CN-121978120-A - Automobile part quality detection method based on multi-sensor fusion

CN121978120ACN 121978120 ACN121978120 ACN 121978120ACN-121978120-A

Abstract

The invention relates to the technical field of automobile part detection, and particularly discloses an automobile part quality detection method based on multi-sensor fusion, which comprises the following steps: the four-station rotary detection device is used for driving the rotary platform to switch the corresponding sensor base to the working position according to the type of the identified part, loading the corresponding detection program and parameters after the sensor is in place to complete detection preparation, and S4 is used for conveying the parts to the detection station, driving the in-place sensor to execute high-precision detection according to the preset path and parameters by the four-station rotary detection device and outputting the result, thereby solving the technical problems that the existing single automobile part quality detection cannot adapt to different types of automobile parts and cannot be dynamically adjusted.

Inventors

WANG YUEQING
JIANG PEIJUN
Xia Zixuan
LI YUJUAN
Shao Kangsheng
SHI QINGFENG
LI SHUANG
Liao Yingxu

Assignees

重庆科技大学

Dates

Publication Date: 20260505
Application Date: 20260130

Claims (10)

1. A method for detecting the quality of automobile parts based on multi-sensor fusion is characterized by comprising the following steps: S1, placing the automobile parts on a conveyor belt, and conveying the automobile parts to a detection area by the conveyor belt. S2, acquiring part images through a camera, and determining specific types and models of the part images through image recognition. And S3, driving the rotary platform to switch the corresponding sensor base to the working position by the four-station rotary detection device according to the identified part type, and loading the corresponding detection program and parameters after the sensor is in place to finish detection preparation. And S4, conveying the parts to a detection station, driving the in-place sensor by the four-station rotary detection device to execute high-precision detection according to a preset path and parameters, and outputting a result.
2. The automobile part quality detection method based on multi-sensor fusion of claim 1, wherein the four-station rotary detection device comprises an industrial camera, the industrial camera is arranged on a conveying path of a conveying belt, a mechanical arm is arranged on the side wall of the conveying belt, a chassis is arranged at the telescopic end of the mechanical arm, a protection cylinder is arranged on the chassis, a first motor is arranged in the protection cylinder, the first motor is fixedly connected with the chassis, a power output shaft of the first motor is fixedly connected with a rotary cylinder, and a plurality of sensor fixing assemblies are arranged on the periphery of the rotary cylinder; one of the plurality of sensor fixing assemblies is used for fixing a high-precision 3D structure light camera so as to perform full-size scanning on the precision casting; The other one of the plurality of sensor fixing assemblies is used for fixing the thermal infrared imager so as to perform nondestructive inspection on the welding seams and the internal structures of the identified welding structural parts; Another one of the plurality of sensor fixing assemblies is used for fixing a high-resolution industrial CCD camera so as to detect surface flaws of the identified stamping appearance piece; Another one of the plurality of sensor fixing assemblies is used for fixing a high-precision 3D line laser profile measuring instrument so as to measure the adhesive tape of the identified adhesive coating component; The chassis is provided with a deriving component, the deriving component is used for deriving one of a plurality of corresponding sensor fixing components, the chassis is provided with a detection module, and the detection module is used for identifying and analyzing acquired image data of the automobile part and generating a detection control instruction based on an analysis result.
3. The method for detecting the quality of the automobile parts based on the multi-sensor fusion of claim 2, wherein the plurality of sensor fixing assemblies comprise bearing cylinders, the bearing cylinders are slidably arranged on the outer wall of the rotating cylinder, the bearing cylinders are used for loading corresponding sensors, and sliding blocks are arranged on the bearing cylinders.
4. The method for detecting the quality of the automobile parts based on the multi-sensor fusion of claim 2, wherein the guiding-out assembly comprises a sliding frame, the sliding frame is arranged on the upper side of the bottom plate, a sliding moving plate is arranged in the sliding frame, a fixed plate is fixedly connected below the sliding frame, a second motor is arranged on the fixed plate, the power output end of the second motor is fixedly connected with a screw rod, one end of the screw rod, which is far away from the second motor, is in threaded connection with the moving plate, a clamping block is fixedly connected below the moving plate, a second sliding groove which can be in sliding contact with the clamping block is formed in the sliding block, and a second opening through which the bearing cylinder can pass is formed in the protecting cylinder.
5. The method for detecting the quality of the automobile parts based on the multi-sensor fusion of claim 4, wherein the plurality of sensor fixing assemblies further comprise sliding strips, the sliding strips are fixedly connected with the bearing cylinder, a first sliding groove is formed in the outer wall of the rotating cylinder, and the sliding strips are slidably arranged in the first sliding groove.
6. The method for detecting the quality of the automobile parts based on the multi-sensor fusion according to claim 2, wherein the detecting module comprises an image recognition and decision-making sub-module, a rotation driving control sub-module and a linear deriving control sub-module, the image recognition and decision-making sub-module is used for recognizing the collected image of the automobile parts and generating a control instruction, the rotation driving control sub-module is used for driving the rotation cylinder to rotate according to the control instruction so as to switch the detecting sensors, and the linear deriving control sub-module is used for controlling the deriving mechanism to position the selected sensors to the detecting stations.
7. The method for detecting the quality of the automobile parts based on the multi-sensor fusion of claim 6, wherein the image recognition and decision sub-module is used for receiving and analyzing the image of the automobile parts shot by the vision acquisition equipment, recognizing the type of the current part according to a preset part feature library, and generating a control instruction containing the station information of the target sensor according to the type of the current part.
8. The method for detecting the quality of the automobile parts based on the multi-sensor fusion of claim 6, wherein the rotary driving control sub-module is used for calculating a target station angle according to a control instruction sent by the image recognition and decision sub-module, controlling the first motor to drive the rotary cylinder and the sensor fixing assembly to rotate, and positioning the bearing cylinder bearing the corresponding detection sensor to a position ready to be led out.
9. The method for detecting the quality of the automobile parts based on the multi-sensor fusion of claim 6, wherein the linear derivation control submodule is used for controlling the second motor to drive the derivation assembly after receiving the rotation in-place signal, and stably pushing out the aligned bearing cylinder and the detection sensor on the bearing cylinder from the protection cylinder to a preset detection station.
10. The method for detecting the quality of the automobile parts based on the multi-sensor fusion of claim 1, further comprising the detection of a controller, wherein the detection comprises a dynamometer rack for simulating different driving conditions, and is characterized in that: the system also comprises a vehicle simulation platform, a dynamics model construction module, a scene simulation module, an upper computer and a real-time machine module; The dynamics model building module is used for building a vehicle dynamics model; the vehicle simulation platform is used for installing the intelligent driving domain controller to be tested; the scene simulation module is used for establishing a 3D simulation scene; The upper computer is used for receiving feedback data in the testing process and controlling the dynamometer rack, the real-time machine module, the dynamics model building module and the vehicle simulation platform according to a preset testing flow and parameters; When the intelligent driving area controller is tested, the real-time machine module and the scene simulation module conduct data interaction to obtain sensor data, the real-time machine module transmits the sensor data to the intelligent driving area controller, the intelligent driving area controller conducts algorithm path planning according to the sensor data and current state information of a vehicle, then an execution instruction is fed back to the real-time machine module, the dynamometer machine rack feeds back working condition information to the real-time machine module, the real-time machine module brings the working condition information and the execution instruction into the vehicle dynamics model to conduct solving operation to obtain response information, the response information is transmitted to the scene simulation module, and the scene simulation module conducts real-time updating of a 3D simulation scene according to the response information.

Description

Automobile part quality detection method based on multi-sensor fusion Technical Field The invention relates to the technical field of automobile part detection, and particularly discloses an automobile part quality detection method based on multi-sensor fusion. Background In the automobile manufacturing industry, the quality detection of parts is a key link for ensuring the safety and reliability of the whole automobile, currently, aiming at the detection of different types of defects (such as out-of-tolerance size, surface defects and welding internal defects), single or fixed multi-sensor schemes are commonly adopted in the industry, and common practice includes that an industrial camera special for two-dimensional visual detection, a structured light scanner for three-dimensional size measurement, a thermal infrared imager for nondestructive inspection and other devices are respectively deployed along a conveyor belt of a production line or at different stations, each device works independently to finish the preset single type detection task, and for complex parts, a plurality of independent detection stations are needed to be sequentially carried out, or workpieces are transferred to different special detection devices to finish the all-round quality assessment. However, the existing detection mode has a certain problem that firstly, a plurality of detection devices or sensors with single functions are required to be installed at different positions of a production line in a scattered way, so that the whole detection system is large in occupied area, the layout of a wire body is complex, the purchase, installation and maintenance costs of the devices are obviously increased, and because the functions and the positions of the sensors are fixed, when the production line is switched to different types of automobile parts, for example, the engine cylinder body which needs three-dimensional scanning is switched to a car body welding part which needs thermal imaging detection, the existing system cannot automatically select the best matched sensor, and the optimal detection visual angle and the sensor cooperative working mode cannot be dynamically adjusted for specific parts, so that the complex detection flow and low efficiency are directly caused, and the requirements of modern flexible manufacturing on quick production change and intelligent detection are difficult to be met. Disclosure of Invention In view of the above, the present invention aims to provide a method for detecting quality of automobile parts based on multi-sensor fusion, so as to solve the technical problem that the existing single quality detection of automobile parts cannot adapt to different types of automobile parts and cannot be dynamically adjusted. In order to achieve the aim, the invention provides the following technical scheme that the automobile part quality detection method based on multi-sensor fusion comprises the following steps: S1, placing the automobile parts on a conveyor belt, and conveying the automobile parts to a detection area by the conveyor belt. S2, acquiring part images through a camera, and determining specific types and models of the part images through image recognition. And S3, driving the rotary platform to switch the corresponding sensor base to the working position by the four-station rotary detection device according to the identified part type, and loading the corresponding detection program and parameters after the sensor is in place to finish detection preparation. And S4, conveying the parts to a detection station, driving the in-place sensor by the four-station rotary detection device to execute high-precision detection according to a preset path and parameters, and outputting a result. In the scheme, a closed loop flow from identification to automatic switching to accurate detection is constructed, the scheme is obviously compared with the prior art, the conventional common device adopts a mode of replacing a detection tool by using a fixed sensor or needing manual intervention, the production replacement efficiency is low, the flexible production requirements of multiple varieties of mixed lines are difficult to deal with, the method enables a single detection device to be matched with the optimal detection scheme of multiple parts independently through real-time image identification of S2 and automatic driving switching of S3, seamless connection from universal conveying, intelligent identification and sensor self-adaptation positioning to high-precision detection is realized, the problems that the traditional detection line is insufficient in flexibility, the beat is slow, manual pain points are relied are effectively solved, and the whole intelligent level and comprehensive detection efficiency of the production line are obviously improved while the detection precision is ensured. Further, the four-station rotary detection device comprises an industrial camera, the industrial camera is arranged