test devices is aligned, and the distance between the two test antennas is compensated.

[0110] Figures 6, 7, and 8 show the aligned image acquisition device 8 and calibration device 7 from different perspectives. Figures 6, 7, and 8 illustrate the alignment process at a single detection point after the optical systems are mounted on the antenna mounting bracket using their respective fixed adapters. After alignment, the Z-axis of the calibration device 7 coincides with and is opposite to the Z-axis of the image acquisition device 8; their Y-axis is parallel and opposite to each other; and their X-axis is parallel and opposite to each other. This represents the ideal situation; however, slight errors are permissible in real-world scenarios.

[0111] As shown in Figure 7, the front of the calibration device 7 includes a Charuco calibration pattern. The Charuco calibration pattern combines a checkerboard pattern and ArUco codes. The checkerboard pattern provides high-precision corner points for positioning, while the ArUco codes provide a unique ID for identification, ensuring stable operation even when the pattern is partially obscured.

[0112] As shown in Figures 6 to 8, the image acquisition device 8 is installed at the end 11 of the first robot 1. The calibration device 7 is installed at the end 21 of the second robot 2. The image acquisition device 8 and the calibration device 7 are precisely installed so that they are in the same direction and position as the two radar horns in the first and second test devices installed on the two collaborative robots during the test. That is, the relative position of the image acquisition device 8 and the calibration device 7 is exactly the same as the relative position of the two radar horn antennas in the first and second test devices. In this way, through calibration, the image acquisition device 8 is aligned with the calibration device 7, and the two radar horn antennas are also aligned during the test.

[0113] As shown in Figure 9, the image acquisition device 8 can be mounted on the shield base 10 of the first test device and the flange of the end 11 of the first robot 1 via the camera bracket 9.

[0114] As shown in Figure 10, the calibration device 7 can be installed on the shield base 40 of the second test device and the flange of the end 21 of the second robot 2.

[0115] It should be noted that the calibration device 7 can be installed in various ways. For example, the calibration device 7 can be directly installed on the microwave antenna bracket used to install the second test device, or directly installed on the TCP of the second robot 2, or the existing tool head (such as the first test device or the second test device) can be replaced by a tool changer (tool quick change device or automatic tool exchange system).

[0116] The method and system provided in this disclosure improve upon traditional alignment methods that rely on "external testing equipment" such as laser trackers. The advantage of the calibration method in this disclosure is that it directly aligns the robot. Therefore, the system and method provided in this disclosure have stronger self-calibration capabilities because they directly measure the characteristic under test. The sensors used (camera and alignment disk) are not part of the measurement system, thus providing an independent alignment means for the end effector axis of the robot or collaborative robot. Furthermore, the process is fully automated and requires virtually no human intervention.

[0117] In related technologies, methods for testing the transmittance of radome / antenna components include: one is using a long-distance laboratory equipped with a weather radar as a detector (test distance required to be 100 meters); the other is using a weather radar as the signal source, with the detector arranged in an arc array. Laboratory testing is not only expensive and time-consuming, but its accuracy is also often controversial, especially considering the differences between different testing methods. The procurement cost of radome testing equipment is also extremely high, as it requires the use of a real avionics weather radar as the signal source and must be placed in a microwave-safe electromagnetic shielded room.

[0118] The radome evaluation and transmittance testing system provided in this disclosure is a small-footprint testing device that uses a low-power signal source. Its potential retail price is approximately one-tenth of existing systems, while maintaining comparable measurement accuracy.

[0119] As shown in Figure 11, this disclosure also provides a radome testing system. The radome testing system includes a data processing terminal 1100, a fixture positioner 1110, a first test device positioner (which can be used to position the antenna, and is therefore also called the first antenna positioner, antenna positioner 1) 1120, a first test device 1130 (which includes a horn antenna 1), a radome 1140, a second test device 1150 (which includes a horn antenna 2), a second test device positioner (which can be used to position the antenna, and is therefore also called the second antenna positioner, antenna positioner 2) 1160, a radome positioner 1170, a transceiver (including a transmitter and a receiver) 1180, and positioner control equipment 1190.

[0120] Exemplarily, the radome may include: a skin, which is the main material layer constituting the radome body, and may use materials with low dielectric constant and low loss tangent, such as fiberglass reinforced plastic (FRP), ceramics, composite materials, or special engineering plastics; a sandwich structure, which, for radomes requiring high strength (such as aircraft nose cones), may employ a sandwich structure or a honeycomb structure. This structure achieves lightweighting while ensuring rigidity and strength, and its electromagnetic properties are carefully designed to optimize transmittance. Exemplarily, the radome may also include a rain-resistant / anti-static coating. Exemplarily, the radome may include a metal frame / mounting base, which is connected and fixed to the aircraft fuselage or robot structure via the metal frame. Exemplarily, when operating normally, the radome may be mounted on the nose cone of a commercial aircraft. The radome / radome is located at the very front of the nose and houses a weather radar antenna array.

[0121] For example, the method provided in this disclosure includes: mounting the radome on a rotating device; rotating the radome by the rotating device to detect the transmittance of predetermined detection points in different regions of the radome by the first testing device and the second testing device.

[0122] For example, to test the radome as a whole, several sub-products are designed, manufactured, and assembled to form the final test bench. After the radome is loaded onto the test bench, it is placed on four mounting points for precise positioning. Each mounting point is physically connected to a rotary drive. That is, the radome is positioned on a rotating turntable, with the center of the radome aligned with the central axis of the turntable. This allows the radome to rotate so that all inspection equipment (radar near-field testing and active thermal imaging) can access the entire radome surface.

[0123] For example, the radome testing system also includes a fixture for precisely clamping and securing the radome 1140. Since the proposed solution is designed for use in MRO (Maintenance, Repair, and Overhaul) plants that handle the maintenance of various radome models, the fixture can accommodate different types of radomes.

[0124] In this embodiment, the clamp positioner 1110 is a mechanism that enables the clamp to perform single-axis or multi-axis rotation / tilt movements to change its spatial angle. It is a rotatable display platform that drives the fixed radome 1140 to rotate, oriented different areas of the radome toward the test antenna, so as to accurately detect predetermined detection points on different areas of the radome. For example, this can be achieved using the rotating device 1200 in Figure 12, thereby expanding the workspace and accessibility of the first and second robots, allowing them to cover different areas of the radome without having to move a long robotic arm.

[0125] This embodiment of the invention simplifies the movement trajectories of the first and second robots by rotating the radome body using a rotating device. This allows the robots to move within a single fan-shaped area without needing to go around to the other side of the radome. Combined with the rotation of the rotating device, the entire radome surface can be covered, reducing the robot's movement range, improving speed and accuracy, and lowering programming complexity. Furthermore, the radome positioner 1170 allows for more precise control of the radome 1140's pose.

[0126] The first test device positioner 1120 can be used to precisely control and change the spatial pointing (azimuth and elevation) of the horn antenna 1, for example, by means of the aforementioned first robot 1.

[0127] The second test device positioner 1160 can be used to precisely control and change the spatial pointing (azimuth and elevation) of the horn antenna 2, for example, by means of the aforementioned second robot 2.

[0128] For example, transceiver 1180 may include a VNA (Vector Network Analyzer). Transceiver 1180 transmits a test signal, which is a radio frequency (RF) signal, to first test device 1300. This test signal passes through radome 1140, is received by second test device 1150, and returns to transceiver 1180. Transceiver 1180 sends the received data to data processing terminal 1100. Data processing terminal 1100 processes the data, for example, to obtain the transmittance of the radome. Data processing terminal 1100 may also send control signals to positioner control device 1190 based on the data processing results. Positioner control device 1190 may send control signals to fixture positioner 1110, radome positioner 1170, first test device positioner 1120, and second test device positioner 1160 respectively to control the orientation of the fixture, radome, horn antenna 1, and horn antenna 2.

[0129] For example, the data processing terminal 1100 and the locator control device 1190 can be separate physical hardware or integrated into the same physical hardware, such as both being part of a controller.

[0130] Figure 11 is a schematic diagram of the radio frequency (RF) section of the radome testing system, which can be divided into two main parts: a motion control module that controls the overall mechanical movement of the system, and a signal processing section that controls signal transmission and data post-processing. The entire system can be driven by a data processing terminal 1100. The motion control module is responsible for all physical movement and precisely positions the first test device 1130 and the second test device 1150. The motion control module includes a positioner control device 1190, which can be a dedicated computer. It receives high-level instructions (e.g., "move to coordinates X, Y, Z") from the data processing terminal 1100 and converts these instructions into precise electrical signals to drive the movement of the fixture positioner 1110, radome positioner 1170, first test device positioner 1120, and second test device positioner 1160, thereby driving the robot arms of the first and second robots and any other positioning mechanisms (such as linear guides). The robot arms of the first and second robots are respectively equipped with horn antenna 1 and horn antenna 2, responsible for performing specific movement tasks.

[0131] For example, data processing terminal 1100 issues the command: "Go to point A for measurement." Positioner control device 1190 calculates the rotation amount of each joint of the first and second robots required to go to point A. Positioner control device 1190 drives the robot arms of the first and second robots to precisely move horn antenna 1 and horn antenna 2 to the predetermined position to measure point A of the radome, ensuring the correct orientation (the line connecting the axes of horn antenna 1 and horn antenna 2 is perpendicular to the radome wall). Positioner control device 1190 feeds back to data processing terminal 1100: "In position." Data processing terminal 1100 controls transceiver 1180 to generate, transmit, and receive radio frequency test signals to complete the measurement.

[0132] For example, transceiver 1180 generates a 9.375 GHz radio frequency signal, which is transmitted to transmitting horn antenna 1 via a cable. Horn antenna 2 receives the signal passing through the radome / antenna radome from horn antenna 1 and transmits it back to VNA via another cable. The internal circuitry of VNA can accurately compare the amplitude (intensity) and phase (time delay) of the transmitted and received test signals. Transceiver 1180 can calculate all the information of the transmittance (from amplitude) and phase offset (from phase) of the measurement point / predetermined detection point based on the comparison results. For example, data processing terminal 110 can store or run path planning algorithms to determine the scanned grid point sequence, i.e., determine the scan path of the radome and the predetermined detection points on it, and generate the corresponding motion trajectories of the first and second robots.

[0133] For example, the data processing terminal 1110 can also receive data returned by the VAN, such as the transmittance of each predetermined detection point, and store the received data together with the spatial coordinates of the predetermined detection point.

[0134] For example, the data processing terminal 1110 can also perform near-field data to far-field data transformation, such as converting near-field data into far-field data.

[0135] For example, the data processing terminal 1110 can also run a weighted average to synthesize thousands of near-field data points into a far-field performance index (such as far-field transmittance) that simulates a real antenna.

[0136] For example, the data processing terminal 1110 can also generate color transmittance heat maps, etc.

[0137] Exemplarily, the data processing terminal 1110 can also provide an operator with a graphical interface for starting tests, monitoring progress, and viewing results. For example, the operator clicks "Start Test" on the software of the data processing terminal 1110. The data processing terminal 1110 commands the motion control module to move the horn antenna 1 and horn antenna 2 to the pose corresponding to the first measurement point. After positioning, the data processing terminal commands the VNA to perform measurements and acquire data. The data processing terminal saves the measured data along with the current spatial coordinates. The above measurement steps are repeated until all grid points have been scanned. After scanning is completed, the data processing terminal 1110 automatically executes the post-processing algorithm to generate the final evaluation report and visualization charts. This embodiment of the disclosure achieves modularity of the automated testing system by clearly separating motion control and signal processing into two major modules, all driven by a unified central data processing terminal, facilitating development and maintenance. Mechanical motion and signal acquisition are perfectly coordinated and highly synchronized. The process is fully automated, requiring no manual intervention. Central control ensures process consistency and data integrity.

[0138] For example, the method provided in this disclosure includes: removing the image acquisition device on the first mounting bracket and removing the calibration device on the second mounting bracket; installing a first testing device on the first mounting bracket and a second testing device on the second mounting bracket; moving the first robot and the second robot according to the motion trajectory and the corrected pose to align the first testing device and the second testing device at predetermined detection points, and detecting the radome at N predetermined detection points using the first testing device and the second testing device.

[0139] In this embodiment, the installation of the calibration plate and the camera is precisely designed to ensure that they are in the exact same normal position and orientation relative to the horn antenna mounted on their respective collaborative robots. After calibration, the radar waves can be vertically incident on the surface of the radome / antenna radome (normal), achieving precise alignment of the polarized radar signal beam.

[0140] In this embodiment, the dual-robot tool center point alignment method based on Charuco patterns can achieve sub-pixel level alignment accuracy, reaching sub-millimeter and sub-arcsecond levels, meeting the stringent requirements of radio frequency testing. Compared to laser trackers that cost hundreds of thousands of dollars, an industrial camera and a printed Charuco board are extremely inexpensive. This embodiment solves the high-cost, high-precision industrial robot calibration problem with a low-cost, highly intelligent vision solution. It simulates and replaces the alignment requirements of horn antennas, transferring the precision of optical alignment to radio frequency alignment through precise mechanical installation. The entire process can be completed automatically by software, requiring no professional personnel, and can be executed quickly before each test, ensuring the system is always in optimal condition and achieving routine high-precision calibration. A "deviation-compensation" database of thousands of points is established throughout the robot's workspace. Subsequently, when the robot performs radome testing, the control system queries this database and automatically compensates for each predetermined pose, thereby achieving continuous high-precision alignment in actual operation.

[0141] For example, the method provided in this disclosure includes: when the radome is not installed between the first robot and the second robot, acquiring a first signal power of a test signal at a predetermined detection point by aligning the first test device and the second test device; installing the radome between the first robot and the second robot, acquiring a second signal power of a test signal passing through the radome at the predetermined detection point by aligning the first test device and the second test device; and obtaining the transmittance of the radome at the predetermined detection point based on the first signal power and the second signal power.

[0142] For example, as shown in Figure 12, without an antenna radome installed between the first robot 1 and the second robot 2, the first signal power is obtained through the first test device 1130 and the second test device 1150. As shown in Figure 13, with an antenna radome installed between the first robot 1 and the second robot 2, the second signal power is obtained through the first test device 1130 and the second test device 1150.

[0143] In this embodiment of the disclosure, as shown in FIG12, the first signal power of the test signal is measured in free space (a reference measurement environment without a radome / antenna). As shown in FIG13, after installing the radome / antenna, the measurement is repeated under the exact same position and conditions to obtain the second signal power of the test signal. Transmittance = (Second signal power with radome / First signal power without radome) × 100%.

[0144] For example, the test signal radiated by the VNA is fed to horn antenna 1 and horn antenna 2 via a high-quality coaxial cable (with low loss, excellent shielding, and stable phase characteristics). The radio frequency signal generated by the VNA is transmitted from port 1 to the transmitting antenna (e.g., horn antenna 1), and the signal captured from the receiving antenna (e.g., horn antenna 2) is transmitted back to port 2 of the VNA. The VNA can generate a clean and frequency-tunable radio frequency signal as a test signal and can accurately measure the amplitude and phase of the input test signal. By comparing the transmitted and received test signals, various characteristics of the radome under test can be analyzed.

[0145] For example, the cable is guided along the robotic arm. In RF testing, a moving cable can generate phase noise and signal fluctuations due to bending and swaying. Securing it along the arm minimizes changes in cable shape, ensuring stable signal transmission and resulting in high-precision, repeatable data.

[0146] For example, the electromagnetic waves radiated by horn antenna 1 and horn antenna 2 are polarized light, and they are fronted by plano-convex focusing lenses to collimate the radiation as much as possible. Horn antennas themselves have a certain directionality, but the beams they emit still have a certain divergence angle. The role of the lenses is to further collimate the divergent beams into a nearly parallel beam.

[0147] For example, the lens is made of polymeric materials such as polytetrafluoroethylene (PTFE), and the horn antenna (referred to as the horn) is made of aluminum. At microwave frequencies, certain polymers such as PTFE have very low dielectric loss, resulting in minimal energy absorption as electromagnetic waves pass through. Aluminum is an ideal material for manufacturing microwave antennas because it is both a good conductor (ensuring efficient electromagnetic wave radiation) and lightweight and easy to process.

[0148] For example, a horn antenna is a waveguide antenna shaped like a horn. This structure can effectively convert electromagnetic waves propagating in the waveguide into directional beams propagating in free space, and vice versa.

[0149] For example, a horn antenna is a polarized horn antenna. That is, the electromagnetic waves transmitted or received by the horn antenna have a specific polarization direction (e.g., linear polarization, such as vertical or horizontal polarization). Specifically, the transmission path is from VNA port 1 to a high-quality coaxial cable, then to polarized horn antenna 1 (which converts the electrical signal into an electromagnetic wave), and then through a plano-convex lens to collimate the diverging electromagnetic wave into a narrow beam directed towards the radome under test. The reception path is from the electromagnetic wave reflected or transmitted back from the radome, through a lens to converge the beam, then through horn antenna 2 to convert the electromagnetic wave into an electrical signal, and back to VNA port 2 via a high-quality coaxial cable.

[0150] In testing, the actual radar antenna on the aircraft is not operational, and may not even be inside the radome. This disclosure uses a standard horn antenna instead, because the performance of horn antennas is known, stable, and repeatable. This ensures consistency and impartiality in testing: regardless of which radome is tested, the signal source and receiver remain constant; only the radome itself changes, thus accurately assessing the radome's quality.

[0151] In this embodiment of the invention, through the calibration of the optical system described above, the test signal emitted by the horn antenna 1 can be incident perpendicularly or normally onto the radome during the radome test process. That is, the direction of electromagnetic wave propagation is completely parallel / coincides with the normal direction of the test point (predetermined detection point) on the radome surface, and the beam is directed perpendicularly onto the radome surface.

[0152] In this embodiment, the first robot and the second robot are two collaborative robots. They are high-precision industrial robot arms programmed to work together safely and flexibly. The first and second robots move synchronously, maintaining the test signal incident normally to achieve high-precision measurement. The end caps of the first and second robots are respectively equipped with corresponding horn antennas and lenses, which move synchronously on the complex curved surface of the radome / antenna cover, always ensuring that the beam is perpendicular to the surface of the measured point.

[0153] Figures 14, 15, 16 and 17 are schematic diagrams of the alignment of the two test antennas.

[0154] As shown in Figure 18, the first testing device 1130 includes a shielding cover 1132, a horn antenna 1131 (e.g., horn antenna 1), and a shielding cover base 10. The rear of the shielding cover base 10 is fixed to the flange of the first robot TCP, which is shown in the figure as being fixed to the end 11 of the first robot 1. The second testing device 1150 includes a shielding cover 1152, a horn antenna 1151 (e.g., horn antenna 2), and a shielding cover base 40. The rear of the shielding cover base 40 is fixed to the flange of the second robot TCP, which is shown in the figure as being fixed to the end 21 of the second robot 2.

[0155] In Figures 14 and 15, the first test device 1130 has a shield 1132, and the second test device 1150 has a shield 1152. Figures 16 and 17 are schematic diagrams showing the alignment of the two horn antennas after the shields have been removed.

[0156] Figures 14, 15, 16, and 17 illustrate examples of installing two horn antennas. The main microwave propagation direction is along the Z-axis, while the polarization direction of the horn antennas is along the Y-axis. The origin for aligning the two horn antennas is the center of their flanges. The calibration target is for the two horn antennas to coincide on their Z-axis but in opposite directions (transmit / receive directions); and coincide on their Y-axis but in opposite directions (same polarization).

[0157] This disclosure provides a radar radome evaluation and transmission test system and a novel method for center point alignment of a dual-robot tool. When two robotic arms each hold a horn antenna, the TCP can be defined as the phase center of each horn antenna, i.e., a theoretical point from which electromagnetic waves appear to be emitted. This is the electrical center of the antenna radiation, not its physical center. Through the collaborative work of two independent robotic arms, the TCPs of the two robots establish a precise, stable, and known spatial relationship in three-dimensional space. As shown in Figures 14 to 17, the alignment goal is to ensure that throughout the scanning process, regardless of changes in the surface curvature of the radome, the two horn antennas always satisfy the ideal test geometry: collinear alignment, the line connecting the TCP of the transmitting horn antenna and the TCP of the receiving horn antenna (the Z-axis shown in Figures 14 to 17) must always be perpendicular to the tangent plane of the radome's test point; and distance alignment, the distance between the two TCPs must always remain constant, for example, 96 mm.

[0158] This disclosure provides a novel method for precise TCP alignment of two robots, optimizing and simplifying the alignment process for robots or collaborative robots requiring high levels of cooperation. It ensures a high degree of parallelism and perpendicularity (normal) between the end effectors of the two robots over a wide workspace. This requires precise control of six degrees of freedom (6DoF): three Cartesian translational degrees of freedom (XYZ) and three rotational degrees of freedom—pitch, yaw, and roll. In a specific application of the test bench provided in this disclosure, each of the two collaborative robots holds a polarized radar horn antenna on its sixth axis (end effector axis). Therefore, not only spatial alignment is required, but also precise matching of the polarization direction (polarization axis, e.g., the Y-axis shown in Figures 14 to 17) and tilt angle (tilt axis, i.e., the Z-axis shown in Figures 14 to 17) of the two antennas.

[0159] For example, a high-precision digital 3D model of the radome under test is acquired through 3D laser scanning or photogrammetry. Based on this model, a scanning path is planned, creating a 3D trajectory on the radome's curved surface. The robotic arm has six rotational joints, giving it six degrees of freedom in 3D space: three degrees of freedom (X, Y, Z) for position control, and three more degrees of freedom (rotation, pitch, yaw) for attitude control. Therefore, it can not only move the horn to a point but also adjust its orientation. For each target point on the path, the normal direction (i.e., the vertical direction) of the radome surface at that point is calculated based on the 3D model. Then, it controls the two robotic arms so that the central axes of the transmitting and receiving horns coincide with this normal. This ensures that the radio waves penetrate the radome wall through normal incidence, a standard and repeatable test condition. The system treats the transmitting and receiving horns as a single unit. They move synchronously not only in direction but also in position, always maintaining a preset fixed distance (e.g., 96 mm) between themselves and the radome surface. In a maintenance and inspection scenario, the accuracy of measuring local material properties is far more important than simulating the global antenna pattern. Synchronous movement of both arms ensures normal incidence at every point, thus standardizing all measurements to the same level and making the data comparable.

[0160] Specifically, this alignment goal can be achieved by combining offline programming and real-time control. In the offline planning phase, a precise digital model of the radome can be obtained through 3D scanning. For each point on the scanning path, the surface normal direction is calculated. Then, based on this normal direction and a set constant measurement distance, the precise positions and attitudes that the two robots' TCPs (Transmission Control Points) need to reach are calculated. This generates a synchronized motion trajectory. In the online execution and compensation phase, due to minor positioning errors and gear backlash inherent in the robots themselves, simply executing the preset program is insufficient to achieve metrological accuracy. Therefore, the alignment or calibration process calibrates and verifies the actual TCP positions of the two robots, ensuring that the "theoretical position" calculated in the software matches the "actual position reached" by the robots. Any minor deviation will be measured and compensated during the system calibration phase. Misalignment or inaccurate alignment will directly introduce measurement errors. If the beam is not incident normally but obliquely at an angle onto the radome, it will cause distortion in the transmittance measurement (e.g., lower readings). It will be impossible to distinguish whether signal attenuation is caused by poor material properties or an incorrect measurement angle. If the distance between the two speakers fluctuates during scanning, it alters the antenna's radiation pattern, affecting signal strength and phase, and causing inconsistencies and incomparability in measurement data from different points. Dual-robot tool center-point alignment, through precise robot control and calibration technology, creates and maintains an "idealized," "unchanging" laboratory-grade testing environment on complex 3D surfaces. This results in extremely high accuracy and repeatability of the measurement data. All data points are acquired under the same standard, allowing for fair comparison. This ensures the generation of accurate and reliable compliance reports.

[0161] Through this precise calibration and compensation, the dual-robot system ensures measurement accuracy, with TCP alignment error <0.1mm and angular error <0.1°; guarantees data consistency, ensuring all measurement points are obtained under identical geometric conditions; achieves true verticality, with the beam always directly incident on the radome surface; and maintains a constant measurement distance, such as precisely controlling a 96mm test distance. This enables the near-field testing system to produce results highly correlated with far-field testing, providing the MRO industry with laboratory-level measurement capabilities while maintaining the convenience and economy of a workshop environment.

[0162] Exemplarily, the optical system includes a camera, a calibration plate, and a fixing adapter (referred to as the first fixing adapter and the second fixing adapter, respectively). Exemplarily, as shown in FIG18, assume that the first test device 1130 includes a horn antenna 1131, and the second test device 1150 includes a horn antenna 1151. Horn antennas 1131 and 1151 have 45-degree mounting brackets 1802 and positioning posts 1801 on their respective corresponding robot TCPs to ensure repeatability of the horn antenna installation.

[0163] During the calibration phase, the horn antennas are removed. The first fixed adapter, containing the camera, is mounted onto the mounting bracket of the first robot. The second fixed adapter, containing the calibration plate, is mounted onto the mounting bracket of the second robot. The three-step optical alignment described earlier is performed. At this point, the system is actually simulating and calibrating the relative pose between the two horn antennas. During the testing phase (RF measurement), the first and second fixed adapters are removed. The horn antennas are reinstalled onto their respective mounting brackets. Due to the presence of the positioning posts and precision interfaces, the actual pose of the horn antennas at this point is highly consistent with the simulated pose during optical alignment. This ensures that the reference for RF measurement (horn antenna phase center) and the reference for optical alignment (camera optical center / target center) are unified through a precision mechanical interface. The positioning post design ensures long-term stable operation of the system without the need for complex calibration after each RF antenna installation. A single optical system calibration can be used long-term to guarantee the accuracy of RF measurements.

[0164] The nose radome is an integral part of an aircraft radar system. It serves not only as a protective shield against damage from airborne radar debris such as ice, freezing rain, static electricity, lightning, hail, and bird strikes, but also as an electromagnetic window for transmitting signals. Radar radome performance evaluation includes environmental and electrical tests such as transmission efficiency, sidelobe level, beamwidth, incident reflection, and beam deflection. Related technologies test radome performance under far-field conditions. In a very large microwave anechoic chamber (walls covered with absorbing material to simulate an open environment), the radome and radar antenna are placed at one end, and a precision receiving antenna is placed at the other end, far away (far field). In this embodiment, far field refers to a distance large enough that the electromagnetic waves reaching the receiving antenna can be considered parallel plane waves. This allows for a realistic simulation of radar wave propagation in the air and accurate measurement of overall radiation characteristics such as beamwidth, sidelobes, and pointing. Far-field testing requires a large test site and is very expensive.

[0165] The testing system provided in this disclosure can perform near-field testing. Instead of directly measuring the formed beam at a distance, it uses a probe (which can be a horn antenna, such as horn antenna 1 or 2) to scan thousands of points on a plane closely attached to the radome surface, recording the signal characteristics at each point. All this near-field data is then converted and synthesized to simulate the true performance of the radar antenna in the far field (including indicators such as transmission efficiency). High-precision testing can be performed in a relatively small room, making it suitable for maintenance and verification in a workshop.

[0166] For example, the method provided in this disclosure includes: installing the radome between the first robot and the second robot; acquiring near-field data of a test signal passing through the radome at a predetermined detection point using the aligned first and second test devices; and obtaining far-field data of the test signal based on the near-field data.

[0167] In the near-field region close to the antenna, near-field data (including amplitude and phase) is measured at thousands of points on a closed surface or plane in front of the radome using a precise sampling probe. This measures the amplitude and phase of the test signal before and after passing through the radome. Then, using a mathematical model based on electromagnetic field theory, this near-field data is converted into a far-field radiation pattern, i.e., far-field data. Thousands of points mean a very dense grid, sufficient to capture any minute performance inhomogeneities on the radome (such as repair / reinforcement areas, aging areas, etc.), avoiding defects missed due to insufficient sampling.

[0168] If the scanning surface is a plane located in front of the antenna, a Fast Fourier Transform (FFT) can be performed on the acquired near-field data. This transformation is equivalent to decomposing the measured field distribution into a superposition of countless plane waves propagating in different directions. This "set of plane waves" is called the angular spectrum. A coordinate transformation is performed on the angular spectrum (from the direction cosine coordinate system of the plane waves to the spherical coordinate system for angles), removing the influence of the probe's directivity. The transformed and compensated angular spectrum is directly the mathematical expression of the far-field radiation pattern. Performing an inverse FFT or direct calculation yields the amplitude and phase radiation patterns of the far field. If the scanning surface is a cylindrical surface surrounding the antenna, a cylindrical wave expansion is performed (Fourier transform in the angular dimension, Fourier-Bessel expansion in the height dimension). The measured field is expanded using cylindrical harmonic functions and then extrapolated to the far field. This yields a complete radiation pattern of the antenna within 360° of the horizontal plane and a certain angle in the vertical plane. If the scanning surface is a sphere completely surrounding the antenna, a spherical wave expansion is performed (using spherical harmonic functions and Legendre polynomials). The measured field is expanded using a set of orthogonal spherical wave functions. Once the coefficients are determined, the field at any point in space (including the far field) can be calculated. A complete three-dimensional radiation pattern across the entire antenna space can be obtained.

[0169] When evaluating a radome, near-field measurements are first performed on the test antenna alone to obtain its far-field radiation pattern in free space. Then, the radome is installed between the test antennas, and measurements are taken again on the identical near-field sampling surface to obtain the far-field radiation pattern after the radome is installed. By comparing the free-space far-field radiation pattern with the radome-installed far-field radiation pattern, the performance indicators introduced by the radome can be precisely quantified, such as transmission efficiency, also known as insertion phase delay, which represents the power loss of the signal after passing through the radome. Near-field data can be used to calculate the field distribution on the antenna aperture surface. If an anomaly occurs in the aperture field after the radome is installed, the specific region on the radome causing the problem can be precisely located (e.g., a poor bonding point, uneven thickness), which is impossible with far-field measurements. The controlled indoor environment avoids external environmental interference, resulting in extremely high measurement accuracy. All tests do not radiate sensitive signals, meeting electromagnetic confidentiality requirements.

[0170] For example, the method provided in this disclosure includes: generating and transmitting the test signal to the first test device or the second test device via a vector network analyzer; and receiving the returned test signal from the second test device or the first test device via the vector network analyzer.

[0171] For example, the measurement distance is less than D represents the aperture of the first or second test device, and λ represents the wavelength of the test signal emitted by the first or second test device.

[0172] In electromagnetism, based on the distance between the observation point and the antenna, the antenna's radiation field can be divided into three regions: the reactive near-field region, the radiating near-field region, and the far-field region. The near field refers to the reactive near-field region and the radiating near-field region closest to the antenna. The reactive near-field region (extremely near-field) is the region closest to the antenna. In this region, electromagnetic energy mainly oscillates (stores and exchanges) between the antenna and the surrounding space, rather than radiating outwards. The radiating near-field region (Fresnel zone) is a slightly farther region. Electromagnetic waves begin to radiate primarily outwards, but the wavefront (equiphase surface) is not yet a plane, but a sphere. The amplitude and phase of the field change drastically in space. In contrast, there is the far field. In the far field region, electromagnetic waves can be considered plane waves, with a plane wavefront, and the radiation pattern (i.e., the spatial distribution of signal strength) is stable and no longer changes with distance. Traditional antenna testing is usually conducted in the far field, but this requires a large test distance. Where D is the antenna aperture and λ is the wavelength.

[0173] The testing system provided in this disclosure is a radome evaluation and transmission testing system based on near-field measurement. It scans and measures the antenna system equipped with a radome in the near-field region (e.g., the radiating near-field region). For large antennas (such as airborne fire control radar), achieving far-field conditions requires distances of hundreds or even thousands of meters, making the construction of such test sites extremely difficult and expensive. Near-field measurement can be completed in a microwave anechoic chamber of only a few meters or even smaller. By accurately measuring the amplitude and phase of the electromagnetic field on a plane in the near-field region, transformations (such as "near-field to far-field transformation") can be used to accurately calculate the antenna's complete radiation characteristics (pattern, gain, sidelobes, etc.) in the far field. The near-field data can be used to deduce the field distribution on the antenna aperture surface, which is very effective for diagnosing subtle defects such as phase distortion and aiming errors introduced by the radome. It can be precisely determined which part of the radome is causing the performance degradation.

[0174] In this embodiment, a robot moves a measurement probe (e.g., a first testing device) on a plane in front of the radome to measure the amplitude and phase of electromagnetic waves emitted from the antenna inside the radome (e.g., a second testing device) and transmitted through the radome point by point. A computer then processes this massive amount of near-field data to evaluate indicators such as the radome's transmission efficiency (insertion loss), beam deflection (aiming error), and pattern distortion.

[0175] For example, the measured distance is not less than 3λ.

[0176] The distance between the two speakers in this system should meet the requirement of being less than [missing information]. And not less than 3λ. Assuming the test RF used is 9.375GHz with a wavelength of 32mm, the measurement distance between the horns should be no less than 96mm. Since the core of the design is to create a small-scale system, the distance between the horns is set close to the lower limit. To improve measurement efficiency, the test machine's radome is scanned and a 3D model is reconstructed in advance to facilitate system motion trajectory planning.

[0177] For example, the frequency f of the test signal is between 7.0 and 11.2 GHz.

[0178] For example, the frequency f of the test signal is 9.375 GHz.

[0179] Where λ = c / f, c = 299792458 m / s, and when f = 9.375 GHz, λ = 32 mm.

[0180] The testing methods for radomes in related technologies require far-field testing, necessitating long-distance laboratories (e.g., 100 meters) to simulate the propagation of radar waves in real air (plane waves). Building and operating large microwave anechoic chambers is extremely costly. Logistics, queuing, and testing cycles can last for weeks or even months. Furthermore, differences in testing setups and procedures between different laboratories can lead to inconsistent results. Additionally, these technologies use real airborne weather radars as the emission source. A real weather radar is itself an expensive avionics device. Moreover, to prevent harm to personnel from high-power microwave radiation, it must be placed in a shielded electromagnetic environment, further increasing costs and space requirements. Therefore, most MRO manufacturers must outsource the testing of repair parts, resulting in long turnaround times and uncontrollable costs.

[0181] For example, the first testing device is a test antenna and the second testing device is a metal plate; or, the first testing device is a metal plate and the second testing device is a test antenna; or, both the first testing device and the second testing device are test antennas; or, both the first testing device and the second testing device are antenna arrays, wherein the antenna array includes multiple test antennas.

[0182] For example, the test antenna is a horn antenna.

[0183] In this embodiment of the disclosure, by employing a collaborative robot (cobot) and a radar horn antenna fed by a VNA, the system is compact and can be safely used in any environment.

[0184] For example, the test antenna is a polarized horn antenna.

[0185] For example, to match the radar's design polarization, the horn antenna is fixed in a single polarization state during testing. The horn antenna and VNA can provide an idealized plane wave (collimated by a lens), and the measurement result is the theoretically optimal value.

[0186] For example, the test antenna is a linearly polarized horn antenna.

[0187] A radar horn antenna is a standard gain antenna, belonging to the waveguide aperture extension structure in the microwave band. It directionally radiates electromagnetic wave energy (transmit mode) or directionally receives it (receive mode), achieving free-space waveguide mode conversion and suppressing sidelobes. As a passive sensor, it requires an external signal source (such as a VNA) to operate. Its power is ≤0.1W (safety level), it is low-cost, and can measure parameters such as transmittance. Horn antennas have good directivity, concentrating energy in one direction for transmission or reception from a primary direction. This helps reduce interference from walls in anechoic chambers, improving the signal-to-noise ratio. Horn antennas provide sufficient gain to ensure a strong signal without the excessively narrow beam of high-gain dish antennas, making them suitable for close-range scanning measurements. Their radiation beam (pattern) is highly regular and symmetrical, and can be accurately predicted through theoretical calculations. A single horn antenna can operate over a wide frequency band, allowing a system to cover multiple test frequencies. The transmitting horn antenna has known, stable, and repeatable performance. It is responsible for transmitting a plane wavefront as a reference to the radome. The receiving probe (horn antenna) is a high-precision sampler, small in size and physical dimensions (aperture), small enough to resolve even the most subtle electromagnetic field changes in the radome's near field. Driven by a robot, it moves precisely in two or three dimensions in front of the radome, sampling point by point. In the near-field testing system, a horn antenna with known performance is used as the transmitter, acting as a reliable signal source. Another horn antenna with known performance (usually smaller) is used as the receiving probe, acting as a precision moving measuring instrument. This "dual-horn" configuration, combined with a VNA, constitutes a complete system capable of accurately measuring amplitude and phase, thus providing raw data for subsequent near-field to far-field conversion and radome performance evaluation.

[0188] The test system of this disclosure uses a horn antenna, including a single-horn mode, a dual-horn mode, and a parallel mode of multiple horns, to verify the transmission efficiency (TE) of the radome. This opens a window for the MRO industry to more easily diagnose and repair radomes.

[0189] Figure 19 shows the single-horn mode. The single-horn mode uses a single horn antenna (which can serve as the first test device 1130) for both transmission and reception. This horn antenna is mounted on a precision robot (e.g., the first robot 1). A metal plate (not shown) is placed inside the radome 1140 on a plane in front of the radome 1140. Moving the plate point-by-point, the double-pass transmission of the radio frequency is measured, completing a gridded scan of the entire area and measuring the transmittance and phase shift. The system is simple in structure and relatively low in cost. The double-pass transmission refers to the test signal transmitted by the horn antenna passing through the radome 1140 for the first time (forward transmission), where the signal is almost 100% reflected by the metal plate. The reflected signal passes through the radome 1140 a second time (reverse transmission) and is received by the horn antenna. By comparing the amplitude and phase of the transmitted and received signals, the double-pass transmittance and double-pass phase shift at that point are calculated.

[0190] Far-field measurements, due to their large measurement dimensions and high cost, are more suitable for radome calibration or qualification than for verification after low-cost maintenance. In the MRO industry, the transmission efficiency of the radome after repair is more important than other electrical performance characteristics. In the workshop, as shown in Figure 19, a single-horn mode can be used to measure the radome's transmittance and phase shift. A first test device 1130 (including a horn antenna) can be placed on the outside of the radome 1140 surface, and a metal sheet or plate can be placed inside the radome 1140 as a second test device to measure the dual transmission of radio frequency.

[0191] Figure 20 shows the dual-horn mode. The dual-horn mode uses two horn antennas, serving as the first test device 1130 and the second test device 1150, respectively. This disclosure proposes and designs an automated radome evaluation and transmission test system to help MRO organizations have a small device at their local factory to inspect and evaluate repaired radomes. The system operates in the near-field range, employing a dual-horn measurement mode. Two horns are placed on both sides of the radome wall 1140, one as a transmitting horn and the other as a receiving horn, to measure the radome wall transmittance. The signal is emitted from the transmitting horn, passes through the radome wall in one go, and is then captured by the receiving horn. The signal penetrates only once, more directly and accurately simulating the penetration process of radar waves in real flight, thus resulting in more accurate and reliable measurement results. This solution is an automated control system with two robotic arms. The robotic arms can position with millimeter-level or even higher precision, completely eliminating errors caused by human hand tremors, inconsistent placement angles, and pressure. The robots can tirelessly and rapidly scan along a preset dense grid path, reducing manual work that would otherwise take hours to just minutes. The robotic arm can move flexibly in six dimensions, ensuring that the two horns remain perpendicular to the complex curved surface of the radome and maintain a constant optimal testing distance. This significantly improves efficiency and reliability, and allows for a global view of the radome surface from a three-dimensional perspective based on near-field data. Furthermore, the system integrates the capability to measure the subsurface of the radome's sandwich structure using flash thermal imaging technology.

[0192] Figure 21 illustrates the multi-speaker mode. The multi-speaker parallel mode uses one or two speaker antenna arrays, such as a row of 8 or 16 fixed receiving probes. Data is collected from an entire row in a single mechanical positioning operation by rapidly switching between them via electronic switches. By reducing mechanical movement, test time can be reduced by an order of magnitude compared to the single-speaker mode, making it ideal for production lines or high-throughput MRO (Maintenance, Repair, and Overhaul) workshops.

[0193] As shown in Figure 21, taking the example where both the first test device 1130 and the second test device 1150 employ horn antenna arrays, that is, multiple horn antennas are integrated into one array. Exemplarily, the number of horn antennas in the first test device 1130 and the number of horn antennas in the second test device 1150 can be equal or unequal. Exemplarily, the number of transmitting horn antennas is less than the number of receiving horn antennas. For example, there can be 2 transmitting horn antennas and 16 receiving horn antennas. The system quickly switches between different transmitting horn antennas, while all receiving horn antennas receive data in parallel. Through this switching and parallel reception, the system can acquire a large number of data points across the entire scanning plane extremely quickly.

[0194] For example, the first testing device and the second testing device are used to perform quality verification on the repaired radome.

[0195] For example, the testing system provided in this disclosure can be applied to verify the transmittance of a composite-repaired radome. It can help MRO (Maintenance, Repair, and Overhaul) companies improve turnaround time and reduce total maintenance costs.

[0196] It should be noted that, since the system provided in this embodiment uses a radar horn antenna rather than a real weather radar, it is impossible to measure the radar sidelobes and their deflection caused by the radome. However, given that the target user is the MRO industry, such in-depth performance qualification is not necessary; accurately assessing the maintenance condition is sufficient to meet their needs.

[0197] Figure 22 shows a comparison of the alignment deviation between two uncalibrated horn antennas and the alignment deviation between two calibrated horn antennas. Here, offXY represents the deviation of the two uncalibrated horn antennas along the X and Y axes, offZ represents the deviation along the Z axis, and offXYZ represents the total deviation of the two uncalibrated horn antennas along the X, Y, and Z axes. offXY_new represents the deviation of the two calibrated horn antennas along the X and Y axes, offZ_new represents the deviation along the Z axis, and offXYZ_new represents the total deviation of the two calibrated horn antennas along the X, Y, and Z axes. As can be seen from Figure 22, the alignment deviation is significantly reduced after calibration.

[0198] For example, the method provided in this disclosure includes: according to the M measurement directions of the antenna under test, according to the elevation angle and azimuth angle scanning sequence, or the azimuth angle and elevation angle scanning sequence, selecting N1 far-field data corresponding to the measurement directions from N far-field data, and weighted synthesizing the far-field transmittance in the corresponding measurement directions; wherein, M and N1 are both positive integers greater than or equal to 1 and less than or equal to N.

[0199] For example, the system output is the radar transmittance rating of the radome. This rating is derived by calculating the transmittance values ​​of the radome at 45 different azimuth and elevation angles, which are derived from near-field data collected at approximately 1,000 measurement points on the radome surface.

[0200] Take a real 700mm aperture radar antenna as an example. It has a parabolic surface. When a plane wave (radar echo) reaches this parabola, every point on its surface reflects the electromagnetic wave. All these reflected waves are precisely focused onto a "feed" at the end of the antenna. The total signal ultimately received by the antenna is the sum of the signals received at all points across its entire aperture. However, it is not a simple addition because electromagnetic waves have phase.

[0201] In some embodiments, the transmittance / far-field transmittance of a real weather radar can be estimated by weighted averaging of values ​​within the actual field of view covered by the real weather radar. This involves near-field to far-field conversion and weighted averaging. The system synthesizes a weighted average of thousands of near-field data points (secondary near-field data) based on the aperture field distribution of the real weather radar antenna (e.g., a weighting function with a heavier center and lighter edges), thereby obtaining an equivalent transmittance / far-field transmittance that represents the performance of the entire radar system.

[0202] This embodiment of the disclosure uses dual robots to perform intensive near-field scanning of the radome, acquiring near-field data from thousands of measurement points on the radome surface. Through weighted averaging, these thousands of near-field data points are synthesized into an equivalent far-field transmittance simulating that of a real radar in 45 different directions (45 azimuth / elevation directions). Exemplarily, a transmittance rating conforming to a standard (e.g., Class B) can also be given based on the results from these 45 directions.

[0203] The radome transmission efficiency (TE) / transmittance is the ratio of the signal power received by the horn antenna with a radome to the signal power received in free space. The captured near-field data is converted to far-field data and matched to 45 measurement positions along the radar gimbal. The radar gimbal is a precision mechanical device that supports and drives the rotation of the radar antenna. It allows the antenna to rotate in azimuth and elevation angles, enabling the radar beam to scan the entire sky. A series of representative angle combinations are obtained (e.g., 9 azimuth angles × 5 elevation angles = 45 positions). These 45 positions cover the core airspace where the radar beam most frequently operates.

[0204] Figure 23 shows an elevation-over-azimuth radar gimbal, one of the motion scan sequences. This radome testing system can adapt to both elevation (E) / azimuth (A) and elevation / azimuth (A) scan sequences. Elevation over Azimuth involves setting an azimuth angle first, then scanning the elevation angle across the entire range; after completion, it steps to the next azimuth angle. Azimuth over Elevation involves setting an elevation angle first, then scanning the azimuth angle across the entire range. The testing system provided in this disclosure can flexibly synthesize and output far-field transmittance data at these 45 positions according to either of these two sequences, based on standard requirements or user settings. It does not mechanically rotate 45 times, but rather, at the data processing level, virtually calculates the performance impact of the radome when the radar beam points in these 45 different directions by adjusting the weighted average calculation angle. The system generates a report listing the transmittance at these 45 locations and compares it with the standard requirements to give a final conclusion on whether the repair is qualified.

[0205] For example, an antenna radome was evaluated and its transmission response (TE) was determined to be Class B. In this test, the horn antenna used a radio frequency of 9.375 GHz. The measurement process was pre-programmed and automatically controlled by software. In the TE measurement section, a collaborative robot holding the horn antenna moved to the same position in free space and with the radome, and then installed the radome on a fixed device. According to the pre-programmed motion path plan, the horn antenna scanned the entire electromagnetic window of the radome, sampling approximately 1200 points. By measuring the signal power with and without the radome, the TE result could be calculated from the captured data. When the elevation-to-azimuth scan sequence was converted into far-field data at 45 different azimuth and elevation positions, approximately 182 sampling points were obtained in each direction, covering most of the radome's electromagnetic window.

[0206] Specifically, in the first round (free space) without the radome installed, the robot, holding a horn, scans along a path of 1200 preset points. The signal power measured at this time is the baseline power under ideal, unobstructed conditions. The system accurately records the spatial coordinates of these 1200 points and the corresponding baseline power (first signal power). In the second round, the radome is installed on the adaptive fixture. Two robots, holding horns, completely replicate the path of the 1200 points from the first round and scan again. The signal power measured at this time (second signal power) is the actual power of the signal after passing through the radome. Only by accurately replicating the path can it be ensured that at each sampling point, the power difference at the same location is compared with and without the radome. For each of the 1200 points, the system calculates its single-point transmittance, obtaining 1200 transmittance data covering the electromagnetic window of the radome. When synthesizing the far-field transmittance of the radar beam pointing in a specific direction (e.g., azimuth 10°, elevation 2°), not all 1200 points are used. From these 1200 points, the 182 points that contribute the most to the direction are selected (these points exactly cover the "projection" area of ​​the simulated antenna aperture on the radome). Then, based on the aperture field distribution function of the real antenna, each of these 182 points is assigned a weight (center points have higher weights, edge points have lower weights), and finally a weighted average is calculated. This weighted average is the equivalent far-field transmittance when the radar beam is pointed in this specific direction. The system repeats the above weighted average calculation process for each of the 45 radar gimbal directions specified in the standard. Ultimately, 45 far-field transmittance values ​​are obtained, representing the impact of the radome on signal strength when the radar scans different airspaces. The report lists these 45 results, and usually takes the minimum or average value as the basis for determining the final radome rating (consistent with the commercial result of Grade B).

[0207] For example, the method provided in this disclosure includes: mapping the transmittance of each predetermined detection point to a three-dimensional model of the radome, and obtaining and displaying a transmittance heat map of the radome.

[0208] For example, a transmittance heatmap as shown in Figure 24 can be obtained. It maps the transmittance value of each measurement point onto a 3D model of the radome, using color (e.g., red for low transmittance, blue for high transmittance) to visually display the performance distribution of the entire "electromagnetic window." Maintenance engineers can easily identify areas of weakness and guide precise rework.

[0209] The system binds the transmittance data from approximately 1200 points to the precise three-dimensional coordinates of predetermined detection points on the radome, using different colors to represent different transmittance value ranges. Figure 24 shows the three-dimensional display of the TE (Transmission Efficiency) results of projecting the transmittance of each sampling point / predetermined detection point onto the radome's three-dimensional model. The system synthesizes the 1200 near-field data into equivalent far-field transmittance in 45 far-field directions and uses these far-field results for final classification. The radome tested by this system is classified as Class B, the same as the radome's classification. The figure clearly shows that the TE results between all sampling points range from 32% to almost 100%. Most sampling points are in the range of 88% to 95%, while some areas in the upper right of the radome show relatively low values, meaning that the transmission effect in this area is not as good as in other areas. The sampling points on the lightning protection strip show very low transmission effect, which is reasonable because metal can block most RF signals. Lightning protection strips are metal strips embedded in the surface of the radome. Their function is to guide the huge current from a lightning strike to the aircraft fuselage, thereby protecting the internal radar. The measurement results correctly show that the transmittance of this area is extremely low, proving the physical accuracy and reliability of the measurement system.

[0210] Test methods in related technologies typically only provide an overall "pass / fail" rating or a single grade. In addition to the final transmittance rating, embodiments of this disclosure also provide a color transmittance heatmap, visually presenting the transmittance distribution across the entire radar window area, thus visualizing performance defects and facilitating the location of problem areas. This provides maintenance, repair, and overhaul (MRO) service providers with unique information unavailable with other equipment.

[0211] For example, the method provided in this disclosure further includes: illuminating the radome with a flash device (e.g., a flash lamp); recording the temperature change of the radome to obtain a thermal image of the radome; and detecting the structural integrity of the radome based on the thermal image.

[0212] The internal structure of the radome can be evaluated using flash thermal imaging with two powerful flash lamps and a high-speed thermal imager. As shown in Figure 25, the surface of the radome / antenna radome 1140 is instantaneously heated using two powerful flash lamps 2520 and 2530. Simultaneously, a high-speed thermal imager 2510 records the cooling process of the surface temperature. If there are internal defects (such as delamination, water accumulation, or collapse of the honeycomb core), the heat dissipation rate of that area will differ from that of intact areas, thus appearing as a "hot spot" or "cold spot" anomaly on the thermal image.

[0213] For example, terahertz imaging technology can be used for non-destructive testing of the internal structure of the radome.

[0214] Exemplarily, this disclosure provides a radome / radome evaluation and transmission testing system, mainly comprising three robotic arms (also referred to as three robots). Two robotic arms (e.g., the first and second robots) hold horn antenna 1 and horn antenna 2 respectively, which are precision actuators for performing radio frequency testing. They are responsible for achieving the previously discussed "always vertical" and "constant measurement distance". Through coordinated movement, they can precisely maintain a "test window" on the complex curved surface of the radome, ensuring that each data point / predetermined detection point is acquired under optimal and consistent conditions. Another robotic arm (e.g., the third robot) holds one or more of the components shown in Figure 25, such as flash lamps 2520 and 2530, a thermal imager 2510, a vector network analyzer (VNA), a motion controller, a data processing terminal, and mechanical structures. The flash lamp and thermal imager are used to perform radome structural defect detection, enabling the system to have dual detection capabilities for electrical performance and structural integrity. The other robotic arm can precisely control the distance and angle between the thermal imager and the radome surface, ensuring that the captured images are clear, consistent, and fully cover the area to be inspected. The motion controller receives motion trajectory instructions from the data processing terminal and converts them into real-time electrical signals that drive the movement of each robot joint, ensuring that the three robotic arms can work smoothly, accurately, and synchronously without interfering with or colliding with each other. The data processing terminal runs the software that controls the entire system, including path planning, instrument control, data acquisition, storage, and the final data processing algorithms (such as near-field to far-field transformation, weighted averaging, and generating 3D color cloud map reports).

[0215] Qualified modern radomes are typically classified into Class A (average transmittance exceeding 90%) or Class B (average transmittance exceeding 87%), with average transmittance exceeding 90% or 87%. This transmittance must be verified after repair. A current problem in transmittance measurement is that transmittance testing of the nose radome is a necessary condition for post-repair verification, but current certification or repair testing is conducted in long-distance anechoic chambers, which are rare. For example, for a large airborne radome (e.g., 1 meter in diameter) operating in the X-band (wavelength 0.03 meters), the required far-field test distance is approximately 67 meters. The testing cost is also quite high. Most composite repair shops do not use transmittance testing equipment. Compared to the 1995 RTCA DO-213 standard, the newly updated RTCA DO-213A document provides more critical measurement requirements. For example, the far-field reference distance is from D... 2 / 2λ was upgraded to 2D 2The required distance has increased to four times the original value ( / λ). This change presents a greater challenge and increases testing costs for far-field radome testing suppliers. Testing radomes using a near-field method requires only an anechoic chamber capable of housing the radome and a small collaborative robot. This makes it possible for repair shops to build their own testing systems. These systems can not only accurately calculate transmittance but also reconstruct the radar's far-field pattern. The near-field system can integrate structural scanning capabilities, enabling "one-stop" inspection.

[0216] In the aviation maintenance and overhaul industry, radomes require quality verification after repair before being reinstalled on the aircraft. This involves checking not only the integrity of the radome's honeycomb structure but also verifying whether its transmittance meets standards. However, traditional methods for testing radome transmittance are time-consuming and require far-field testing ranges or anechoic chambers, which repair shops often lack access to expensive, specific transmittance testing equipment. This disclosure presents a small, low-cost near-field testing system compliant with the updated RTCA DO-213A standard. This system can assess both the structural integrity of the radome's interior and verify the transmittance of the repaired radome. The interior of a radome typically has a honeycomb structure, which is both lightweight and robust. After repair, these "cells" need to be inspected for flattening, damage, or water ingress. This system also integrates non-destructive testing techniques such as ultrasonic or terahertz scanning, simultaneously checking the transmittance and scanning the internal honeycomb structure to ensure no hidden damage is present.

[0217] For example, absorbing foam covers all fixtures, supports, and equipment in the work area. In RF testing, one of the biggest sources of interference is multipath reflection—electromagnetic waves reflected back from metal clamps, supports, or walls, interfering with the main signal. These reflections contaminate the data, causing measurement distortion. Covering with absorbing foam, much like in a microwave anechoic chamber, absorbs these stray reflected waves, ensuring that the VNA measures only the signal that directly penetrates the radome, thus improving measurement accuracy and reliability.

[0218] For example, the test antenna is a precisely calibrated linearly polarized antenna, ensuring the stability of relevant signal parameters under test conditions. The precisely calibrated linearly polarized antenna serves as both a transmitting and receiving horn; the performance of the horn antenna itself (such as VSWR and gain) is known and stable, and its systematic errors can be corrected during data processing. Linear polarization clearly defines the polarization mode of the electromagnetic wave (such as vertical or horizontal polarization), which is a prerequisite for maintaining consistency with the actual operating mode of airborne radar. Polarization mismatch will lead to measurement errors.

[0219] For example, the motion of all motion devices is calibrated using a third-party ranging system to ensure that the captured data is attributed to a finite error range of the hardware components. This third-party ranging system calibration is used to calibrate the actual position of all motion devices. The robot arm's own positioning accuracy may have errors at the micrometer level. Using a higher-precision external measurement system (such as a laser tracker) to calibrate the robot's actual motion position quantifies and minimizes the spatial positioning error of the entire system. This means that the system not only knows where the robot "should" be, but also where it "actually" is, ensuring that each data point has accurate spatial coordinates and ultimately limiting the overall error introduced by the hardware to a known, acceptable range.

[0220] The testing system provided in this disclosure solves the problems of positioning and repeatability through robotics, addresses environmental electromagnetic interference through foam absorption, mitigates instrument inherent errors through calibrated antennas, and resolves mechanical motion accuracy issues through third-party calibration. This system can generate near-laboratory-grade, repeatable, and highly reliable test data in an MRO workshop environment.

[0221] Without an internal testing platform, MRO companies must transport the radome to a third-party laboratory for compliance testing after completing repairs. This results in significant time delays: arranging transportation, transit time, waiting in line for testing, obtaining reports, and then returning it to the MRO workshop. The entire process can easily consume weeks or even months. With the testing system provided in this disclosure, the radome can be tested immediately in the workshop after repair. If the initial test fails, engineers can immediately pinpoint the problem, make on-site corrections, and retest within minutes or hours. This avoids the repeated cycles of "repair-external testing-failure-rework," each potentially wasting weeks. Without waiting for external reports, decisions to continue subsequent processes or release the aircraft can be made quickly based on real-time data. Aircraft downtime is significantly reduced, allowing MRO companies to serve more customers, improving asset turnover, and reducing total costs.

[0222] The radome testing system provided in this disclosure uses a collaborative robot and a horn antenna fed by a VNA (Vehicle Navigation Array) to replace the actual radar antenna. The VNA is a general-purpose, low-power testing instrument, significantly cheaper than a dedicated weather radar, and is very safe. Because a low-power VNA is used as the signal source, the system poses no radiation safety risk and can be used in open workshops without the need for a shielded room. The collaborative robot is more flexible and less expensive than customized heavy-duty industrial robots and is suitable for shared workspaces. The testing system provided in this disclosure replaces fixed far-field testing with robotic near-field scanning, allowing it to be deployed in ordinary MRO (Maintenance, Repair, and Overhaul) workshops without requiring special facilities, requiring minimal floor space, and can be used to evaluate radome component repairs, precisely targeting the MRO market.

[0223] This disclosure provides a radome evaluation and transmission (transmissivity) testing system. This device is a testing apparatus used to determine the radar transmissivity (the percentage of radar waves that can penetrate the radome) of the foreground of a commercial aircraft fuselage. Exemplarily, the testing method provided in this disclosure utilizes a low-power test signal generated by a VNA to transmit a signal at a frequency of 9.375 GHz to a pair of polarized radar horns equipped with plano-convex collimating lenses. These radar horn antennas are mounted on the end axes of two collaborative robots. The two collaborative robots are programmed to move synchronously along the normal direction of the radome profile (i.e., maintaining a perpendicular incident angle), allowing electromagnetic waves to be incident perpendicularly onto the radome surface. Simultaneously, the use of a small-diameter (e.g., 100 mm) alignment plate / calibration device reduces the risk of robot collisions. Exemplarily, the radome is mounted on a rotary table / rotation device to enable the collaborative robots to move along a minimum path. For example, the cable connecting the radar horn antenna and the VNA is laid and guided along the collaborative robot arm. The system records readings in free space (without the radome) and with the radome, and calculates the transmittance percentage based on the power values ​​obtained from the two readings. For example, the transmittance performance of the weather radar can be estimated by calculating the average transmittance value within the actual coverage area (measuring point) of the weather radar. In addition, the system generates a color "heat map" of transmittance covering the radar window area. The entire device is controlled by proprietary software developed in-house.

[0224] It is understood that the calibration system and calibration method provided in this disclosure can be applied not only to the calibration of radome testing systems, but also to any field that requires dual robots (automated equipment) or multiple robots to achieve precise alignment and high-precision collaborative operation in a multi-axis system, such as similar application scenarios in the automotive or aerospace manufacturing industries, such as precision assembly, welding, and inspection.

[0225] The testing method provided in this disclosure includes two processes: unshielded benchmark testing and shielded wave transmission testing. Each test can be initiated by an operator via a single button press on a control console. Exemplarily, the test bench base is constructed of aluminum profiles (relatively easy to procure), and the wave transmission measurement unit includes a collaborative robot (cobot), a radar horn antenna, a VNA, and a computer. The system has few moving parts and high reliability. By utilizing the calibration system and method provided in this disclosure, the robot or collaborative robot can achieve precise alignment and coordinated operation, exhibiting near-perfect parallelism and azimuth consistency. This technology is of critical value in radome test bench applications—ensuring that two radar horn antennas point towards each other with high-precision alignment, their beam directions forming a specific angle with the sixth axis, thereby maximizing signal transmission. The radar horn carried by the collaborative robot is also oriented in the plane of rotation to ensure the alignment accuracy of the polarized radar beam.

[0226] This disclosure provides a near-field, small-footprint test bench for manufacturing and testing, which can simulate far-field test results and complete radome testing within 10 hours. A radio frequency (RF) signal (9.375 GHz) is fed from a low-power VNA to the radar horn. Radar transmittance measurements are performed in two phases: with the radome closed and open. Recorded values ​​are converted into percentages of power and transmittance, with over 1000 points measured. The transmittance of these points is weighted and averaged to simulate the signal's reception by a full-size (e.g., 700 mm aperture) airborne radar antenna.

[0227] The system provided in this disclosure features a miniaturized design, a small footprint (e.g., only 2m × 3m), allowing for deployment in limited spaces within a repair shop and easy mobility. Instead of using weather radar, a Vector Network Analyzer (VNA) is employed, resulting in lower test bench costs and safer operation in open environments. The cost is relatively low, approximately 1 / 10 of traditional solutions. Scanning the radome generates approximately 1000 data points, providing a practical radome performance distribution map for MRO (Maintenance, Repair, and Overhaul) entities. A high-resolution color transmittance heatmap visually displays the radome transmittance distribution. Low-power VNA signals are used, with strictly controlled power; full measurement is completed in a single shift (e.g., within 8 hours), significantly improving maintenance efficiency. A modular architecture reduces supply chain risks. Low maintenance requirements reduce downtime costs. Single-button operation enables fully automated testing with a single click, lowering the operational threshold. For example, feature point recognition can be used to automatically identify radome surface features via machine vision, adapting to different models. Collision avoidance path planning is also possible. For example, dynamic tilt control can also be implemented to adjust the pitch / yaw angle of the test antenna in real time to maintain the beam perpendicular incidence. This enables robot displacement control, achieving alignment optimization and simplification. For example, the antenna attitude can be optimized in real time using a proportional-integral algorithm to suppress accumulated errors.

[0228] The entire radome testing process takes approximately 5 hours. Testing a larger radome can be completed in about 9 hours. From an efficiency standpoint, the proposed new radome evaluation and transmission testing system is significantly more efficient than traditional far-field measurement methods, especially when MRO companies cannot afford expensive far-field RF measurement equipment and must ship radomes to other cities or even other countries for valid certification. The system proposed in this disclosure can help the MRO industry perform maintenance and verification within the workshop. This system improves the efficiency of the MRO industry's production processes.

[0229] The radome evaluation and transmission test system based on near-field measurements proposed in this disclosure is based on the newly updated RTCA DO-213A guidelines. It allows the MRO industry to have a low-cost, small-size radome transmission measurement facility that can be equipped in the workshop. The entire system has been built and calibrated. Two commercially applicable radomes have been tested. The overall fabrication time for these radomes is within 10 hours, less than two shifts in the factory. Test results show that the system meets the standard ranging range. The system is significantly more efficient and effective than traditional methods currently used in the MRO industry. Because this approach is specifically designed for measurements of repaired radomes, particularly for TE result verification in the MRO industry, it is not suitable for radome certification.

[0230] When using radar horns in the near field or Fresnel region (as opposed to the far field), the coherence length of the radiation must be taken into account. The coherence length is found in the near field and is given as approximately 400 mm.

[0231] The method provided in the embodiment of Figure 26 can be executed by a controller or any electronic device. As shown in Figure 26, the method provided in this disclosure includes the following steps.

[0232] In S210, the motion trajectories of the first robot and the second robot are obtained. The motion trajectories include N predetermined detection points, where N is an integer greater than 1. The N predetermined detection points include the first predetermined detection point and the second predetermined detection point.

[0233] In S220, the first robot with an image acquisition device installed at its end and the second robot with a calibration device installed at its end are controlled to move to a first predetermined pose relative to the first predetermined detection point, and the image acquisition device is used to acquire a first calibration image of the calibration device.

[0234] In S230, based on the first calibration image, a first relative pose between the calibration device and the image acquisition device is obtained at the first predetermined detection point.

[0235] In S240, based on the first relative pose, the first robot and / or the second robot are controlled to adjust until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point.

[0236] In S250, the corrected poses of the first robot and the second robot relative to the first predetermined detection point are recorded.

[0237] In S260, the first robot and the second robot are controlled to move to a second predetermined pose relative to the second predetermined detection point until N corrected poses of N predetermined detection points are obtained.

[0238] For example, the motion trajectory is on a curved surface. The method provided in this embodiment further includes: determining a first normal to the curved surface at the first predetermined detection point; when the angles between the axis of the image acquisition device and the axis of the calibration device and the first normal are both less than a first angle threshold, the angle between the first horizontal axis of the image acquisition device and the first horizontal axis of the calibration device is less than a second angle threshold, and the deviation between the origin of the image acquisition device and the origin of the calibration device in the planes of the first and second horizontal axes is less than a predetermined threshold, then it is determined that at the first predetermined detection point, the first relative pose between the calibration device and the image acquisition device meets the alignment condition.

[0239] For example, the method provided in this disclosure includes: obtaining a first transformation matrix from the image acquisition device to its first tool center point and optical correction parameters of the image acquisition device; obtaining a second transformation matrix from the calibration device to its second tool center point; and obtaining a first relative pose based on the first calibration image, the first transformation matrix, the optical correction parameters, and the second transformation matrix.

[0240] For example, obtaining the first transformation matrix from the image acquisition device to its first tool center point and the optical correction parameters of the image acquisition device includes: controlling the second robot to move to a first position to keep the calibration device fixed; controlling the first robot to move to multiple different first preset poses, and controlling the image acquisition device to acquire the first image of the calibration device in different first preset poses respectively; and obtaining the first transformation matrix and the optical correction parameters based on the first images in different first preset poses.

[0241] For example, obtaining a second transformation matrix from the calibration device to its second tool center point includes: controlling the first robot to move to a second position to keep the image acquisition device fixed; controlling the second robot to move to multiple different second preset poses, and controlling the image acquisition device to acquire second images of the calibration device in different second preset poses respectively; and obtaining the second transformation matrix based on the second images in different second preset poses.

[0242] Other aspects of the embodiment shown in Figure 26 can be found in other embodiments, and will not be repeated here.

[0243] The method provided in this disclosure can also be extended to multi-robot systems.

[0244] Figures 27 and 28 illustrate three robots. In the embodiment shown in Figure 27, the first robot 271 has a first image acquisition device 81 installed at its end, the second robot 272 has a calibration device 7 installed at its end, and the third robot 273 has a second image acquisition device 82 installed at its end. That is, two of the three robots have cameras installed, and one robot has a calibration plate installed. Alternatively, as shown in Figure 28, the first robot 271 has a first calibration device 71 installed at its end, the second robot 272 has an image acquisition device 8 installed at its end, and the third robot 273 has a second calibration device 72 installed at its end. That is, two of the three robots have calibration plates installed, and one robot has a camera installed.

[0245] In this embodiment of the disclosure, calibration for any configuration of a multi-robot system can be performed according to the following steps, which mainly consist of three steps:

[0246] The first step is to calculate the second transformation matrix T2 from the optical calibration plate / calibration device to the second robot TCP. This step selects a camera on another robot as the camera for eye-hand calibration, achieving eye-on-hand calibration, where the camera remains stationary while the calibration plate moves to multiple positions to capture images. This step can reduce the measurement error from the calibration plate to the second robot TCP.

[0247] The second step is to calculate the first transformation matrix T1 from the camera to the robot's TCP. In this step, a calibration board from another robot is selected as the calibration board for hand-eye calibration, achieving eye-on-hand calibration. That is, the calibration board remains stationary while the camera moves to multiple positions to capture images. This step can reduce the measurement error in calculating the camera-to-TCP.

[0248] The third step is to correct the system's working posture deviation. At this point, the robots are moved to the working position. The posture of one robot is fixed as needed, while the other robots are moved so that the standard board and the camera are close to the target transformation matrix T3 (i.e., the ideal pose), thus achieving system alignment.

[0249] For multi-robot systems, calibration plates can be installed on one or more robots (TCPs), while cameras are installed on the other robots (TCPs). Depending on the requirements, one robot (either a camera or a calibration plate) can be selected as the fixed target, and then the relative poses of the other robots are corrected and compensated. If multiple robots need to align a target in space (e.g., the reference point pose of a workpiece), additional calibration plates can be placed on the workpiece. First, select one or more robots with cameras to compensate for their pose deviation relative to the workpiece calibration plate. Then, fix one or more robots (with cameras) in place, and then compensate the other robots (with calibration plates).

[0250] As shown in Figure 29, this embodiment of the present disclosure provides a calibration device 100, including: an acquisition unit 110, configured to acquire the motion trajectories of a first robot and a second robot, wherein the motion trajectories include N predetermined detection points, where N is an integer greater than 1, and the N predetermined detection points include a first predetermined detection point and a second predetermined detection point; a control unit 120, configured to control the first robot with an image acquisition device installed at its end cap and the second robot with a calibration device installed at its end cap to move to a first predetermined pose relative to the first predetermined detection point, and to acquire a first calibration image of the calibration device using the image acquisition device; and a processing unit 130, configured to obtain, based on the first calibration image, a calibration image of the first robot with an image acquisition device installed at its end cap and a second robot with a calibration device installed at its end cap. The control unit 120 is further configured to control the first robot and / or the second robot to adjust according to the first relative pose, until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point; the processing unit 130 is further configured to record the corrected pose between the first robot and the second robot relative to the first predetermined detection point; the control unit 120 is further configured to control the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point, until N corrected poses of N predetermined detection points are obtained.

[0251] Other aspects of the embodiment shown in Figure 29 can be found in the above embodiments and will not be repeated here.

[0252] This disclosure also provides a calibration system, including: a first robot and a second robot; an image acquisition device installed at the end of the first robot, and a calibration device installed at the end of the second robot; a controller, the controller being configured to obtain the motion trajectories of the first robot and the second robot, the motion trajectories including N predetermined detection points, where N is an integer greater than 1, and the N predetermined detection points including a first predetermined detection point and a second predetermined detection point; the controller being configured to control the first robot with the image acquisition device installed at its end and the second robot with the calibration device installed at its end to move to a first predetermined pose relative to the first predetermined detection point, and to use the image acquisition device to acquire a first calibration image of the calibration device; based on the first calibration image, to obtain a first relative pose between the calibration device and the image acquisition device at the first predetermined detection point; based on the first relative pose, to control the first robot and / or the second robot to adjust until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point; to record a corrected pose between the first robot and the second robot relative to the first predetermined detection point; and to control the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point until N corrected poses of the N predetermined detection points are obtained.

[0253] For example, the controller is further configured to: control the second robot to move to a first position to keep the calibration device fixed; control the first robot to move to multiple different first preset poses, and control the image acquisition device to acquire first images of the calibration device in different first preset poses respectively; based on the first images in different first preset poses, obtain a first transformation matrix from the image acquisition device to its first tool center point and optical correction parameters of the image acquisition device; control the first robot to move to a second position to keep the image acquisition device fixed; control the second robot to move to multiple different second preset poses, and control the image acquisition device to acquire second images of the calibration device in different second preset poses respectively; based on the second images in different second preset poses, obtain a second transformation matrix from the calibration device to the second tool center point of the second robot; and obtain the first relative pose according to the first calibration image, the first transformation matrix, the optical correction parameters, and the second transformation matrix.

[0254] For example, the calibration device includes a checkerboard-coded hybrid marking calibration pattern.

[0255] For example, the first robot end effector includes a first mounting bracket, and the second robot end effector includes a second mounting bracket. The image acquisition device is mounted on the first mounting bracket via a first fixing adapter, and the calibration device is mounted on the second mounting bracket via a second fixing adapter.

[0256] For example, the first mounting bracket is further used to mount a first testing device, and the second mounting bracket is further used to mount a second testing device. The controller is also used to move the first robot and the second robot according to the motion trajectory and the corrected pose, so as to align the first testing device and the second testing device at predetermined detection points; the first testing device and the second testing device are used to detect the radome at N predetermined detection points.

[0257] For example, the calibration system further includes: a flash device for illuminating the radome; a recording device for recording the temperature change of the radome to obtain a thermal image of the radome; wherein the controller is further configured to detect the structural integrity of the radome based on the thermal image.

[0258] For example, the calibration system further includes a rotating device for mounting and rotating the radome.

[0259] By way of example, embodiments of this disclosure also provide an electronic device, including: one or more processors; and a memory configured to store one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the methods described in any embodiment of this disclosure.

[0260] By way of example, embodiments of this disclosure also provide a computer-readable storage medium storing a computer program that, when run on a computer, causes the computer to perform the methods described in any embodiment of this disclosure.

[0261] By way of example, this disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the methods described in any embodiment of this disclosure.

Claims

1. A calibration method, comprising: An image acquisition device is installed at the end of the first robot, and a calibration device is installed at the end of the second robot. Obtain the motion trajectories of the first robot and the second robot. The motion trajectories correspond to N predetermined detection points, where N is an integer greater than 1. The N predetermined detection points include the first predetermined detection point and the second predetermined detection point. Control the first robot and the second robot to move to a first predetermined pose relative to the first predetermined detection point, and use the image acquisition device to acquire the first calibration image of the calibration device; Based on the first calibration image, a first relative pose between the calibration device and the image acquisition device is obtained at the first predetermined detection point. Based on the first relative pose, control the first robot and / or the second robot to adjust until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point. Record the corrected poses of the first robot and the second robot relative to the first predetermined detection point; Control the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point, until N corrected poses of N predetermined detection points are obtained.

2. The method as described in claim 1, wherein the motion trajectory lies on a curved surface; wherein, The method further includes: Determine the first normal line on the surface where the first predetermined detection point is located; When the angles between the axis of the image acquisition device and the axis of the calibration device and the first normal are both less than a first angle threshold, the angle between the first horizontal axis of the image acquisition device and the first horizontal axis of the calibration device is less than a second angle threshold, and the deviation between the origin of the image acquisition device and the origin of the calibration device in the plane of the first horizontal axis and the second horizontal axis is less than a predetermined threshold, it is determined that at the first predetermined detection point, the first relative pose between the calibration device and the image acquisition device meets the alignment condition.

3. The method of claim 1, comprising: Obtain the first transformation matrix from the image acquisition device to its first tool center point and the optical correction parameters of the image acquisition device; Obtain the second transformation matrix from the calibration device to its second tool center point; The first relative pose is obtained based on the first calibration image, the first transformation matrix, the optical correction parameters, and the second transformation matrix.

4. The method of claim 3, wherein obtaining the first transformation matrix from the image acquisition device to its first tool center point and the optical correction parameters of the image acquisition device comprises: The second robot is controlled to move to the first position to keep the calibration device fixed. The first robot is controlled to move to multiple different first preset poses, and the image acquisition device is controlled to acquire the first image of the calibration device in different first preset poses respectively; Based on the first image with different first preset poses, the first transformation matrix and the optical correction parameters are obtained.

5. The method of claim 3, wherein obtaining the second transformation matrix from the calibration device to its second tool center point comprises: Control the first robot to move to the second position to keep the image acquisition device fixed; The second robot is controlled to move to multiple different second preset poses, and the image acquisition device is controlled to acquire second images of the calibration device in different second preset poses respectively; The second transformation matrix is ​​obtained based on the second image with different second preset poses.

6. The method of claim 1, wherein the calibration device includes a checkerboard-coded hybrid marking calibration pattern.

7. The method of claim 1, wherein the first robot end effector includes a first mounting bracket, and the second robot end effector includes a second mounting bracket; wherein, An image acquisition device is installed at the end effector of the first robot, and a calibration device is installed at the end effector of the second robot, including: Remove the antenna cover between the first robot and the second robot; The image acquisition device is mounted on the first mounting bracket via a first fixed adapter, and the calibration device is mounted on the second mounting bracket via a second fixed adapter.

8. The method of claim 7, wherein obtaining the motion trajectories of the first robot and the second robot comprises: Obtain a three-dimensional model of the radome; Based on the three-dimensional model, the scanning path of the radome is planned, and N predetermined detection points on the scanning path and the normal direction of each predetermined detection point are determined. Based on the normal direction and the measured distance between the first test device and the second test device, the arrival poses of the first tool center point of the first robot and the second tool center point of the second robot are generated to generate the motion trajectory.

9. The method of claim 7, comprising: Remove the image acquisition device from the first mounting bracket and remove the calibration device from the second mounting bracket; A first testing device is mounted on the first mounting bracket, and a second testing device is mounted on the second mounting bracket; The first robot and the second robot are moved according to the motion trajectory and the corrected pose to align the first test device and the second test device at predetermined detection points, and the radome is detected at N predetermined detection points by the first test device and the second test device.

10. The method of claim 9, comprising: When the radome is not installed between the first robot and the second robot, the first signal power of the test signal at the predetermined detection point is collected by the aligned first test device and the second test device. The radome is installed between the first robot and the second robot. The second signal power of the test signal passing through the radome at the predetermined detection point is collected by the aligned first and second test devices. The transmittance of the antenna radome at the predetermined detection point is obtained based on the first signal power and the second signal power.

11. The method of claim 10, comprising: The transmittance of each predetermined detection point is mapped onto the three-dimensional model of the radome to obtain and display the transmittance heat map of the radome.

12. The method of claim 9, comprising: The radome is installed between the first robot and the second robot. Near-field data of the test signal passing through the radome at the predetermined detection point is collected by aligning the first test device and the second test device. The far-field data of the test signal is obtained based on the near-field data.

13. The method of claim 12, comprising: Based on the M measurement directions of the antenna under test, according to the elevation and azimuth scanning sequence, or the azimuth and elevation scanning sequence, select N1 far-field data corresponding to the measurement directions from the N far-field data, and weightedly synthesize the far-field transmittance in the corresponding measurement directions. Where M and N1 are both positive integers greater than or equal to 1 and less than or equal to N.

14. The method of any one of claims 10 to 13, comprising: The test signal is generated and transmitted to the first test device or the second test device by a vector network analyzer; The vector network analyzer receives the returned test signal from the second test device or the first test device.

15. The method of claim 8, wherein the measuring distance is less than D represents the aperture of the first or second test device, and λ represents the wavelength of the test signal emitted by the first or second test device.

16. The method of claim 15, wherein the measured distance is not less than 3λ.

17. The method of claim 15, wherein the frequency of the test signal is between 7.0 and 11.2 GHz.

18. The method of claim 17, wherein the frequency of the test signal is 9.375 GHz.

19. The method of claim 8, wherein the first testing device is a test antenna and the second testing device is a metal plate; or, the first testing device is a metal plate and the second testing device is a test antenna; or, Both the first test device and the second test device are test antennas; Alternatively, both the first test device and the second test device may be antenna arrays, wherein the antenna arrays include multiple test antennas.

20. The method of claim 19, wherein the test antenna is a horn antenna.

21. The method of claim 20, wherein the test antenna is a polarized horn antenna.

22. The method of claim 21, wherein the test antenna is a linearly polarized horn antenna.

23. The method of claim 9, wherein the first testing device and the second testing device are used to perform quality verification on the repaired radome.

24. The method of claim 9, further comprising: Illuminate the radome with a flash device; Record the temperature change of the radome to obtain a thermal image of the radome; The structural integrity of the radome is detected based on the thermal image.

25. The method of claim 9, comprising: The radome is mounted on the rotating device; The radome is rotated by the rotating device to detect the transmittance of predetermined test points in different areas of the radome using the first test device and the second test device.

26. A calibration method, comprising: Obtain the motion trajectories of the first robot and the second robot. The motion trajectories include N predetermined detection points, where N is an integer greater than 1. The N predetermined detection points include the first predetermined detection point and the second predetermined detection point. The first robot with an image acquisition device installed at its end and the second robot with a calibration device installed at its end are controlled to move to a first predetermined pose relative to the first predetermined detection point, and the image acquisition device is used to acquire a first calibration image of the calibration device. Based on the first calibration image, a first relative pose between the calibration device and the image acquisition device is obtained at the first predetermined detection point. Based on the first relative pose, control the first robot and / or the second robot to adjust until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point. Record the corrected poses of the first robot and the second robot relative to the first predetermined detection point; Control the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point, until N corrected poses of N predetermined detection points are obtained.

27. The method of claim 26, wherein the motion trajectory is on a curved surface; wherein, The method further includes: Determine the first normal line on the surface where the first predetermined detection point is located; When the angles between the axis of the image acquisition device and the axis of the calibration device and the first normal are both less than a first angle threshold, the angle between the first horizontal axis of the image acquisition device and the first horizontal axis of the calibration device is less than a second angle threshold, and the deviation between the origin of the image acquisition device and the origin of the calibration device in the plane of the first horizontal axis and the second horizontal axis is less than a predetermined threshold, it is determined that at the first predetermined detection point, the first relative pose between the calibration device and the image acquisition device meets the alignment condition.

28. The method of claim 26, comprising: Obtain the first transformation matrix from the image acquisition device to its first tool center point and the optical correction parameters of the image acquisition device; Obtain the second transformation matrix from the calibration device to its second tool center point; The first relative pose is obtained based on the first calibration image, the first transformation matrix, the optical correction parameters, and the second transformation matrix.

29. The method of claim 28, wherein obtaining the first transformation matrix from the image acquisition device to its first tool center point and the optical correction parameters of the image acquisition device comprises: The second robot is controlled to move to the first position to keep the calibration device fixed. The first robot is controlled to move to multiple different first preset poses, and the image acquisition device is controlled to acquire the first image of the calibration device in different first preset poses respectively; Based on the first image with different first preset poses, the first transformation matrix and the optical correction parameters are obtained.

30. The method of claim 28, wherein obtaining the second transformation matrix from the calibration device to its second tool center point comprises: Control the first robot to move to the second position to keep the image acquisition device fixed; The second robot is controlled to move to multiple different second preset poses, and the image acquisition device is controlled to acquire second images of the calibration device in different second preset poses respectively; The second transformation matrix is ​​obtained based on the second image with different second preset poses.

31. A calibration device, comprising: The acquisition unit is used to acquire the motion trajectory of the first robot and the second robot. The motion trajectory includes N predetermined detection points, where N is an integer greater than 1. The N predetermined detection points include the first predetermined detection point and the second predetermined detection point. The control unit is used to control the first robot with an image acquisition device installed at its end and the second robot with a calibration device installed at its end to move to a first predetermined pose relative to the first predetermined detection point, and to use the image acquisition device to acquire a first calibration image of the calibration device. The processing unit is configured to obtain, based on the first calibration image, a first relative pose between the calibration device and the image acquisition device at the first predetermined detection point; The control unit is further configured to control the first robot and / or the second robot to adjust according to the first relative pose until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point. The processing unit is also used to record the corrected pose of the first robot and the second robot relative to the first predetermined detection point; The control unit is also used to control the first robot and the second robot to move to a second predetermined pose relative to the second predetermined detection point, until N corrected poses of N predetermined detection points are obtained.

32. A calibration system, comprising: First robot and second robot; An image acquisition device installed at the end of the first robot, and a calibration device installed at the end of the second robot; A controller is used to obtain the motion trajectories of the first robot and the second robot. The motion trajectory includes N predetermined detection points, where N is an integer greater than 1. The N predetermined detection points include a first predetermined detection point and a second predetermined detection point. It is also used to control the first robot with an image acquisition device installed at its end and the second robot with a calibration device installed at its end to move to a first predetermined pose relative to the first predetermined detection point, and to use the image acquisition device to acquire a first calibration image of the calibration device; Based on the first calibration image, a first relative pose between the calibration device and the image acquisition device is obtained at the first predetermined detection point; based on the first relative pose, the first robot and / or the second robot are controlled to adjust until the first relative pose between the calibration device and the image acquisition device meets the alignment condition at the first predetermined detection point; the corrected pose between the first robot and the second robot relative to the first predetermined detection point is recorded; the first robot and the second robot are controlled to move to a second predetermined pose relative to the second predetermined detection point until N corrected poses of N predetermined detection points are obtained.

33. The system of claim 32, wherein the controller is further configured to: The second robot is controlled to move to the first position to keep the calibration device fixed. The first robot is controlled to move to multiple different first preset poses, and the image acquisition device is controlled to acquire the first image of the calibration device in different first preset poses respectively; Based on the first image with different first preset poses, obtain the first transformation matrix from the image acquisition device to its first tool center point and the optical correction parameters of the image acquisition device; Control the first robot to move to the second position to keep the image acquisition device fixed; The second robot is controlled to move to multiple different second preset poses, and the image acquisition device is controlled to acquire second images of the calibration device in different second preset poses respectively; Based on the second images with different second preset poses, a second transformation matrix is ​​obtained from the calibration device to the second tool center point of the second robot; The first relative pose is obtained based on the first calibration image, the first transformation matrix, the optical correction parameters, and the second transformation matrix.

34. The system of claim 32, wherein the calibration device includes a checkerboard-coded hybrid marking calibration pattern.

35. The system of claim 32, wherein the first robot end effector includes a first mounting bracket, and the second robot end effector includes a second mounting bracket; in, The image acquisition device is mounted on the first mounting bracket via a first fixed adapter, and the calibration device is mounted on the second mounting bracket via a second fixed adapter.

36. The system of claim 32, wherein the first mounting bracket is further configured to mount the first testing device, and the second mounting bracket is further configured to mount the second testing device; The controller is also used to move the first robot and the second robot according to the motion trajectory and the corrected pose, so as to align the first test device and the second test device at a predetermined detection point. The first testing device and the second testing device are used to test the radome at N predetermined testing points.

37. The system of claim 36, further comprising: A flashing device for illuminating the radome; A recording device is used to record the temperature change of the radome and obtain a thermal image of the radome. The controller is also used to detect the structural integrity of the radome based on the thermal image.

38. The system of claim 36, further comprising: A rotating device for mounting and rotating the radome.

39. An electronic device comprising: One or more processors; A memory configured to store one or more programs that, when executed by one or more processors, cause the electronic device to perform the method of any one of claims 1 to 25, or the method of any one of claims 26 to 30.

40. A computer-readable storage medium storing a computer program that, when run on a computer, causes the computer to perform the method of any one of claims 1 to 25, or the method of any one of claims 26 to 30.

41. A computer program product comprising a computer program that, when executed by a processor, implements the method of any one of claims 1 to 25, or the method of any one of claims 26 to 30.

Images