Satellite complex truss part assembly relation guiding mark automatic pasting system and method

Abstract

The invention provides an automatic pasting system and method for guiding marks in the assembly relation of satellite complex truss parts. The automatic pasting system comprises a high-performance computer 101, an automatic guiding mark pasting control system 102, a stereo camera module 103, a visual recognition module 104, a part grabbing module 105, a mark pasting module 106, a workbench 107, a guiding mark printer 108 and the guiding marks 109. The method can automatically analyze the complex truss assembly relation and generate a guide mark information database, automatically grabs truss parts, adjusts the poses and identifies the truss parts, automatically prints assembly relation guide marks and pastes the guide marks at the designated positions of the parts, automatic operation of the guide mark pasting process is achieved, and the production efficiency is improved. The development efficiency and the digital level of the satellite truss are improved, and support is provided for one-time successful assembly of the satellite complex truss.

Description

Technical Field

[0001] The present invention belongs to the technical field of satellite truss assembly, and in particular relates to a system and method for automatically pasting guide marks for assembly relationships of complex satellite truss parts. Background Art

[0002] Truss structures are a common structural form in satellites. Primarily composed of carbon fiber rods and multi-way joints, they offer advantages such as light weight, high rigidity, diverse forms, and flexible design. They are widely used in satellite structures both domestically and internationally, such as the Keyhole and Lacrosse satellites. With the continuous development of payloads such as high-resolution, high-sensitivity space cameras and satellite antennas, satellite truss structures are becoming larger and more complex, with envelope dimensions exceeding 3000mm, topological configurations exhibiting spatial asymmetry, fully closed rod systems, and hundreds of rods, joints, and other components. Ensuring a successful, one-time assembly is becoming increasingly challenging.

[0003] Currently, truss parts are mainly identified by manually comparing three-dimensional models, and then the assembly relationship guide marks are pasted. Due to the large number of truss parts and the high degree of similarity of some parts, the pasting of guide marks is a large workload and there is a risk of misjudgment, which may lead to assembly errors in complex truss structures. The patent document "An assembly guidance method, system and application based on augmented reality" (CN112734945A) discloses a method and system for assembly guidance based on the three-dimensional edge contour feature information of parts. However, it directly uses contour information as guidance information through camera recognition, and does not improve the marking method of the guidance information.

[0004] Therefore, a method is needed to automatically paste the guide marks of the assembly relationship of complex satellite truss parts, realize the automation of the guide mark pasting process, improve the efficiency and digitalization level of satellite truss development, and provide support for ensuring the successful one-time assembly of complex satellite trusses. Summary of the Invention

[0005] In view of the defects in the prior art, the purpose of the present invention is to provide a system and method for automatically attaching guide marks for the assembly relationship of complex satellite truss parts.

[0006] The system for automatically pasting guide marks for the assembly relationship of satellite complex truss parts provided by the present invention includes: a high-performance computer 101, a guide mark automatic pasting control system 102, a stereo camera module 103, a visual recognition module 104, a part grasping module 105, a mark pasting module 106, a workbench 107, a guide mark printer 108 and a guide mark 109.

[0007] A high-performance computer analyzes the assembly relationship based on the satellite's complex truss 3D model and assembly process information, generates assembly relationship guidance mark information for all m parts, and establishes a parts assembly relationship guidance mark database;

[0008] Traversing all m parts, the stereo camera module photographs part i (i = 1, 2, ..., m) to obtain point cloud data, and a high-performance computer analyzes and calculates the point cloud data to locate part i to obtain positioning information;

[0009] The part grasping module grasps the part i according to the positioning information and brings it into the field of view of the visual recognition module;

[0010] The visual recognition module performs multi-view imaging on part i to obtain multiple images of part i;

[0011] High-performance computers detect part labels based on multiple images, calculate label poses, and determine occlusions;

[0012] The part grasping module adjusts the position of part i according to the label position;

[0013] The visual recognition module images the label of part i and obtains the label image;

[0014] A high-performance computer processes the label image, identifies the label image content, and retrieves the corresponding assembly relationship guidance mark information from the parts assembly relationship guidance mark database;

[0015] Traverse all guide marks of part i, and according to the guide mark information of the assembly relationship, the guide mark printer prints guide mark j;

[0016] Where j = 1, 2, ..., n, n represents the number of guide marks of part i;

[0017] The mark pasting module picks up the guide mark j;

[0018] The high-performance computer plans the pasting path based on the pose of part i and guide marker j and determines the path solution;

[0019] The mark pasting module pastes the guide mark j at the specified position of part i, determines the task progress, and ends the pasting.

[0020] Preferably, the high-performance computer analyzes and calculates point cloud data and label poses through the SGBM stereo matching algorithm, detects part labels using the YOLO target detection algorithm, recognizes label image content using the optical character recognition algorithm, and plans the pasting path using the fast exploration random tree algorithm.

[0021] The multi-view imaging has no fewer than four viewing angles, and obtains no fewer than four multiple images of part i.

[0022] The occlusion judgment is to judge whether the label of part i is blocked. If so, the part grasping module puts down part i and repositions part i to change the grasping position. If not, the part grasping module adjusts the posture of part i into place.

[0023] The posture of the part i is adjusted to be in place so that the normal vector at the center of the part label coincides with the optical axis of the camera lens of the visual recognition module and is within the depth of field of the visual recognition module.

[0024] The mark sticking module picks up the guide mark by vacuum adsorption and sticks the guide mark by flexible pressure.

[0025] The determination of the path solution is to determine whether the pasting path has a solution. If so, the marking pasting module adjusts the posture of the guide mark j. If not, the part grasping module adjusts the posture of the part i and replans the pasting path.

[0026] The judgment task process is to judge whether all guide marks of part i have been traversed. If not, set a=j, j=a+1, and the guide mark printer continues to print the next guide mark j. If so, judge whether all parts of the truss have been traversed. If so, end the pasting. If not, set a=i, i=a+1, and the high-performance computer continues to locate the next part i.

[0027] Preferably, the high-performance computer includes a host, a display, a keyboard and a mouse, and the host is electrically connected to the display, the keyboard and the mouse respectively.

[0028] The guide mark automatic pasting control system is electrically connected to a high-performance computer, and includes a stereo camera module control circuit, a visual recognition module control circuit, a part grasping module control circuit, and a mark pasting module control circuit, which are electrically connected to the stereo camera module, the visual recognition module, the part grasping module, and the mark pasting module respectively.

[0029] The stereo camera module includes an image sensor, a lens, a laser and a laser galvanometer.

[0030] The visual recognition module includes an industrial camera, a camera lens, an illumination light source and a mounting bracket.

[0031] The part grasping module includes a grasping industrial robot arm and a grasping end effector.

[0032] The marking pasting module includes a pasting industrial robot arm and a pasting end effector.

[0033] The workbench provides an operating space for the parts grabbing module and the marking pasting module.

[0034] Preferably, the main frequency of the processor of the host is not less than 3.0 GHz, the memory is not less than 32 GB, the video memory of the independent graphics card is not less than 16 GB, and the hard disk is not less than 1 TB.

[0035] The display size is not less than 27 inches.

[0036] The image sensor is used to generate digital images, has a resolution of not less than 10 million pixels, and has no less than 2 image sensors.

[0037] The lenses are used to transform light, have a field of view of no less than 50°, and their number is consistent with the number of image sensors.

[0038] The laser is used to generate line lasers, and the number of line lasers is no less than 7.

[0039] The laser galvanometer is used to scan the surface of the object to be measured with a line laser, and the scanning frequency is not less than 1 Hz.

[0040] The industrial cameras are used to generate digital images with a resolution of not less than 10 million pixels and the number is not less than 4.

[0041] The camera lens is used to transform light, has a field of view angle of not less than 40°, and the number is consistent with that of industrial cameras.

[0042] The illumination light source is a white LED light source, which is used to illuminate the target.

[0043] The mounting bracket is used to fix the industrial camera.

[0044] The grabbing industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 10 kg, and is used to adjust the position and posture of the grasped parts.

[0045] The maximum grasping diameter of the grasping end effector is not less than Φ100mm, and the maximum grasping mass is not less than 6kg, and it is used to grasp truss parts.

[0046] The pasting industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 5 kg, and is used to adjust the position and posture of the guide mark.

[0047] The pasting end effector is used to pick up and paste the guide mark.

[0048] Preferably, the high-performance computer is used to set up, manage and analyze the automatic pasting process of the guide marks for the assembly relationship of complex satellite truss parts.

[0049] The guide mark automatic pasting control system is used to control the stereo camera module, the visual recognition module, the part grasping module, and the mark pasting module to achieve imaging control, grasping control, and pasting control.

[0050] The stereo camera module is used to perform laser scanning and multi-view imaging on parts.

[0051] The visual recognition module is used to image the parts and their labels.

[0052] The part grasping module is used to grasp and adjust the position of parts.

[0053] The marker pasting module is used to pick up, adjust the position and paste the guide marker.

[0054] The workbench is used to place truss parts, a high-performance computer and a guide marking printer, and to install a stereo camera module, a visual recognition module, a part grabbing module and a marking pasting module.

[0055] The guide mark printer is used to print assembly relationship guide marks that can be pasted.

[0056] The guide mark is used to display assembly relationship guide information.

[0057] According to the present invention, a method for automatically attaching guide marks for assembly relationships of complex satellite truss parts comprises:

[0058] Step S1: Analyze the assembly relationship based on the satellite complex truss 3D model information and assembly process information in a high-performance computer;

[0059] Step S2: Instruct a high-performance computer to generate assembly relationship guidance mark information for all m parts, establish a parts assembly relationship guidance mark database, and traverse all m parts;

[0060] Step S3: The stereo camera module is used to photograph part i (i=1, 2, ..., m) to obtain point cloud data, and a high-performance computer is used to analyze and calculate the point cloud data to locate part i and obtain positioning information;

[0061] Step S4: Instruct the part grasping module to grasp the part i according to the positioning information and bring it into the field of view of the visual recognition module;

[0062] Step S5: Instructing the visual recognition module to perform multi-view imaging on part i to obtain multiple images of part i;

[0063] Step S6: Instruct a high-performance computer to detect part labels based on multiple images and calculate label poses;

[0064] Step S7: Instructing a high performance computer to determine occlusion;

[0065] Step S8: Instruct the part grasping module to adjust the position of part i according to the label position;

[0066] Step S9: Instruct the visual recognition module to image the label of part i to obtain a label image;

[0067] Step S10: Processing the label image in a high-performance computer to identify the label image content;

[0068] Step S11: Instructing a high-performance computer to retrieve corresponding assembly relationship guide mark information from a parts assembly relationship guide mark database;

[0069] Step S12: traverse all guide marks of part i, and the guide mark printer prints guide mark j according to the assembly relationship guide mark information;

[0070] Where j = 1, 2, ..., n, n represents the number of guide marks of part i;

[0071] Step S13: instructing the mark pasting module to pick up the guide mark j;

[0072] Step S14: Instructing a high-performance computer to plan a pasting path based on the position and orientation of the part i and the guide mark j and determine the path solution;

[0073] Step S15: instructing the mark pasting module to paste the guide mark j at the designated position of the part i;

[0074] Step S16: Determine the task progress and end the pasting.

[0075] Preferably, the high-performance computer analyzes and calculates point cloud data and label poses through the SGBM stereo matching algorithm, detects part labels using the YOLO target detection algorithm, recognizes label image content using the optical character recognition algorithm, and plans the pasting path using the fast exploration random tree algorithm.

[0076] The mark sticking module picks up the guide mark by vacuum adsorption and sticks the guide mark by flexible pressure.

[0077] In step S5 , the viewing angles of the multi-view imaging are no less than 4, and no less than 4 multiple images of the part i are obtained.

[0078] In step S7, it is determined whether the label of part i is blocked. If so, the part grasping module puts down part i and executes step S3. If not, step S8 is executed.

[0079] In step S8, the part grasping module makes the normal vector at the center of the part label coincide with the optical axis of the camera lens of the visual recognition module and be within the depth of field of the visual recognition module.

[0080] In step S14, it is determined whether the pasting path has a solution. If so, step S15 is executed. If not, the part grasping module adjusts the posture of part i and executes step S14 to re-plan the pasting path.

[0081] In step S16, it is determined whether all guide marks of part i have been traversed. If not, a=j, j=a+1 are set and step S12 is executed. If yes, it is determined whether all parts of the truss have been traversed.

[0082] Determine whether all parts of the truss have been traversed. If so, end the pasting. If not, set a=i, i=a+1 and execute step S3.

[0083] Preferably, the high-performance computer includes a host, a display, a keyboard and a mouse, and the host is electrically connected to the display, the keyboard and the mouse respectively.

[0084] The guide mark automatic pasting control system is electrically connected to a high-performance computer, and includes a stereo camera module control circuit, a visual recognition module control circuit, a part grasping module control circuit, and a mark pasting module control circuit, which are electrically connected to the stereo camera module, the visual recognition module, the part grasping module, and the mark pasting module respectively.

[0085] The stereo camera module includes an image sensor, a lens, a laser and a laser galvanometer.

[0086] The visual recognition module includes an industrial camera, a camera lens, an illumination light source and a mounting bracket.

[0087] The part grasping module includes a grasping industrial robot arm and a grasping end effector.

[0088] The marking pasting module includes a pasting industrial robot arm and a pasting end effector.

[0089] The workbench provides an operating space for the parts grabbing module and the marking pasting module.

[0090] Preferably, the main frequency of the processor of the host is not less than 3.0 GHz, the memory is not less than 32 GB, the video memory of the independent graphics card is not less than 16 GB, and the hard disk is not less than 1 TB.

[0091] The display size is not less than 27 inches.

[0092] The image sensor is used to generate digital images, has a resolution of not less than 10 million pixels, and has no less than 2 image sensors.

[0093] The lenses are used to transform light, have a field of view of no less than 50°, and their number is consistent with the number of image sensors.

[0094] The laser is used to generate line lasers, and the number of line lasers is no less than 7.

[0095] The laser galvanometer is used to scan the surface of the object to be measured with a line laser, and the scanning frequency is not less than 1 Hz.

[0096] The industrial cameras are used to generate digital images with a resolution of not less than 10 million pixels and the number is not less than 4.

[0097] The camera lens is used to transform light, has a field of view angle of not less than 40°, and the number is consistent with that of industrial cameras.

[0098] The illumination light source is a white LED light source, which is used to illuminate the target.

[0099] The mounting bracket is used to fix the industrial camera.

[0100] The grabbing industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 10 kg, and is used to adjust the position and posture of the grasped parts.

[0101] The maximum grasping diameter of the grasping end effector is not less than Φ100mm, and the maximum grasping mass is not less than 6kg, and it is used to grasp truss parts.

[0102] The pasting industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 5 kg, and is used to adjust the position and posture of the guide mark.

[0103] The pasting end effector is used to pick up and paste the guide mark.

[0104] Preferably, the high-performance computer is used to set up, manage and analyze the automatic pasting process of the guide marks for the assembly relationship of complex satellite truss parts.

[0105] The guide mark automatic pasting control system is used to control the stereo camera module, the visual recognition module, the part grasping module, and the mark pasting module to achieve imaging control, grasping control, and pasting control.

[0106] The stereo camera module is used to perform laser scanning and multi-view imaging on parts.

[0107] The visual recognition module is used to image the parts and their labels.

[0108] The part grasping module is used to grasp and adjust the position of parts.

[0109] The marker pasting module is used to pick up, adjust the position and paste the guide marker.

[0110] The workbench is used to place truss parts, a high-performance computer and a guide marking printer, and to install a stereo camera module, a visual recognition module, a part grabbing module and a marking pasting module.

[0111] The guide mark printer is used to print assembly relationship guide marks that can be pasted.

[0112] The guide mark is used to display assembly relationship guide information.

[0113] Compared with the prior art, the present invention has the following beneficial effects:

[0114] 1. The present invention can automatically analyze complex truss assembly relationships and generate a guidance mark information database, thereby improving the efficiency and digitalization level of satellite truss development.

[0115] 2. The present invention can automatically grab truss parts, adjust their postures and identify them, automatically print assembly relationship guide marks and paste them at designated positions on the parts, thus realizing the automation of the guide mark pasting process.

[0116] 3. Compared with traditional methods, the present invention simplifies the workload and reduces the risk of misjudgment in the pasting of guidance marks, providing support for ensuring the successful one-time assembly of complex satellite trusses. BRIEF DESCRIPTION OF THE DRAWINGS

[0117] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0118] Figure 1 Schematic diagram of the automatic pasting system for guiding the assembly relationship of complex truss parts of satellites.

[0119] Figure 2 Flowchart of the method for automatically pasting assembly relationship guide marks for complex satellite truss parts. DETAILED DESCRIPTION

[0120] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0121] The present invention can automatically analyze the complex truss assembly relationship and generate a guide mark information database, automatically grab the truss parts to adjust the posture and identify them, automatically print the assembly relationship guide mark and paste it at the specified position of the part, realize the automation of the guide mark pasting process, and improve the efficiency and digital level of satellite truss development. According to the present invention, a satellite complex truss parts assembly relationship guide mark automatic pasting system is provided. Figure 1For example, it specifically includes: a high-performance computer 101, a guide mark automatic pasting control system 102, a stereo camera module 103, a visual recognition module 104, a part grasping module 105, a mark pasting module 106, a workbench 107, a guide mark printer 108, and a guide mark 109.

[0122] The high-performance computer is used to set up, manage, and analyze the automatic application of guide markers to complex satellite truss component assembly relationships. This high-performance computer and the guide marker automatic application control system enable parameter setting, process management, data analysis, and motion control, enabling human-computer interaction, model analysis, label detection, pose estimation, component recognition, grasping analysis, application analysis, and data management.

[0123] In further preferred embodiments, the high-performance computer comprises: a host computer, a display, a keyboard, and a mouse. The host computer is electrically connected to the display, keyboard, and mouse, respectively. The host computer processor has a main frequency of no less than 3.0 GHz, a memory of no less than 32 GB, an independent graphics card with no less than 16 GB of video memory, a hard disk of no less than 1 TB, and a display size of no less than 27 inches.

[0124] The guide mark automatic pasting control system is used to control the stereo camera module, visual recognition module, part grasping module, and mark pasting module to achieve imaging control, grasping control, pasting control, etc.

[0125] The guide mark automatic pasting control system includes a stereo camera module control circuit, a visual recognition module control circuit, a part grasping module control circuit, and a mark pasting module control circuit, which are electrically connected to the stereo camera module, the visual recognition module, the part grasping module, and the mark pasting module respectively. The guide mark automatic pasting control system is electrically connected to a high-performance computer.

[0126] The stereo camera module is used to perform laser scanning and multi-view imaging on the truss parts, and the stereo camera module is used to realize the positioning of the truss parts, and to realize the acquisition of part point cloud data and the positioning of the parts.

[0127] The stereo camera module includes an image sensor, a lens, a laser, and a laser galvanometer. The image sensor is used to generate digital images with a resolution of at least 10 megapixels and is provided in at least two pieces. The lens is used to transform light and has a field of view of at least 50°. The number of lenses matches the number of image sensors. The laser is used to generate line lasers, with at least seven line lasers. The laser galvanometer is used to scan the line lasers across the surface of the object being measured at a scanning frequency of at least 1 Hz.

[0128] The visual recognition module is used to image the truss parts and their labels, and the visual recognition module is used to recognize the truss parts, obtain the part images and label images, and recognize the label content.

[0129] The visual recognition module includes an industrial camera, a camera lens, an illumination source, and a mounting bracket. The industrial camera is used to generate digital images, with a resolution of at least 10 megapixels, and there are at least four of them. The camera lens is used to transform light, with a field of view of at least 40°, and the number of lenses is the same as that of the industrial camera. The illumination source is a white LED light source, which is used to illuminate the target. The mounting bracket is used to secure the industrial camera.

[0130] The part grasping module is used to grasp and adjust the posture of the truss parts. The part grasping module is used to adjust the posture of the truss parts, realize the posture adjustment during the part recognition process, and realize the posture adjustment and part fixation during the guide mark pasting process.

[0131] The part-grabbing module includes a grasping industrial robot arm and a grasping end effector. The grasping industrial robot arm is used to adjust the position and posture of the grasped truss parts. It is a six-degree-of-freedom serial industrial robot arm with a load capacity of no less than 10kg. The grasping end effector is used to grasp the truss parts, with a maximum grasping diameter of no less than Φ100mm and a maximum grasping mass of no less than 6kg.

[0132] The mark pasting module is used to pick up, adjust the position and paste the guide mark. The mark pasting module is used to paste the guide mark and realize the pasting of the assembly relationship guide mark at the specified position of the truss part.

[0133] The marker application module includes an industrial robotic arm and an end effector. The industrial robotic arm is a six-degree-of-freedom, serial-type industrial robotic arm with a payload of no less than 5 kg, used to adjust the position and posture of the guide marker. The end effector is used to pick up and apply the guide marker, using vacuum suction to pick up the marker and flexible pressure to apply the marker.

[0134] The workbench is used to place truss parts, high-performance computers, guide marking printers, install stereo camera modules, visual recognition modules, parts grasping modules, and marking and pasting modules, and provide operating space for the parts grasping modules and the marking and pasting modules;

[0135] The guide mark printer is used to print assembly relationship guide marks that can be pasted;

[0136] The guide mark is used to display assembly relationship guidance information to facilitate operators to assemble complex trusses.

[0137] In more preferred embodiments, a high-performance computer analyzes the assembly relationship based on the satellite complex truss three-dimensional model information and assembly process information, generates assembly relationship guidance mark information for all m parts, and establishes a parts assembly relationship guidance mark database;

[0138] Traversing all m parts, the stereo camera module shoots part i (i=1, 2, ..., m) to obtain point cloud data, and a high-performance computer analyzes and calculates the point cloud data to locate part i to obtain positioning information.

[0139] The part grasping module grasps part i according to the positioning information and brings it into the field of view of the visual recognition module.

[0140] The visual recognition module performs multi-view imaging on part i to obtain multiple images of part i.

[0141] High-performance computers detect part labels based on multiple images, calculate label poses, and determine occlusions.

[0142] The part grasping module adjusts the posture of part i according to the label posture.

[0143] The visual recognition module images the label of part i to obtain a label image.

[0144] A high-performance computer processes the label image, identifies the label image content, and retrieves the corresponding assembly relationship guidance mark information from the parts assembly relationship guidance mark database;

[0145] Traverse all guide marks of part i, and according to the guide mark information of the assembly relationship, the guide mark printer prints guide mark j;

[0146] Wherein, j=1, 2, ..., n, and n represents the number of guide marks of part i.

[0147] The marker pasting module picks up the guide marker j.

[0148] The high-performance computer plans the pasting path based on the pose of part i and guide marker j and determines the path solution;

[0149] The mark pasting module pastes the guide mark j at the specified position of part i, determines the task progress, and ends the pasting.

[0150] Specifically, the high-performance computer uses the SGBM stereo matching algorithm to analyze and calculate point cloud data and label poses, adopts the YOLO target detection algorithm to detect part labels, uses the optical character recognition algorithm to identify label image content, and adopts the fast exploration random tree algorithm to plan the pasting path.

[0151] The multi-view imaging has no fewer than four viewing angles, and obtains no fewer than four multiple images of part i.

[0152] The occlusion judgment is to judge whether the label of part i is blocked. If so, the part grasping module puts down part i and repositions part i to change the grasping position. If not, the part grasping module adjusts the posture of part i into place.

[0153] The posture of the part i is adjusted to be in place so that the normal vector at the center of the part label coincides with the optical axis of the camera lens of the visual recognition module and is within the depth of field of the visual recognition module.

[0154] The mark sticking module picks up the guide mark by vacuum adsorption and sticks the guide mark by flexible pressure.

[0155] The determination of the path solution is to determine whether the pasting path has a solution. If so, the marking pasting module adjusts the posture of the guide mark j. If not, the part grasping module adjusts the posture of the part i and replans the pasting path.

[0156] The judgment task process is to judge whether all guide marks of part i have been traversed. If not, set a=j, j=a+1, and the guide mark printer continues to print the next guide mark j. If so, judge whether all parts of the truss have been traversed. If so, end the pasting. If not, set a=i, i=a+1, and the high-performance computer continues to locate the next part i.

[0157] According to the present invention, a method for automatically pasting guide marks for the assembly relationship of complex satellite truss parts is provided. Figure 2 For example, the following steps are included:

[0158] Step S1: Open the satellite complex truss 3D model on a high-performance computer and analyze the assembly relationship using a secondary developed plug-in based on the model's own information and assembly process information;

[0159] Step S2: Automatically generate assembly relationship guidance mark information for all m truss parts, establish a truss part assembly relationship guidance mark database, and traverse all m parts of the satellite complex truss;

[0160] Step S3: The stereo camera module captures truss part i (i=1, 2, ..., m, where m is the number of parts) to obtain point cloud data. The point cloud data is analyzed and calculated using the SGBM (Semi-Global Block Matching) stereo matching algorithm on a high-performance computer to locate the truss part i.

[0161] Step S4: Based on the positioning information of the truss part i, the part grasping module automatically grasps the truss part i and places it within the field of view of the visual recognition module;

[0162] Step S5: Because part labels need to be identified to retrieve the corresponding assembly relationship, and part labels may be attached to any surface of the part, the visual recognition module is required to perform multi-view imaging of the truss part i, with at least four viewing angles, to obtain multiple images of the truss part i, with at least four images, so that the imaging range can cover any surface of the part and ensure that the part label image is obtained;

[0163] Step S6: Based on the multiple images of the truss part i, the YOLO (You Only Look Once) target detection algorithm is used on a high-performance computer to automatically detect the part labels in the image, and the SGBM stereo matching algorithm is used to calculate the label pose;

[0164] Step S7: Determine whether the label of the truss part i is blocked. If so, the part grabbing module puts down the truss part i and executes step S3 to change the grabbing position. If not, execute step S8.

[0165] Step S8: Since the optimal imaging range of the visual recognition module is limited, it is necessary to adjust the part pose according to the part label pose to better image the part label. Specifically, the part grasping module adjusts the pose of the truss part i so that the normal vector at the center of the part label coincides with the optical axis of the camera lens of the visual recognition module. At the same time, the part label is within the depth of field of the visual recognition module to ensure that the visual recognition module can obtain a vertical and clear image of the part label.

[0166] Step S9: the visual recognition module images the label of the truss part i to obtain a label image;

[0167] Step S10: Processing the label image of the truss part i in a high-performance computer, and automatically recognizing the label content using an optical character recognition (OCR) algorithm;

[0168] Step S11: according to the recognition result of the tag content of the truss part i, the corresponding assembly relationship guidance mark information is retrieved from the truss part assembly relationship guidance mark database;

[0169] Step S12: Traversing all guide marks of truss part i and printing guide mark j (j=1, 2, ..., n, where n is the number of guide marks of truss part i) according to the guide mark information of the assembly relationship;

[0170] Step S13: the mark sticking module automatically picks up the guide mark j through vacuum adsorption;

[0171] Step S14: Based on the pose of truss part i and the pose of guide marker j, a rapid-exploring random tree (RRT) algorithm is used in a high-performance computer to plan a pasting path and determine whether the pasting path has a solution. If so, step S15 is executed. If not, the part grasping module adjusts the pose of truss part i and executes step S14 to replan the pasting path.

[0172] Step S15: the mark pasting module adjusts the position of the guide mark j and automatically pastes the guide mark j at the designated position of the truss part i by flexible pressure;

[0173] Step S16: Determine whether all guide marks of truss part i have been traversed. If so, determine whether all parts of the truss have been traversed. If not, proceed to the next guide mark and execute step S12.

[0174] The process of determining whether all parts of the truss have been traversed is completed. If so, the process ends. If not, the process continues with the next part and executes step S3.

[0175] In more preferred examples, the high-performance computer includes a host, a display, a keyboard and a mouse, and the host is electrically connected to the display, the keyboard and the mouse respectively.

[0176] The guide mark automatic pasting control system is electrically connected to a high-performance computer, and includes a stereo camera module control circuit, a visual recognition module control circuit, a part grasping module control circuit, and a mark pasting module control circuit, which are electrically connected to the stereo camera module, the visual recognition module, the part grasping module, and the mark pasting module respectively.

[0177] The stereo camera module includes an image sensor, a lens, a laser and a laser galvanometer.

[0178] The visual recognition module includes an industrial camera, a camera lens, an illumination light source and a mounting bracket.

[0179] The part grasping module includes a grasping industrial robot arm and a grasping end effector.

[0180] The marking pasting module includes a pasting industrial robot arm and a pasting end effector.

[0181] The workbench provides an operating space for the parts grabbing module and the marking pasting module.

[0182] Specifically, the high-performance computer is used to set up, manage and analyze the automatic pasting process of guiding marks for the assembly relationship of complex satellite truss parts.

[0183] The guide mark automatic pasting control system is used to control the stereo camera module, the visual recognition module, the part grasping module, and the mark pasting module to achieve imaging control, grasping control, and pasting control.

[0184] The stereo camera module is used to perform laser scanning and multi-view imaging on parts.

[0185] The visual recognition module is used to image the parts and their labels.

[0186] The part grasping module is used to grasp and adjust the position of parts.

[0187] The marker pasting module is used to pick up, adjust the position and paste the guide marker.

[0188] The workbench is used to place truss parts, a high-performance computer and a guide marking printer, and to install a stereo camera module, a visual recognition module, a part grabbing module and a marking pasting module.

[0189] The guide mark printer is used to print assembly relationship guide marks that can be pasted.

[0190] The guide mark is used to display assembly relationship guide information.

[0191] In more preferred examples, the main frequency of the host processor is not less than 3.0 GHz, the memory is not less than 32 GB, the independent graphics card memory is not less than 16 GB, and the hard disk is not less than 1 TB.

[0192] The display size is not less than 27 inches.

[0193] The image sensor is used to generate digital images, has a resolution of not less than 10 million pixels, and has no less than 2 image sensors.

[0194] The lenses are used to transform light, have a field of view of no less than 50°, and their number is consistent with the number of image sensors.

[0195] The laser is used to generate line lasers, and the number of line lasers is no less than 7.

[0196] The laser galvanometer is used to scan the surface of the object to be measured with a line laser, and the scanning frequency is not less than 1 Hz.

[0197] The industrial cameras are used to generate digital images with a resolution of not less than 10 million pixels and the number is not less than 4.

[0198] The camera lens is used to transform light, has a field of view angle of not less than 40°, and the number is consistent with that of industrial cameras.

[0199] The illumination light source is a white LED light source, which is used to illuminate the target.

[0200] The mounting bracket is used to fix the industrial camera.

[0201] The grabbing industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 10 kg, and is used to adjust the position and posture of the grasped parts.

[0202] The maximum grasping diameter of the grasping end effector is not less than Φ100mm, and the maximum grasping mass is not less than 6kg, and it is used to grasp truss parts.

[0203] The pasting industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 5 kg, and is used to adjust the position and posture of the guide mark.

[0204] The pasting end effector is used to pick up and paste the guide mark.

[0205] During the process of pasting guide marks for the assembly relationship of complex satellite truss parts, the present invention realizes parameter setting, process management, data analysis and motion control through a high-performance computer and a guide mark automatic pasting control system, utilizes a stereo camera module to realize the positioning of truss parts, utilizes a visual recognition module to realize the recognition of truss parts, utilizes a part grasping module to realize the adjustment of the truss part posture, utilizes a mark pasting module to realize the pasting of guide marks, and finally realizes the automation of the guide mark pasting process, thereby improving the efficiency and digital level of satellite truss development.

[0206] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. A satellite complex truss parts assembly relationship guide mark automatic pasting system, characterized in that: include: High performance computer 101, guide mark automatic pasting control system 102, stereo camera module 103, visual recognition module 104, part grabbing module 105, mark pasting module 106, workbench 107, guide mark printer 108 and guide mark 109; The high-performance computer analyzes the assembly relationship based on the satellite complex truss 3D model information and assembly process information, generates assembly relationship guidance mark information for all m parts, and establishes a parts assembly relationship guidance mark database; Traverse all m parts, the stereo camera module shoots part i (i=1, 2, ..., m) to obtain point cloud data, and the high-performance computer analyzes and calculates the point cloud data to locate part i to obtain positioning information; The part grabbing module grabs part i into the field of view of the visual recognition module according to the positioning information; The visual recognition module performs multi-view imaging on part i to obtain multiple images of part i; High-performance computers detect part labels based on multiple images, calculate label poses, and determine occlusions; The part grasping module adjusts the position of part i according to the label position; The visual recognition module images the label of part i to obtain a label image; The high-performance computer processes the label image, identifies the label image content, and retrieves the corresponding assembly relationship guidance mark information in the parts assembly relationship guidance mark database; Traverse all guide marks of part i, and according to the guide mark information of the assembly relationship, the guide mark printer prints the guide mark j; Wherein, j = 1, 2, ..., n, n represents the number of guide marks of part i; The mark pasting module picks up the guide mark j; The high-performance computer plans the pasting path and determines the path solution according to the position and posture of the part i and the guide mark j; The mark pasting module pastes the guide mark j at the specified position of part i, determines the task progress, and ends the pasting.

2. The satellite complex truss parts assembly relationship guide mark automatic pasting system according to claim 1 is characterized in that: The high-performance computer analyzes and calculates point cloud data and label poses through the SGBM stereo matching algorithm, detects part labels using the YOLO target detection algorithm, recognizes label image content using the optical character recognition algorithm, and plans the pasting path using the fast exploration random tree algorithm; The multi-view imaging has no fewer than 4 viewing angles, and no fewer than 4 multiple images of part i are obtained; The occlusion determination is to determine whether the label of part i is blocked. If so, the part grasping module puts down part i and repositions part i to change the grasping position. If not, the part grasping module adjusts the position of part i into place. The position of the part i is adjusted to be in place so that the normal vector at the center of the part label coincides with the optical axis of the camera lens of the visual recognition module and is within the depth of field of the visual recognition module; The mark pasting module picks up the guide mark by vacuum adsorption and pastes the guide mark by flexible pressurization; The determining path solution is to determine whether the pasting path has a solution, if so, the marking pasting module adjusts the posture of the guide mark j, if not, the part grasping module adjusts the posture of the part i to re-plan the pasting path; The judgment task process is to judge whether all guide marks of part i have been traversed. If not, set a=j, j=a+1, and the guide mark printer continues to print the next guide mark j. If so, judge whether all parts of the truss have been traversed. If so, end the pasting. If not, set a=i, i=a+1, and the high-performance computer continues to locate the next part i.

3. The satellite complex truss parts assembly relationship guide mark automatic pasting system according to claim 1 is characterized in that: The high-performance computer comprises a host, a display, a keyboard and a mouse, wherein the host is electrically connected to the display, the keyboard and the mouse respectively; The guide mark automatic pasting control system is electrically connected to the high-performance computer, including a stereo camera module control circuit, a visual recognition module control circuit, a part grabbing module control circuit, and a mark pasting module control circuit, which are electrically connected to the stereo camera module, the visual recognition module, the part grabbing module, and the mark pasting module respectively; The stereo camera module includes an image sensor, a lens, a laser and a laser galvanometer; The visual recognition module includes an industrial camera, a camera lens, a lighting source and a mounting bracket; The part grabbing module includes a grabbing industrial robot arm and a grabbing end effector; The marking pasting module includes a pasting industrial robot arm and a pasting end effector; The workbench provides an operating space for the parts grabbing module and the marking pasting module.

4. The satellite complex truss parts assembly relationship guide mark automatic pasting system according to claim 3 is characterized in that: The host has a processor frequency of no less than 3.0 GHz, a memory of no less than 32 GB, an independent graphics card memory of no less than 16 GB, and a hard disk of no less than 1 TB; The display size is not less than 27 inches; The image sensor is used to generate digital images, has a resolution of not less than 10 million pixels, and the number is not less than 2; The lenses are used to transform light, have a field of view of no less than 50°, and their number is consistent with the number of image sensors; The laser is used to generate line lasers, and the number of line lasers is not less than 7; The laser galvanometer is used to scan the line laser on the surface of the object to be measured, and the scanning frequency is not less than 1 Hz; The industrial cameras are used to generate digital images, with a resolution of not less than 10 million pixels, and the number is not less than 4; The camera lenses are used to transform light, with a field of view of not less than 40°, and the number is consistent with that of industrial cameras; The illumination light source is a white LED light source, which is used to illuminate the target; The mounting bracket is used to fix the industrial camera; The grabbing industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 10kg, and is used to adjust the position and posture of the grabbed parts; The maximum grasping diameter of the grasping end effector is not less than Φ100mm, and the maximum grasping mass is not less than 6kg, and it is used to grasp truss parts; The pasting industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 5kg, and is used to adjust the position and posture of the guide mark; The pasting end effector is used to pick up and paste the guide mark.

5. The satellite complex truss parts assembly relationship guide mark automatic pasting system according to claim 3 is characterized in that: The high-performance computer is used to set, manage and analyze the automatic pasting process of the guide mark of the complex satellite truss parts assembly relationship; The guide mark automatic pasting control system is used to control the stereo camera module, the visual recognition module, the part grabbing module, and the mark pasting module to achieve imaging control, grabbing control, and pasting control; The stereo camera module is used to perform laser scanning and multi-view imaging on parts; The visual recognition module is used to image the parts and their labels; The part grasping module is used to grasp and adjust the position and posture of the parts; The marker pasting module is used to pick up, adjust the position and paste the guide marker; The workbench is used to place truss parts, high-performance computers and guide marking printers, and to install a stereo camera module, a visual recognition module, a part grabbing module and a marking pasting module; The guide mark printer is used to print assembly relationship guide marks that can be pasted; The guide mark is used to display assembly relationship guide information.

6. A method for automatically pasting guide marks for assembly relations of complex satellite truss parts, characterized in that: include: Step S1: analyzing the assembly relationship based on the satellite complex truss three-dimensional model information and assembly process information in a high-performance computer; Step S2: Instruct a high-performance computer to generate assembly relationship guidance mark information of all m parts, establish a parts assembly relationship guidance mark database, and traverse all m parts; Step S3: The stereo camera module is used to photograph part i (i=1, 2, ..., m) to obtain point cloud data, and a high-performance computer is used to analyze and calculate the point cloud data to locate part i to obtain positioning information; Step S4: instructing the part grabbing module to grab part i into the field of view of the visual recognition module according to the positioning information; Step S5: Instruct the visual recognition module to perform multi-view imaging on part i to obtain multiple images of part i; Step S6: Instruct a high-performance computer to detect part labels based on multiple images and calculate label poses; Step S7: Let the high performance computer determine the occlusion; Step S8: Instruct the part grabbing module to adjust the position of part i according to the label position; Step S9: Instruct the visual recognition module to image the label of part i to obtain a label image; Step S10: Processing the label image in a high-performance computer to identify the label image content; Step S11: enabling a high-performance computer to retrieve corresponding assembly relationship guide mark information from a parts assembly relationship guide mark database; Step S12: traverse all guide marks of part i, and the guide mark printer prints guide mark j according to the guide mark information of the assembly relationship; Wherein, j = 1, 2, ..., n, n represents the number of guide marks of part i; Step S13: instructing the mark pasting module to pick up the guide mark j; Step S14: Instruct the high-performance computer to plan the pasting path according to the position and posture of the part i and the guide mark j and determine the path solution; Step S15: instructing the mark pasting module to paste the guide mark j at the designated position of the part i; Step S16: Determine the task progress and end pasting.

7. The method for automatically pasting guide marks for assembly relations of satellite complex truss parts according to claim 6 is characterized in that: The high-performance computer analyzes and calculates point cloud data and label poses through the SGBM stereo matching algorithm, detects part labels using the YOLO target detection algorithm, recognizes label image content using the optical character recognition algorithm, and plans the pasting path using the fast exploration random tree algorithm; The mark pasting module picks up the guide mark by vacuum adsorption and pastes the guide mark by flexible pressurization; In step S5, the viewing angles of the multi-view imaging are not less than 4, and the multiple images of the part i obtained are not less than 4; In the step S7, it is determined whether the label of the part i is blocked. If so, the part grabbing module puts down the part i and executes the step S3. If not, the step S8 is executed. In step S8, the part grabbing module makes the normal vector at the center of the part label coincide with the optical axis of the camera lens of the visual recognition module and is within the depth of field of the visual recognition module; In the step S14, it is determined whether the pasting path has a solution. If so, step S15 is executed. If not, the part grabbing module adjusts the posture of part i and step S14 is executed to re-plan the pasting path. In the step S16, it is determined whether all guide marks of part i have been traversed. If not, a=j, j=a+1 are set and step S12 is executed. If yes, it is determined whether all parts of the truss have been traversed. Determine whether all parts of the truss have been traversed. If so, end the pasting. If not, set a=i, i=a+1 and execute step S3.

8. The method for automatically pasting guide marks for assembly relations of satellite complex truss parts according to claim 6 is characterized in that: The high-performance computer comprises a host, a display, a keyboard and a mouse, wherein the host is electrically connected to the display, the keyboard and the mouse respectively; The guide mark automatic pasting control system is electrically connected to the high-performance computer, including a stereo camera module control circuit, a visual recognition module control circuit, a part grabbing module control circuit, and a mark pasting module control circuit, which are electrically connected to the stereo camera module, the visual recognition module, the part grabbing module, and the mark pasting module respectively; The stereo camera module includes an image sensor, a lens, a laser and a laser galvanometer; The visual recognition module includes an industrial camera, a camera lens, a lighting source and a mounting bracket; The part grabbing module includes a grabbing industrial robot arm and a grabbing end effector; The marking pasting module includes a pasting industrial robot arm and a pasting end effector; The workbench provides an operating space for the parts grabbing module and the marking pasting module.

9. The method for automatically pasting guide marks for assembly relations of satellite complex truss parts according to claim 8 is characterized in that: The host has a processor frequency of no less than 3.0 GHz, a memory of no less than 32 GB, an independent graphics card memory of no less than 16 GB, and a hard disk of no less than 1 TB; The display size is not less than 27 inches; The image sensor is used to generate digital images, has a resolution of not less than 10 million pixels, and the number is not less than 2; The lenses are used to transform light, have a field of view of no less than 50°, and their number is consistent with the number of image sensors; The laser is used to generate line lasers, and the number of line lasers is not less than 7; The laser galvanometer is used to scan the line laser on the surface of the object to be measured, and the scanning frequency is not less than 1 Hz; The industrial cameras are used to generate digital images, with a resolution of not less than 10 million pixels, and the number is not less than 4; The camera lenses are used to transform light, with a field of view of not less than 40°, and the number is consistent with that of industrial cameras; The illumination light source is a white LED light source, which is used to illuminate the target; The mounting bracket is used to fix the industrial camera; The grabbing industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 10kg, and is used to adjust the position and posture of the grabbed parts; The maximum grasping diameter of the grasping end effector is not less than Φ100mm, and the maximum grasping mass is not less than 6kg, and it is used to grasp truss parts; The pasting industrial robot arm is a six-degree-of-freedom serial industrial robot arm with a load of no less than 5kg, and is used to adjust the position and posture of the guide mark; The pasting end effector is used to pick up and paste the guide mark.

10. The method for automatically pasting guide marks for assembly relations of satellite complex truss parts according to claim 8, characterized in that: The high-performance computer is used to set, manage and analyze the automatic pasting process of the guide mark of the complex satellite truss parts assembly relationship; The guide mark automatic pasting control system is used to control the stereo camera module, the visual recognition module, the part grabbing module, and the mark pasting module to achieve imaging control, grabbing control, and pasting control; The stereo camera module is used to perform laser scanning and multi-view imaging on parts; The visual recognition module is used to image the parts and their labels; The part grasping module is used to grasp and adjust the position and posture of the parts; The marker pasting module is used to pick up, adjust the position and paste the guide marker; The workbench is used to place truss parts, high-performance computers and guide marking printers, and to install a stereo camera module, a visual recognition module, a part grabbing module and a marking pasting module; The guide mark printer is used to print assembly relationship guide marks that can be pasted; The guide mark is used to display assembly relationship guide information.

Images