Welding quality control system and method, electronic equipment and storage medium

Abstract

The invention provides a welding quality control system and method, electronic equipment and a storage medium, and relates to the technical field of welding quality. Multi-source monitoring data in the welding process is collected in real time through a sensing module, and a decision module carries out space-time alignment, feature extraction and pre-training model analysis on the data to generate an adjusting instruction; the closed-loop control module adjusts process parameters of the welding executing mechanism in real time according to the instruction, and therefore the problems that in the prior art, welding parameters are fixed, variables in the welding process cannot be adapted in real time, and welding defects occur frequently and efficiency is low due to the fact that manual adjustment is relied on can be solved. The technical effects that the welding defect rate is reduced, the welding efficiency is improved, and dependence on manual experience is reduced are achieved.

Description

Technical Field

[0001] This disclosure relates to the field of welding quality technology, and in particular to a welding quality control system and method, electronic equipment and storage medium. Background Technology

[0002] Welding technology, as a core process in advanced manufacturing, is widely used in scenarios with stringent requirements for structural integrity, such as aerospace and nuclear power equipment. With development, welding quality control has evolved from traditional manual experience to automated systems. Existing technologies have built a basic process system through fixed parameter control and single sensor monitoring.

[0003] However, existing welding quality control systems directly employ fixed parameter adjustment modes without establishing a spatiotemporal alignment mechanism for multi-source heterogeneous data. This may result in variables such as assembly errors (e.g., misalignment ±2mm) and material thermal conductivity fluctuations (±15%) not being compensated for in real time. The contradiction between the high-end manufacturing sector's requirements for welding quality stability (e.g., aerospace component defect rate <0.1%) and the response speed of existing technologies is particularly prominent, necessitating the development of a more precise quality control method. Summary of the Invention

[0004] This disclosure provides a welding quality control system and method, electronic equipment, and storage medium. Its main objective is to at least partially solve one of the technical problems in the related art.

[0005] According to a first aspect of this disclosure, a welding quality control system is provided, comprising: a sensing module, a decision module, and a closed-loop control module; The sensing module is used to collect multi-source monitoring data of the welding robot in real time during the welding process; The decision module is communicatively connected to the sensing module and is used to perform fusion analysis on the multi-source monitoring data and generate welding parameter adjustment instructions. The closed-loop control module is communicatively connected to the decision module and is used to adjust the process parameters of the welding robot's welding actuator in real time according to the welding parameter adjustment command. The sensing module transmits the collected monitoring data to the decision module in real time. After performing spatiotemporal alignment and feature extraction on the received data, the decision module analyzes the welding status through a pre-trained welding quality recognition model and generates parameter adjustment instructions. The closed-loop control module receives the welding parameter adjustment instructions and drives the welding actuator to complete the real-time adjustment of the welding parameters.

[0006] Optionally, the sensing module includes: An image acquisition unit is configured to acquire molten pool morphology features and weld seam tracking data. The image acquisition unit includes a combination of an industrial camera and a laser contour sensor. An acoustic acquisition unit is configured to acquire the acoustic characteristics of an electric arc and the acoustic signature data of a molten droplet transition. The acoustic acquisition unit includes a microphone array. A temperature acquisition unit is configured to acquire temperature field distribution data of the welding area, and the temperature acquisition unit includes an infrared thermal imager.

[0007] Optionally, the decision module includes: The data fusion unit is configured to perform spatiotemporal registration and feature extraction on multi-source monitoring data; The parameter optimization unit is configured to analyze the fused features based on the welding quality recognition model and output defect warning information and process parameter correction values. The welding quality recognition model is a CNN-LSTM hybrid network architecture, in which the CNN network is used to extract image spatial features and the LSTM network is used to process time series data. The data fusion unit transmits the processed feature vector to the parameter optimization unit, which generates adjustment instructions based on the parameter-quality mapping relationship established by historical welding quality data.

[0008] Optionally, the closed-loop control module includes: An attitude adjustment unit is configured to adjust the spatial position and welding attitude of the welding torch, and the attitude adjustment unit controls the position and posture of the welding torch through the welding robot. The wire feeding control unit is configured to dynamically adjust the wire feeding speed, and the wire feeding control unit adjusts the wire feeding speed through a servo motor; The motion control unit is configured to control the walking trajectory of the welding robot; Each control unit receives adjustment commands from the decision module and uses a servo drive system to synchronously adjust welding parameters.

[0009] Optionally, it also includes: a digital twin module, which is bidirectionally connected to the decision module and configured as follows: Construct a virtual simulation model of the welding process; Predict welding quality under different parameter combinations based on the current welding status; The optimized parameter suggestions are fed back to the decision-making module.

[0010] Optionally, the digital twin module uses real-time data to drive a virtual model, simulating the molten pool metallurgical reaction process and thermal stress distribution, predicting welding defects and optimizing process parameters.

[0011] According to a second aspect of this disclosure, a welding quality control method is provided, comprising: Real-time acquisition of multi-source monitoring data during the welding process; Receive the multi-source monitoring data, perform fusion analysis on the data, and generate welding parameter adjustment instructions based on the analysis results; The welding parameter adjustment command is received, and the welding actuator is controlled to adjust the process parameters in real time to form a closed-loop control process.

[0012] According to a third aspect of this disclosure, an electronic device is provided, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the second aspect above.

[0013] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described in the second aspect above.

[0014] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in the second aspect above.

[0015] The welding quality control system and method, electronic equipment, and storage medium disclosed herein acquire multi-source monitoring data of the welding process in real time through a sensing module. The decision module performs spatiotemporal alignment, feature extraction, and pre-trained model analysis on the data to generate adjustment instructions. The closed-loop control module adjusts the process parameters of the welding actuator in real time according to the instructions. Therefore, it can solve the problems of frequent welding defects and low efficiency caused by fixed welding parameters, inability to adapt to variables in the welding process in real time, and reliance on manual adjustment in the prior art. It achieves the technical effects of reducing welding defect rate, improving welding efficiency, and reducing reliance on human experience.

[0016] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0017] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein: Figure 1 This is a schematic diagram of the structure of a welding quality control system provided in an embodiment of the present disclosure; Figure 2 This is a schematic diagram of another welding quality control system provided in an embodiment of the present disclosure; Figure 3 A schematic flowchart of a welding quality control method provided in this embodiment of the present disclosure; Figure 4 A schematic block diagram of an example electronic device provided for embodiments of this disclosure. Detailed Implementation

[0018] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0019] The welding quality control system and method, electronic equipment, and storage medium of this disclosure are described below with reference to the accompanying drawings.

[0020] Figure 1 This is a schematic diagram of the structure of a welding quality control system provided in an embodiment of the present disclosure.

[0021] like Figure 1 As shown, the welding quality control system includes: a sensing module 11, a decision module 12, and a closed-loop control module 13; The sensing module 11 is used to collect multi-source monitoring data of the welding robot in real time during the welding process.

[0022] In the embodiments of this disclosure, in the intelligent adaptive welding quality control system, the sensing module 11 serves as the core unit for sensing the welding process status. Its core function is to collect multi-source monitoring data generated by the welding robot during welding operations in real time. This multi-source monitoring data needs to comprehensively reflect the key status information in the welding process, providing basic data support for the subsequent dynamic control of welding quality. Here, "multi-source" aims to cover different dimensions of parameters affecting quality during the welding process, ensuring that the collected data can characterize the real-time welding status from multiple perspectives. "Real-time collection" ensures the synchronization of data with the welding operation process, avoiding misjudgment or delay in control of the welding status due to data lag. As one implementation, the sensing module 11 may include a visual sensing component for capturing visual information related to weld morphology and molten pool status, an acoustic sensing component for collecting acoustic signals during the welding process, and a temperature sensing component for monitoring the temperature distribution in the welding area. It may collect specific data such as weld groove position deviation, molten pool size, arc sound, and welding area temperature, but is not limited to the above component types and data types.

[0023] The sensing module 11 can overcome the limitations of single-dimensional or non-real-time data acquisition in traditional welding by collecting multi-source monitoring data in real time. It can comprehensively and timely obtain key status information of the welding process, providing a reliable data foundation for the subsequent decision-making module to carry out data fusion analysis and generate precise parameter adjustment instructions. This avoids the problem of insufficient welding quality control caused by incomplete or delayed data, and lays a data-level support for improving the stability of welding quality.

[0024] The decision module 12 is communicatively connected to the sensing module 11 and is used to perform fusion analysis on the multi-source monitoring data and generate welding parameter adjustment instructions.

[0025] In the embodiments of this disclosure, in the intelligent adaptive welding quality control system, the decision module 12, as the core analysis unit, establishes a communication connection with the sensing module 11 to achieve effective data transmission and interaction. Its core function is to receive multi-source monitoring data collected by the sensing module 11, perform fusion analysis on the multi-source monitoring data, and finally generate appropriate welding parameter adjustment instructions based on the analysis results. The "fusion analysis" aims to integrate the information value of monitoring data from different dimensions, eliminate the limitations of single data in judging the welding state, and ensure a comprehensive and accurate assessment of the current welding process state. The "generation of welding parameter adjustment instructions" is based on the above accurate assessment, providing a clear basis for subsequent adjustment of welding actuator parameters to adapt to dynamic changes in the welding process. As one implementation method, this module can first perform spatiotemporal alignment of multi-source data, then construct a defect warning model through a deep learning model (such as a CNN-LSTM hybrid network) to judge the welding state, and simultaneously generate a specific correction scheme by combining a "parameter-quality" mapping relationship library constructed from historical high-quality data. However, the above specific implementation method is not a limitation on its function.

[0026] The decision module 12, through the fusion analysis of multi-source monitoring data, can avoid the one-sidedness and subjectivity of relying on single data or human experience to judge the welding status in traditional welding, and ensure the accuracy of welding status assessment. At the same time, the welding parameter adjustment instructions generated based on the analysis results provide a scientific basis for subsequent real-time adjustment of welding process parameters, effectively supporting the dynamic optimization of the welding process, and helping to reduce welding defects and improve welding efficiency.

[0027] The closed-loop control module 13 is communicatively connected to the decision module 12 and is used to adjust the process parameters of the welding actuator of the welding robot in real time according to the welding parameter adjustment command. The sensing module 11 transmits the collected monitoring data to the decision module 12 in real time. After performing spatiotemporal alignment and feature extraction on the received data, the decision module 12 analyzes the welding status through a pre-trained welding quality recognition model and generates parameter adjustment commands. The closed-loop control module 13 receives the welding parameter adjustment commands and drives the welding actuator to complete the real-time adjustment of the welding parameters.

[0028] In the embodiments of this disclosure, in the intelligent adaptive welding quality control system, the closed-loop control module 13, as the core execution unit for realizing dynamic control of the welding process, maintains a stable communication connection with the decision module 12 to ensure the accurate and timely transmission of welding parameter adjustment commands. Its core function is to adjust the process parameters corresponding to the welding actuator of the welding robot in real time based on the welding parameter adjustment commands output by the decision module 12, thereby constructing a complete control link of "data-driven - command generation - parameter correction". The operation of this module relies on the collaboration of various modules within the system: First, the sensing module 11 transmits the collected multi-source monitoring data of the welding process to the decision module 12 in real time. The decision module 12 first performs spatiotemporal alignment on the received data to ensure data correlation, then performs feature extraction to filter key information, and then analyzes the current welding state through a pre-trained welding quality recognition model and generates appropriate parameter adjustment commands. Finally, the closed-loop control module 13 receives the command and drives the welding actuator to complete the adjustment of process parameters, ensuring the synchronization of the welding process and parameter control. As one implementation method, the welding actuator can be equipped with a high-precision six-axis welding robot, with a servo motor integrated into the end welding torch. The closed-loop control module 13 can achieve data reception, command response and parameter correction within 10ms, forming a millisecond-level closed loop. However, the specific actuator type and response speed mentioned above are not limitations on the function of this module.

[0029] The closed-loop control module 13 solves the problem of untimely correction of deviations caused by fixed parameters and delayed manual adjustments in traditional welding by constructing a real-time closed-loop control logic of "sensing-decision-control". It ensures that parameter deviations are corrected as soon as they occur during the welding process, effectively reducing welding defects. At the same time, it eliminates the need for repeated manual intervention in parameter adjustments, which not only improves welding efficiency but also reduces the reliance on the experience of senior welders, thus ensuring welding quality stability and production flexibility.

[0030] The welding quality control system disclosed herein collects multi-source monitoring data of the welding process in real time through a sensing module. The decision module performs spatiotemporal alignment, feature extraction, and pre-trained model analysis on the data to generate adjustment instructions. The closed-loop control module adjusts the process parameters of the welding actuator in real time according to the instructions. Therefore, it can solve the problems of frequent welding defects and low efficiency caused by fixed welding parameters, inability to adapt to variables in the welding process in real time, and reliance on manual adjustment in the prior art. It achieves the technical effects of reducing welding defect rate, improving welding efficiency, and reducing reliance on human experience.

[0031] Furthermore, in one possible implementation of this embodiment, such as Figure 2 As shown, the sensing module 11 includes: Image acquisition unit 111 is configured to acquire molten pool morphology features and weld seam tracking data. The image acquisition unit 111 includes a combination of an industrial camera and a laser contour sensor. The acoustic acquisition unit 112 is configured to acquire the acoustic characteristics of the electric arc and the acoustic signature data of the molten droplet transition. The acoustic acquisition unit 112 includes a microphone array. Temperature acquisition unit 113 is configured to acquire temperature field distribution data of the welding area, and the temperature acquisition unit 113 includes an infrared thermal imager.

[0032] Specifically, such as Figure 2 As shown, it comprises three functional subunits: an image acquisition unit 111, an acoustic acquisition unit 112, and a temperature acquisition unit 113. The image acquisition unit 111 employs a hardware device combining an industrial camera and a laser contour sensor. During welding operations, this combination captures the morphological characteristics of the molten pool in real time (such as the width, height, and boundary clarity of the molten pool) using the industrial camera, while simultaneously acquiring precise tracking data of the weld seam (such as the positional deviation of the weld bevel, misalignment, and gap size) using the laser contour sensor, achieving synchronous acquisition of key visual information related to welding. The acoustic acquisition unit 112 is specifically configured as a microphone array. This microphone array needs to be deployed near the welding area, avoiding direct arc light and excessive dust accumulation, to collect real-time electrical signals generated during the welding process. The system includes arc acoustic characteristics (such as high-frequency acoustic signals during stable arc combustion and changes in acoustic frequency during arc fluctuations) and droplet transition acoustic data (such as characteristic acoustic patterns when droplets detach from the end of the welding wire and transition to the molten pool). The temperature acquisition unit 113 uses an infrared thermal imager as its core hardware. This infrared thermal imager can perform non-contact temperature monitoring of the welding area and generate and output temperature field distribution data of the welding area in real time (such as the temperature of the center of the molten pool, the temperature gradient of the heat-affected zone, and the temperature values ​​of local overheating points). The three sub-units maintain the synchronization of data acquisition during operation and transmit the monitoring data they acquire to the decision module 12, which is connected to the sensing module 11.

[0033] By clearly defining the hardware configuration and acquisition targets of each acquisition subunit, the relevance and accuracy of multi-source monitoring data are ensured. For example, laser contour sensors improve the accuracy of weld seam tracking data, and microphone arrays avoid interference from arc smoke and dust on acoustic data. Furthermore, through the coordinated acquisition of visual, acoustic, and temperature dimensions, key status information affecting quality during the welding process is comprehensively covered. This provides high-quality data support for the subsequent decision-making module 12 to perform precise data fusion analysis and defect early warning, effectively reducing the risk of misjudgment of welding quality due to incomplete or inaccurate data.

[0034] Furthermore, in one possible implementation of this embodiment, such as Figure 2 As shown, the decision module 12 includes: The data fusion unit 121 is configured to perform spatiotemporal registration and feature extraction on multi-source monitoring data; The parameter optimization unit 122 is configured to analyze the fused features based on the welding quality recognition model and output defect warning information and process parameter correction values. The welding quality recognition model is a CNN-LSTM hybrid network architecture, in which the CNN network is used to extract image spatial features and the LSTM network is used to process time series data. The data fusion unit 121 transmits the processed feature vector to the parameter optimization unit 122, which generates adjustment instructions based on the parameter-quality mapping relationship established by historical welding quality data.

[0035] Specifically, such as Figure 2As shown, it mainly consists of a data fusion unit 121 and a parameter optimization unit 122, and the two units maintain data interaction. Among them, the core configuration function of the data fusion unit 121 is to perform spatiotemporal registration and feature extraction operations on the multi-source monitoring data transmitted by the sensing module 11 (including the molten pool morphology / weld tracking data of the image acquisition unit 111, the arc sound / molten droplet transition acoustic texture data of the acoustic acquisition unit 112, and the temperature field distribution data of the temperature acquisition unit 113). The spatiotemporal registration step requires the data from different sources to be associated and aligned according to the timestamp and spatial location to ensure that the visual, acoustic, and temperature data of the same welding area at the same time form a corresponding relationship. The feature extraction step then selects key information representing the welding state from the registered data, such as extracting the molten pool width and residual height features from the image data, extracting the arc sound frequency fluctuation features from the acoustic data, and extracting the temperature gradient features of the heat-affected zone from the temperature data, and finally forming a standardized feature vector. The configuration function of parameter optimization unit 122 is to analyze the above feature vectors based on a preset welding quality recognition model. The welding quality recognition model specifically adopts a CNN-LSTM hybrid network architecture. The CNN network is responsible for deep extraction of spatial features of image data (such as the spatial distribution of molten pool morphology and the geometric features of weld groove), and the LSTM network is responsible for dynamic processing of data with temporal correlation such as acoustics and temperature (such as the changing trend of arc sound with welding process and the temporal fluctuation of temperature field). After the parameter optimization unit 122 analyzes the data through the model, it outputs defect warning information (such as predicting the types of defects such as porosity and cracks that may occur at a specific time node) on the one hand, and outputs process parameter correction values ​​on the other hand. The generation of correction values ​​needs to be based on the parameter-quality mapping relationship established by historical welding quality data (i.e., the combination of current, voltage and welding speed corresponding to defect-free welds). Finally, the parameter optimization unit 122 integrates the process parameter correction values ​​into welding parameter adjustment commands and transmits them to the closed-loop control module 13.

[0036] The spatiotemporal registration operation of the data fusion unit 121 avoids the analysis bias caused by the time and space misalignment of multi-source data, ensuring the accuracy of feature extraction. At the same time, the welding quality recognition model with CNN-LSTM hybrid network architecture not only takes into account the accurate extraction of image spatial features, but also realizes the dynamic analysis of time series data, which greatly improves the timeliness of defect warning and the adaptability of process parameter correction values. The parameter-quality mapping relationship based on historical data further ensures the scientific nature of adjustment instructions and effectively reduces the probability of welding defects.

[0037] Furthermore, in one possible implementation of this embodiment, such as Figure 2 As shown, the closed-loop control module 13 includes: The attitude adjustment unit 131 is configured to adjust the spatial position and welding attitude of the welding torch. The attitude adjustment unit 131 controls the position and posture of the welding torch through the welding robot. The wire feeding control unit 132 is configured to dynamically adjust the wire feeding speed, and the wire feeding control unit 132 adjusts the wire feeding speed through a servo motor; The motion control unit 133 is configured to control the walking trajectory of the welding robot; Each control unit receives adjustment commands from the decision module 12 and uses a servo drive system to synchronously adjust welding parameters.

[0038] Specifically, such as Figure 2 As shown, it consists of an attitude adjustment unit 131, a wire feeding control unit 132 and a motion control unit 133, and each unit establishes an instruction receiving link with the decision module 12 and relies on the servo drive system to realize parameter adjustment. The core function of the attitude adjustment unit 131 is to adjust the spatial position and welding posture of the welding torch. This unit is directly linked to the actuator of the welding robot. By controlling the joint movement of the welding robot, it corrects the spatial position of the welding torch (such as the vertical height and lateral offset of the welding torch to the weld) and the welding posture (such as the angle and tilt direction between the welding torch and the workpiece surface) in real time to ensure that the welding torch is always aligned with the center of the weld bevel. The wire feeding control unit 132 is configured to dynamically adjust the wire feeding speed. It integrates a servo motor connected to the wire feeding mechanism. According to the adjustment command output by the decision module 12, it can change the speed of the servo motor to precisely control the rate at which the welding wire is fed to the molten pool, avoiding excessive welding due to excessive wire feeding or insufficient molten pool filling due to excessively slow wire feeding. The motion control unit 133 is responsible for controlling the walking trajectory of the welding robot. This unit drives the welding robot to move smoothly along the preset weld trajectory by parsing the path parameters in the adjustment command. At the same time, it fine-tunes the walking speed according to the real-time welding status (such as appropriately reducing the speed when welding thick plates to ensure penetration). During operation, each control unit synchronously receives adjustment commands from the decision module 12. The servo drive system coordinates the timing of actions of each unit to ensure that the parameter adjustments of welding torch posture, wire feeding speed and travel trajectory are executed synchronously, avoiding welding quality fluctuations caused by the lag in the adjustment of a single parameter.

[0039] The posture adjustment unit 131 precisely controls the welding torch position, effectively solving the problem of weld positioning deviation caused by workpiece assembly errors; the wire feeding control unit 132 uses a servo motor to dynamically adjust the speed, ensuring the stability of the molten pool deposition amount; and each unit is synchronously adjusted through the servo drive system, avoiding welding defects (such as incomplete fusion and porosity) caused by asynchronous parameter adjustments, significantly improving the stability of the welding process and the consistency of welding quality.

[0040] Furthermore, in one possible implementation of this embodiment, such as Figure 2As shown, it also includes: a digital twin module 14, which is bidirectionally connected to the decision module 12 and configured as follows: Construct a virtual simulation model of the welding process; Predict welding quality under different parameter combinations based on the current welding status; The optimized parameter suggestions are fed back to decision module 12.

[0041] Specifically, the welding quality control system also includes a digital twin module 14, which establishes a bidirectional communication connection with the decision module 12, enabling bidirectional data interaction and collaborative work. The core configuration functions of the digital twin module 14 are implemented as follows: When constructing a virtual simulation model of the welding process, it uses the current welding basic data transmitted by the decision module 12 (including workpiece material properties, welding robot structural parameters, initial welding process parameters, etc.) and combines it with real-time multi-source monitoring data collected by the sensing module 11 (such as molten pool morphology and temperature field distribution). Through 3D modeling and physical simulation algorithms, it recreates the equipment, workpiece, and welding dynamic process of the actual welding scene in a 1:1 ratio within the virtual environment. When predicting the welding quality under different parameter combinations based on the current welding state, this module receives the current welding state data sent by the decision module 12 (such as current current, voltage, welding speed, and corresponding weld parameters). The system receives status scores and inputs multiple different combinations of process parameters (such as different combinations of current and welding speed) into the virtual simulation model. It simulates the molten pool metallurgical reaction, temperature field changes, and weld formation process under each set of parameters. Through the built-in quality assessment algorithm (such as comparing the size deviation and defect rate of the simulated weld with the standard weld), it predicts the welding quality level corresponding to each set of parameters. When the optimized parameter suggestions are fed back to the decision module 12, the digital twin module 14 will select the optimal parameter combination that meets the welding quality standards (such as the defect rate being lower than the preset threshold and the weld strength meeting the standard) from the prediction results and transmit it to the decision module 12 in the form of data signals to provide a reference for generating the final welding parameter adjustment instructions.

[0042] By constructing a virtual simulation model through the digital twin module 14, the effectiveness of parameters can be verified without conducting a large number of actual trial welds, effectively reducing the cost and time of trial welds. Based on the current state, the welding quality of different parameters can be predicted in advance to avoid actual welding defects caused by improper parameters. Furthermore, the feedback of optimization parameter suggestions to the decision module 12 further improves the accuracy of the adjustment instructions generated by the decision module, helping to achieve the dual guarantee of "pre-optimization + real-time control" in the welding process, and significantly improving the stability of welding quality and production efficiency.

[0043] Furthermore, in one possible implementation of this embodiment, the digital twin module 14 uses real-time data to drive a virtual model, simulate the molten pool metallurgical reaction process and thermal stress distribution, predict welding defects, and optimize process parameters.

[0044] Specifically, this module relies on a two-way communication link with the decision module 12 to acquire real-time multi-source welding process data (including molten pool width, weld reinforcement height, welding area temperature field data, arc acoustic characteristics, etc.) collected by the sensing module 11 and preliminarily processed by the decision module 12. This real-time data serves as the core driving source for the virtual model, constructing a simulation model dynamically synchronized with the actual welding scenario within the virtual environment. When simulating the molten pool metallurgical reaction process, the module combines real-time acquired molten pool temperature, current, and voltage parameters with preset metallurgical characteristic data such as the chemical composition and phase transformation temperature of the welding wire and base material. Through a thermodynamic and metallurgical kinetic coupling algorithm, it accurately simulates the microscopic processes of molten droplets detaching from the welding wire, merging into the molten pool, element diffusion within the molten pool, and subsequent solidification and crystallization. When simulating thermal stress distribution, the module calculates the temperature gradient at different locations in the welding area based on real-time temperature field data. Combined with physical property parameters such as the thermal expansion coefficient and elastic modulus of the workpiece material, it generates a thermal stress distribution cloud map inside the workpiece through a structural mechanics simulation algorithm, clearly presenting stress concentration areas. Meanwhile, the digital twin module 14 compares the simulated molten pool metallurgical state (such as crystallization rate) and thermal stress distribution results with the preset defect-free welding quality standard thresholds (such as the crystallization rate range and maximum allowable thermal stress value of qualified welds). If the simulation results are found to exceed the threshold range, the module predicts possible welding defects (such as cold cracks caused by excessively fast crystallization and hot cracks caused by stress concentration), and iteratively adjusts the process parameters in the virtual model (such as fine-tuning the welding current and changing the walking speed) to select the optimal parameter combination that can avoid defects. The optimized parameters are then fed back to the decision module 12.

[0045] By using real-time data to drive the virtual model, the dynamic consistency between the simulation process and the actual welding is ensured, which greatly improves the accuracy of simulating the metallurgical reaction and thermal stress distribution of the molten pool. By predicting welding defects in advance, parameters can be optimized to avoid actual defects before they occur, which significantly reduces the defect rate of actual welding. Furthermore, the parameter iterative optimization based on virtual simulation does not require a large number of actual trial welds, which effectively reduces the cost and time of trial welds. At the same time, it makes the feedback process parameters more in line with the actual welding requirements, which further improves the stability of welding quality.

[0046] Figure 3 This is a schematic flowchart of a welding quality control method provided in an embodiment of the present disclosure.

[0047] like Figure 3 As shown, the method includes the following steps: Step 201: Collect multi-source monitoring data of the welding process in real time.

[0048] Specifically, the core function of the data acquisition step, which forms the foundation of the entire control process, is to collect multi-source monitoring data in real time during the welding process and accurately transmit this data to the subsequent intelligent analysis steps. "Multi-source monitoring data" aims to cover different dimensions of parameters that may affect welding quality during the welding process, comprehensively reflecting the real-time welding status and avoiding the incomplete status representation problem caused by relying on single-dimensional data in traditional welding. "Real-time acquisition" requires that the data acquisition rhythm be synchronized with the welding operation progress to prevent data lag from causing a disconnect between subsequent intelligent analysis and the actual welding status, ensuring that the analysis results accurately match the current welding conditions. After data acquisition is achieved through a suitable acquisition unit, this step needs to establish a stable data transmission link to transmit the collected multi-source monitoring data to the intelligent analysis steps without delay or distortion, providing basic data support for subsequent data fusion, status judgment, and parameter adjustment command generation. In one implementation, step 201 can acquire visual data such as the shape of the molten pool and the position of the weld through the visual acquisition unit, acquire acoustic data such as the sound of the electric arc and the sound of the molten droplet transition through the acoustic acquisition unit, acquire temperature field distribution data of the welding area through the temperature acquisition unit, and transmit these data to the intelligent analysis step simultaneously. However, the specific acquisition method mentioned above is not a limitation of this step.

[0049] By collecting multi-source monitoring data of the welding process in real time, the problem of single data collection dimension and poor timeliness in traditional welding is effectively solved, ensuring that the data transmitted to the intelligent analysis step is comprehensive and synchronous. Reliable data transmission lays a solid foundation for the subsequent intelligent analysis step to accurately judge the welding status and generate scientific parameter adjustment instructions. It ensures the effectiveness of the entire welding quality control process from the data source and helps to reduce welding defects caused by incomplete or delayed data.

[0050] Step 202: Receive the multi-source monitoring data, perform fusion analysis on the data, and generate welding parameter adjustment instructions based on the analysis results.

[0051] Specifically, as the core analysis step connecting data acquisition and parameter adjustment, its primary function is to receive the multi-source monitoring data transmitted in step 201, ensuring that the data is received without loss or deviation, providing a complete data foundation for subsequent analysis. Subsequently, this step conducts a fusion analysis on the received multi-source monitoring data. This "fusion analysis" aims to break down the isolation of data from different dimensions. By performing correlation processing and feature extraction on the data, it integrates the information value of data from various dimensions, avoiding the bias and subjectivity of assessments based on single data or manual experience in traditional welding, ensuring a comprehensive and accurate identification of the current welding state (e.g., whether there are potential defects, whether the process parameters are suitable for the current working conditions). After completing the fusion analysis and clarifying the welding state, step 202 needs to generate welding parameter adjustment instructions based on the analysis results. The instruction generation must combine welding quality standards with current welding requirements, ensuring that the instructions can specifically address existing state deviations, providing a clear basis for subsequent real-time adjustments to welding parameters. As one implementation method, step 202 can first perform spatiotemporal alignment on the multi-source data, analyze the welding status and predict defects through a deep learning model, and then combine the parameter-quality mapping relationship constructed by historical high-quality welding data to generate adjustment instructions including parameters such as current, voltage, and welding speed. However, the specific analysis and instruction generation methods mentioned above are not limitations of this step.

[0052] By integrating and analyzing multi-source monitoring data, the problem of one-sided condition assessment and reliance on human experience in traditional welding is effectively solved, and the accuracy of welding condition judgment is improved. The welding parameter adjustment instructions generated based on the analysis results can accurately match the current welding conditions and provide scientific guidance for real-time adjustment of process parameters in subsequent closed-loop control steps. From the analysis level, the rationality of welding parameter optimization is guaranteed, which helps to reduce welding defects caused by improper parameters and improve the stability of welding quality.

[0053] Step 203: Receive the welding parameter adjustment command and control the welding actuator to adjust the process parameters in real time to form a closed-loop control process.

[0054] Specifically, as the core execution and closed-loop process for achieving dynamic optimization of welding parameters, its primary function is to accurately receive the welding parameter adjustment commands generated in step 202, ensuring that the commands are received without distortion or delay, thus providing a precise basis for subsequent parameter adjustments. Subsequently, based on the received adjustment commands, this step drives the adapted control unit to act on the welding actuator, achieving real-time adjustment of the welding process parameters. This "real-time adjustment" requires the adjustment action to be synchronized with the welding operation process, avoiding the accumulation of deviations caused by fixed parameters or delayed manual adjustments in traditional welding, and ensuring that the process parameters can adapt to variables in the welding process (such as workpiece assembly errors, environmental interference, etc.) in a timely manner. Simultaneously, this step constructs a closed-loop control process of "acquisition-analysis-adjustment-feedback" through feedback data of the welding status after parameter adjustment (linked with the data acquisition step in step 201), enabling continuous optimization of welding parameter adjustments based on actual welding results, rather than a single, one-time adjustment. In one implementation, step 203 can adjust the welding torch position by the attitude control unit, adjust the wire feeding speed by the wire feeding control unit, and correct the walking trajectory by the motion control unit, so as to realize the synchronous adjustment of process parameters such as current, voltage, and welding speed, and form a closed loop by feeding back the adjusted welding status data through the sensing module. However, the specific control method mentioned above is not a limitation of this step.

[0055] By adjusting process parameters in real time through real-time control of the welding actuator, the problem of fixed process parameters and lagging manual adjustment in traditional welding is effectively solved, ensuring that parameter deviations occurring during welding can be corrected in a timely manner. The formation of a closed-loop control process enables continuous optimization of welding parameter adjustments, further improving the adaptability of welding process parameters to actual working conditions. This reduces the occurrence of welding defects at the execution level, while also reducing reliance on the experience of senior welders and improving welding efficiency and quality stability.

[0056] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0057] Figure 4 A schematic block diagram of an example electronic device 300 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0058] like Figure 4As shown, the electronic device 300 includes a computing unit 301, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 302 or a computer program loaded from storage unit 308 into RAM (Random Access Memory) 303. The RAM 303 may also store various programs and data required for the operation of the electronic device 300. The computing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An I / O (Input / Output) interface 305 is also connected to the bus 304.

[0059] Multiple components in electronic device 300 are connected to I / O interface 305, including: input unit 306, such as keyboard, mouse, etc.; output unit 307, such as various types of displays, speakers, etc.; storage unit 308, such as disk, optical disk, etc.; and communication unit 309, such as network card, modem, wireless transceiver, etc. Communication unit 309 allows electronic device 300 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0060] The computing unit 301 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 301 performs the various methods and processes described above, such as welding quality control methods. For example, in some embodiments, the welding quality control method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 300 via ROM 302 and / or communication unit 309. When the computer program is loaded into RAM 303 and executed by the computing unit 301, one or more steps of the methods described above may be performed. Alternatively, in other embodiments, the computing unit 301 may be configured to perform the aforementioned welding quality control method by any other suitable means (e.g., by means of firmware).

[0061] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0062] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0063] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0064] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0065] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0066] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service system that addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0067] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0068] The various numerical designations such as "first," "second," etc., used in this disclosure are merely for ease of description and are not intended to limit the scope of the embodiments of this disclosure, nor do they indicate a sequential order.

[0069] At least one of the features described in this disclosure can also be described as one or more, and multiple features can be two, three, four or more, and this disclosure does not impose any limitations. In the embodiments of this disclosure, for a technical feature, the technical features in that technical feature are distinguished by "first", "second", "third", "A", "B", "C" and "D", etc., and there is no sequential order or size order among the technical features described by "first", "second", "third", "A", "B", "C" and "D".

[0070] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0071] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A weld quality control system characterized by, Comprise: Sensing module, decision module, closed loop control module; The sensing module is used for real-time acquisition of multi-source monitoring data of the welding robot in the welding process; The decision module is in communication connection with the sensing module, and is used for fusion analysis on the multi-source monitoring data and generation of a welding parameter adjustment instruction; The closed loop control module is in communication connection with the decision module, and is used for real-time adjustment of process parameters of a welding execution mechanism of the welding robot according to the welding parameter adjustment instruction; Wherein, the sensing module transmits the collected monitoring data to the decision module in real time, the decision module performs time-space alignment and feature extraction on the received data, analyzes the welding state through a pre-trained welding quality identification model, and generates a parameter adjustment instruction, and the closed loop control module receives the welding parameter adjustment instruction and drives the welding execution mechanism to complete real-time adjustment of the welding parameters.

2. The system of claim 1, wherein, The sensing module comprises: An image acquisition unit configured to acquire molten pool shape features and weld tracking data, the image acquisition unit comprising a combination of an industrial camera and a laser profile sensor; An acoustic acquisition unit configured to acquire arc acoustic features and droplet transfer acoustic print data, the acoustic acquisition unit comprising a microphone array; A temperature acquisition unit configured to acquire weld area temperature field distribution data, the temperature acquisition unit comprising an infrared thermal imager.

3. The system of claim 1, wherein, The decision module comprises: A data fusion unit configured to perform time-space registration and feature extraction on the multi-source monitoring data; A parameter optimization unit configured to analyze the fused features based on the welding quality identification model, output defect warning information and process parameter correction values, the welding quality identification model being a CNN-LSTM hybrid network architecture, wherein the CNN network is used to extract image spatial features, and the LSTM network is used to process time series data; Wherein, the data fusion unit transmits the processed feature vectors to the parameter optimization unit, and the parameter optimization unit generates adjustment instructions based on a parameter-quality mapping relationship established based on historical welding quality data.

4. The system of claim 1, wherein, The closed loop control module comprises: A posture adjustment unit configured to adjust the spatial position and welding posture of the welding torch, the posture adjustment unit adjusting the welding torch pose through the welding robot; A wire feeding control unit configured to dynamically adjust the wire feeding speed, the wire feeding control unit adjusting the wire feeding speed through a servo motor; A motion control unit configured to control the walking trajectory of the welding robot; Wherein, each control unit receives the adjustment instruction of the decision module, and realizes synchronous adjustment of the welding parameters through a servo drive system.

5. The system of claim 1, wherein, Further comprising a digital twin module in bidirectional communication connection with the decision module, configured to: Build a virtual simulation model of the welding process; Predict the welding quality under different parameter combinations based on the current welding state; Feedback the optimized parameters to the decision module.

6. The system of claim 5, wherein, The digital twin module drives the virtual model through real-time data, simulates the molten pool metallurgical reaction process and thermal stress distribution, predicts welding defects and optimizes process parameters.

7. The system of claim 1, wherein, Further comprising a human-computer interaction module in communication connection with the decision module, configured to: Receive input settings of welding quality standards; Display real-time welding curve and defect warning information; Support the input and execution of manual intervention instructions.

8. A method of weld quality control, characterized by, Comprise: Real-time acquisition of multi-source monitoring data of the welding process; Receive the multi-source monitoring data, perform fusion analysis on the data, and generate welding parameter adjustment instructions based on the analysis results; Receive the welding parameter adjustment instructions, control the welding execution mechanism to adjust the process parameters in real time, and form a closed-loop control process.

9. An electronic device, comprising: Comprise: At least one processor; And A memory in communication connection with the at least one processor; wherein The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method of claim 8.

10. A non-transitory computer-readable storage medium having stored thereon computer instructions, wherein, The computer instructions are used to make the computer execute the method according to claim 8. The computer instructions are used to make the computer execute the method according to claim 8.

Images