Welding gun fault prediction method and device based on edge calculation and electronic equipment

Abstract

The invention discloses a welding gun fault prediction method and device based on edge calculation and electronic equipment, and the method comprises the steps: obtaining welding arc parameters and physical state data of a target welding gun, and transmitting the welding arc parameters and the physical state data to an edge calculation layer; the welding arc parameters and the physical state data are input into a pre-trained welding gun fault prediction model issued by a cloud service layer through an edge calculation layer, a fault prediction result is obtained, and a corresponding maintenance plan is generated according to the fault prediction result; actual maintenance data of a target welding gun is obtained through the edge calculation layer, a target sample is generated according to the welding arc parameters, the physical state data and the actual maintenance data, the target sample is uploaded to the cloud service layer, and a cloud server conducts incremental training on a welding gun fault prediction model according to the target sample. The method improves the accuracy and timeliness of welding gun fault prediction, reduces the shutdown risk and maintenance cost of a production line, and can be applied to the technical field of automobile production.

Description

Technical Field

[0001] This invention relates to the field of automotive manufacturing technology, and in particular to a welding torch fault prediction method, device, and electronic device based on edge computing. Background Technology

[0002] In automotive body welding production, the welding torch is a core piece of equipment. A production line typically has 20 to 30 welding torches. Problems such as electrode wear, abnormal current, and cooling system malfunctions can all lead to welding quality defects (such as incomplete welds or missed welds). If these defects are not detected in time, they can result in defective products or even burn out the welding torch, causing the entire production line to shut down and resulting in huge losses for automakers. Traditional solutions involve regular manual inspections and post-production maintenance. However, this method relies on human experience for prediction, which is not very accurate and can easily lead to delayed fault detection, thus still posing significant production risks.

[0003] The above problems urgently need to be addressed. Summary of the Invention

[0004] The purpose of this invention is to at least partially solve one of the technical problems existing in the prior art.

[0005] Therefore, one objective of this invention is to provide a welding torch fault prediction method based on edge computing, which improves the accuracy and timeliness of welding torch fault prediction and reduces the downtime risk and maintenance cost of the production line.

[0006] Another objective of this invention is to provide a welding torch fault prediction device based on edge computing.

[0007] To achieve the above-mentioned technical objectives, the technical solutions adopted in the embodiments of the present invention include: On one hand, embodiments of the present invention provide a welding torch fault prediction method based on edge computing, comprising the following steps: The welding arc parameters and physical state data of the target welding torch are acquired, and the welding arc parameters and physical state data are transmitted to the edge computing layer. The edge computing layer inputs the welding arc parameters and the physical state data into the pre-trained welding torch fault prediction model issued by the cloud service layer to obtain the fault prediction result, and generates the corresponding maintenance plan based on the fault prediction result. The edge computing layer obtains the actual maintenance data of the target welding torch, generates a target sample based on the welding arc parameters, the physical state data, and the actual maintenance data, and uploads the target sample to the cloud service layer, so that the cloud server can incrementally train the welding torch fault prediction model based on the target sample.

[0008] Furthermore, in one embodiment of the present invention, the step of acquiring the welding arc parameters and physical state data of the target welding torch, and transmitting the welding arc parameters and physical state data to the edge computing layer, specifically includes: The welding current timing data and welding voltage timing data of the target welding torch are obtained through current sensors and voltage sensors; Temperature timing data of the nozzle and transformer of the target welding torch are obtained through a temperature sensor; Vibration timing data of the robotic arm of the target welding torch are obtained by using an accelerometer. The electrode pressure timing data of the target welding torch is obtained by a piezoelectric force sensor; The welding current timing data and the welding voltage timing data are used as the welding arc parameters, and the temperature timing data, the vibration timing data, and the electrode pressure timing data are used as the physical state data. The welding arc parameters and the physical state data are transmitted to the edge computing layer through the edge gateway.

[0009] Furthermore, in one embodiment of the present invention, the cloud service layer pre-trains the welding torch fault prediction model through the following steps: Obtain historical welding arc parameters and historical physical state data of the sample welding torch; Training samples are constructed based on the historical welding arc parameters and the historical physical state data, and welding gun fault labels of the training samples are determined by manual annotation. The training samples are input into a pre-constructed CNN-LSTM hybrid neural network to obtain predicted fault labels; A first loss value is determined based on the predicted fault label and the welding torch fault label; The parameters of the CNN-LSTM hybrid neural network are updated based on the first loss value to obtain the welding torch fault prediction model; The historical welding arc parameters include historical welding current time series data and historical welding voltage time series data. The historical physical state data includes historical temperature time series data, historical vibration time series data, and historical electrode pressure time series data. The welding torch fault label includes the fault type and the corresponding fault severity.

[0010] Furthermore, in one embodiment of the present invention, the CNN-LSTM hybrid neural network includes a current convolutional branch network, a voltage convolutional branch network, a temperature convolutional branch network, a vibration convolutional branch network, a pressure convolutional branch network, an LSTM branch network, a feature fusion layer, and an output layer. The step of inputting the training samples into the pre-constructed CNN-LSTM hybrid neural network to obtain predicted fault labels specifically includes: The historical welding current time series data is input into the current convolutional branch network to obtain local current features; The historical welding voltage time series data is input into the voltage convolutional branch network to obtain local voltage features; The historical temperature time series data is input into the temperature convolutional branch network to obtain local temperature features; The historical vibration time series data is input into the vibration convolutional branch network to obtain local vibration features; The historical electrode pressure time series data is input into the pressure convolutional branch network to obtain local pressure features; The historical welding arc parameters and the historical physical state data are input into the LSTM branch network to obtain multimodal long-term dependency features. The feature fusion layer performs feature fusion on the local current features, local voltage features, local temperature features, local vibration features, local pressure features, and long-term multimodal dependence features based on an attention mechanism to obtain global spatiotemporal fusion features; The output layer maps the global spatiotemporal fusion features to the predicted fault labels.

[0011] Furthermore, in one embodiment of the present invention, the step of generating a corresponding maintenance plan based on the fault prediction result specifically includes: Obtain the fault maintenance mapping table issued by the cloud service layer; Determine the target fault type and fault severity corresponding to the fault prediction results; Based on the target fault type and the fault severity, the corresponding target maintenance strategy is obtained by querying the fault maintenance mapping table; The maintenance plan is generated based on the target maintenance strategy, and the maintenance plan is pushed to the management terminal.

[0012] Furthermore, in one embodiment of the present invention, the step of generating a target sample based on the welding arc parameters, the physical state data, and the actual maintenance data, and uploading the target sample to the cloud service layer, specifically includes: The target sample is constructed based on the welding arc parameters and the physical state data; The fault prediction results are corrected based on the actual maintenance data to obtain the target fault label; The target sample is labeled according to the target fault label, and the labeled target sample is uploaded to the cloud service layer.

[0013] Furthermore, in one embodiment of the present invention, the incremental training of the welding torch fault prediction model based on the target sample specifically includes: The target sample is input into the welding torch fault prediction model to obtain the sample prediction result; A second loss value is determined based on the sample prediction results and the target fault label; The parameters of the output layer of the welding torch fault prediction model are updated based on the second loss value to obtain the optimized welding torch fault prediction model.

[0014] On the other hand, embodiments of the present invention provide a welding torch fault prediction device based on edge computing, comprising: The data acquisition module is used to acquire the welding arc parameters and physical state data of the target welding torch, and transmit the welding arc parameters and physical state data to the edge computing layer; The fault prediction module is used to input the welding arc parameters and the physical state data into the pre-trained welding torch fault prediction model issued by the cloud service layer through the edge computing layer, obtain the fault prediction result, and generate the corresponding maintenance plan based on the fault prediction result. The model optimization module is used to obtain the actual maintenance data of the target welding torch through the edge computing layer, generate target samples based on the welding arc parameters, the physical state data and the actual maintenance data, and upload the target samples to the cloud service layer, so that the cloud server can incrementally train the welding torch fault prediction model based on the target samples.

[0015] On the other hand, embodiments of the present invention provide an electronic device, including: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the above-described edge computing-based welding torch fault prediction method.

[0016] On the other hand, embodiments of the present invention also provide a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the above-described edge computing-based welding torch fault prediction method.

[0017] On the other hand, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the above-described edge computing-based welding torch fault prediction method.

[0018] The advantages and beneficial effects of the present invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention: In this embodiment of the invention, the welding arc parameters and physical state data of the target welding torch are acquired and transmitted to the edge computing layer. The edge computing layer then inputs the welding arc parameters and physical state data into a pre-trained welding torch fault prediction model distributed by the cloud service layer to obtain fault prediction results. Based on the fault prediction results, a corresponding maintenance plan is generated. The edge computing layer acquires the actual maintenance data of the target welding torch, generates target samples based on the welding arc parameters, physical state data, and actual maintenance data, and uploads the target samples to the cloud service layer. This allows the cloud server to incrementally train the welding torch fault prediction model based on the target samples. This invention extends computing power to the production site through an edge computing layer, effectively solving problems such as high data transmission latency, high cloud dependence, and untimely local response in traditional industrial scenarios. By deploying a cloud-trained welding torch fault prediction model at the edge computing layer, welding torch faults can be predicted based on real-time collected welding arc parameters and physical state data, and corresponding maintenance plans can be generated. This enables real-time monitoring, early warning, and precise maintenance of welding torch faults, improving the accuracy and timeliness of welding torch fault prediction. At the same time, it transforms passive maintenance into proactive prevention, significantly reducing the downtime risk and maintenance costs of the production line. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments of the present invention are described below. It should be understood that the drawings described below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating the steps of a welding torch fault prediction method based on edge computing, as provided in an embodiment of the present invention. Figure 2 A structural block diagram of a welding torch fault prediction device based on edge computing provided in an embodiment of the present invention; Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0023] The edge computing-based welding torch fault prediction method provided in this invention can be applied to terminals, servers, or software running on either terminal or server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited to these. The server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network. The software can be an application implementing the edge computing-based welding torch fault prediction method, but is not limited to the above forms.

[0024] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0025] It should be noted that in various specific embodiments of the present invention, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of the present invention require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to a confirmation page. Only after obtaining the user's separate permission or consent is the necessary user-related data for the normal operation of the embodiments of the present invention acquired.

[0026] Reference Figure 1 This invention provides a welding torch fault prediction method based on edge computing, specifically including the following steps: S101. Obtain the welding arc parameters and physical state data of the target welding torch, and transmit the welding arc parameters and physical state data to the edge computing layer; S102. The welding arc parameters and physical state data are input into the pre-trained welding torch fault prediction model issued by the cloud service layer through the edge computing layer to obtain the fault prediction results, and the corresponding maintenance plan is generated based on the fault prediction results. S103. Obtain the actual maintenance data of the target welding torch through the edge computing layer, generate target samples based on welding arc parameters, physical state data and actual maintenance data, and upload the target samples to the cloud service layer so that the cloud server can incrementally train the welding torch fault prediction model based on the target samples.

[0027] This invention adopts a cloud-edge-device collaborative architecture, including a device perception layer, an edge computing layer, and a cloud service layer, to achieve a complete closed loop from data collection to decision execution, wherein: 1) Equipment perception layer: mainly responsible for data acquisition and preliminary processing. It deploys various sensors on the welding torch and related equipment to collect welding arc parameters and physical state data in real time, and uses industrial-grade wireless protocols to transmit the data to the edge computing layer.

[0028] 2) Edge computing layer: It undertakes data processing, feature extraction and fault diagnosis tasks. It uses industrial-grade edge computers for real-time data analysis and processing, runs lightweight AI models distributed from the cloud for fault identification and prediction, triggers local alarms in real time and executes preliminary decisions, and uploads key data to the cloud.

[0029] 3) Cloud service layer: mainly responsible for global data management and model optimization, storing all historical operating data, fault records and maintenance records of welding torches, training and updating models based on big data, optimizing prediction algorithms, generating fault maintenance mapping tables and pushing them to the edge computing layer.

[0030] This invention extends computing power to the production site through an edge computing layer, effectively solving problems such as high data transmission latency, high cloud dependence, and untimely local response in traditional industrial scenarios. By deploying a cloud-trained welding torch fault prediction model at the edge computing layer, welding torch faults can be predicted based on real-time collected welding arc parameters and physical state data, and corresponding maintenance plans can be generated. This enables real-time monitoring, early warning, and precise maintenance of welding torch faults, improving the accuracy and timeliness of welding torch fault prediction. At the same time, it transforms passive maintenance into proactive prevention, significantly reducing the downtime risk and maintenance costs of the production line.

[0031] As a further optional implementation, the welding arc parameters and physical state data of the target welding torch are acquired, and the welding arc parameters and physical state data are transmitted to the edge computing layer, specifically including: S1011. Obtain welding current timing data and welding voltage timing data of the target welding torch through current sensor and voltage sensor; S1012. Obtain the temperature timing data of the nozzle and transformer of the target welding torch through a temperature sensor; S1013. Obtain vibration timing data of the robotic arm of the target welding torch through an accelerometer; S1014. Obtain the electrode pressure timing data of the target welding torch through a piezoelectric force sensor; S1015. Use welding current timing data and welding voltage timing data as welding arc parameters, and use temperature timing data, vibration timing data and electrode pressure timing data as physical state data. Transmit the welding arc parameters and physical state data to the edge computing layer through the edge gateway.

[0032] Specifically, the selection and installation of multi-source sensors in this embodiment of the invention are as follows: 1) Temperature monitoring: Deploy thermocouples or infrared sensors at key points such as welding torch nozzles and transformers, with a sampling frequency ≥100Hz and an accuracy of ±1℃ (refer to the water-cooled welding torch temperature control standard).

[0033] 2) Vibration monitoring: A triaxial accelerometer (such as PCB 356A15) is mounted on the welding gun robotic arm to capture vibrations in the 10Hz–10kHz frequency band and identify bearing wear or structural loosening.

[0034] 3) Electrical parameter monitoring: Welding arc parameters are acquired in real time through current / voltage Hall sensors with an accuracy of ±0.5%, and transient fluctuations (such as short-circuit peak values) are recorded simultaneously.

[0035] 4) Pressure monitoring: A piezoelectric force sensor is integrated into the welding torch clamping mechanism to monitor electrode pressure deviation (within ±5% of the target value).

[0036] Edge computing gateway deployment: Industrial intelligent gateways (such as YC-DPM-ECT) are used to achieve multi-protocol compatibility (Profibus DP / EtherCAT), with data synchronization latency ≤1ms and real-time early warning support.

[0037] As an optional implementation, the cloud service layer pre-trains a welding torch failure prediction model through the following steps: S201. Obtain historical welding arc parameters and historical physical state data of the sample welding torch; S202. Construct training samples based on historical welding arc parameters and historical physical state data, and determine the welding gun fault labels of the training samples through manual annotation. S203. Input the training samples into a pre-constructed CNN-LSTM hybrid neural network to obtain the predicted fault labels; S204. Determine the first loss value based on the predicted fault label and the welding torch fault label; S205. Update the parameters of the CNN-LSTM hybrid neural network based on the first loss value to obtain the welding torch fault prediction model; The historical welding arc parameters include historical welding current time series data and historical welding voltage time series data. The historical physical state data includes historical temperature time series data, historical vibration time series data, and historical electrode pressure time series data. The welding torch fault labels include fault type and corresponding fault severity.

[0038] Specifically, historical welding arc parameters and historical physical state data of the sample welding torch are obtained as training samples. The corresponding fault type and severity are determined by manual criteria. Fault types include various known possible fault types such as nozzle blockage, transformer overheating, and electrode aging, as well as no fault. Fault severity includes mild, moderate, and severe. The training samples are input into a pre-constructed CNN-LSTM hybrid neural network to obtain predicted fault labels. A first loss value is determined based on the predicted fault labels and the welding torch fault labels. The parameters of the CNN-LSTM hybrid neural network are updated based on the first loss value to complete one round of iterative training. When the number of iterations reaches a preset threshold or the loss value is lower than the preset threshold, training stops, and the welding torch fault prediction model is obtained.

[0039] As a further optional implementation, the CNN-LSTM hybrid neural network includes current convolutional branch networks, voltage convolutional branch networks, temperature convolutional branch networks, vibration convolutional branch networks, pressure convolutional branch networks, LSTM branch networks, feature fusion layers, and an output layer. Training samples are input into the pre-constructed CNN-LSTM hybrid neural network to obtain predicted fault labels, specifically including: S2031. Input the historical welding current time series data into the current convolutional branch network to obtain the local current features; S2032. Input the historical welding voltage time series data into the voltage convolutional branch network to obtain the voltage local features; S2033. Input historical temperature time series data into a temperature convolutional branch network to obtain local temperature features; S2034. Input historical vibration time series data into the vibration convolutional branch network to obtain local vibration features; S2035. Input the historical electrode pressure time series data into the pressure convolutional branch network to obtain local pressure features; S2036. Input historical welding arc parameters and historical physical state data into the LSTM branch network to obtain multimodal long-term dependency features. S2037. Based on the attention mechanism, the feature fusion layer performs feature fusion on local current features, local voltage features, local temperature features, local vibration features, local pressure features, and long-term multimodal dependence features to obtain global spatiotemporal fusion features. S2038. The global spatiotemporal fusion features are mapped to predicted fault labels through the output layer.

[0040] Specifically, the CNN-LSTM hybrid neural network of this embodiment includes convolutional branch networks corresponding to each sensor mode, used to extract local features of a single sensor mode (such as current spikes and vibration spectrum peaks), and also includes an LSTM branch network, used to capture long-term dependencies between multiple sensor modes (such as the effect of temperature gradient on voltage). The feature fusion layer fuses the local features of each sensor mode and the long-term dependency features of multiple sensor modes to obtain global spatiotemporal fusion features. The output layer can then map these global spatiotemporal fusion features into predicted fault labels.

[0041] As an optional implementation method, a corresponding maintenance plan is generated based on the fault prediction results, which specifically includes: S1021. Obtain the fault maintenance mapping table issued by the cloud service layer; S1022. Determine the target fault type and fault severity corresponding to the fault prediction results; S1023. Based on the target fault type and fault severity, query the fault maintenance mapping table to obtain the corresponding target maintenance strategy; S1024. Generate a maintenance plan based on the target maintenance strategy and push the maintenance plan to the management terminal.

[0042] Specifically, the cloud service layer generates and updates a fault maintenance mapping table based on the stored historical operating data, fault records, and maintenance records of the welding torch, and then periodically distributes it to the edge computing layer. This fault maintenance mapping table stores maintenance strategies corresponding to different fault types and different fault severity levels. The edge computing layer matches and searches the fault maintenance mapping table according to the target fault type and fault severity corresponding to the fault prediction results to obtain the corresponding target maintenance strategy, thereby generating a maintenance plan and pushing it to the management terminal.

[0043] After viewing the fault warning information and maintenance plan on the management terminal, managers can promptly perform fault maintenance on the target welding torch to prevent it from malfunctioning and affecting production.

[0044] As a further optional implementation, a target sample is generated based on welding arc parameters, physical state data, and actual maintenance data, and the target sample is uploaded to the cloud service layer, specifically including: S1031. Construct a target sample based on welding arc parameters and physical state data; S1032. Correct the fault prediction results based on actual maintenance data to obtain the target fault label; S1033. Label the target sample according to the target fault label, and upload the labeled target sample to the cloud service layer.

[0045] Specifically, after the management personnel perform fault maintenance on the target welding torch, the actual fault maintenance operation is recorded to obtain actual maintenance data. By comparing the actual maintenance data with the fault prediction results, it can be determined whether the fault prediction results are accurate. Correcting the fault prediction results will yield the actual target fault label of the target welding torch.

[0046] As a further optional implementation, the welding torch fault prediction model is incrementally trained based on the target samples, which specifically includes: S1034. Input the target sample into the welding torch fault prediction model to obtain the sample prediction result; S1035. Determine the second loss value based on the sample prediction results and the target fault label; S1036. Update the parameters of the output layer of the welding torch fault prediction model according to the second loss value to obtain the optimized welding torch fault prediction model.

[0047] Specifically, when the cloud service layer incrementally trains the welding torch fault prediction model, it keeps the parameters of each convolutional branch network, LSTM branch network, and feature fusion layer unchanged, and updates the parameters of the output layer (i.e., the classification head) according to the loss value, thereby fine-tuning and optimizing the welding torch fault prediction model and improving the accuracy of fault classification.

[0048] The method steps of the embodiments of the present invention have been described above. It can be understood that the embodiments of the present invention, by extending computing power to the production site through an edge computing layer, can effectively solve problems such as high data transmission latency, high cloud dependence, and untimely local response in traditional industrial scenarios. Deploying a cloud-trained welding torch fault prediction model at the edge computing layer enables welding torch fault prediction based on real-time collected welding arc parameters and physical state data, and generates corresponding maintenance plans. This achieves real-time monitoring, early warning, and precise maintenance of welding torch faults, improving the accuracy and timeliness of welding torch fault prediction. Simultaneously, it transforms passive maintenance into proactive prevention, significantly reducing production line downtime risks and maintenance costs.

[0049] Reference Figure 2 This invention provides a welding torch fault prediction device based on edge computing, comprising: The data acquisition module is used to acquire the welding arc parameters and physical state data of the target welding torch, and transmit the welding arc parameters and physical state data to the edge computing layer; The fault prediction module is used to input welding arc parameters and physical state data into a pre-trained welding torch fault prediction model issued by the cloud service layer through the edge computing layer, obtain fault prediction results, and generate corresponding maintenance plans based on the fault prediction results. The model optimization module is used to obtain the actual maintenance data of the target welding torch through the edge computing layer, generate target samples based on welding arc parameters, physical state data and actual maintenance data, and upload the target samples to the cloud service layer, so that the cloud server can incrementally train the welding torch fault prediction model based on the target samples.

[0050] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0051] Reference Figure 3 This invention provides an electronic device, comprising: At least one processor; At least one memory for storing at least one program; When the above-mentioned at least one program is executed by the above-mentioned at least one processor, the above-mentioned at least one processor implements the above-mentioned edge computing-based welding torch fault prediction method.

[0052] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0053] This invention also provides a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the aforementioned edge computing-based welding torch fault prediction method.

[0054] This invention provides a computer-readable storage medium that can execute a welding torch fault prediction method based on edge computing provided in the method embodiments of this invention. It can execute any combination of the implementation steps of the method embodiments and has the corresponding functions and beneficial effects of the method.

[0055] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described edge computing-based welding torch fault prediction method.

[0056] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0057] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0058] The embodiments described in this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0059] The terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0060] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the aforementioned blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0061] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the aforementioned functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0062] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0063] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0064] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the aforementioned program can be printed, because the aforementioned program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0065] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0066] In the foregoing description of this specification, references to terms such as "one embodiment," "another embodiment," or "some embodiments" indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0067] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0068] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A welding torch fault prediction method based on edge computing, characterized in that, Includes the following steps: The welding arc parameters and physical state data of the target welding torch are acquired, and the welding arc parameters and physical state data are transmitted to the edge computing layer. The edge computing layer inputs the welding arc parameters and the physical state data into the pre-trained welding torch fault prediction model issued by the cloud service layer to obtain the fault prediction result, and generates the corresponding maintenance plan based on the fault prediction result. The edge computing layer obtains the actual maintenance data of the target welding torch, generates a target sample based on the welding arc parameters, the physical state data, and the actual maintenance data, and uploads the target sample to the cloud service layer, so that the cloud server can incrementally train the welding torch fault prediction model based on the target sample.

2. The welding torch fault prediction method based on edge computing according to claim 1, characterized in that, The process of acquiring the welding arc parameters and physical state data of the target welding torch, and transmitting the welding arc parameters and physical state data to the edge computing layer, specifically includes: The welding current timing data and welding voltage timing data of the target welding torch are obtained through current sensors and voltage sensors; Temperature timing data of the nozzle and transformer of the target welding torch are obtained through a temperature sensor; Vibration timing data of the robotic arm of the target welding torch are obtained by using an accelerometer. The electrode pressure timing data of the target welding torch is obtained by a piezoelectric force sensor; The welding current timing data and the welding voltage timing data are used as the welding arc parameters, and the temperature timing data, the vibration timing data, and the electrode pressure timing data are used as the physical state data. The welding arc parameters and the physical state data are transmitted to the edge computing layer through the edge gateway.

3. The welding torch fault prediction method based on edge computing according to claim 1, characterized in that, The cloud service layer pre-trains the welding torch fault prediction model through the following steps: Obtain historical welding arc parameters and historical physical state data of the sample welding torch; Training samples are constructed based on the historical welding arc parameters and the historical physical state data, and welding gun fault labels of the training samples are determined by manual annotation. The training samples are input into a pre-constructed CNN-LSTM hybrid neural network to obtain predicted fault labels; A first loss value is determined based on the predicted fault label and the welding torch fault label; The parameters of the CNN-LSTM hybrid neural network are updated based on the first loss value to obtain the welding torch fault prediction model; The historical welding arc parameters include historical welding current time series data and historical welding voltage time series data. The historical physical state data includes historical temperature time series data, historical vibration time series data, and historical electrode pressure time series data. The welding torch fault label includes the fault type and the corresponding fault severity.

4. The welding torch fault prediction method based on edge computing according to claim 3, characterized in that, The CNN-LSTM hybrid neural network includes current convolutional branch networks, voltage convolutional branch networks, temperature convolutional branch networks, vibration convolutional branch networks, pressure convolutional branch networks, LSTM branch networks, feature fusion layers, and output layers. The step of inputting the training samples into the pre-constructed CNN-LSTM hybrid neural network to obtain predicted fault labels specifically includes: The historical welding current time series data is input into the current convolutional branch network to obtain local current features; The historical welding voltage time series data is input into the voltage convolutional branch network to obtain local voltage features; The historical temperature time series data is input into the temperature convolutional branch network to obtain local temperature features; The historical vibration time series data is input into the vibration convolutional branch network to obtain local vibration features; The historical electrode pressure time series data is input into the pressure convolutional branch network to obtain local pressure features; The historical welding arc parameters and the historical physical state data are input into the LSTM branch network to obtain multimodal long-term dependency features. The feature fusion layer performs feature fusion on the local current features, local voltage features, local temperature features, local vibration features, local pressure features, and long-term multimodal dependence features based on an attention mechanism to obtain global spatiotemporal fusion features; The output layer maps the global spatiotemporal fusion features to the predicted fault labels.

5. The welding torch fault prediction method based on edge computing according to claim 1, characterized in that, The step of generating a corresponding maintenance plan based on the fault prediction results specifically includes: Obtain the fault maintenance mapping table issued by the cloud service layer; Determine the target fault type and fault severity corresponding to the fault prediction results; Based on the target fault type and the fault severity, the corresponding target maintenance strategy is obtained by querying the fault maintenance mapping table; The maintenance plan is generated based on the target maintenance strategy, and the maintenance plan is pushed to the management terminal.

6. The welding torch fault prediction method based on edge computing according to claim 1, characterized in that, The step of generating a target sample based on the welding arc parameters, the physical state data, and the actual maintenance data, and uploading the target sample to the cloud service layer, specifically includes: The target sample is constructed based on the welding arc parameters and the physical state data; The fault prediction results are corrected based on the actual maintenance data to obtain the target fault label; The target sample is labeled according to the target fault label, and the labeled target sample is uploaded to the cloud service layer.

7. The welding torch fault prediction method based on edge computing according to claim 6, characterized in that, The incremental training of the welding torch fault prediction model based on the target sample specifically includes: The target sample is input into the welding torch fault prediction model to obtain the sample prediction result; A second loss value is determined based on the sample prediction results and the target fault label; The parameters of the output layer of the welding torch fault prediction model are updated based on the second loss value to obtain the optimized welding torch fault prediction model.

8. A welding torch fault prediction device based on edge computing, characterized in that, include: The data acquisition module is used to acquire the welding arc parameters and physical state data of the target welding torch, and transmit the welding arc parameters and physical state data to the edge computing layer; The fault prediction module is used to input the welding arc parameters and the physical state data into the pre-trained welding torch fault prediction model issued by the cloud service layer through the edge computing layer, obtain the fault prediction result, and generate the corresponding maintenance plan based on the fault prediction result. The model optimization module is used to obtain the actual maintenance data of the target welding torch through the edge computing layer, generate target samples based on the welding arc parameters, the physical state data and the actual maintenance data, and upload the target samples to the cloud service layer, so that the cloud server can incrementally train the welding torch fault prediction model based on the target samples.

9. An electronic device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a welding torch fault prediction method based on edge computing as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements a welding torch fault prediction method based on edge computing as described in any one of claims 1 to 7.

Images