Intelligent management system capable of tracing whole process of distribution and recovery of welding materials

Abstract

The invention provides an intelligent management system capable of tracing the whole process of distribution and recovery of welding materials, and relates to the technical field of welding material management, and the system comprises a dispatching system which is used for generating an electronic material picking list with a two-dimensional code according to the information of a welding task, and issuing the electronic material picking list after the online approval of a picking application submitted by a receiver is passed; the state monitoring module is used for continuously collecting storage data of undistributed welding materials and processing the storage data to obtain an availability prediction result; the welding material issuing module is used for calling the availability prediction result according to the to-be-received welding material data so that issuing personnel can prepare to-be-issued welding materials which can be issued, and identifying and counting through to-be-issued welding material images obtained through shooting; and the welding material recovery module is used for carrying out residual material identification according to the residual welding material image shot by the recovery personnel, binding the recovery data obtained by identification with an electronic material getting list and then synchronizing the electronic material getting list to the dispatching system. The method has the beneficial effects that the whole process can be traced, and the welding quality is ensured while the welding material receiving and recycling process time is shortened.

Description

Technical Field

[0001] This invention relates to the field of welding material management technology, and in particular to an intelligent management system that allows for the traceability of the entire process of welding material distribution and recycling. Background Technology

[0002] As a core component of third-generation nuclear power technology, the CAP1000 series nuclear island demands extremely stringent welding quality during installation. The reliability of the welding process directly determines the structural integrity and operational safety of the nuclear island equipment. Welding consumables, as the core consumables in welding operations, require standardized management processes to ensure welding quality. Currently, in CAP1000 nuclear island installation practice, welding consumable management still predominantly employs a traditional offline manual method, which presents the following prominent problems: 1. The approval process for welding material requisition is cumbersome and inefficient. The installation of welding materials in the nuclear island requires a wide variety of types, and the approval process for their requisition necessitates the completion of a welding material requisition form. This form must meticulously record over ten key pieces of information, including the welding material furnace batch number, specifications, rated quantity, application location (such as the main loop piping of the nuclear island or pressure vessel connectors), welding procedure qualification number (WPS number), and welder's certificate number. Due to the lack of digital filling tools, all of the above information relies on manual writing by technicians, which is not only time-consuming and labor-intensive but also prone to errors such as missing furnace batch numbers or confusing specifications, and incomplete information (such as failing to indicate the nuclear safety level of the application location). Meanwhile, the approval process requires multiple levels of manual signature confirmation, which seriously affects the continuity of welding operations. Especially in emergency repair scenarios, delays in approval may lead to construction stoppages.

[0003] 2. The process of distributing and recycling welding materials relies on manual labor, resulting in weak risk management capabilities. The process of issuing welding materials requires the material administrator to manually check whether the information on the requisition form matches the actual materials, including key parameters such as welding material specifications, quantity, and furnace batch number. However, manual verification is prone to oversights (such as issuing the wrong specifications or missing the furnace batch number). The welding material recycling process also relies on manual records. After the work is completed, any remaining welding materials must be returned to the materials warehouse by the welder. The administrator manually records the quantity and condition of the recycled materials (such as whether they have been opened and the remaining length) and updates the ledger. This process suffers from problems such as untimely record-keeping and inaccurate data. For example, the moisture content and oxidation level of the recycled welding materials are not promptly noted, making it impossible to accurately determine whether the welding materials meet the standards when redistributing them later. In addition, the manual recycling process makes it difficult to track the entire usage trajectory of the welding materials. If welding quality defects occur, it is impossible to quickly locate the flow path of the problematic welding materials.

[0004] Traditional offline models lack a unified digital management platform. Data on the entire lifecycle of welding materials, from warehousing, requisition, distribution, use to recycling, is scattered and stored in paper ledgers, failing to form a complete data chain. When welding quality issues arise, a significant amount of time must be spent sifting through paper records to trace welding material information, and there are risks such as lost ledgers and illegible handwriting, making it difficult to meet the nuclear power industry's mandatory requirements for full traceability and precise location of quality.

[0005] In summary, the existing offline, manual management model for welding materials in the CAP1000 nuclear island suffers from problems such as cumbersome processes, low efficiency, poor risk control, and insufficient traceability. It cannot meet the stringent requirements for nuclear power welding quality and is ill-suited to support the trend of digital transformation in nuclear power engineering. Therefore, there is an urgent need to construct a traceable management system for the entire process of welding material issuance and recycling, achieving digital approval for requisition, automation of issuance and recycling, and precise quality traceability, thus providing a reliable guarantee for the welding quality of CAP1000 nuclear island installation. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides an intelligent management system for the entire process of welding material issuance and recycling that is traceable, comprising: The dispatching system is used to generate an electronic material requisition form with a QR code based on the welding task information, and to issue the electronic material requisition form after the requisition application submitted by the requisitioner is approved online. The QR code contains the welding material data to be requisitioned. The status monitoring module is connected to the dispatching system and is used to continuously collect the stored data of unissued welding materials, process the stored data to obtain the availability prediction result of the unissued welding materials, and synchronize it to the dispatching system. The welding material dispensing module connects to a scanner, a high-speed document scanner, and the dispatching system. It is used to retrieve the availability prediction results based on the welding material data to be requisitioned obtained by the dispensing personnel through scanning the electronic material requisition form with the scanner, so that the dispensing personnel can prepare the welding materials to be dispensed. It also identifies and counts the welding materials to be dispensed by the image of the welding materials to be dispensed obtained by the high-speed document scanner, and synchronizes the specification data of the welding materials to be dispensed to the electronic material requisition form, the requisitioning personnel, and the identification and counting results to the dispatching system. The welding material recycling module is connected to the high-speed scanner and is used to identify residual materials based on the images of residual welding materials taken by the recycling personnel through the high-speed scanner. The identified recycling data is then bound to the electronic material requisition form and synchronized to the dispatching system.

[0007] Preferably, the status monitoring module includes: The data acquisition unit is used to continuously collect environmental parameters of the storage location of the welding material to be issued and the storage time of the welding material to be issued as the storage data; The availability prediction unit, connected to the data acquisition unit, is used to input the stored data into the pre-trained welding material performance prediction model to obtain the availability prediction result of the welding material to be issued and synchronize it to the dispatching system. The availability prediction results include at least one risk type and its corresponding confidence level.

[0008] Preferably, the data acquisition unit includes a temperature and humidity sensor and an oxidation potential sensor deployed at the storage location, used to collect the temperature and humidity data of the storage location and the surface potential value of the undistributed welding material as the environmental parameters, respectively. The risk types include one or more of the following: damp electrode coating, hardening of electrode coating, and oxidation of welding wire surface.

[0009] Preferably, the dispatching system includes a status warning module, used to mark the corresponding undistributed welding materials as distributable when the confidence level corresponding to each of the risk types is less than a first threshold; When the confidence level corresponding to any of the aforementioned risk types is not less than the first threshold and less than the second threshold, the corresponding unissued welding materials are marked as temporarily suspended, and an environmental warning prompt containing the environmental parameters, the availability prediction results, and the specification data of the corresponding unissued welding materials is generated and pushed to the relevant technical personnel for review. And when the confidence level corresponding to any of the aforementioned risk types is not less than the second threshold, the corresponding unissued welding materials are marked as unissued and the relevant technical personnel are notified to conduct on-site confirmation.

[0010] Preferably, the dispatching system further includes a status unlocking module, comprising: The first unlocking unit is used for technicians to view the environmental warning prompts online and provide disposal suggestions, and then sign an online consent form for distribution and push it to the warehouse management personnel. After the warehouse management personnel execute the disposal suggestions, they upload a disposal completion instruction, and then adjust the temporarily suspended distribution status to a distributable status according to the disposal completion instruction. The second unlocking unit is used for technicians to sign and upload a confirmation form after on-site re-inspection, and then adjust the unissued state to the issueable state based on the confirmation form.

[0011] Preferably, the welding material dispensing module includes: The specification recognition unit is used to identify the type of welding material specification contained in the image of the welding material to be distributed; The contour recognition unit, connected to the specification recognition unit, is used to identify the overall contour of each weld material for each type of weld material specification, calculate the contour overlap rate, and count the overall contour to obtain a first counting result. The density estimation unit, connected to the specification identification unit, is used to calculate the overall occupied area of ​​each weld material contained therein and the variance of the pixel density of the weld material area for each type of weld material specification, and to calculate the ratio between the overall occupied area and the standard area of ​​a single weld material of the corresponding weld material specification to obtain a second counting result. The result calibration unit is connected to the contour recognition unit and the density estimation unit respectively. When the first counting result and the second counting result are inconsistent, the unit configures the weights according to the contour overlap rate and the pixel density variance, and weights the first counting result and the second counting result to obtain the recognition counting result.

[0012] Preferably, the contour overlap rate is the ratio of the weld contour area of ​​the adhesion region to the total contour area for each type of weld material specification. The pixel density is the ratio of the number of pixels in each grid within the welding material region corresponding to each welding material specification type to the total number of pixels in the region, and the variance is the variance of the pixel density of all grids within the welding material region corresponding to each welding material specification type.

[0013] Preferably, the result calibration unit includes a configuration subunit, configured to configure a first weight for the first counting result and a second weight for the second counting result when the contour overlap rate is less than a third threshold, wherein the first weight is greater than the second weight. When the contour overlap rate is between the third and fourth thresholds, a first weight is assigned to the first counting result and a second weight is assigned to the second counting result, wherein the first weight is less than the second weight. When the contour overlap rate is greater than the fourth threshold, the first weight is configured to 0 and the second weight is configured to 1.

[0014] Preferably, the welding material recycling module includes: The scrap material recognition unit is used to input the image of the remaining welding material into the pre-trained scrap material recognition model to obtain the remaining length and surface condition of each scrap material contained therein. The waste material management unit, connected to the waste material identification unit, is used to mark the corresponding waste material as reusable when the remaining length and the surface condition meet the preset reuse conditions, otherwise as non-reusable, and to count the quantity of waste material as the recycling data.

[0015] Preferably, the dispatching system further includes a report generation module, used to generate and save electronic reports based on the welding material data to be requisitioned, the specification data of the welding materials to be issued, the identification and counting results, and the recycling data associated with the electronic material requisition form.

[0016] The above technical solution has the following advantages or beneficial effects: 1) By integrating digital approval, full-process traceability, automatic counting, and QR code verification technologies, the traditional pain points of CAP1000 nuclear island welding material management have been completely solved, the level of welding material management has been improved, human error has been effectively controlled, the efficiency of the entire process has been improved, and full-process traceability has been achieved. The time for welding material requisition and recycling has been shortened while ensuring welding quality, which in turn shortens the construction period and improves economic benefits. 2) Through the online approval and electronic material requisition form generation functions of the dispatch system, the entire requisition process can be digitally controlled. The main professional work group can apply for welders online in advance, the welding foreman can review and dispatch the work online, and the technician can approve and trigger the generation of electronic material requisition forms online. The entire approval process does not require manual transmission of paper documents, which significantly shortens the approval time and is especially suitable for the rapid response needs in emergency repair scenarios. In addition, the electronic material requisition form comes with a QR code containing the data of the welding materials to be requisitioned, which does not require manual refilling. Moreover, the data of the welding materials to be requisitioned is synchronized to the subsequent issuance and recycling stages, realizing the paperless management of welding materials throughout the entire process. 3) When welding materials are issued, the welding material issuance module simultaneously captures images of the issued welding materials using a high-speed scanner and binds the image data with the electronic material requisition form, the recipient information, and the identification and counting results; when welding materials are recycled, the welding material recycling module captures images of the remaining welding materials using a high-speed scanner and stores them in conjunction with the original electronic material requisition form. This can completely restore the recipient's identity, the physical state of the materials at the time of issuance, and the characteristics of the remaining materials at the time of recycling for each batch of welding materials. When welding quality defects occur, the entire circulation path of the problematic welding materials can be quickly traced, effectively improving traceability efficiency. 4) The counting results and recycling data are obtained through automatic image recognition during the distribution and recycling of welding materials, which changes the traditional manual counting method, greatly saves the time of manual counting and avoids the risk of counting errors; 5) By scanning the QR code with a barcode scanner, the information of the recipient can be retrieved quickly and accurately, which changes the traditional way of receiving welding rods and prevents mis-receiving. 6) The status monitoring module continuously collects the stored data of unissued welding materials and generates availability prediction results, which can identify the risk of welding material performance degradation in advance and avoid the issuance of unqualified welding materials. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the structure of an intelligent management system for the traceability of the entire process of welding material distribution and recycling, which is a preferred embodiment of the present invention. Detailed Implementation

[0018] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The present invention is not limited to this embodiment; other embodiments that conform to the spirit of the present invention may also fall within the scope of the present invention.

[0019] In a preferred embodiment of the present invention, based on the above-mentioned problems existing in the prior art, an intelligent management system for the entire process of welding material issuance and recycling that is traceable is provided, such as... Figure 1 As shown, it includes: The dispatch system 1 is used to generate electronic material requisition forms with QR codes based on welding task information, and to issue electronic material requisition forms after the requisition application submitted by the requisitioner is approved online. The QR code contains the data of welding materials to be requisitioned. Status monitoring module 2 is connected to dispatch system 1 and is used to continuously collect the stored data of unissued welding materials, process the stored data to obtain the availability prediction results of unissued welding materials and synchronize them to dispatch system 1. The welding material distribution module 3 is connected to the scanner 4, the high-speed scanner 5 and the dispatching system 1. It is used to call the availability prediction results based on the welding material data to be issued obtained by the distributor through scanning the electronic material requisition form with the scanner, so that the distributor can prepare the welding materials to be issued. It also recognizes and counts the images of the welding materials to be issued obtained by the high-speed scanner, and synchronizes the specification data of the welding materials to be issued to the dispatching system after binding the electronic material requisition form, the recipient and the recognition and counting results. The welding material recycling module 6 is connected to the high-speed scanner 5. It is used to identify the remaining welding materials based on the images of the remaining welding materials taken by the recycling personnel through the high-speed scanner, and then synchronize the identified recycling data to the electronic material requisition form and the dispatch system 1.

[0020] Specifically, in this embodiment, the welding task information includes, but is not limited to, weld joint number, required welding material specifications and quantity, and task type such as nuclear island main loop welding. This welding task information is preferably entered into the dispatch system 1 by the work group. Subsequently, the dispatch system 1 can generate an electronic requisition form with a QR code based on the welding task information. The requisition form contains welding material data including welding material specifications, quantity, and approval node. It can also be associated with weld joint number, welding material furnace batch number range, and usage time limit requirements, which completely solves the problem of task and welding material mismatch when manually filling out requisition forms, and ensures that each electronic requisition form complies with the technical specifications of nuclear island welding.

[0021] Among them, the dispatch system 1 is preferably customized and developed according to the actual process on site. It can be installed on a computer or mobile terminal. The technicians of the work group can directly enter the task information and submit the requisition application on the welding site without having to return to the office to fill out paper forms. The approvers (team leaders, quality engineers, and material managers) can receive approval notifications in real time through the system and view the details of the requisition application and related task data online anytime and anywhere, so as to achieve instant reception and instant review, significantly shortening the time for requisition and approval processes.

[0022] Furthermore, when the applicant submits a requisition application in the dispatch system 1, the application should preferably include the applicant's welder qualification certification information so that the dispatch system 1 can confirm whether the welder's qualification certification information matches the welding task requirements. For example, for a nuclear-grade weld welding task, the dispatch system 1 will verify whether the applicant holds a Class I qualification certificate, and whether the certificate is valid and whether the permitted welding materials are consistent with the current task. If the qualifications do not match, the dispatch system 1 will directly issue a pop-up warning and reject the application, avoiding the risk of unlicensed operation caused by oversights in qualification verification during traditional manual approval.

[0023] After the electronic material requisition form is approved, the welding material distribution process can begin. When the distribution personnel use barcode scanner 4 to scan the QR code on the electronic material requisition form presented by the recipient, the dispatch system 1 will automatically retrieve the welding material data to be requisitioned and the availability prediction results from the status monitoring module 2. The distribution interface will clearly display the specifications, quantity, furnace batch number, and availability status of the welding materials to be issued. At the same time, it will compare the data with the actual labels of the welding materials stored in the warehouse. For example, if the furnace batch number barcode on the welding material packaging is scanned, the dispatch system 1 will automatically verify whether it matches the data to be requisitioned. If the actual specifications do not match the application, the furnace batch number is not within the allowed range, or the availability is predicted to be unavailable, the system will immediately issue a pop-up warning and prohibit the issuance operation, replacing the traditional method of manually checking documents and physical items, and significantly reducing the error rate.

[0024] Subsequently, the personnel placing the welding materials to be distributed into the shooting area of ​​the high-speed scanner 5, which then sends the captured image of the welding materials to the welding material distribution module 3 for identification and counting. After distribution, the welding material distribution module 3 automatically binds the specifications, furnace batch number, and actual quantity of the welding materials to be distributed with the electronic material requisition form, the recipient's information (including facial image, simultaneously captured by the high-speed scanner), and the AI ​​counting results, forming a complete distribution record, which is then synchronized to the dispatch system 1 in real time. It can be seen that this distribution record not only includes the traditional paper ledger entries of who received the materials and how much was received, but also adds details such as the welding material status at the time of receipt, images of the counting process, and the receipt time accurate to the second. If subsequent traceability is required, the complete distribution process data can be quickly retrieved through the dispatch system 1, effectively improving traceability efficiency.

[0025] If there are any leftover welding materials after use, a recycling process is required. The recycling personnel place the remaining welding materials (such as unused single welding rods or cut welding wires) under the high-speed scanner 5. The high-speed scanner 5 sends the captured image of the remaining welding materials to the welding material recycling module 6 for residual material identification. The identified recycling data is then bound to the electronic material requisition form and synchronized to the dispatch system 1 to achieve a data closed loop of requisition, issuance, and recycling.

[0026] As can be seen, through the collaborative work of the dispatch system 1, status monitoring module 2, welding material distribution module 3, welding material recycling module 6, and supporting equipment, the entire welding material management system has achieved digital approval for requisition, precise distribution, intelligent recycling management, and full-process traceability. This not only solves the pain points of traditional manual management, such as low efficiency, high risk, and difficulty in traceability, but also technically ensures the stability of welding quality in the CAP1000 nuclear island, while supporting the digital transformation needs of nuclear power engineering. The results are then synchronized to the dispatch system. In a preferred embodiment of the present invention, the status monitoring module 2 includes: The data acquisition unit 21 is used to continuously collect environmental parameters of the storage location of the welding materials to be issued and the storage time of the welding materials to be issued as storage data. The availability prediction unit 22 is connected to the data acquisition unit 21 and is used to input the stored data into the pre-trained welding material performance prediction model to obtain the availability prediction results of the welding materials to be issued and synchronize them to the dispatching system. Availability prediction results include at least one risk type and its corresponding confidence level.

[0027] Specifically, in this embodiment, the status monitoring module 2 performs availability prediction on the stored unissued welding materials, thereby shifting the performance control of welding materials from post-event inspection to pre-event prediction. The status monitoring module 2 synchronizes the availability prediction results to the dispatching system 1 in real time and binds them to the furnace batch number and specifications of the corresponding batch of welding materials within the system. When the welding material issuance module 3 calls the data of welding materials to be issued, the prediction results are presented to the issuance personnel along with the data, clearly indicating which welding materials can be issued directly, which need to be re-inspected before issuance, and which are prohibited from issuance, thus preventing unqualified welding materials from entering the work site and providing a pre-emptive guarantee for welding quality.

[0028] More specifically, the data acquisition unit 21 includes a temperature and humidity sensor and an oxidation potential sensor deployed at the storage location, used to collect temperature and humidity data and the surface potential value of undistributed welding materials as environmental parameters. The temperature and humidity sensors are preferably deployed in the center of each layer of the welding material storage rack, in the corners inside the sealed cabinet, and at warehouse ventilation openings to ensure comprehensive data coverage and avoid the problem of undetected localized temperature and humidity anomalies caused by traditional single-point collection. The oxidation potential sensor is preferably a miniature patch type, which can be directly attached to the surface of the welding wire reel or a metal probe inside the welding rod box. The aforementioned storage time can be linked to the welding material furnace batch number and storage time through the warehouse management function of the dispatch system 1. The data acquisition unit 21 automatically reads this time and accumulates the storage time, accurate to the hour, avoiding omissions and errors in manually recording storage time.

[0029] Furthermore, the above-mentioned welding material performance prediction model preferably adopts a fusion model of gradient boosting tree (XGBoost) and long short-term memory network (LSTM). XGBoost is used to process static features such as temperature and humidity and storage time, while LSTM is used to capture the dynamic trend of oxidation potential changes over time, and finally outputs the risk type and confidence level.

[0030] The aforementioned risk types include one or more of the following: electrode coating dampness, electrode coating hardening, and welding wire surface oxidation. For example, for a batch of E5015 electrodes, with input data of "temperature 28℃, humidity 70% RH, storage for 4 days, oxidation potential -250mV", the model outputs a 68% confidence level for electrode coating dampness, a 12% confidence level for electrode coating hardening, and a 5% confidence level for welding wire surface oxidation.

[0031] In a preferred embodiment of the present invention, the dispatching system 1 includes a status warning module 11, which is used to mark the corresponding undistributed welding materials as distributable when the confidence level corresponding to each risk type is less than a first threshold. When the confidence level corresponding to any risk type is not less than the first threshold and less than the second threshold, the corresponding unissued welding materials are marked as temporarily suspended, and an environmental early warning prompt containing environmental parameters, availability prediction results and the specification data of the corresponding unissued welding materials is generated and pushed to the relevant technical personnel for review. And when the confidence level corresponding to any risk type is not less than the second threshold, the corresponding unissued welding materials are marked as unissued and the relevant technical personnel are notified to conduct on-site confirmation.

[0032] Specifically, in this embodiment, the first threshold is a low-risk threshold, which is set based on the premise that the risk does not affect the basic performance of the welding material. When the confidence level corresponding to each risk type is less than the first threshold, it is preferable to display a green icon in the dispatching system 1 to indicate that it is in a disbursable state, so that the normal welding material disbursement process is not hindered.

[0033] The second threshold mentioned above is a high-risk threshold, set based on the risk that may lead to the scrapping of welding materials. If the confidence level of any risk type is between the first threshold and the second threshold, it means that the current storage environment has caused the welding materials to have damp coating, hardened coating, or oxidized surface of welding wire, but the risk level is not high. In the dispatch system 1, a yellow mark can be displayed to indicate that it is in a state of temporary suspension. Even if intervention is carried out, it may be repairable. For example, a batch of welding materials was found to be damp in time, and after drying, the reuse rate reached 90%, reducing waste.

[0034] If the confidence level of any risk type is higher than the second threshold, it indicates that the corresponding unissued welding material is a high-risk welding material. In the dispatch system 1, a red mark is displayed to indicate that it is not to be issued, thus effectively intercepting high-risk welding materials. At the same time, an emergency on-site confirmation notice is pushed to the technicians to make it clear that the technicians should go to the warehouse for on-site re-inspection in a short period of time.

[0035] In a preferred embodiment of the present invention, the dispatching system 1 further includes a status unlocking module 12, comprising: The first unlocking unit 121 is used for technicians to view environmental warning prompts online and provide disposal suggestions. Then, they can sign an online consent form for distribution and push it to the warehouse management personnel. After the warehouse management personnel implement the disposal suggestions, they can upload the disposal completion instruction and then adjust the temporarily suspended distribution status to the distribution status according to the disposal completion instruction. The second unlocking unit 122 is used for technicians to sign and upload a confirmation form after on-site re-inspection, and then adjust the unissued status to issueable status based on the confirmation form.

[0036] Specifically, in this embodiment, for the temporarily suspended issuance status, after reviewing the warning prompts and disposal suggestions, the technicians fill out a disposal opinion confirmation form in the system. If they agree to dry at 60℃ for 2 hours, the oxidation potential needs to be retested after drying, and an electronic signature is completed. After receiving the confirmation form, the warehouse management personnel execute the disposal operation. After the disposal is completed, the disposal record needs to be uploaded, such as the temperature-time curve of the drying equipment and the retested oxidation potential data (-220mV). After the system verifies that the data is qualified, it automatically adjusts the temporarily suspended issuance to the issuance status, avoiding the waste caused by the traditional one-size-fits-all scrapping. The entire disposal process is online, without the need for manual transmission of paper documents. If quality problems occur later, it is possible to quickly locate whether there are any omissions in the disposal process.

[0037] The aforementioned treatment suggestions can also be pre-configured according to nuclear power welding material handling specifications. Technicians can provide treatment suggestions simply by selecting options, further improving treatment efficiency. For example, the treatment suggestion for damp flux coating is to dry it with hot air at 60°C for 2 hours, and the treatment suggestion for slight oxidation is to lightly sand it with sandpaper to remove the oxide film.

[0038] For materials that cannot be issued, technicians need to bring portable testing equipment (such as a flux moisture content meter and an oxide film thickness meter) to the warehouse to conduct item-by-item testing on the unissued welding materials. For example, for surface oxidation of the welding wire, the oxide film thickness needs to be measured (required to be <10μm); for flux hardening, the flux hardness needs to be tested (required to be <HV75). After passing the re-inspection, the technicians need to fill out an on-site re-inspection report, including the test data and the signature of the re-inspection personnel, and upload it to the system. The system needs to be approved by both the technical manager and the quality engineer. After approval, a confirmation of use opinion form is generated. The warehouse manager uses this form to unlock the non-issue status and adjust it to issueable, ensuring that the unlocked welding materials meet the usage standards. All re-inspection data and approval records are stored in the system, forming a complete traceability chain of early warning, re-inspection, and unlocking.

[0039] In a preferred embodiment of the present invention, the welding material dispensing module 3 includes: Specification recognition unit 31 is used to identify the type of welding material specifications contained in the image of welding material to be issued; The contour recognition unit 32 is connected to the specification recognition unit 31, which is used to identify the overall contour of each welding material for each type of welding material specification, calculate the contour overlap rate, and count the overall contour to obtain the first counting result. The density estimation unit 33, connected to the specification identification unit 31, is used to calculate the overall occupied area of ​​each weld material contained therein and the variance of the pixel density of the weld material area for each weld material specification type, and to calculate the ratio between the overall occupied area and the standard area of ​​a single weld material of the corresponding weld material specification type to obtain a second counting result. The result calibration unit 34 is connected to the contour recognition unit 32 and the density estimation unit 33 respectively. When the first counting result and the second counting result are inconsistent, the unit configures the weights according to the contour overlap rate and the pixel density variance, and weights the first counting result and the second counting result to obtain the recognition counting result.

[0040] Specifically, considering the complex scenarios that may occur during actual distribution, such as mixed placement of multiple specifications and slight stacking, this embodiment effectively improves counting accuracy through multi-feature cross-validation. Here, multi-feature refers to counting based on contour features and density features respectively, and then weighting and summing the results of both to further improve counting accuracy.

[0041] Before counting, it is necessary to first determine the type of welding material specifications contained in the image of the welding material to be issued. Specifically, the type of welding material specifications can be identified through the built-in feature library of commonly used welding material specifications in the nuclear island, such as: Based on their morphological characteristics, there is a significant difference between the long, straight cylindrical shape of welding wire and the layered cylindrical shape of welding electrode coating and core.

[0042] Based on size characteristics, the diameter or length can be extracted using the minimum bounding rectangle algorithm, or further distinguished by combining the diameter-to-length ratio.

[0043] Based on texture characteristics, there is a significant morphological difference between the rough texture of the electrode coating and the smooth metal surface of the welding wire.

[0044] During identification, the extracted features can be matched with the features stored in the welding material specification feature library to obtain the welding material specification type.

[0045] For each identified welding material specification type, multi-feature cross-validation counting is performed, specifically including: For contour features: The Canny edge detection algorithm is preferably used to extract all contours. Target specifications are filtered by the minimum aspect ratio of the bounding rectangle, such as Φ2.0 welding wire, where the aspect ratio of the bounding rectangle is approximately 5:1, to distinguish it from welding rods. Then, discrete contours belonging to the same specification are merged into a single-specification overall contour region. Preferably, a mask operation can be used to preserve the target contour, with the remaining areas set as background. Furthermore, the area after dilation and the area after erosion are calculated using morphological dilation-erosion methods. The difference between the two is the welding material contour area of ​​the adhesion region. The contour overlap rate is the ratio of the welding material contour area of ​​the adhesion region to the total contour area for each welding material specification type.

[0046] For density features: pixel density is the ratio of the number of pixels in each grid in the welding material area corresponding to each welding material specification type to the total number of pixels in the area, and variance is the variance of the pixel density of all grids in the welding material area corresponding to each welding material specification type.

[0047] Furthermore, the aforementioned contour overlap rate can directly reflect the stacking state of the welding materials. If low overlap prioritizes contour counting and high overlap prioritizes density estimation, the variance of pixel density can reflect the uniformity of density distribution. The larger the variance, the higher the weight of density estimation.

[0048] In a preferred embodiment of the present invention, the result calibration unit 34 includes a configuration subunit 341, configured to configure a first weight for a first counting result and a second weight for a second counting result when the contour overlap rate is less than a third threshold, wherein the first weight is greater than the second weight. When the contour overlap rate is between the third and fourth thresholds, a first weight is assigned to the first counting result and a second weight is assigned to the second counting result, wherein the first weight is less than the second weight. When the contour overlap rate is greater than the fourth threshold, the first weight is set to 0 and the second weight is set to 1.

[0049] Specifically, in this embodiment, the third threshold is set to 0.1 and the fourth threshold is set to 0.3. For example, it can include the following three scenarios: 1. Low overlap scene In this scenario, the contour overlap rate is less than the third threshold. The main principle is to focus on contour counting, meaning the first weight is greater than the second weight. The specific values ​​of the first and second weights can be further subdivided as follows: If the variance is less than the third threshold, the density is considered uniform and the profile is completely reliable. The first weight can be set to 1 and the second weight to 0.

[0050] If the variance is between the third and fourth thresholds, the density is considered to be slightly uneven. In this case, the contour is the main factor and the density is used as an auxiliary factor for correction. The first weight can be set to 0.9 and the second weight can be set to 0.1.

[0051] If the variance is greater than the fourth threshold, the density is considered to be severely uneven. In this case, the contour is the main factor and the density is used as an auxiliary factor to eliminate interference. The first weight can be set to 0.7 and the second weight can be set to 0.3.

[0052] 2. Overlapping scenes In this scenario, the contour overlap rate is between the third and fourth thresholds. The main principle is to use density counting as the core, meaning the first weight is less than the second weight. The specific values ​​of the first and second weights can be further subdivided as follows: If the variance is less than the third threshold, the density is considered uniform and completely reliable. The first weight can be set to 0 and the second weight to 1.

[0053] If the variance is between the third and fourth thresholds, the density is considered to be slightly uneven. In this case, density is the main factor, and the contour is used as an auxiliary correction. The first weight can be set to 0.1 and the second weight to 0.9.

[0054] If the variance is greater than the fourth threshold, the density is considered to be severely uneven. In this case, density is the primary factor, and density is used as an auxiliary factor to eliminate interference. The first weight can be configured as 0.2 and the second weight as 0.8. Prioritize the areas of difference for verification. This can remind the distribution personnel to conduct manual verification or rearrange the items.

[0055] 3. High overlap scenes In this scenario, the contour overlap rate is greater than the fourth threshold. The main principle is to completely dominate the density count. The first weight can be configured as 0 and the second weight as 1.

[0056] Furthermore, if the first counting result and the second counting result are the same, the above multi-feature cross-validation can be skipped, and the first counting result or the second counting result can be directly used as the identification counting result.

[0057] In a preferred embodiment of the present invention, the welding material recycling module 6 includes: The scrap material recognition unit 61 is used to input the image of the remaining welding material into the pre-trained scrap material recognition model to obtain the remaining length and surface condition of each scrap material contained therein. The waste material management unit 62 is connected to the waste material identification unit 61. When the remaining length and surface condition meet the preset reuse conditions, the corresponding waste material is marked as reusable, otherwise it is marked as unreusable, and the quantity of waste material is counted as recycling data.

[0058] Specifically, in this embodiment, to address potential issues such as bending, breakage, and end wear in recycled scrap, the scrap identification model preferably performs morphological skeleton extraction on the image of the remaining welding material to obtain the centerline of the scrap. Even if the scrap is bent, its true outline can be restored. The length of the centerline is calculated using a pixel-to-actual-size conversion factor, which is the remaining length of the scrap. To address visual length deviations caused by electrode tip wear and wire tip oxidation, the model automatically identifies the wear area (grayscale value 20% lower than normal) and corrects the length by half the maximum width of the wear area to ensure measurement accuracy.

[0059] Furthermore, the scrap identification model can also accurately identify surface defects that affect welding quality, specifically including: Electrode surface condition: coating peeling (area percentage), coating cracking (crack length / width), corrosion (corrosion area percentage), moisture (surface reflectivity); Welding wire surface condition: oxidation (image features corresponding to oxidation potential), scratches (number of scratches with a depth > 0.1 mm), and oil stains (percentage of oil stain area).

[0060] Furthermore, the remaining length and surface condition of the scrap material can be determined by preset reuse conditions to determine whether it can be reused. If it can be reused, it can be marked as reusable in the dispatching system 1 and assigned a unique reuse code. This unique reuse code is preferably associated with the original electronic material requisition number to achieve traceability. Otherwise, it is marked as unreusable and the reason for unreusability is noted accordingly.

[0061] It is evident that the recycled data not only includes reusable and non-reusable quantities, but also records in detail the length distribution, surface defect types and proportions of leftover materials of various specifications, and automatically binds to the corresponding electronic material requisition forms, realizing a data closed loop of requisition-issuance-use-recycling.

[0062] In a preferred embodiment of the present invention, the dispatching system 1 further includes a report generation module 13, which is used to generate and save electronic reports based on the welding material data to be requisitioned, the specification data of welding materials to be issued, the identification and counting results, and the recycling data associated with the electronic material requisition form.

[0063] The above description is merely a preferred embodiment of the present invention and does not limit the implementation and protection scope of the present invention. Those skilled in the art should realize that any equivalent substitutions and obvious changes made using the content of this specification and illustrations should be included within the protection scope of the present invention.

Claims

1. A smart management system for welding material issuing and recycling whole-process traceability, characterized in that, The system comprises: a dispatch system for generating an electronic material requisition form with a two-dimensional code according to welding task information, and issuing the electronic material requisition form after the requisition application submitted by the requisition person is approved online, wherein the two-dimensional code contains welding material data to be used; a state monitoring module connected to the dispatch system for continuously collecting storage data of unissued welding materials, processing the storage data to obtain availability prediction results of the unissued welding materials, and synchronizing the results to the dispatch system; a welding material issuing module connected to a scanning gun, a high-speed camera, and the dispatch system for calling the availability prediction results according to the welding material data to be used obtained by the issuing personnel scanning the electronic material requisition form, preparing the unissued welding materials to be issued, identifying and counting the unissued welding materials to be issued through the high-speed camera, and synchronizing the specification data of the unissued welding materials to the electronic material requisition form, the requisition person, and the identification and counting results to the dispatch system; a welding material recycling module connected to the high-speed camera for identifying the remaining materials according to the remaining material images obtained by the recycling personnel through the high-speed camera, and synchronizing the identified recycling data to the electronic material requisition form to the dispatch system.

2. The intelligent management system of claim 1, wherein, The state monitoring module comprises: a data collection unit for continuously collecting environmental parameters of the storage location of the unissued welding materials and storage duration of the unissued welding materials as the storage data; an availability prediction unit connected to the data collection unit for inputting the storage data into a pre-trained welding material performance prediction model to obtain the availability prediction results of the unissued welding materials, and synchronizing the results to the dispatch system; The availability prediction results include at least one risk type and its corresponding confidence.

3. The intelligent management system of claim 2, wherein, The data collection unit comprises a temperature and humidity sensor and an oxidation potential sensor deployed at the storage location for collecting temperature and humidity data of the storage location and surface potential value of the unissued welding materials as the environmental parameters, respectively; The risk types include one or more of electrode coating dampness, electrode coating hardening, and welding wire surface oxidation.

4. The intelligent management system of claim 2, wherein, The dispatch system comprises a state warning module for marking the corresponding unissued welding materials as issuable when the confidence corresponding to each risk type is less than a first threshold value; marking the corresponding unissued welding materials as temporarily issuable when the confidence corresponding to any risk type is not less than a first threshold value and less than a second threshold value, and generating an environmental warning prompt containing the environmental parameters, the availability prediction results, and the specification data of the corresponding unissued welding materials to push to the corresponding technical personnel for viewing; and marking the corresponding unissued welding materials as unissuable when the confidence corresponding to any risk type is not less than a second threshold value, and notifying the corresponding technical personnel for on-site confirmation.

5. The intelligent management system of claim 4, wherein, The dispatch system further comprises a state unlocking module comprising: A first unlocking unit is configured to allow a technician to view the environmental early warning and give a treatment suggestion online, and then sign an agreement to issue an opinion sheet online and push the opinion sheet to a warehouse manager, and allow the warehouse manager to upload a treatment completion instruction after executing the treatment suggestion, and then adjust the temporary suspension state to a deliverable state according to the treatment completion instruction; A second unlocking unit is configured to allow the technician to sign and upload a confirmation use opinion sheet after on-site re-inspection, and then adjust the undeliverable state to the deliverable state according to the confirmation use opinion sheet.

6. The intelligent management system of claim 1, wherein, The welding material delivery module comprises: A specification identification unit is configured to identify a welding material specification type contained in the welding material image to be delivered; A contour identification unit connected to the specification identification unit is configured to identify an overall contour of each welding material for each welding material specification type, calculate a contour overlap rate, and count the overall contour to obtain a first count result; A density estimation unit connected to the specification identification unit is configured to calculate an overall occupied area of each welding material and a variance of pixel density of a welding material area for each welding material specification type, and calculate a ratio between the overall occupied area and a single welding material standard area of the corresponding welding material specification type to obtain a second count result; A result calibration unit connected to the contour identification unit and the density estimation unit is configured to perform weight configuration according to the contour overlap rate and the variance of pixel density when the first count result and the second count result are inconsistent, and perform weighting on the first count result and the second count result to obtain the identification count result.

7. The intelligent management system of claim 6, wherein, The contour overlap rate is a ratio of a welding material contour area of a corresponding adhesion area to a total contour area for each welding material specification type; The pixel density is a ratio of a number of pixels of each grid contained in a corresponding welding material area to a total number of area pixels for each welding material specification type, and the variance is a variance of the pixel density of all the grids contained in the corresponding welding material area for each welding material specification type.

8. The intelligent management system of claim 6, wherein, The result calibration unit comprises a configuration subunit configured to configure a first weight for the first count result and a second weight for the second count result when the contour overlap rate is less than a third threshold value, and the first weight is greater than the second weight, configure a first weight for the first count result and a second weight for the second count result when the contour overlap rate is between the third threshold value and a fourth threshold value, and the first weight is less than the second weight, configure the first weight as 0 and the second weight as 1 when the contour overlap rate is greater than the fourth threshold value.

9. The intelligent management system of claim 1, wherein, The welding material recovery module comprises: A surplus material identification unit is configured to input the remaining welding material image into a pre-trained surplus material identification model to obtain a remaining length and a surface state of each surplus material contained therein. A surplus management unit is connected to the surplus identification unit, configured to mark the corresponding surplus as reusable if the remaining length and the surface state satisfy a preset reuse condition, and as non-reusable otherwise, and count the number of reusable surplus as the recovery data.

10. The intelligent management system of claim 1, wherein, The dispatch system further comprises a report generation module configured to generate an electronic report based on the to-be-used welding material data associated with the electronic material requisition, the specification data of the to-be-issued welding material, the identification counting result, and the recovery data, and save the electronic report.

Images