Intelligent detection method for launch vehicle cabin section

Abstract

The application provides a launch vehicle cabin section intelligent detection method, which is characterized in that the method uses laser three-dimensional scanning technology, image processing technology and intelligent control technology, and comprises the following steps: step S1: automatic acquisition of laser three-dimensional scanning point cloud data; step S2: automatic splicing of up-and-down station three-dimensional point cloud data; step S3: self-adaptive cutting of three-dimensional point cloud data; step S4: automatic identification and extraction of three-dimensional point cloud data edge contour; step S5: automatic identification of heading support image; and step S6: automatic generation of a detection report. The measurement process adopts a non-contact automatic measurement method, the system operation and tolerance evaluation method are fixed through programming, the detection reliability is improved, the personnel demand is reduced, the detection efficiency is improved, a digital measurement mode is adopted, a digital inspection report is automatically generated, the measurement data can be stored in a structured way, and necessary conditions are provided for inspection data analysis and digital factory construction.

Description

Technical Field

[0001] This invention relates to launch vehicle testing technology, and more particularly to an intelligent testing method for launch vehicle modules. Background Technology

[0002] To ensure that the dimensions of the launch vehicle modules meet the requirements for processing and assembly, that the internal supports are installed without errors or omissions, and that the holes and slots are correctly positioned, it is necessary to measure the distance between the front and rear end frames of the modules, parallelism, perpendicularity of the axis to the end face, flatness of the end face, roundness, reference torsion, and the dimensions of the supports, bow clamps, holes, and slots within the modules. Previously, traditional platform measurement methods or laser trackers were used to check the end face dimensions of the modules, while the installation positions, headings, and quantities of supports, brackets, and openings within the modules were checked manually using visual inspection, steel tape measures, and measuring tapes.

[0003] Existing methods require manual handheld measurement, resulting in poor measurement reliability, low efficiency, high labor intensity, low level of information-based measurement data, and the inability to automate the process. Visual inspection also carries a significant risk of errors and omissions. Summary of the Invention

[0004] To address the shortcomings of existing testing technologies, this invention provides an intelligent testing method for launch vehicle modules. The intelligent testing method for launch vehicle modules provided by this invention will be further described in detail below through specific embodiments and with reference to the accompanying drawings:

[0005] The specific advantages of this invention are as follows:

[0006] 1) The intelligent inspection method for launch vehicle modules described in this invention adopts a non-contact automatic measurement method in the measurement process. The system operation and tolerance evaluation methods are solidified through programming, realizing geometric dimension measurement and automatic judgment of errors and omissions, improving inspection coverage, reducing the influence of human factors, and improving inspection reliability.

[0007] 2) The intelligent inspection method for launch vehicle modules described in this invention employs intelligent measurement technology, overcoming the shortcomings of previous manual handheld and visual inspection methods that could not persistently store data. This reduces personnel requirements and improves inspection efficiency. The measurement process reduces the number of times modules need to be lifted and turned, eliminating the need for personnel to work at heights and improving inherent safety.

[0008] 3) The intelligent inspection method for launch vehicle modules described in this invention employs digital measurement, enabling visualized display of measurement results and automatic generation of digital inspection reports. The measurement data can be stored in a structured manner, providing necessary conditions for inspection data analysis and the construction of digital factories.

[0009] This invention provides an automatic measurement and judgment method for parameters such as the distance between the front and rear end frames of launch vehicle sections, the diameter of the fitted circle of the docking hole, roundness, perpendicularity, reference torsion, installation position, heading, and installation accuracy of internal supports and slots. Specifically, it is applied to the measurement of the geometric dimensions of launch vehicle sections and the determination of the presence, misinstallation, or omission of mounting supports. This measurement system and method can also be used for the intelligent measurement of similar structural products. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the measurement process for the intelligent detection method of launch vehicle sections according to the present invention.

[0011] Figure 2 This is a schematic diagram of the composition of the intelligent measurement system for launch vehicle sections according to the present invention.

[0012] Explanation of reference numerals in the attached figures:

[0013] 1-Laser 3D scanning system; 2-Lifting mechanism; 3-Launch vehicle section

[0014] 4-Channel support fixtures 5-Sunken foundation 6-Ground Detailed Implementation

[0015] The present invention will be described in detail below with reference to specific implementation examples. The following embodiments will help those skilled in the art to further understand the present invention, but do not limit the present invention in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention.

[0016] The core technology of this invention is to achieve intelligent inspection of launch vehicle sections using a detection method that integrates laser three-dimensional scanning technology, image processing technology, and intelligent control technology.

[0017] The intelligent inspection method for launch vehicle modules provided by this invention uses laser 3D scanning technology, image processing technology, and intelligent control technology. It completes comprehensive point cloud and image data acquisition through automatic station switching. After automatically stitching and cropping the point cloud data from different stations, it calculates the spatial distance of cluster features through automatic detection of point cloud edge contours and feature extraction. It also uses model matching methods to identify heading features, realizing automatic measurement and judgment of the distance between the front and rear end frames of the module, the diameter of the fitting circle of the docking hole, roundness, perpendicularity, reference torsion, installation position, heading, and installation accuracy of the internal support and slots. It has the advantages of high detection efficiency and good consistency of measurement results.

[0018] Figure 1 This is a schematic diagram of the measurement process for the intelligent detection method of launch vehicle sections according to the present invention. Figure 1As shown, the intelligent detection method for launch vehicle sections of the present invention includes the following steps:

[0019] Step S1: Automatic acquisition of laser 3D scanning point cloud data.

[0020] The automatic control lidar range measurement and scanning method automatically acquires point cloud data. Specifically, it automatically controls the laser transmitter to emit a modulated continuous wave towards the target. After the continuous wave reaches the target, it is reflected back. The distance between the measuring device and the target is calculated based on the wavelength of the modulated wave and the phase delay between the transmitted and received waves. Then, the lidar pan-tilt unit automatically scans 360 degrees to obtain the three-dimensional coordinates of the point cloud data.

[0021] Step S2: Automatic stitching of 3D point cloud data of upper and lower stations.

[0022] The circle center fitting algorithm is used to calculate the plane coordinates of the circle center of multiple cross sections with different distance intervals in two sets of point clouds. At the same time, the edge data of the cylinder at the starting scan position of the same cross section is obtained. The point cloud containing the overlapping area of ​​the circle center and the edge of the cylinder is matched point by point by point using the two-point distance matching method. The rotation and translation matrix and the translation matrix are calculated by the least squares method, which is then extended to all point clouds to complete the overall automatic stitching of the two sets of point clouds.

[0023] Step S3: Adaptive cropping of 3D point cloud data.

[0024] Bounding box technology is used to achieve adaptive cutting and trimming based on axis alignment. Considering that the measurement parameters are all calculated using point cloud data of the upper and lower end faces, the cutting and trimming of the upper and lower end faces of the cabin is completed by setting the boundary values ​​of the bounding box.

[0025] Step S4: Automatic identification and extraction of the edge contours of the 3D point cloud data.

[0026] Automatic edge contour extraction is achieved by searching for closed feature points in cylindrical point cloud data. A cylindrical curve network is constructed on the cabin point cloud. The point cloud is clustered into point clusters using the normal estimation region growing method to further reduce the point cloud size. A graph of these point clusters is built and closed feature lines matching the clusters are extracted using first-order vector feature segmentation. The area where the cabin surface contour can be defined is found, and the spatial distance of the cluster features is calculated.

[0027] Step S5: Automatic recognition of heading support image.

[0028] Automatic image recognition of heading support based on machine vision technology is achieved by using model matching technology. First, a target characteristic model to be identified is established, taking into account the changes in the position of feature edges. The image to be identified is compared with the known model to obtain a similarity measure between the two, thereby identifying and classifying the target.

[0029] Step S6: Automatic generation of the test report.

[0030] By constructing and evaluating the measured point cloud, geometric dimension templates such as the distance between the front and rear end frames of the compartment, parallelism, perpendicularity of the axis to the end face, flatness of the end face, roundness, and reference torsion are obtained according to the definition. The digital model is imported and compared to automatically generate a measurement and inspection report on the dimensions of the brackets, bow clamps, and hole slots in the compartment, as well as the judgment of mis-installation.

[0031] This invention can be used to measure the distance between the front and rear end frames of launch vehicle sections, parallelism, perpendicularity of the axis to the end face, flatness of the end face, roundness, reference torsion, and the dimensions of supports, bow clamps, holes, and slots within the section. Specifically, it is applied to the geometric dimension measurement of launch vehicle sections and the determination of the presence, misinstallation, or omission of mounting brackets. This measurement system and method can also be used for the intelligent measurement of similar structural products.

[0032] Figure 2 This is a schematic diagram of the composition of the intelligent measurement system for launch vehicle sections of the present invention, as shown below. Figure 2 As shown, the system of the present invention mainly consists of a laser three-dimensional scanning system 1, a lifting mechanism 2, and a section support fixture 4. The laser three-dimensional scanning system 1 is installed on top of the lifting mechanism 2, which is installed in a sunken foundation 5. At the start of the measurement, the launch vehicle section 3 is placed on the support fixture 4, the lifting mechanism 2 is located in the sunken foundation 5, and the optical scanning sensor is below the ground 6. During the measurement, the lifting mechanism 2 drives the optical scanning measurement sensor 1 to rise to each station for measurement.

[0033] Before measurement, a testing and operation plan is prepared in advance according to the specifications of the compartment. The entire testing system operation and data processing process can be completed with one click. During measurement, the lifting mechanism drives the laser 3D scanning system to complete multi-station scanning and point acquisition; the software system automatically stitches the 3D point cloud data from multiple stations, performs adaptive cropping and recognition extraction of the Sian Cloud data according to the set rules, and automatically recognizes the images of the internal support structure.

[0034] It should be noted that the above description is merely illustrative and explanatory of the present invention. Those skilled in the art should understand that any modifications and substitutions to the present invention fall within the scope of protection of the present invention.

Claims

1. An intelligent detection method for launch vehicle modules, characterized in that, It utilizes laser 3D scanning technology, image processing technology, and intelligent control technology, and includes the following steps: Step S1: Automatic acquisition of laser 3D scanning point cloud data; Step S2: Automatic stitching of 3D point cloud data of upper and lower stations; Step S3: Adaptive cropping of 3D point cloud data; Step S4: Automatic identification and extraction of edge contours from 3D point cloud data; Step S5: Automatic image recognition of the heading support; Step S6: Automatic generation of the test report. Step S1 includes: automatically controlling the ranging and scanning of the lidar to automatically acquire point cloud data, that is, automatically controlling the laser emitter to emit a modulated continuous wave towards the target to be measured. After the continuous wave reaches the target to be measured, it is returned. The distance between the measuring device and the target to be measured is calculated based on the wavelength of the modulated wave and the phase delay between the emitted wave and the received wave. Then, the three-dimensional coordinates of the point cloud data can be obtained by the lidar pan-tilt unit automatically scanning 360 degrees. Step S2 includes: using a center-fitting algorithm to calculate the center plane coordinates of multiple cross-sections with different distance intervals in two sets of point clouds, and simultaneously obtaining the edge data of the cylinder at the starting scanning position of the same cross-section. The two sets of point clouds containing the overlapping areas of the center and the edge of the cylinder are matched point-to-point using the two-point distance matching method. The rotation and translation matrix and the translation matrix are calculated using the least squares method, thereby extending to all point clouds to complete the overall automatic stitching of the two sets of point clouds. The bounding box technique is used to achieve adaptive cutting and trimming based on axis alignment. Considering that the measurement parameters are all calculated using point cloud data of the upper and lower end faces, the cutting and trimming of the upper and lower end faces of the cabin is completed by setting the bounding box boundary values. Step S4 includes: automatically extracting edge contours by searching for closed feature points in cylindrical point cloud data; constructing a cylindrical curve network on the compartment point cloud; clustering the point cloud into point clusters using the normal estimation region growing method; further reducing the point cloud size; establishing a graph of these point clusters; and extracting closed feature lines matching the clusters using first-order vector feature segmentation; finding the definable region of the compartment surface contour; and calculating the spatial distance of the cluster features. The intelligent measurement system for launch vehicle sections used in the intelligent detection method for launch vehicle sections includes a laser 3D scanning system, a lifting mechanism, and section support fixtures. The laser 3D scanning system is installed on top of the lifting mechanism, which is installed within a sunken foundation. At the start of the measurement, the launch vehicle section is placed on the support fixture, the lifting mechanism is located within the sunken foundation, and the optical scanning sensor is below ground level. During measurement, the lifting mechanism raises the optical scanning sensor to each station for measurement. Before measurement, a testing and operation plan is prepared in advance according to the specifications of the compartment. During measurement, the lifting mechanism drives the laser three-dimensional scanning system to complete multi-station scanning and point acquisition. The software system automatically stitches the three-dimensional point cloud data from multiple stations, performs adaptive cropping and recognition extraction of the point cloud data according to the set rules, and automatically recognizes the images of the support structure inside the compartment.

2. The intelligent detection method for launch vehicle sections according to claim 1, characterized in that, Step S5 includes: automatic recognition of heading support images based on machine vision technology, using model matching technology to first establish a target characteristic model to be recognized, considering the changes in the position of feature edges, comparing the image to be recognized with the known model, thereby obtaining a similarity measure between the two, and thus recognizing and classifying the target.

3. The intelligent detection method for launch vehicle sections according to claim 2, characterized in that, Step S6 includes: constructing and evaluating the measured point cloud, obtaining a geometric dimension template according to the definition, importing and comparing the digital model, and automatically generating a dimension measurement and mis-installation judgment and detection report for the internal support, bow clamp, and slot positions of the compartment.

4. The intelligent detection method for launch vehicle sections according to claim 3, characterized in that, In step S6, the geometric dimension template includes geometric dimension templates for the distance between the front and rear end frames of the compartment, parallelism, perpendicularity of the axis to the end face, flatness of the end face, roundness, and reference torsion; the dimension measurement and misassembly judgment include dimension measurement and misassembly judgment of the internal support, bow-shaped clamp, and slot positions of the compartment.

Images