Automatic production line control system and method for stamping and welding vehicle body parts

Abstract

The invention discloses an automatic production line control system and method for stamping and welding vehicle body parts, and relates to the technical field of vehicle body part production. The system comprises a welding robot module, an integrated visual detection module, an image processing server, a control PLC, a welding parameter regulator, a defective workpiece processing module, a data tracing cloud platform, a multispectral illumination module, a deep learning acceleration card and a welding spot temperature infrared sensor. On-line full inspection of the welding spot quality is achieved, the omission ratio is extremely low, and the integrated vision module maintains the imaging definition in a splashing environment through a nanometer coated lens and a piezoelectric self-cleaning technology; the multi-spectral lighting module solves the problem of reflective interference of an aluminum vehicle body, the defect recognition accuracy is high, the real-time closed-loop response is fast, the pseudo soldering recurrence rate is low through dynamic current compensation, the pneumatic marking positioning precision is high, accurate coordinates are provided for repair, the system supports high-speed detection of welding spots, and the high-beat requirement of a vehicle body part punching and welding automatic production line is met.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle body component production, and in particular relates to a control system and method for an automated punching and welding production line of vehicle body components. Background Art

[0002] The automated stamping and welding production line for body components (such as battery pack housings, aluminum door sill beams, and carbon fiber roof frames) is a core component of vehicle manufacturing. This system typically consists of a stamping unit, a welding robot cluster, a material conveyor line, a fixture station, and a central control system. The welding stations utilize multiple industrial robots equipped with servo welding guns (resistance welding / laser welding), executing high-precision trajectory motion and welding parameters under PLC control. The core challenges are: the extensive use of mixed materials such as aluminum alloy, high-strength steel, and composite materials in the vehicle body results in complex weld nugget formation mechanisms, and the lightweight design results in dense weld spacing, placing extremely high demands on quality control. Currently, mainstream production lines use two types of weld quality monitoring technologies: welding parameter monitoring and offline spot checks. These technologies inherently lack the ability to fully inspect the physical form of welds in real time, preventing them from establishing a closed loop of "detection-analysis-compensation." This has become a bottleneck restricting improvements in body welding yields. To this end, we provide a control system and method for an automated production line for punching and welding of vehicle body parts to solve the above problems. Summary of the Invention

[0003] The purpose of the present invention is to provide a control system and method for an automated production line for punching and welding of vehicle body parts. By integrating an integrated visual inspection module with a deep learning accelerator card at the millisecond level and coordinating a dynamic closed-loop control chain for welding parameters, the present invention solves the problems in the prior art of the inability to fully inspect hidden defects in welds in real time and the resulting mass scrapping caused by delayed quality feedback.

[0004] To solve the above technical problems, the present invention is implemented through the following technical solutions.

[0005] The present invention is a control system and method for an automated production line for punching and welding of vehicle body parts, comprising a welding robot module, an integrated visual inspection module, an image processing server, a control PLC, a welding parameter regulator, a defective workpiece processing module, a data tracing cloud platform, a multispectral lighting module, a deep learning accelerator card, a solder joint temperature infrared sensor, an electrode wear monitoring module, an audible and visual alarm, an industrial switch, an edge computing gateway, a cooling circulation system, and a safety interlock module. The integrated visual inspection module is fixedly connected to the welding robot module, the multispectral lighting module is fixedly connected to the integrated visual inspection module, the integrated visual inspection module is unidirectionally electrically connected to the image processor, the deep learning accelerator card is bidirectionally electrically connected to the image processor, the welding parameter regulator, the defective workpiece processing module, the solder joint temperature infrared sensor, the electrode wear monitoring module, the audible and visual alarm, the industrial switch, the edge computing gateway, the cooling circulation system, and the safety interlock module are all unidirectionally electrically connected to the control PLC, and the data tracing cloud platform is radio-connected to the edge computing gateway.

[0006] The present invention is further configured such that the integrated visual inspection module includes a fast camera, an LED light source and a lens self-cleaning unit. The fast camera is equipped with a nano-anti-splash coating lens. After the welding of the welding robot module is completed, the lens self-cleaning unit is started to clean the lens of the fast camera. The anti-splash coating lens can protect the lens of the fast camera to prevent splashing dust particles from scratching the lens of the fast camera.

[0007] The present invention is further configured such that the image processor includes a weld spot positioning submodule, a feature extraction submodule, and a deep learning defect classification submodule. The weld spot positioning submodule matches the weld spot coordinates based on the welding gun posture data. The weld core feature extraction submodule calculates the weld core diameter, indentation depth, and spatter area. The deep learning defect classification submodule uses the YOLOv5 model to output the defect type (cold weld / spatter / excessive indentation / crack) in real time.

[0008] The present invention is further configured such that the welding parameter regulator includes a dynamic current compensation unit, an electrode pressure adaptive unit and a welding cycle fine-tuner. The electrode pressure adaptive unit adjusts the pressure according to the amount of electrode wear. The dynamic current compensation unit outputs an analog quantity to the welding robot module through a DAC module based on a defect type lookup table compensation, thereby compensating for energy shortages caused by electrode wear in real time. The electrode pressure adaptive unit receives the amount of electrode wear and adjusts the pressure to maintain effective metal plastic deformation. The welding cycle fine-tuner generates high-precision pulses, controls the power-on time, and suppresses uneven weld nuggets caused by power grid fluctuations.

[0009] The present invention is further configured such that the defective workpiece processing module includes a pneumatic marking machine and a three-axis AGV rejection robot arm. The three-axis AGV rejection robot arm has a maximum load of 50 kg and a repeat positioning accuracy of ±0.1 mm. The solenoid valve of the pneumatic marking machine drives the tungsten steel needle to impact the surface of the workpiece to form a deep QR code and engrave a traceable mark containing the workpiece ID, weld point serial number and defect type. The three-axis AGV rejection robot arm locates the workpiece based on the conveyor chain encoder, and the vacuum suction cup grabs it and moves it into the waste area to replace manual sorting.

[0010] The present invention is further configured such that the electrode wear monitoring module includes a laser displacement sensor, a thermocouple and a wear prediction algorithm library. The wear prediction algorithm library estimates the remaining life based on an LSTM model. The laser displacement sensor emits a laser beam and measures the depth of the depression on the electrode end surface through triangulation to quantify the wear amount in real time. The wear prediction algorithm library: the LSTM model inputs a sequence (number of welds, average current, temperature gradient) and outputs the remaining life.

[0011] The present invention is further configured such that the safety interlock module includes a light curtain sensor, an emergency stop relay and a dual-channel safety PLC. The response time of the emergency stop relay is ≤10ms. The dual-channel safety PLC has two independent processors that execute code synchronously, meeting the SIL3 level and preventing false operations. The emergency stop relay is a magnetic latching relay that self-locks contacts through permanent magnets when power is off.

[0012] The present invention is further configured such that the welding robot module includes an anti-electromagnetic interference cover, which adopts a double-layer copper mesh shielding layer to attenuate welding arc radiation by ≥60dB.

[0013] The present invention is further configured such that the welding parameter regulator includes a dynamic current compensation unit, which has a built-in temperature compensation curve. When the ambient temperature rises, the metal resistivity increases, and the heat generation at the same welding current increases significantly, which can easily lead to defects such as spattering and excessive indentation. The dynamic current compensation unit adjusts the output current in real time through a temperature-current negative feedback mechanism to ensure that the energy of the molten core formation is constant.

[0014] The method for controlling the control system of an automated production line for punching and welding of vehicle body parts comprises the following steps: a. When the welding gun of the welding robot module leaves the workpiece surface, the welding point coordinates are sent to the vision module, and the controller integrated in the welding robot module synchronously records the welding parameters; b. Automatically select the light source based on the workpiece material: aluminum parts use blue light + infrared, and steel parts use white light. The integrated visual inspection module captures images at a shutter speed of 1 / 5000s within 0.3s after welding and simultaneously triggers the lens self-cleaning unit to vibrate three times for lens cleaning. c. Welding point positioning: Matching the welding point position based on the welding robot module coordinates and the workpiece CAD model, Feature extraction: nugget diameter: Canny operator edge detection + minimum circumscribed circle fitting, indentation depth: laser triangulation method to reconstruct 3D morphology, spatter area: Otsu threshold segmentation + connected domain analysis, defect classification: YOLOv5 model output defect category and confidence level; d. Qualified solder joints: data is stored in the local cache, and every 50 solder joints are uploaded to the data traceability cloud platform; Unqualified solder joints: The sound and light alarm flashes a red warning light and beeps at 95dB. The pneumatic marking machine engraves a QR code containing the workpiece ID and solder joint serial number 5mm from the defect point. The three-axis AGV rejection robot locates the workpiece according to the encoder signal of the conveyor chain, and the vacuum suction cup grabs the workpiece and moves it into the scrap bin. e. When the laser displacement sensor detects that the electrode is sunken ≥0.2mm, or the temperature gradient suddenly changes by >15℃ / s, the system locks the welding robot module and pops up a replacement prompt, and uploads the wear data to the data tracing cloud platform.

[0015] The present invention has the following beneficial effects.

[0016] 1. This invention realizes full online inspection of solder joint quality with an extremely low missed detection rate. The integrated vision module uses nano-coated lenses and piezoelectric self-cleaning technology to maintain imaging clarity in a splash environment. The multi-spectral lighting module solves the interference of reflective light from the aluminum body, with high defect recognition accuracy and fast real-time closed-loop response. Dynamic current compensation reduces the recurrence rate of cold solder joints. The pneumatic marking has high positioning accuracy, providing precise coordinates for repair. The system supports high-speed inspection of solder joints and meets the high-beat requirements of the automated production line for punching and welding of body parts.

[0017] 2. The electrode wear monitoring module of the present invention combines laser displacement and temperature monitoring to provide early warning of electrode failure risks, thereby improving electrode life. The three-level confidence mechanism integrates infrared temperature verification to avoid misjudgment leading to line stoppage. The data traceability cloud platform constructs a solder joint quality map, and analyzes the correlation between defects and welding parameters, ambient temperature and humidity in real time to promote process optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments.

[0019] Figure 1 This is the system schematic diagram of the control system and method for the automated production line for punching and welding of vehicle body parts.

[0020] Figure 2 This is the system schematic diagram of the integrated visual inspection module in the control system and method of the automated production line for punching and welding of body parts.

[0021] Figure 3 This is the system schematic diagram of the image processor in the control system and method of the automated production line for punching and welding of body parts.

[0022] Figure 4 This is the system schematic diagram of the welding parameter regulator in the control system and method of the automated punching and welding production line for body parts.

[0023] Figure 5 This is the system schematic diagram of the defective workpiece processing module in the control system and method of the automated production line for punching and welding of body parts.

[0024] Figure 6 This is the system schematic diagram of the electrode wear monitoring module in the control system and method of the automated punching and welding production line for body parts.

[0025] Figure 7 This is the system schematic diagram of the safety interlock module in the control system and method of the automated production line for punching and welding of body parts. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present invention will be described below in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Example 1

[0027] See also Figure 1-7 The present invention is a control system and method for an automated production line for punching and welding of vehicle body parts, comprising a welding robot module, an integrated visual inspection module, an image processing server, a control PLC, a welding parameter regulator, a defective workpiece processing module, a data tracing cloud platform, a multispectral lighting module, a deep learning accelerator card, a solder joint temperature infrared sensor, an electrode wear monitoring module, an audible and visual alarm, an industrial switch, an edge computing gateway, a cooling circulation system, and a safety interlock module. The integrated visual inspection module is fixedly connected to the welding robot module, the multispectral lighting module is fixedly connected to the integrated visual inspection module, the integrated visual inspection module is unidirectionally electrically connected to the image processor, the deep learning accelerator card is bidirectionally electrically connected to the image processor, the welding parameter regulator, the defective workpiece processing module, the solder joint temperature infrared sensor, the electrode wear monitoring module, the audible and visual alarm, the industrial switch, the edge computing gateway, the cooling circulation system, and the safety interlock module are all unidirectionally electrically connected to the control PLC, and the data tracing cloud platform is radio-connected to the edge computing gateway.

[0028] Specifically: The welding robot module performs welding operations and outputs the welding gun posture coordinates in real time. The integrated visual inspection module collects welding spot images after welding and adapts to the optical properties of workpieces of different materials through the multi-spectral lighting module. The welding spot temperature infrared sensor synchronously captures the temperature field distribution of the weld core to assist in verifying defects. The laser displacement sensor in the electrode wear monitoring module measures the depth of the electrode end surface depression in real time. The image processing server receives the visual image and completes the defect classification and weld core size calculation in a short time through the deep learning accelerator card. For suspected defects with high confidence, the infrared temperature secondary verification is started. The edge computing gateway pre-processes the electrode wear data, compresses it and uploads it to the data traceability cloud platform through the industrial switch. The control PLC receives the analysis results and triggers a triple response. The cooling circulation system maintains the temperature of the welding robot module to prevent overheating drift. The data traceability cloud platform associates the welding parameters, test results, and environmental data of the welding spot to build a full life cycle quality map; through machine learning, the root cause is excavated and process optimization instructions are automatically issued. Example 2

[0029] See also Figure 1-7 Based on the first embodiment, the integrated visual inspection module includes a fast camera, an LED light source and a lens self-cleaning unit. The fast camera is equipped with a nano-anti-splash coating lens. The image processor includes a solder joint positioning submodule, a feature extraction submodule and a deep learning defect classification submodule. The solder joint positioning submodule matches the solder joint coordinates based on the welding gun posture data. The welding parameter regulator includes a dynamic current compensation unit, an electrode pressure adaptation unit and a welding cycle fine-tuner. The electrode pressure adaptation unit adjusts the pressure according to the amount of electrode wear. The defective workpiece processing module includes a pneumatic marking machine and a three-axis AGV rejection robot arm. The maximum load of the culling robot arm is 50kg, and the repeatability accuracy is ±0.1mm. The electrode wear monitoring module includes a laser displacement sensor, a thermocouple and a wear prediction algorithm library. The wear prediction algorithm library estimates the remaining life based on the LSTM model. The safety interlock module includes a light curtain sensor, an emergency stop relay and a dual-channel safety PLC. The response time of the emergency stop relay is ≤10ms. The welding robot module includes an anti-electromagnetic interference cover. The anti-electromagnetic interference cover uses a double-layer copper mesh shielding layer to attenuate welding arc radiation ≥60dB. The welding parameter regulator includes a dynamic current compensation unit, and the dynamic current compensation unit has a built-in temperature compensation curve.

[0030] Specifically: After the welding robot module completes welding, the lens self-cleaning unit is started to clean the lens of the fast camera. The anti-splash coated lens can protect the lens of the fast camera to prevent flying dust particles from scratching the lens of the fast camera. The weld core feature extraction submodule calculates the weld core diameter, indentation depth, and spatter area. The deep learning defect classification submodule uses the YOLOv5 model to output the defect type (cold weld / spatter / too deep indentation / crack) in real time. The dynamic current compensation unit outputs the analog quantity to the welding robot module through the DAC module based on the defect type lookup table compensation, and compensates for the energy shortage caused by electrode wear in real time. The electrode pressure adaptive unit receives the electrode wear amount and adjusts the pressure to maintain effective metal plastic deformation. The welding cycle fine-tuner generates high-precision pulses, controls the power-on time, and suppresses the uneven weld core caused by power grid fluctuations. The solenoid valve of the pneumatic marking machine drives the tungsten steel needle to impact the workpiece surface, forming a deep QR code and engraving the workpiece ID. , traceability marking of weld serial number and defect type, three-axis AGV rejection robot arm locates workpiece based on conveyor chain encoder, vacuum suction cup grabs and moves to scrap area, replacing manual sorting, laser displacement sensor emits laser beam, measures the depth of depression on electrode end face through triangulation, and quantifies wear in real time, wear prediction algorithm library: LSTM model input sequence (number of welds, average current, temperature gradient), outputs remaining life, dual-channel safety PLC, two independent processors execute code synchronously, meet SIL3 level, eliminate false operation, emergency stop relay: magnetic holding relay, self-locking contact through permanent magnet when power is off, anti-electromagnetic interference cover uses double-layer copper mesh shielding to attenuate welding arc radiation, when ambient temperature rises, metal resistivity increases, heat generation increases significantly under the same welding current, which can easily lead to defects such as spatter and excessive indentation, dynamic current compensation unit adjusts output current in real time through temperature-current negative feedback mechanism to ensure constant energy of weld nugget formation.

[0031] A method for controlling a control system of an automated punching and welding production line for vehicle body parts, characterized in that it comprises the following steps: a. When the welding gun of the welding robot module leaves the workpiece surface, the welding point coordinates are sent to the vision module, and the controller integrated in the welding robot module synchronously records the welding parameters; b. Automatically select the light source based on the workpiece material: aluminum parts use blue light + infrared, and steel parts use white light. The integrated visual inspection module captures images at a shutter speed of 1 / 5000s within 0.3s after welding and simultaneously triggers the lens self-cleaning unit to vibrate three times for lens cleaning. c. Welding point positioning: Matching the welding point position based on the welding robot module coordinates and the workpiece CAD model, Feature extraction: nugget diameter: Canny operator edge detection + minimum circumscribed circle fitting, indentation depth: laser triangulation method to reconstruct 3D morphology, spatter area: Otsu threshold segmentation + connected domain analysis, defect classification: YOLOv5 model output defect category and confidence level; d. Qualified solder joints: data is stored in the local cache, and every 50 solder joints are uploaded to the data traceability cloud platform; Unqualified solder joints: The sound and light alarm flashes a red warning light and beeps at 95dB. The pneumatic marking machine engraves a QR code containing the workpiece ID and solder joint serial number 5mm from the defect point. The three-axis AGV rejection robot locates the workpiece according to the encoder signal of the conveyor chain, and the vacuum suction cup grabs the workpiece and moves it into the scrap bin. e. When the laser displacement sensor detects that the electrode is sunken ≥0.2mm, or the temperature gradient suddenly changes by >15℃ / s, the system locks the welding robot module and pops up a replacement prompt, and uploads the wear data to the data tracing cloud platform.

[0032] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all details in detail, nor do they limit the invention to only the specific implementation methods described. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can better understand and utilize the present invention.

Claims

1. The control system for the automated punching and welding production line of vehicle body parts includes a welding robot module, an integrated visual inspection module, an image processing server, a control PLC, a welding parameter regulator, a defective workpiece processing module, a data traceability cloud platform, a multispectral lighting module, a deep learning accelerator card, a solder point temperature infrared sensor, an electrode wear monitoring module, an audible and visual alarm, an industrial switch, an edge computing gateway, a cooling circulation system, and a safety interlock module. It features: The integrated visual inspection module is fixedly connected to the welding robot module, the multispectral lighting module is fixedly connected to the integrated visual inspection module, the integrated visual inspection module is unidirectionally electrically connected to the image processor, the deep learning accelerator card is bidirectionally electrically connected to the image processor, the welding parameter regulator, defective workpiece processing module, solder point temperature infrared sensor, electrode wear monitoring module, sound and light alarm, industrial switch, edge computing gateway, cooling circulation system and safety interlock module are all unidirectionally electrically connected to the control PLC, and the data tracing cloud platform is radio-connected to the edge computing gateway.

2. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The integrated visual inspection module includes a fast camera, an LED light source and a lens self-cleaning unit. The fast camera is equipped with a nano-anti-splash coating lens.

3. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The image processor includes a solder joint positioning submodule, a feature extraction submodule and a deep learning defect classification submodule. The solder joint positioning submodule matches the solder joint coordinates based on the welding gun posture data.

4. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The welding parameter regulator includes a dynamic current compensation unit, an electrode pressure self-adapting unit and a welding cycle fine-tuner. The electrode pressure self-adapting unit adjusts the pressure according to the amount of electrode wear.

5. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The defective workpiece processing module includes a pneumatic marking machine and a three-axis AGV rejection robot arm. The three-axis AGV rejection robot arm has a maximum load of 50kg and a repeat positioning accuracy of ±0.1mm.

6. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The electrode wear monitoring module includes a laser displacement sensor, a thermocouple and a wear prediction algorithm library, and the wear prediction algorithm library estimates the remaining life based on the LSTM model.

7. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The safety interlock module includes a light curtain sensor, an emergency stop relay and a dual-channel safety PLC, and the response time of the emergency stop relay is ≤10ms.

8. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The welding robot module includes an anti-electromagnetic interference cover, which adopts a double-layer copper mesh shielding layer to attenuate welding arc radiation by ≥60dB.

9. The control system for the automated punching and welding production line of vehicle body parts according to claim 1, characterized in that: The welding parameter regulator comprises a dynamic current compensation unit, and the dynamic current compensation unit has a built-in temperature compensation curve.

10. The method for controlling a vehicle body component punching and welding automated production line according to any one of claims 1 to 9, characterized in that: The following steps are involved: a. When the welding gun of the welding robot module leaves the workpiece surface, the welding point coordinates are sent to the vision module, and the controller integrated in the welding robot module synchronously records the welding parameters; b. Automatically select the light source based on the workpiece material: aluminum parts use blue light + infrared, and steel parts use white light. The integrated visual inspection module captures images at a shutter speed of 1 / 5000s within 0.3s after welding and simultaneously triggers the lens self-cleaning unit to vibrate three times for lens cleaning. c. Welding point positioning: Matching the welding point position based on the welding robot module coordinates and the workpiece CAD model, Feature extraction: nugget diameter: Canny operator edge detection + minimum circumscribed circle fitting, indentation depth: laser triangulation method to reconstruct 3D morphology, spatter area: Otsu threshold segmentation + connected domain analysis, defect classification: YOLOv5 model output defect category and confidence level; d. Qualified solder joints: data is stored in the local cache, and every 50 solder joints are uploaded to the data traceability cloud platform; Unqualified solder joints: The sound and light alarm flashes a red warning light and beeps at 95dB. The pneumatic marking machine engraves a QR code containing the workpiece ID and solder joint serial number 5mm from the defect point. The three-axis AGV rejection robot locates the workpiece according to the encoder signal of the conveyor chain, and the vacuum suction cup grabs the workpiece and moves it into the scrap bin. e. When the laser displacement sensor detects that the electrode is sunken ≥0.2mm, or the temperature gradient suddenly changes by >15℃ / s, the system locks the welding robot module and pops up a replacement prompt, and uploads the wear data to the data tracing cloud platform.

Images