Motorcycle muffler intelligent manufacturing process based on laser cutting and automatic welding

Abstract

The application discloses a kind of intelligent manufacturing processes of motorcycle muffler based on laser cutting and automatic welding, and relates to the technical field of motorcycle muffler manufacturing;The process includes: intelligent laser cutting and bevel prefabrication, cylinder automatic roll and group, multi-robot collaborative adaptive welding, on-line quality detection and data tracing four core steps.By laser cutting synchronous prefabrication bevel and positioning mark hole, accurate positioning is realized by combining visual positioning and modular fixture;Adopt double-robot collaborative welding, the first robot scans weld and adaptive tack weld, the second robot carries out cover surface welding;After welding, quality detection is carried out by three-dimensional scanning, and full-process data tracing system is established.The application realizes the full-process intelligentization, flexible production from sheet metal to finished product, significantly improves production efficiency, product quality consistency and traceability.

Description

Technical Field

[0001] This invention relates to the technical field of motorcycle muffler manufacturing, and in particular to an intelligent manufacturing process and system for motorcycle mufflers that integrates laser precision cutting, automated assembly, adaptive welding and intelligent quality inspection. Background Technology

[0002] As a key component of the exhaust system, the manufacturing quality of the motorcycle muffler directly affects the vehicle's performance, noise level, and service life. Traditional manufacturing processes typically separate processes such as blanking, rolling, and welding, resulting in low production efficiency, loss of precision due to inter-process transfers, high dependence on operator skills, and difficulty in ensuring product quality consistency.

[0003] Currently, some automated equipment has emerged in the industry in an attempt to improve the above-mentioned problems. For example: 1. Chinese patent document CN106392428B discloses an automatic welding machine for automobile muffler bodies. This machine uses a fixed mold and a movable mold to clamp the body and employs a welding torch that can move along X and Y axis guide rails for welding, thus automating the welding process and improving welding quality. However, this equipment focuses solely on the welding process and is not integrated with the upstream cutting and blanking process. Furthermore, its mold system lacks adaptability for rapid changeovers in the production of diverse, small-batch motorcycle mufflers.

[0004] 2. Chinese patent document CN119525893A discloses a high-efficiency welding device for automotive muffler ports. It achieves automatic clamping and rotation welding of end caps through a clever clamping and rotating unit, improving the efficiency of port welding. However, it is also an independent welding unit, not involving the preparation and cutting of the cylinder itself. Furthermore, its structure is mainly designed for automotive mufflers, and its applicability to the more complex and varied specifications of motorcycle mufflers is questionable.

[0005] 3. Two Chinese patent documents, CN119368944B and CN215034556U respectively, focus on laser cutting technology. The former is specifically used to cut grooves at specific angles on composite muffler plates to facilitate subsequent welding, while the latter provides a general-purpose laser cutting machine for steel structures. These technologies solve the problem of precise material cutting, but the cut workpieces still need to be manually or using other equipment to transport them to the welding station, failing to form a continuous manufacturing flow.

[0006] In summary, existing technical solutions mostly focus on automating a single step in the muffler manufacturing process, such as cutting or welding, resulting in isolated automation systems. A comprehensive solution that organically combines laser cutting with automated welding, and achieves intelligent and flexible production throughout the entire process from sheet metal to finished product, tailored to the specific characteristics of motorcycle mufflers, remains a pressing technical challenge in this field. Summary of the Invention

[0007] Therefore, it is necessary to provide a smart manufacturing process and system for motorcycle mufflers based on laser cutting and automated welding to overcome the shortcomings of existing technologies. This solution achieves efficient, high-quality, and flexible production from sheet metal to finished mufflers through integrated innovation of processes and equipment.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A smart manufacturing process for motorcycle mufflers based on laser cutting and automated welding includes the following steps: S1. Intelligent laser cutting and beveling: Using an integrated laser cutting system, the metal sheet is precisely cut according to the product model data issued by the host computer to cut out the cylinder unfolding plate, end cap and partition; welding bevels are prefabricated simultaneously on both sides of the longitudinal seam of the cylinder unfolding plate; positioning mark holes for subsequent visual recognition are processed on the sheet. S2. Automatic rolling and assembly of the cylinder: The cut cylinder unfolding plate is rolled into a cylindrical shape by a rolling machine, and the positioning mark hole is captured by a vision recognition system to achieve precise positioning of the cylinder in the fixture; then, the robot transports the end caps, partitions and other internal components to the designated position in the cylinder and pre-fixes them; S3. Multi-robot collaborative adaptive welding: The cylinder is held by a servo headstock and rotates at a constant speed; The first welding robot is equipped with a laser vision sensor to scan the longitudinal and circumferential seams of the cylinder, identify the weld trajectory and bevel shape in real time, and adaptively adjust the welding path and process parameters to perform the root pass welding. The second welding robot performs cover welding based on the data scanned and shared by the first robot, achieving high-quality welding of the longitudinal seam and the circumferential seam; S4. Online Quality Inspection and Data Traceability: After welding is completed, a 3D laser scanner is used to scan the key welds, generate 3D point cloud data and compare it with the standard model to detect defects such as weld reinforcement height, width and undercut. Each product is assigned a unique ID, and all production data, such as cutting parameters, welding parameters, and inspection results, are associated with this ID and stored in the database to achieve full life cycle quality traceability.

[0009] Furthermore, the integrated laser cutting system and the multi-robot collaborative adaptive welding system are integrated into the same flexible manufacturing unit. The two are connected by a programmable servo turntable, and the cut workpiece is automatically transferred to the welding station.

[0010] Furthermore, the welding station adopts a modular magnetic clamp with an RFID chip built into the clamp base to record the clamp model and the corresponding welding program; when the system identifies the installed clamp model, it automatically calls the corresponding robot welding program and parameters.

[0011] The present invention also provides an intelligent manufacturing system that applies the above-described process, the system comprising: Integrated laser cutting workstation; Cylindrical roll and visual positioning station; A multi-robot collaborative welding workstation, which includes a servo head and tailstock, welding robots, laser vision sensors and welding torches; Online quality inspection station; And a central control system that unifies the control of the aforementioned stations.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Integration and high efficiency: By integrating laser cutting and welding and automating the flow between workstations, the stagnation and transfer between processes are eliminated, greatly improving the overall production efficiency, which is especially suitable for the small-batch, multi-batch production mode of motorcycle mufflers.

[0013] 2. High quality and high consistency: Pre-fabrication of the bevel combined with adaptive welding technology ensures the uniformity of weld formation and internal quality; online inspection promptly removes defective products, ensuring the reliability of products leaving the factory.

[0014] 3. High flexibility: The combination of modular fixtures and RFID identification technology allows for switching product models simply by changing the fixture and automatically calling up the program, resulting in extremely short changeover time and fast market response.

[0015] 4. Intelligent and traceable: Full-process data collection and monitoring not only realizes closed-loop optimization of process parameters, but also establishes a complete product quality archive, providing data support for subsequent process improvement and problem tracing. Attached Figure Description

[0016] Figure 1 This is an overall layout diagram of the intelligent manufacturing system shown in this invention; Figure 2 This is a schematic diagram of the process flow for laser cutting and beveling in this invention; Figure 3 This is a flowchart of the cylinder rolling and visual positioning process in this invention; Figure 4This is a diagram of the multi-robot collaborative welding system shown in this invention; Figure 5 This is a system diagram of the modular magnetic clamp shown in this invention; Figure 6 This is a flowchart of the online quality inspection and data traceability process in this invention; Figure 7 This is a flowchart illustrating the overall process flow of the present invention. Figure 8 This is a system diagram for machine learning welding optimization according to the present invention. Detailed Implementation

[0017] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings and examples. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0018] Example 1: This example uses a typical motorcycle cylindrical muffler to illustrate its manufacturing process.

[0019] I. System Composition The intelligent manufacturing system mainly includes: 1. Laser cutting workstation: It adopts a 2000W fiber laser, is equipped with a commercially available operating system, and is integrated into the end of a six-axis robot arm, enabling three-dimensional cutting.

[0020] 2. Rolling and Assembly Station: A three-roll rolling machine is used, and a positioning system based on a conventional vision sensor is configured.

[0021] 3. Collaborative Welding Workstation: This includes an adjustable-speed servo headstock held by a three-jaw chuck; a commercially available ARCMate120iC / 10L robot as the primary welding robot, with a laser vision sensor and welding torch integrated at its arm end; and a commercially available ARCMate120iC / 7L robot as the secondary welding robot. The laser vision sensor uses a commercially available ScanTracker model or another similar sensor.

[0022] 4. Control System: A commercially available S7-1500 series PLC from a common brand is used as the lower-level machine to handle logic control; the upper-level machine uses an industrial computer running a pre-set MES client to handle order management, formula distribution, and data collection.

[0023] II. Process Execution 1. Intelligent laser cutting: The operator selects the "Model A" muffler production order in the MES system.

[0024] The laser cutting robot receives instructions, retrieves a 2mm thick 304 stainless steel plate from the material warehouse, first cuts out an 80mm end cap, and then completes the cutting of all holes on a cylindrical unfolding plate with dimensions of 250mm*600mm.

[0025] Key step: When cutting the two long sides of the cylindrical unfolded plate, the robot deflects the cutting head by 30 degrees and simultaneously cuts a single-sided V-shaped bevel on the edge of the plate.

[0026] Finally, cut 3mm diameter circular positioning holes at the four corners of the board.

[0027] 2. Automatic cylinder rolling and assembly: The cut cylindrical unfolded plate is conveyed to the rolling station by a conveyor belt.

[0028] The vision system identifies four positioning holes, guiding the robot to accurately grasp the sheet material and feed it into the rolling machine.

[0029] After rolling, the cylinder is transferred to a combination fixture at the welding station. The positioning pins on the fixture engage with the positioning holes on the cylinder to achieve precise positioning. The robot then inserts the pre-cut end caps and partitions into the cylinder and temporarily secures them using positioning welds.

[0030] 3. Multi-robot collaborative adaptive welding: The servo head clamps the cylinder and begins to rotate at a constant speed of 10 rpm.

[0031] The first welding robot starts up, and the laser vision sensor at the end of its arm scans the longitudinal seam of the cylinder. The sensor compares the detected actual weld with the theoretical model, calculates the path deviation, and records it as ΔX, ΔY, and ΔZ respectively; and calculates the bevel cross-sectional area.

[0032] The robot controller calculates the required welding heat input in real time based on the cross-sectional area of ​​the bevel, and guides the welding torch to perform precise root pass welding by adjusting current, voltage, etc., and compensating for path deviation.

[0033] The second welding robot obtains weld position and cross-sectional data from the first robot via Ethernet, and performs cover welding with slightly larger welding parameters in the same coordinated motion mode to ensure a full and smooth weld transition.

[0034] After the longitudinal seam welding is completed, the system completes the welding of the circumferential seams at both ends using a similar process.

[0035] 4. Online quality inspection and data traceability: The welded muffler is then sent to the inspection station. A Gocator 2350 series line laser scanner scans the critical circumferential seam.

[0036] The point cloud data obtained from the scan is compared with the standard CAD model to automatically calculate the weld reinforcement height, which is 0.5-1.5mm, and the width, which is 4-6mm. It also detects whether there are defects such as undercut and weld beads.

[0037] All data, including laser power and speed during cutting, current, voltage and speed during welding, and the detected weld size, are bound to the silencer's unique QR code ID and stored in the database.

[0038] III. Feasibility and Technical Effect Demonstration To verify the effectiveness of this invention, the following comparative experiments were conducted: Experimental group: Using the integrated intelligent manufacturing process and system described in this invention.

[0039] Control group: The traditional decentralized process was adopted, which involves CNC punching blanking -> manual rolling and spot fixing -> operator completing welding on a standard welding positioner.

[0040] The comparison results are shown in the table below:

[0041] The comparison of the above objective data shows that the intelligent manufacturing process provided by this invention is significantly superior to the traditional decentralized manufacturing mode in terms of production efficiency, product quality, production flexibility and intelligence level, which fully demonstrates the technical feasibility, advancement and outstanding substantive features of this invention. Example 2

[0042] Based on Example 1, Example 2 further introduces a machine learning-based welding defect prediction and process parameter self-optimization function, aiming to achieve predictive quality control and process self-evolution.

[0043] I. System Enhancement Configuration The intelligent manufacturing system disclosed in Embodiment 2 is enhanced based on the hardware of Embodiment 1 as follows: 1. Data Acquisition System: A high-speed industrial camera is integrated near the laser vision sensor of the first welding robot to capture dynamic images of the weld pool in real time. Simultaneously, the current and voltage signals of the welding power source are acquired at high frequencies, such as 1kHz.

[0044] 2. Calculation Unit: A process parameter self-optimization module is added to the S7-1500 PLC or host industrial control computer. This module is equipped with a trained machine learning model.

[0045] II. Model Training and Deployment 1. Data preparation and training: In the initial stage of the system, a large amount of data on normal and abnormal welding processes are collected, including: molten pool images, current and voltage waveforms, and corresponding post-weld three-dimensional scanning detection results, such as whether there are defects such as porosity, undercut, and lack of fusion.

[0046] Using this data, a convolutional neural network-long short-term memory network, or simply CNN-LSTM hybrid model, is trained. The CNN branch is responsible for extracting spatial features from the molten pool image, such as the shape and brightness of the molten pool, while the LSTM branch is responsible for extracting temporal features from the current and voltage sequence data.

[0047] The model outputs the probability of various defects and suggestions for adjusting process parameters, such as current fine-tuning and welding speed variation.

[0048] 2. Online prediction and adaptive optimization: During the welding process, real-time images of the molten pool and electrical signals are input into a deployed machine learning model.

[0049] The model performs real-time reasoning. If it predicts that the probability of a certain type of defect, such as porosity, exceeds a preset threshold (e.g., 30%), the system will immediately send a command to the welding robot controller to adaptively adjust the welding parameters. For example, when a porosity risk is predicted, the system will automatically slightly increase the welding current or decrease the welding speed to increase the penetration depth and promote gas escape.

[0050] All adjustment records, predicted data, and actual final detection results are linked to the product ID and fed back into the database for continuous model learning and optimization, forming an intelligent closed loop of perception-prediction-decision-execution-feedback.

[0051] III. Demonstration of Technical Effects By introducing the machine learning optimization function of this embodiment, the following further technical effects are achieved compared with the system that only has the function of Embodiment 1: 1. Defect prevention capability: It has achieved a leap from post-weld inspection to in-weld prediction and intervention, and can nip potential defects in the bud during the welding process.

[0052] 2. Process self-optimization: The system can autonomously adapt to uncertainties such as slight fluctuations in the composition of the plate material and gradual changes in equipment status, maintain the best welding quality, and reduce the dependence on the precise setting of initial parameters.

[0053] 3. Continuous improvement: The system has the ability to become smarter with use. As production data accumulates, the model's prediction accuracy and optimization effect continue to improve.

[0054] 4. By comparing production line data, after applying this embodiment, the welding qualification rate was further increased from 99.5% to over 99.9%. At the same time, the rework rate caused by unstable welding quality was reduced by 70%, which fully demonstrates the effectiveness and advancement of this embodiment 2.

[0055] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0056] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A smart manufacturing process for motorcycle mufflers based on laser cutting and automated welding, characterized in that, It includes the following steps: S1. Intelligent laser cutting and beveling: Using an integrated laser cutting system, the metal sheet is precisely cut according to the product model data to cut out the cylinder unfolding plate, end cap and partition; welding bevels are prefabricated on both sides of the longitudinal seam of the cylinder unfolding plate at the same time, and positioning mark holes for subsequent visual recognition are processed on the sheet. S2. Automatic rolling and assembly of the cylinder: The cut cylinder unfolding plate is rolled into a cylindrical shape by a rolling machine, and the positioning mark hole is captured by a vision recognition system to achieve precise positioning; then, the robot transports the end cap and partition to the designated position inside the cylinder and pre-fixes them; S3. Multi-robot collaborative adaptive welding: The cylinder is held by a servo headstock and rotates at a constant speed; the first welding robot is equipped with a laser vision sensor to scan the longitudinal and circumferential seams of the cylinder, identify the weld trajectory and bevel shape in real time, and adaptively adjust the welding path and process parameters to perform the root pass welding. The second welding robot performs cover welding based on the data scanned and shared by the first robot. S4. Online quality inspection and data traceability: After welding is completed, a 3D laser scanner is used to scan the key welds, generate 3D point cloud data and compare it with the standard model to detect the weld quality; each product is assigned a unique ID, and all production data is associated with this ID and stored in the database.

2. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 1, characterized in that: The integrated laser cutting system and the multi-robot collaborative adaptive welding system are integrated in the same flexible manufacturing unit. The two are connected by a programmable servo turntable, and the workpiece is automatically transferred to the welding station after cutting.

3. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 1, characterized in that: The welding station uses a modular magnetic clamp with an RFID chip built into the clamp base to record the clamp model and the corresponding welding program. When the system identifies the installed clamp model, it automatically calls the corresponding robot welding program and parameters.

4. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 1, characterized in that: In step S3, the first welding robot detects the path deviation between the actual weld and the theoretical model using a laser vision sensor, and records them as ΔX, ΔY, and ΔZ respectively. It also calculates the bevel cross-sectional area and calculates the required welding heat input in real time based on the bevel cross-sectional area. The path deviation is compensated by adjusting the current and voltage parameters.

5. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 1, characterized in that: In step S1, the beveling prefabrication is achieved by deflecting the laser cutting head by 30° to simultaneously cut a single-sided V-shaped bevel on the edge of the plate.

6. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 1, characterized in that: In step S4, the weld quality inspection includes weld reinforcement inspection, with a standard of 0.5-1.5mm; weld width inspection, with a standard of 4-6mm; and inspection of undercut and weld bead defects.

7. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 1, characterized in that, The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding also includes steps for predicting welding defects and optimizing process parameters based on machine learning, specifically including: Real-time acquisition of molten pool images and welding electrical parameters during the welding process; The collected data is input into a pre-trained machine learning model, which is a hybrid model of convolutional neural network and long short-term memory network. The model outputs the probability of various defects occurring and suggestions for adjusting process parameters in real time. When the predicted defect probability exceeds a preset threshold, the system automatically adjusts the welding parameters to intervene.

8. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 7, characterized in that: The sampling frequency of the real-time acquired molten pool image is not less than 100Hz, and the acquisition frequency of the welding electrical parameters is not less than 1kHz.

9. The intelligent manufacturing process for motorcycle mufflers based on laser cutting and automated welding according to claim 7 or 8, characterized in that: The pre-trained machine learning model is trained using supervised learning. The training data comes from a large number of molten pool images and welding electrical parameter sequences collected during the historical production process and which have been correlated with the final detection results. Furthermore, the system uses the process data of each welding process and the actual detection results as new samples to perform periodic incremental training on the model in order to achieve continuous optimization of the model.

Images