CN-122016807-A - Automatic measurement method and system for white parts of automobile

CN122016807ACN 122016807 ACN122016807 ACN 122016807ACN-122016807-A

Abstract

The invention discloses an automatic measurement method and system for automobile white parts, which relate to the technical field of automobile detection and comprise the steps of positioning and clamping an automobile white part in a clamping tool, and forming a rough positioning clamping state based on a first clamping mechanism; the method comprises the steps of identifying and outputting coarse positioning coordinates through a machine vision system arranged on a detection station, planning a target gesture of an automobile body-in-white for optimal detection based on the coarse positioning coordinates, controlling a plurality of second clamping mechanisms to conduct height adjustment according to the target gesture and form a fine positioning clamping state in cooperation with the first clamping mechanisms, constructing a detection path according to the actual gesture of the automobile body-in-white, controlling a flexible executing mechanism carrying a high-precision sensor to conduct accurate detection analysis on the automobile body-in-white along the detection path, and outputting detection data. Through coarse positioning clamping, image acquisition and preliminary analysis, gesture planning and conversion, detection path construction and accurate detection analysis, the problems of detection blind areas and low efficiency caused by fixed gestures are solved.

Inventors

WANG SHUAI
LIU JU

Assignees

广州特斯克仪器科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260129

Claims (10)

1. The automatic measurement method for the automobile white part is characterized by being suitable for an automobile white part clamping tool, the clamping tool comprises a workbench, two groups of first clamping mechanisms symmetrically distributed on two sides of the workbench and a second clamping mechanism arranged between the two groups of first clamping mechanisms, and the measurement method comprises the following steps: Positioning and clamping the automobile white body in the clamping tool, and carrying out initial positioning and clamping on the automobile white body based on a first clamping mechanism to form a rough positioning and clamping state; the method comprises the steps of performing image acquisition and preliminary analysis on the integral structural profile and the connecting position of the automobile white body in the rough positioning clamping state through a machine vision system arranged on a detection station, and identifying and outputting rough positioning coordinates of more than one defect area to be confirmed; planning a target attitude of the white automobile body for optimizing detection based on the rough positioning coordinates, controlling a plurality of second clamping mechanisms to adjust the heights according to the target attitude, and converting the white automobile body from the rough positioning clamping state to a fine positioning clamping state corresponding to the target attitude in cooperation with the first clamping mechanisms; according to the rough positioning coordinates and the actual body-in-white posture of the automobile in the fine positioning clamping state, a detection path is constructed; And controlling a flexible executing mechanism carrying a high-precision sensor to accurately detect and analyze the white automobile body of the automobile along the detection path, and outputting detection data.
2. The automated measurement method of automotive white body parts according to claim 1, wherein the step of performing image acquisition and preliminary analysis on the overall structural profile and the connection position of the automotive white body in the rough positioning and clamping state by a machine vision system disposed on the detection station, and the step of identifying and outputting rough positioning coordinates of one or more defect areas to be confirmed comprises: acquiring a car body panoramic image through a global scanning camera, and matching the car body panoramic image with a preset car body CAD model; The method comprises the steps of projecting a CAD model onto a matched car body panoramic image, and searching an actual edge corresponding to a projection contour and feature points in the car body panoramic image by utilizing an edge detection algorithm to obtain key feature points of the car body contour of the white car body; on the basis of identifying the outline and key characteristic points of the automobile body, preliminary defect detection is carried out on the surface of the white body of the automobile, the area possibly with the defect is screened out, and the three-dimensional rough positioning coordinates of the area to be confirmed in the initial coordinate system are output.
3. The automated measurement method of automotive body-in-white components of claim 1, wherein planning for optimal detection of an automotive body-in-white target pose based on the coarse positioning coordinates comprises: grouping the coarse positioning coordinates based on a K-means clustering algorithm, and dividing distribution data of the coarse positioning coordinates on a white body of an automobile; And extracting concentrated distribution positions of the defect region to be detected from the distribution data of the coarse positioning coordinates, and selecting a corresponding target gesture from a preset gesture library based on the concentrated distribution positions, wherein the target gesture comprises one of a forward tilting gesture, a backward tilting gesture, a side tilting gesture and a compound tilting gesture.
4. The automated measurement method of automotive white parts according to claim 1, wherein controlling the plurality of second clamping mechanisms to adjust the heights according to the target attitude, and cooperating with the first clamping mechanism, the converting the automotive white body from the rough positioning clamping state to the fine positioning clamping state corresponding to the target attitude includes: Calculating the target height required by each second clamping mechanism according to the selected target gesture and the CAD model of the white automobile body; Controlling the servo lifting columns of the second clamping mechanisms to move to the target height, and monitoring and adjusting the clamping force of each second clamping mechanism in real time through a force sensor on each clamping mechanism in the adjusting process; and clamping and fixing the white automobile body of the automobile based on a plurality of second clamping mechanisms to form a precise positioning and clamping state.
5. The automated measurement method of automotive white parts according to claim 1, wherein the constructing a detection path according to the rough positioning coordinates and the actual attitude of the automotive white body in the fine positioning and clamping state comprises: Calculating a space transformation parameter of the actual posture of the vehicle body based on the height value of each second clamping mechanism in the fine positioning clamping state; converting the coarse positioning coordinates to the current robot coordinate system by utilizing the space transformation parameters; and carrying out reachability analysis and collision detection simulation in the converted coordinate point set, and planning out an optimal motion sequence and a smooth track of the flexible executing mechanism.
6. The automated measurement method of automotive white parts according to claim 5 wherein calculating the spatial transformation parameters of the actual pose of the vehicle body based on the height values of the respective second clamping mechanisms in the precisely positioned clamping state comprises: When the white body of the automobile is subjected to fine positioning clamping, acquiring the pose change data of the white body of the automobile, and acquiring the theoretical height value of each second clamping mechanism according to the pose change data of the automobile; constructing a sensitivity matrix according to the mapping relation between the vehicle body pose change data and the theoretical height value; and acquiring the actual height value of each second clamping mechanism, and calculating the space transformation parameter of the actual posture of the vehicle body by combining the theoretical height value and the sensitivity matrix.
7. The automated measurement method of automotive white body parts of claim 6, wherein converting the coarse positioning coordinates to the current robot coordinate system using the spatial transformation parameters comprises: Correcting the coarse positioning coordinates based on the space transformation parameters to obtain a plurality of corresponding coordinate points of the coarse positioning coordinates on a white body digital model of the automobile; Acquiring a relative position relation of a white body digital model of the tail end detection tool under a robot coordinate system according to an installation mode of the tail end detection tool of the robot; and converting the corresponding coordinate points into motion node coordinates of a robot coordinate system by combining the relative position relations.
8. The automated measurement method of automotive body-in-white components of claim 1, wherein controlling the flexible actuator carrying the high-precision sensor to perform accurate detection analysis on the automotive body-in-white along the detection path and outputting detection data comprises: Dynamically selecting at least two fixed cameras with optimal visual angles to synchronously acquire images according to the actual gesture and the defect position, and acquiring fixed detection data; simultaneously controlling an end sensor carried by the flexible executing mechanism to move to a supplementary observation position for data acquisition, and acquiring dynamic detection data; and mapping the fixed detection data and the dynamic detection data on the surface features of the same three-dimensional vehicle body model for fusion analysis, and outputting the detection data of the white body of the vehicle.
9. The automated measurement method of automotive white body parts of claim 1, further comprising: collecting and recording key dimension measurement values, defect statistical data and corresponding process parameters of each automobile body in white; Drawing a continuous production data control chart based on the critical dimension measurement value, the defect statistical data and corresponding process parameters; And acquiring a detection change trend of the automatic measurement system based on the connection production data control diagram, and generating real-time early warning information if the detection change trend is abnormal.
10. An automated measurement system for automotive white parts, wherein the measurement system is configured to perform the automated measurement method of any one of claims 1 to 9, the measurement system comprising: the rough positioning clamping assembly is used for positioning and clamping the automobile white body in the clamping tool, and performing initial positioning clamping on the automobile white body based on the first clamping mechanism to form a rough positioning clamping state; the coarse positioning analysis component is used for carrying out image acquisition and preliminary analysis on the integral structure profile and the connecting position of the automobile white body in the coarse positioning clamping state through a machine vision system arranged on the detection station, and identifying and outputting coarse positioning coordinates of more than one defect area to be confirmed; the fine positioning clamping assembly is used for planning a target attitude of the white automobile body for optimizing detection based on the rough positioning coordinates, controlling a plurality of second clamping mechanisms to adjust the height according to the target attitude, and converting the white automobile body of the automobile from the rough positioning clamping state to a fine positioning clamping state corresponding to the target attitude in cooperation with the first clamping mechanism; The detection path planning component is used for constructing a detection path according to the rough positioning coordinates and the actual body-in-white posture of the automobile in the fine positioning clamping state; the detection component is used for controlling the flexible executing mechanism carrying the high-precision sensor to accurately detect and analyze the white automobile body of the automobile along the detection path and outputting detection data.

Description

Automatic measurement method and system for white parts of automobile Technical Field The invention relates to the technical field of automobile detection, in particular to an automatic detection and system for white parts of an automobile. Background In the automobile manufacturing process, the quality detection of the white body of the automobile is a core link for guaranteeing the structural integrity and the safety performance of the whole automobile. Currently, the industry generally adopts a combination mode of a fixed clamping tool matched with a photoelectric sensor and a shooting camera to detect the connection position of parts of an automobile white body and the surface defects of an integral structure. The fixed tool locks the white body of the automobile in a single gesture, so that the detection operation is limited to a preset path, and the gesture of the automobile body cannot be dynamically adjusted according to the actual defect distribution. The limitation causes the detection of complex curved surface areas, internal structures or hidden parts to have obvious blind areas, so that the full coverage is difficult to realize, and the missed detection or misjudgment of surface defects is easy to cause. Meanwhile, the fixed posture severely limits the observation view angle of the sensor, and reduces the accuracy and consistency of defect identification. In addition, when detecting different regions, operating personnel need repeatedly dismantle, relocation frock or automobile body, have not only prolonged single detection cycle, have still introduced extra manual operation error, and whole detection efficiency is low, is difficult to adapt to modern car production line and detects the urgent demand of high accuracy, high efficiency. These problems directly limit the automation level and quality control capabilities of the detection process. Disclosure of Invention The invention aims to overcome the defects of the prior art, provides an automatic measurement method and system for white parts of an automobile, and aims to solve the problems of the prior art by coarse positioning and clamping, image acquisition and preliminary analysis, gesture planning and conversion, detection path construction and accurate detection analysis, the problem of fixed gesture detection blind area and inefficiency that leads to has been solved, has the dynamic gesture adjustment that realizes the white body of car, reduces the detection blind area, promotes the advantage of detection precision and efficiency. The invention provides an automatic measurement method of automobile white body parts, which is suitable for an automobile white body clamping tool, wherein the clamping tool comprises a workbench, two groups of first clamping mechanisms symmetrically distributed on two sides of the workbench and a second clamping mechanism arranged between the two groups of first clamping mechanisms, and the measurement method comprises the following steps: Positioning and clamping the automobile white body in the clamping tool, and carrying out initial positioning and clamping on the automobile white body based on a first clamping mechanism to form a rough positioning and clamping state; the method comprises the steps of performing image acquisition and preliminary analysis on the integral structural profile and the connecting position of the automobile white body in the rough positioning clamping state through a machine vision system arranged on a detection station, and identifying and outputting rough positioning coordinates of more than one defect area to be confirmed; planning a target attitude of the white automobile body for optimizing detection based on the rough positioning coordinates, controlling a plurality of second clamping mechanisms to adjust the heights according to the target attitude, and converting the white automobile body from the rough positioning clamping state to a fine positioning clamping state corresponding to the target attitude in cooperation with the first clamping mechanisms; according to the rough positioning coordinates and the actual body-in-white posture of the automobile in the fine positioning clamping state, a detection path is constructed; And controlling a flexible executing mechanism carrying a high-precision sensor to accurately detect and analyze the white automobile body of the automobile along the detection path, and outputting detection data. Further, the image acquisition and the preliminary analysis are performed on the integral structure profile and the connection position of the white body of the automobile in the rough positioning clamping state through a machine vision system arranged on the detection station, and the identification and output of the rough positioning coordinates of more than one defect area to be confirmed comprise: acquiring a car body panoramic image through a global scanning camera, and matching the car body panoramic image with a preset car body CAD mo