Welding defect monitoring method and system based on production business
By collecting workpiece and weld information before welding, and combining it with an IoT system and a welding defect prediction model, the system achieves precise location of welding defects and real-time display and storage of monitoring results. This solves the problem of insufficient integration between welding defect monitoring and production operations in existing technologies, and improves welding quality and production efficiency.
Patent Information
- Application Number
- CN202511527386.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-23
- Publication Date
- 2026-02-06
AI Technical Summary
Existing welding defect monitoring technologies cannot be deeply integrated with production operations, and cannot achieve a one-to-one correspondence between welding defect monitoring results and workpiece information and weld information. This results in limited welding quality inspection effectiveness and a lack of judgment on welding process quality and data mining value.
By collecting workpiece and weld information before welding begins, and using a welding defect prediction model for online prediction, combined with IoT system and robot register values, the system can accurately locate welding defects and display and store monitoring results in real time, generate defect reports, and push them to the client of quality inspection and maintenance personnel.
It achieves a one-to-one correspondence between welding defect monitoring results and workpiece and weld information, provides guidance for weld inspection and rework, saves data storage space, and improves welding quality and production efficiency.
Smart Images
Figure CN121481937A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of welding defect monitoring technology, specifically relating to a welding defect monitoring method and system based on production operations. Background Technology
[0002] Welded structural components constitute a significant portion of engineering machinery products, accounting for over 50% of the total weight. The quality of these welded components directly impacts the product's quality, performance, and reliability. Therefore, renowned domestic and international engineering machinery manufacturers consider structural component welding one of the most crucial manufacturing processes, continuously adopting new technologies, processes, and equipment to improve welding quality. Currently, the inspection of structural component welding quality primarily relies on post-weld visual inspection, penetrant testing, magnetic particle testing, and ultrasonic testing—all post-weld inspections with limited effectiveness in improving welding quality and analyzing the causes of defects. Addressing these issues, welding process parameter monitoring systems have emerged. Their main function is to record information such as welding current, voltage, gas flow rate, and molten pool images during the welding process, as well as some welding defect prediction methods. However, these methods and system functions are relatively isolated and cannot be integrated with the workpiece welding production process. Consequently, molten pool image monitoring in the workshop has gradually become a dispensable monitoring tool, failing to play its role in welding defect monitoring. Current and voltage acquisition systems only record welding process parameters, lacking further analysis, making it impossible to determine the quality of the welding process or trace specific workpieces and weld locations, thus losing the value of data mining. Existing welding monitoring technologies lack operational support for structural component welding production operations. Summary of the Invention
[0003] To address the shortcomings of existing technologies, this invention provides a welding defect monitoring method and system based on production operations. This method integrates welding defect monitoring with production operations, achieving a one-to-one correspondence between welding defect monitoring results and workpiece and weld information. It provides guidance for weld inspection and rework, and a data foundation for process improvement, thus achieving deep integration with production processes.
[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows:
[0005] In a first aspect, a welding defect monitoring method based on production operations is provided, comprising: after the workpiece to be welded is loaded and before welding begins, collecting workpiece information and weld information of the workpiece to be welded; and executing predetermined welding defect monitoring logic based on the identification result of the weld information.
[0006] Furthermore, the weld information includes a weld number, which is represented by setting a robot register value in the welding program. When the register value is a six-digit number, the meaning of the six digits is as follows: the first two digits represent whether it is a critical weld, with 11 representing a critical weld and 10 representing a non-critical weld; the middle two digits represent all weld numbers of a workpiece, numbered sequentially from 01 to 99; the last two digits represent which layer of the weld, with 00 representing the first layer, 01 representing the second layer, and so on.
[0007] Furthermore, based on the identification results of the weld information, a predetermined welding defect monitoring logic is executed, including: when the first two digits of the weld number are 10, only welding parameters are collected, and no welding defect prediction is performed; when the first two digits of the weld number are 11, and welding has started, the collected welding parameters are input into a preset welding defect prediction model to perform online prediction of welding defects, output the welding defect type, and classify the defect severity level.
[0008] Furthermore, the pre-defined welding defect prediction models include welding defect prediction models applicable to numerical signals and / or welding defect prediction models applicable to image information.
[0009] Furthermore, based on the identification results of the weld information, the predetermined welding defect monitoring logic is executed, which also includes: taking the weld start end as the zero point, calculating the defect location according to the welding speed and welding start time, and generating defect warning information, including time, workpiece information, weld number, defect location and defect severity level; the defect warning information and workpiece monitoring report are simultaneously sent to the mobile client of the quality inspection and maintenance personnel, so that the quality inspection and maintenance personnel can promptly specify rework plans and improve welding processes.
[0010] Furthermore, based on the identification results of the weld information, the predetermined welding defect monitoring logic is executed, which also includes: when a set defect severity level occurs, a stop signal is generated and sent to the welding workstation to control the welding robot to stop welding.
[0011] Secondly, a welding defect monitoring system based on production operations is provided, comprising: a data acquisition module for acquiring workpiece information and weld information of the workpiece to be welded after the workpiece is loaded and before welding begins; and a welding defect monitoring module for executing predetermined welding defect monitoring logic based on the identification results of the weld information.
[0012] Furthermore, the data acquisition module collects workpiece information and weld information of the workpiece to be welded from the IoT system through the API interface. The workpiece information includes the workpiece's stamp number, and the weld information includes the weld number.
[0013] Furthermore, the welding defect monitoring module transmits defect warning information and workpiece monitoring reports to the IoT system via Kafka. The IoT system then sends the monitoring reports to the mobile client of quality inspection and maintenance personnel via message sending function, enabling them to promptly specify rework plans and improve welding processes.
[0014] Furthermore, it also includes a display and storage module, which has at least the following functions: a) real-time display of welding parameters, including welding current, welding voltage, gas flow rate, and molten pool image; b) real-time display and updating of welding defect early warning information; c) after the entire workpiece welding is completed, displaying the generated welding defect report, including welding equipment information, workpiece information, welding defect information, and monitoring data corresponding to the occurrence of the defect; d) saving and managing the original data and defect early warning information monitored during the welding process for later retrieval.
[0015] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:
[0016] (1) The present invention collects the workpiece information and weld information of the workpiece to be welded after the workpiece is loaded and before welding begins; based on the identification result of the weld information, a predetermined welding defect monitoring logic is executed; the welding defect monitoring result is realized to correspond one-to-one with the workpiece information and weld information, and the monitoring report is directly pushed to the production personnel client to provide guidance for weld inspection and rework and to provide a data basis for process improvement.
[0017] (2) This invention obtains the workpiece stamp number (workpiece information) and weld number through IoT (Internet of Things: refers to connecting any object to the network through information sensing devices according to the agreed protocol, and the object exchanges and communicates information through information transmission media to realize intelligent identification, positioning, tracking and supervision functions). Welding parameters are collected during the welding process, and monitoring data is processed and defect prediction is performed according to whether it is a critical weld. Defect information and shutdown signal are generated, and a workpiece defect monitoring report is formed and sent to the quality inspection and maintenance personnel's client APP.
[0018] (3) The present invention proposes a workpiece weld seam information collection method and defect prediction logic to realize weld seam defect location and weld seam prediction on demand, which not only meets the welding production requirements, but also saves data storage space: by setting the welding robot register value, the weld seam number information collection and key / non-key weld seam distinction are realized. Based on the weld seam number and weld bead number, arc start and end status and welding speed, the welding defect is accurately located. Data processing and defect prediction are only performed on key weld seams. For non-key weld seams, only the original data is displayed and no calculation or prediction is performed, which meets the production requirements and saves data storage space. Attached Figure Description
[0019] Figure 1This is a schematic diagram of the system framework of a welding defect monitoring system based on production operations provided in an embodiment of the present invention;
[0020] Figure 2 This is a schematic diagram of the data acquisition process in an embodiment of the present invention;
[0021] Figure 3 This is a schematic diagram of the main process for predicting welding defects in an embodiment of the present invention. Detailed Implementation
[0022] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.
[0023] Example 1
[0024] A welding defect monitoring method based on production operations includes: collecting workpiece information and weld information of the workpiece to be welded after the workpiece is loaded and before welding begins; and executing predetermined welding defect monitoring logic based on the identification result of the weld information.
[0025] like Figures 1-3 As shown, this invention closely follows the production process of "material loading-welding-repair welding-quality inspection and rework" for structural component welding, realizing a one-to-one correspondence between welding defect monitoring results and product and weld information. Monitoring reports are directly pushed to the production personnel's client, providing guidance for weld inspection and rework, providing a data basis for process improvement, and achieving a deep integration with the production business process.
[0026] This invention comprises four steps: First, before welding begins at the welding workstation after loading, workpiece and weld information is collected via an IoT system and merged with data collected by devices such as current and gas flow sensors and molten pool cameras; second, based on the welding status of the workstation, the collected data is calculated and welding defects are predicted, and a shutdown signal is generated for severe defects; third, the real-time data collected during the welding process and the generated alarm information are centrally displayed, stored, and managed through data transmission; finally, after the workpiece welding is completed, the system generates a workpiece welding defect monitoring report, which is pushed to the client devices of quality inspection and maintenance personnel and process engineers via the IoT system to guide workpiece rework and welding process improvement. Figure 1 As shown.
[0027] (1) Weld information and welding parameter acquisition: such as Figure 2As shown, before welding begins after the workpiece is loaded, information such as the workpiece stamp number (workpiece information), weld number, welding start and end status, and welding speed are collected from the IoT system via the API interface. Sensors such as current, voltage, gas flow, and molten pool camera are installed on the welding machine and welding torch to collect the fluctuation of welding process parameters and the morphology of the molten pool. The workpiece information and welding parameter information are merged to bind the welding process monitoring data (current, voltage, gas flow, molten pool image) with the information of the workpiece being welded. This is the basis for tracing welding defects back to the "workpiece-weld-location".
[0028] In this process, the weld seam number is represented by setting robot register values in the welding program. The register value can generally be set to a 6-digit number, such as 110800, with the following meanings: The first two digits represent whether it is a critical weld seam, 11 indicates that critical weld seams require defect prediction, and 10 indicates that non-critical weld seams do not require defect prediction, in order to meet the actual production needs of non-critical weld seams that do not require online monitoring, and also to save data storage space; The middle two digits represent the weld seam numbers of a workpiece, which can be numbered from 01 to 99; The last two digits represent which layer of the weld seam, 00 represents the first layer, 01 represents the second layer, and so on, to meet the production needs of multi-layer and multi-pass welding.
[0029] (2) Based on the identification results of the weld information, execute the predetermined welding defect monitoring logic, including: when the first two digits of the weld number are 10, only the welding parameters are collected and no welding defect prediction is performed; when the first two digits of the weld number are 11 and welding has started, the collected welding parameters are input into the preset welding defect prediction model to perform online prediction of welding defects, output the welding defect type, and classify the severity level of the defects.
[0030] Online welding defect prediction is processed according to production requirements. The welding process defect prediction logic is as follows: Figure 3 As shown:
[0031] a) Determine whether prediction is needed based on the weld number (check if the first two digits of the weld number are 11), and further determine the welding status (check if the welding status is 1). If both are "yes", proceed to the data processing and defect prediction process; otherwise, only display the original data and do not perform data calculation or analysis (i.e., do not perform welding defect prediction).
[0032] b) After entering the data processing and defect prediction process, the monitoring data is input into a preset welding defect prediction model to comprehensively determine the type of welding defect and classify the severity level of the defect. The preset welding defect prediction models include welding defect prediction models suitable for numerical signals and / or welding defect prediction models suitable for image information. The specific welding defect prediction model used can be selected and configured from existing technologies according to production needs.
[0033] c) Using the weld start point as the zero point, calculate the location of the defect based on the welding speed and welding start time;
[0034] d) Generate defect warning information, including time, stamp number, weld number, defect location, defect level, etc.
[0035] e) For severe welding defects, a shutdown signal is generated and transmitted to the welding workstation via IoT to control the robot to stop, reducing the continuous generation of defects. For other defects, early warning information is saved and continuously monitored.
[0036] (3) Monitoring results shall be displayed and stored in real time, including at least the following:
[0037] a) Real-time display of welding parameters, including welding current, welding voltage, gas flow rate, and molten pool image;
[0038] b) Real-time display and updating of welding defect early warning information;
[0039] c) After the entire workpiece welding is completed, the generated welding defect report is displayed, including welding equipment information, workpiece information, welding defect information, and monitoring data corresponding to the occurrence of the defect;
[0040] d) Save and manage the raw data and defect warning information monitored during the welding process for later retrieval.
[0041] (4) Monitoring Report Generation and Push: Defect warning information and workpiece monitoring reports are transmitted to the IoT system via Kafka (a high-throughput distributed publish-subscribe messaging system that can process all action stream data of consumers on the website, supports Hadoop parallel loading mechanism to unify online and offline message processing, and also provides real-time messages through clusters). The IoT system sends the monitoring reports to the mobile client of quality inspection and maintenance personnel through the message sending function. Maintenance personnel receive the welding defect report before the product arrives at the maintenance station, and quickly rework the weld according to the location of the welding defect marked in the monitoring report, improving the rework efficiency, solving the hidden danger of defect omission caused by blind spots in traditional weld flaw detection, and improving the weld pass rate. Based on the IoT welding defect monitoring data, process personnel analyze the high-incidence areas of welding defects, the corresponding welding process parameters, etc., to improve the workpiece welding process, improve the workpiece welding quality, and reduce the number of welding defects.
[0042] For welding process parameter monitoring, simply collecting welding current / voltage parameters cannot be linked to workpiece type or weld location, rendering the collected data useless. Furthermore, the parameter monitoring and analysis system is disconnected from the production system. In actual production, not every weld process requires monitoring; monitoring non-critical welds is neither necessary for quality control nor a waste of data storage space. This invention, based on an automated welding station, uses an IoT-enabled system to acquire the workpiece stamp number and weld number. Welding parameters are collected during the welding process, and monitoring data is processed and defect predictions are performed based on weld importance. Defect information and stop signals are generated, creating workpiece defect monitoring reports, which are sent to a client app for quality inspection and maintenance personnel. This provides guidance for weld inspection and rework, achieving a one-to-one correspondence between welding defect monitoring results and product / weld information, and directly pushing monitoring reports to production personnel's clients, achieving on-demand monitoring and direct application of results.
[0043] Example 2
[0044] Based on the welding defect monitoring method based on production operations described in Embodiment 1, this embodiment provides a welding defect monitoring system based on production operations, including:
[0045] The data acquisition module is used to collect workpiece information and weld information of the workpiece to be welded after the workpiece is loaded and before welding begins.
[0046] The welding defect monitoring module is used to execute predetermined welding defect monitoring logic based on the identification results of the weld information.
[0047] The data acquisition module collects workpiece information and weld information of the workpiece to be welded from the IoT system through the API interface. The workpiece information includes the workpiece's stamp number, and the weld information includes the weld number.
[0048] The welding defect monitoring module transmits defect warning information and workpiece monitoring reports to the IoT system via Kafka. The IoT system then sends the monitoring reports to the mobile client of quality inspection and maintenance personnel via message sending function, enabling them to promptly specify rework plans and improve welding processes.
[0049] It also includes a display storage module, which has at least the following functions:
[0050] a) Real-time display of welding parameters, including welding current, welding voltage, gas flow rate, and molten pool image;
[0051] b) Real-time display and updating of welding defect early warning information;
[0052] c) After the entire workpiece welding is completed, the generated welding defect report is displayed, including welding equipment information, workpiece information, welding defect information, and monitoring data corresponding to the occurrence of the defect;
[0053] d) Save and manage the raw data and defect warning information monitored during the welding process for later retrieval.
[0054] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A welding defect monitoring method based on production operations, characterized in that, include: After the workpiece to be welded is loaded and before welding begins, collect the workpiece information and weld information. Based on the identification results of the weld information, a predetermined welding defect monitoring logic is executed.
2. The welding defect monitoring method based on production operations according to claim 1, characterized in that, The weld information includes the weld number, which is represented by setting the robot register value in the welding program. When the register value is a six-digit number, the meaning of the six digits is as follows: the first two digits represent whether it is a critical weld, with 11 representing a critical weld and 10 representing a non-critical weld; the middle two digits represent all the weld numbers of a workpiece, numbered sequentially from 01 to 99; the last two digits represent which layer of the weld, with 00 representing the first layer, 01 representing the second layer, and so on.
3. The welding defect monitoring method based on production operations according to claim 2, characterized in that, Based on the identification results of the weld information, a predetermined welding defect monitoring logic is executed, including: When the first two digits of the weld number are 10, only welding parameters are collected, and no welding defect prediction is performed. When the first two digits of the weld number are 11 and welding has started, the collected welding parameters are input into the preset welding defect prediction model to perform online prediction of welding defects, output the welding defect type, and classify the severity level of the defects.
4. The welding defect monitoring method based on production operations according to claim 3, characterized in that, Pre-defined welding defect prediction models include welding defect prediction models applicable to numerical signals and / or welding defect prediction models applicable to image information.
5. The welding defect monitoring method based on production operations according to claim 3, characterized in that, Based on the identification results of the weld information, the predetermined welding defect monitoring logic is executed, which also includes: taking the weld start end as the zero point, calculating the defect location according to the welding speed and welding start time, and generating defect warning information, including time, workpiece information, weld number, defect location and defect severity level. Defect warning information and workpiece monitoring reports are simultaneously sent to the mobile client of quality inspection and maintenance personnel, so that they can promptly specify rework plans and improve welding processes.
6. The welding defect monitoring method based on production operations according to claim 5, characterized in that, Based on the identification results of the weld information, the predetermined welding defect monitoring logic is executed, which also includes: when a set defect severity level occurs, a stop signal is generated and sent to the welding workstation to control the welding robot to stop welding.
7. A welding defect monitoring system based on production operations, characterized in that, include: The data acquisition module is used to collect workpiece information and weld information of the workpiece to be welded after the workpiece is loaded and before welding begins. The welding defect monitoring module is used to execute predetermined welding defect monitoring logic based on the identification results of the weld information.
8. The welding defect monitoring system based on production operations according to claim 7, characterized in that, The data acquisition module collects workpiece information and weld information of the workpiece to be welded from the IoT system through the API interface. The workpiece information includes the workpiece's stamp number, and the weld information includes the weld number.
9. The welding defect monitoring system based on production operations according to claim 7, characterized in that, The welding defect monitoring module transmits defect warning information and workpiece monitoring reports to the IoT system via Kafka. The IoT system then sends the monitoring reports to the mobile client of quality inspection and maintenance personnel via message sending function, enabling them to promptly specify rework plans and improve welding processes.
10. The welding defect monitoring system based on production operations according to claim 7, characterized in that, It also includes a display storage module, which has at least the following functions: a) Real-time display of welding parameters, including welding current, welding voltage, gas flow rate, and molten pool image; b) Real-time display and updating of welding defect early warning information; c) After the entire workpiece welding is completed, the generated welding defect report is displayed, including welding equipment information, workpiece information, welding defect information, and monitoring data corresponding to the occurrence of the defect; d) Save and manage the raw data and defect warning information monitored during the welding process for later retrieval.