Welding current optimization method and device suitable for multiple material thicknesses

Abstract

The invention discloses a welding current optimization method and device suitable for multiple material thicknesses, and relates to the technical field of artificial intelligence, the method comprises the steps that welding process parameters are collected, and the welding process parameters at least comprise the welding pressure, the welding time and the sample piece thickness; establishing an artificial neural network model, and training the artificial neural network model by using the welding pressure, the welding time and the sample piece thickness; a Step Decay learning rate adjustment method is adopted to optimize the convergence of the artificial neural network model; using the trained artificial neural network model to predict the welding current in real time according to the material thickness and the welding parameters; and the welding process is dynamically adjusted through a real-time feedback mechanism.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to an optimization method and device for welding current suitable for multiple material thicknesses. Background Art

[0002] As an important connection technology, welding occupies a pivotal position in the manufacturing industry. Especially in high-demand industrial fields such as automobile manufacturing, aerospace, and shipbuilding, welding quality is directly related to product performance, safety, and service life. With the rapid development of intelligent manufacturing technology, welding technology has also ushered in new changes, gradually developing from traditional welding methods that rely on manual experience to intelligent and automated directions.

[0003] At present, the parameter settings in the welding process (such as current, pressure, time, etc.) mainly rely on the experience and skill level of the operator. Although this method can meet the welding needs to a certain extent, it has many limitations. Summary of the invention

[0004] The purpose of the present application is to provide a method and device for optimizing welding current suitable for multiple material thicknesses.

[0005] To achieve the above objectives, this application provides the following solutions: In a first aspect, the present application provides a method for optimizing welding current applicable to multiple material thicknesses, comprising: Collecting welding process parameters, wherein the welding process parameters at least include: welding pressure, welding time, and sample thickness; Establishing an artificial neural network model, and training the artificial neural network model using the welding pressure, the welding time, and the sample thickness; Adopting the Step Decay learning rate adjustment method to optimize the convergence of the artificial neural network model; Use the trained artificial neural network model to predict the welding current in real time based on material thickness and welding parameters; Dynamically adjust the welding process through real-time feedback mechanism.

[0006] Optionally, the step of collecting welding process parameters includes: An artificial neural network model with multiple hidden layers is constructed, and each hidden layer uses the ReLU activation function for nonlinear mapping.

[0007] Optionally, the step of optimizing the convergence of the artificial neural network model by using the Step Decay learning rate adjustment method comprises: The Step Decay learning rate adjustment method is used to gradually reduce the learning rate in multiple training cycles to optimize the convergence of the model and avoid falling into the local optimal solution.

[0008] Optionally, the step of using the trained artificial neural network model to predict the welding current in real time according to the material thickness and welding parameters includes: The real-time welding parameters are input into the trained artificial neural network model to obtain the predicted optimal welding current, and the welding process is adjusted through a real-time feedback mechanism.

[0009] Optionally, the step of adjusting the welding process by a real-time feedback mechanism comprises: The welding process is monitored through a real-time feedback mechanism, and the monitoring data is fed back to the artificial neural network model for dynamic optimization.

[0010] Optionally, the step of dynamically adjusting the welding process through a real-time feedback mechanism includes: According to the predicted optimal welding current, the current parameters during the welding process are dynamically adjusted to improve the welding quality.

[0011] In a second aspect, the present application provides an optimization device for welding currents of multiple material thicknesses, comprising: An acquisition module is used to collect welding process parameters, wherein the welding process parameters at least include: welding pressure, welding time and sample thickness; A training module, used for establishing an artificial neural network model, and training the artificial neural network model using the welding pressure, the welding time and the sample thickness; A Step Decay learning rate adjustment device is used to optimize the convergence of the artificial neural network model; A prediction module for predicting welding current in real time based on material thickness and welding parameters using a trained artificial neural network model; Optimization module for dynamically adjusting the welding process through real-time feedback mechanisms.

[0012] Optionally, the acquisition module is further used to: An artificial neural network model with multiple hidden layers is constructed, and each hidden layer uses the ReLU activation function for nonlinear mapping.

[0013] Optionally, the training module is further used to: The Step Decay learning rate adjustment device is used to gradually reduce the learning rate in multiple training cycles to optimize the model convergence and avoid falling into the local optimal solution.

[0014] Optionally, the prediction module is further used to: The real-time welding parameters are input into the trained artificial neural network model to obtain the predicted optimal welding current, and the welding process is adjusted through a real-time feedback mechanism.

[0015] Optionally, the optimization module is further used to: The welding process is monitored through a real-time feedback mechanism, and the monitoring data is fed back to the artificial neural network model for dynamic optimization.

[0016] Optionally, the optimization module is further used to: According to the predicted optimal welding current, the current parameters during the welding process are dynamically adjusted to improve the welding quality.

[0017] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of any one of the above-mentioned methods for optimizing welding current applicable to welding of multiple material thicknesses.

[0018] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for optimizing welding current applicable to multiple material thicknesses.

[0019] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for optimizing welding current applicable to multiple material thicknesses.

[0020] According to the specific embodiments provided in this application, this application discloses the following technical effects: The present application provides a method and device for optimizing welding current for multiple material thicknesses. By using an artificial neural network model to automatically predict and adjust welding process parameters, the limitations of traditional manual experience adjustment are avoided, thereby improving the accuracy and stability of welding current control. Secondly, the adoption of the Step Decay learning rate adjustment method optimizes the convergence of the model, enabling it to adapt to the welding requirements of different materials and thicknesses, and enhancing the generalization ability and adaptability of the model. In addition, the application of the real-time feedback mechanism enables the welding process to be dynamically adjusted according to the predicted results and the actual welding situation, ensuring the continuous stability of the welding quality. Finally, the implementation of this series of intelligent measures not only greatly reduces the adjustment time and labor cost of the welding process, but also reduces the rework rate due to the improvement of welding quality and stability, thereby significantly improving production efficiency. In summary, these steps realize the automation and closed-loop control of the welding process by introducing intelligent technology and real-time feedback mechanism, which not only improves the welding quality and efficiency, but also reduces the production cost, providing strong support for the intelligent transformation in the field of welding. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 A schematic flow chart of a method for optimizing welding current applicable to multiple material thicknesses provided in one embodiment of the present application; Figure 2 A network diagram of a method for optimizing welding current for multiple material thicknesses provided in one embodiment of the present application; Figure 3 A schematic diagram of functional modules of a device for optimizing welding current for various material thicknesses provided in one embodiment of the present application; Figure 4 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0024] like Figure 1 As shown, some embodiments of the present application provide a method for optimizing welding current applicable to multiple material thicknesses, and in the embodiments of the present application, the method includes the following steps 101 to 105. Among them: Step 101, collecting welding process parameters, wherein the welding process parameters at least include: welding pressure, welding time and sample thickness.

[0025] In the embodiments of the present application, welding pressure refers to the pressure applied to the weld joint during welding, which directly affects the quality and performance of the weld joint. Welding time refers to the duration of the welding operation, which is an important factor in controlling welding heat input and weld morphology. Sample thickness refers to the thickness of the workpiece to be welded. Workpieces of different thicknesses have different requirements for welding process parameters, so it is also one of the factors that must be considered.

[0026] When collecting these parameters, it is necessary to ensure the accuracy and completeness of the data so that subsequent model training can achieve good results.

[0027] Step 102: Establish an artificial neural network model, and train the artificial neural network model using the welding pressure, the welding time and the sample thickness.

[0028] In the embodiment of the present application, firstly, an artificial neural network model needs to be established according to the collected welding process parameters. The artificial neural network is a mathematical model that imitates the structure and function of biological neural networks and has powerful nonlinear mapping and learning capabilities.

[0029] After the model is established, the model is trained using the collected data on welding pressure, welding time, and sample thickness. The training process is to gradually reduce the error between the model output and the actual welding current by continuously adjusting the internal parameters of the model (such as weights and biases). Through a large amount of training data, the model can learn the complex relationship between welding process parameters and welding current.

[0030] Step 103: using the Step Decay learning rate adjustment method to optimize the convergence of the artificial neural network model.

[0031] In the embodiment of the present application, when training an artificial neural network model, the learning rate is a very important hyperparameter. It determines the step size of the model to update the parameters in each iteration. A learning rate that is too large may cause the model to fail to converge, and a learning rate that is too small will cause the training process to be too slow. Step Decay is a commonly used learning rate adjustment method. This method periodically reduces the learning rate according to a certain step size during the training process. In this way, the optimal solution can be quickly approached in the early stage of training, and then the model parameters can be fine-tuned by reducing the learning rate in the later stage of training, thereby improving the convergence and final performance of the model.

[0032] Step 104, using the trained artificial neural network model to predict the welding current in real time according to the material thickness and welding parameters.

[0033] In the embodiment of the present application, after the model training is completed, it can be used to perform real-time prediction of welding current. In the actual welding process, only the current material thickness and welding parameters (such as welding pressure and welding time) need to be input, and the model can quickly output the corresponding welding current prediction value.

[0034] This real-time prediction capability is of great significance for realizing the automation and intelligence of the welding process. It can help welders or welding robots adjust the welding current according to specific circumstances, thereby ensuring welding quality and improving production efficiency.

[0035] Step 105, dynamically adjusting the welding process through a real-time feedback mechanism.

[0036] In the embodiments of the present application, the welding process is dynamically adjusted through a real-time feedback mechanism, which generally involves real-time monitoring and feedback control of various parameters in the welding process.

[0037] Specifically, the predicted welding current can be compared with the actual welding current. If there is a difference, the actual welding current is brought close to the predicted value by adjusting the welding process parameters (such as welding pressure, welding time, etc.). This dynamic adjustment process can be carried out in real time to ensure that the welding process is always kept in the best state.

[0038] In addition, the model parameters or training data can be further adjusted according to the welding quality monitoring results to improve the prediction accuracy and adaptability of the model. In this way, closed-loop control of the welding process can be achieved, thereby improving welding quality and production efficiency.

[0039] The embodiment of the present application uses an artificial neural network model to automatically predict and adjust the welding process parameters, avoiding the limitations of traditional manual experience adjustment, thereby improving the accuracy and stability of welding current control. Secondly, the adoption of the Step Decay learning rate adjustment method optimizes the convergence of the model, enabling it to adapt to the welding requirements of different materials and thicknesses, and enhancing the generalization ability and adaptability of the model. In addition, the application of the real-time feedback mechanism enables the welding process to be dynamically adjusted according to the prediction results and the actual welding situation, ensuring the continuous stability of the welding quality. Finally, the implementation of this series of intelligent measures not only greatly reduces the adjustment time and labor cost of the welding process, but also reduces the rework rate due to the improvement of welding quality and stability, thereby significantly improving production efficiency. In summary, these steps realize the automation and closed-loop control of the welding process by introducing intelligent technology and real-time feedback mechanism, which not only improves the welding quality and efficiency, but also reduces the production cost, providing strong support for the intelligent transformation of the welding field.

[0040] Optionally, the step 101 includes: constructing an artificial neural network model including multiple hidden layers, each hidden layer using a ReLU activation function for nonlinear mapping.

[0041] In the embodiments of the present application, an artificial neural network is a mathematical model that imitates the structure and function of a biological neural network and is used to process complex data and pattern recognition tasks. ANN consists of multiple layers, including an input layer, a hidden layer, and an output layer. The input layer receives data, the hidden layer processes the data, and the output layer gives the final result.

[0042] In a multi-layer neural network, the number of hidden layers is an important design decision. More hidden layers usually mean that the model can learn more complex patterns, but it may also lead to overfitting and increased computational cost. Each hidden layer contains a certain number of neurons (or nodes), which are connected to the neurons in the previous layer through weights.

[0043] Each layer of neurons is connected to the neurons of the previous layer through weights, forming a fully connected layer (also called a dense layer). In addition to the weight connection, each neuron also contains a bias term to adjust the activation threshold of the neuron. Each hidden layer uses the ReLU activation function for nonlinear mapping The activation function is a very important component of a neural network, which determines how a neuron responds to its input. The activation function introduces nonlinearity, allowing the neural network to learn and represent complex nonlinear relationships.

[0044] ReLU (Rectified Linear Unit) is one of the most popular activation functions. Its mathematical expression is f(x) = max(0, x). When the input x is greater than 0, the ReLU function returns x; when x is less than or equal to 0, the ReLU function returns 0. The ReLU function has the advantages of simple calculation and gradient is not easy to disappear (compared with sigmoid and tanh activation functions), so it is widely used in deep learning.

[0045] Without an activation function, a multi-layer neural network is actually just a combination of linear transformations and cannot learn nonlinear relationships. By applying the ReLU activation function after each hidden layer, the neural network is able to introduce nonlinearity, thereby having the ability to learn and represent complex nonlinear patterns.

[0046] Optionally, step 103 includes: using a Step Decay learning rate adjustment method to gradually reduce the learning rate in multiple training cycles to optimize model convergence and avoid falling into a local optimal solution.

[0047] In the embodiments of the present application, Step Decay is a commonly used learning rate adjustment strategy, which aims to optimize the convergence of the model by gradually reducing the learning rate during the training process. The basic idea of ​​this method is to divide the training process into multiple stages, using a different learning rate for each stage, and usually gradually reducing the learning rate as the training progresses.

[0048] Before training begins, you need to set an initial learning rate. The choice of this initial learning rate is usually based on experience or hyperparameter tuning. During training, the learning rate will be reduced by a fixed decay factor at certain training cycles (such as a certain number of epochs or iterations). For example, if the initial learning rate is 0.1 and the decay factor is 0.5, then the learning rate will be halved after each preset training cycle. The choice of decay cycle and decay factor has a great impact on the training effect of the model. A shorter decay cycle and a larger decay factor may cause the learning rate to drop to a very small value too early, thereby affecting the convergence speed of the model; while a longer decay cycle and a smaller decay factor may cause the learning rate to remain large in the later stages of training, making it difficult for the model to converge stably.

[0049] By adopting the Step Decay learning rate adjustment method, a larger learning rate can be used in the early stage of training to quickly update model parameters and accelerate convergence. As training progresses, gradually reducing the learning rate can help the model perform a more detailed search near the global optimal solution of the loss function, thereby improving the convergence and final performance of the model. In deep learning, the problem of local optimal solutions may be encountered during model training. A local optimal solution refers to a solution found by the model during training. Although the loss function value at this solution is small, it is not the global minimum. The Step Decay learning rate adjustment method can prevent the model from falling into the local optimal solution to a certain extent. This is because in the later stage of training, as the learning rate gradually decreases, the search step size of the model in the parameter space will also decrease accordingly, making it more likely that the model will search near the global optimal solution instead of converging to the local optimal solution too early.

[0050] The optimization algorithm used is selected from the gradient descent method. Specifically, the gradient descent method is used to minimize the loss function of the model. The model parameters are updated in the opposite direction of the gradient by calculating the gradient of the loss function with respect to the model parameters.

[0051] Specifically, the update rule of the gradient descent method is as follows: θ=θ-η·∂L(θ) / ∂θ (1) Where θ represents the parameters of the neural network model (such as weights and biases). η represents the learning rate, which is used to control the step size of each parameter update. ∂L(θ) / ∂θ represents the loss function L(θ) with respect to the parameter θ.

[0052] Through multiple iterations, the gradient descent method gradually optimizes the model parameters so that the loss function value gradually approaches the global or local minimum.

[0053] In practical applications, the Step Decay learning rate adjustment method is often used in combination with other optimization algorithms (such as SGD, Adam, etc.). In deep learning frameworks such as PyTorch, the Step Decay learning rate adjustment method can be implemented by setting the corresponding learning rate scheduler (such as torch.optim.lr_scheduler.StepLR).

[0054] Optionally, the step 104 includes: inputting real-time welding parameters into the trained artificial neural network model to obtain a predicted optimal welding current, and adjusting the welding process through a real-time feedback mechanism.

[0055] In the embodiment of the present application, during the welding process, there are a series of parameters that affect the welding quality, such as welding speed, welding material type, electrode diameter, welding voltage, etc. These parameters are called real-time welding parameters. These parameters are acquired in real time by sensors or other measuring devices and provided as input data to the trained artificial neural network model.

[0056] In the previous step, the artificial neural network has been trained using historical welding data (including welding parameters and the corresponding optimal welding current). This training process aims to allow the neural network to learn the complex relationship between welding parameters and optimal welding current. Once trained, the neural network model is able to predict the optimal welding current based on new input data (i.e. real-time welding parameters).

[0057] When real-time welding parameters are input into the trained neural network model, the model uses what it has learned to calculate the optimal welding current predicted under the current welding conditions. This prediction is an estimate of the optimal welding current by the neural network model, aiming to optimize the welding quality.

[0058] Once the predicted optimal welding current is obtained, this value is immediately used to adjust the welding process. This is usually achieved through a real-time feedback mechanism that receives the prediction results of the neural network and converts them into actual control instructions for the welding equipment. For example, if the predicted optimal welding current is higher than the currently set welding current, the feedback mechanism will increase the welding current accordingly.

[0059] The key to the real-time feedback mechanism is its rapid response capability, which can continuously monitor and adjust welding parameters during the welding process to ensure that the welding quality is always maintained at the optimal level. This real-time adjustment capability is an important feature of the automated welding system, which makes the welding process more stable, reliable and efficient.

[0060] In practical applications, this step is usually closely integrated with welding automation equipment to form a closed-loop control system. The welding automation equipment is responsible for performing welding operations and collecting welding parameters in real time; the artificial neural network model is responsible for predicting the optimal welding current based on these parameters; and the real-time feedback mechanism is responsible for converting the prediction results into actual control instructions to optimize the welding process.

[0061] Optionally, after step 105, the method further includes: monitoring the welding process through a real-time feedback mechanism, and feeding back the monitoring data to the artificial neural network model for dynamic optimization.

[0062] In the embodiment of the present application, the real-time feedback mechanism plays a vital role in the welding process. It uses sensors and other monitoring equipment to collect various parameters and data in the welding process in real time, such as welding current, welding voltage, welding speed, weld quality, etc. These data provide real-time status information of the welding process, which is the basis for subsequent optimization and adjustment.

[0063] The real-time feedback mechanism usually includes the following parts: Sensors are used to measure physical quantities during the welding process, such as current, voltage, temperature, etc. The data acquisition system converts the data measured by the sensors into digital signals and transmits them to the processing system. The monitoring software processes and analyzes the collected data in real time to extract useful information.

[0064] Once the real-time feedback mechanism has collected the monitoring data of the welding process, the data is transferred to the trained artificial neural network model. The model has previously learned the relationship between welding parameters and welding quality through historical data. Now, it uses this new real-time data to update and optimize its predictions and decisions. The raw monitoring data is cleaned and normalized to meet the requirements of the model input. The processed data is used as the input of the model and fed into the artificial neural network. Based on the input data, the model predicts the current state or future trends of the welding process, such as weld quality, welding efficiency, etc.

[0065] After the artificial neural network model makes predictions based on the real-time feedback data, the welding process can be dynamically optimized based on the prediction results. The optimization may involve adjusting welding parameters (such as current, voltage, speed, etc.), improving welding processes, changing welding materials, etc. The goal of dynamic optimization is to optimize the welding process, improve welding quality and efficiency, and reduce resource waste and costs. The optimization process may be automatic or semi-automatic, depending on the complexity and degree of automation of the system.

[0066] In practical applications, dynamic optimization is usually combined with a control system to form a closed-loop control loop. The control system issues control instructions based on the prediction results of the model and adjusts the operating parameters of the welding equipment, thereby achieving real-time control and optimization of the welding process.

[0067] Optionally, the step 105 includes: dynamically adjusting current parameters during the welding process according to the predicted optimal welding current to improve welding quality.

[0068] In the embodiment of the present application, the value of the optimal welding current under the current welding conditions is predicted by some method (such as an artificial neural network model). This predicted value is obtained based on the study and analysis of historical welding data and the consideration of the real-time parameters of the current welding process. The optimal welding current refers to the current under which the welding process can achieve the best welding quality, such as weld strength, appearance, penetration, etc.

[0069] Once the predicted optimal welding current is obtained, the next step is to dynamically adjust the current parameters during the welding process based on this predicted value. This usually involves the following steps: During the welding process, the current welding current is monitored in real time by sensors or other monitoring equipment. This is the basis for ensuring that the current can be accurately adjusted. The real-time monitored welding current is compared with the predicted optimal welding current. If there is a deviation between the current current and the optimal current, an adjustment is required. Based on the results of the comparison, the welding current is dynamically adjusted through the control system of the welding equipment to make it close to or reach the predicted optimal welding current. This adjustment process may be continuous or staged, depending on the response speed of the welding equipment and the complexity of the welding process. After adjusting the current, the welding current is monitored again and compared with the predicted value. If there is still a deviation, further adjustments are made until the welding current matches the predicted optimal welding current.

[0070] By dynamically adjusting the current parameters during the welding process, the welding process can be carried out under optimal conditions, thereby improving the welding quality. Specifically, optimizing the welding current can bring the following benefits: The appropriate welding current can ensure good fusion of the weld and reduce the generation of defects such as pores and cracks. By accurately controlling the welding current, unnecessary energy waste can be avoided and the welding speed can be increased. Optimizing the welding process can reduce the scrap rate and rework rate, thereby reducing production costs.

[0071] In practical applications, dynamic adjustment of welding current is usually combined with automated welding systems. These systems can monitor welding parameters in real time and automatically adjust them according to preset algorithms or models. With the development of artificial intelligence and machine learning technology, the ability to predict the optimal welding current is also constantly improving, providing stronger support for the automation and intelligence of the welding process.

[0072] Based on the same inventive concept, the embodiment of the present application also provides a device for optimizing welding current for multiple material thicknesses for implementing the above-mentioned method for optimizing welding current for multiple material thicknesses. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more embodiments of the device for optimizing welding current for multiple material thicknesses provided below can refer to the limitations of the method for optimizing welding current for multiple material thicknesses above, and will not be repeated here.

[0073] In an exemplary embodiment, Figure 3 As shown, a device 20 for optimizing welding current suitable for multiple material thicknesses is provided, comprising: An acquisition module 201 is used to collect welding process parameters, which at least include: welding pressure, welding time and sample thickness; A training module 202, for establishing an artificial neural network model, and training the artificial neural network model using the welding pressure, the welding time and the sample thickness; A Step Decay learning rate adjustment device is used to optimize the convergence of the artificial neural network model; Prediction module 203, used to predict welding current in real time according to material thickness and welding parameters using a trained artificial neural network model; The optimization module 204 is used to dynamically adjust the welding process through a real-time feedback mechanism.

[0074] Optionally, the acquisition module 201 is further used to: An artificial neural network model with multiple hidden layers is constructed, and each hidden layer uses the ReLU activation function for nonlinear mapping.

[0075] Optionally, the training module 202 is further used to: The Step Decay learning rate adjustment device is used to gradually reduce the learning rate in multiple training cycles to optimize the model convergence and avoid falling into the local optimal solution.

[0076] Optionally, the prediction module 203 is further used for: The real-time welding parameters are input into the trained artificial neural network model to obtain the predicted optimal welding current, and the welding process is adjusted through a real-time feedback mechanism.

[0077] Optionally, the optimization module 204 is further configured to: The welding process is monitored through a real-time feedback mechanism, and the monitoring data is fed back to the artificial neural network model for dynamic optimization.

[0078] Optionally, the optimization module 204 is further configured to: According to the predicted optimal welding current, the current parameters during the welding process are dynamically adjusted to improve the welding quality.

[0079] The embodiment of the present application uses an artificial neural network model to automatically predict and adjust the welding process parameters, avoiding the limitations of traditional manual experience adjustment, thereby improving the accuracy and stability of welding current control. Secondly, the adoption of the Step Decay learning rate adjustment method optimizes the convergence of the model, enabling it to adapt to the welding requirements of different materials and thicknesses, and enhancing the generalization ability and adaptability of the model. In addition, the application of the real-time feedback mechanism enables the welding process to be dynamically adjusted according to the prediction results and the actual welding situation, ensuring the continuous stability of the welding quality. Finally, the implementation of this series of intelligent measures not only greatly reduces the adjustment time and labor cost of the welding process, but also reduces the rework rate due to the improvement of welding quality and stability, thereby significantly improving production efficiency. In summary, these steps realize the automation and closed-loop control of the welding process by introducing intelligent technology and real-time feedback mechanism, which not only improves the welding quality and efficiency, but also reduces the production cost, providing strong support for the intelligent transformation of the welding field.

[0080] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store optimization data for welding currents suitable for multiple material thicknesses. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for optimizing welding currents suitable for multiple material thicknesses is implemented.

[0081] Those skilled in the art will understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0082] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0083] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0084] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0085] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0086] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0087] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0088] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0089] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for optimizing welding current for various material thicknesses, characterized in that: The optimization method for welding current applicable to multiple material thicknesses includes: Collecting welding process parameters, wherein the welding process parameters at least include: welding pressure, welding time, and sample thickness; Establishing an artificial neural network model, and training the artificial neural network model using the welding pressure, the welding time, and the sample thickness; Adopting the Step Decay learning rate adjustment method to optimize the convergence of the artificial neural network model; Use the trained artificial neural network model to predict the welding current in real time based on material thickness and welding parameters; Dynamically adjust the welding process through real-time feedback mechanism.

2. The method for optimizing welding current applicable to multiple material thicknesses according to claim 1, characterized in that: The step of collecting welding process parameters comprises: An artificial neural network model with multiple hidden layers is constructed, and each hidden layer uses the ReLU activation function for nonlinear mapping.

3. The method for optimizing welding current applicable to multiple material thicknesses according to claim 1, characterized in that: The step of optimizing the convergence of the artificial neural network model by using the Step Decay learning rate adjustment method comprises: The Step Decay learning rate adjustment method is used to gradually reduce the learning rate in multiple training cycles to optimize the convergence of the model and avoid falling into the local optimal solution.

4. The method for optimizing welding current applicable to multiple material thicknesses according to claim 1, characterized in that: The step of using the trained artificial neural network model to predict the welding current in real time according to the material thickness and welding parameters includes: The real-time welding parameters are input into the trained artificial neural network model to obtain the predicted optimal welding current, and the welding process is adjusted through a real-time feedback mechanism.

5. The method for optimizing welding current applicable to multiple material thicknesses according to claim 1, characterized in that: The step of adjusting the welding process through the real-time feedback mechanism comprises: The welding process is monitored through a real-time feedback mechanism, and the monitoring data is fed back to the artificial neural network model for dynamic optimization.

6. The method for optimizing welding current applicable to multiple material thicknesses according to claim 1, characterized in that: The step of dynamically adjusting the welding process through the real-time feedback mechanism includes: According to the predicted optimal welding current, the current parameters during the welding process are dynamically adjusted to improve the welding quality.

7. An optimization device for welding current of various material thicknesses, characterized in that: The device for optimizing welding current suitable for multiple material thicknesses comprises: An acquisition module is used to collect welding process parameters, wherein the welding process parameters at least include: welding pressure, welding time and sample thickness; A training module, used for establishing an artificial neural network model, and training the artificial neural network model using the welding pressure, the welding time and the sample thickness; A Step Decay learning rate adjustment device is used to optimize the convergence of the artificial neural network model; A prediction module for predicting welding current in real time based on material thickness and welding parameters using a trained artificial neural network model; Optimization module for dynamically adjusting the welding process through real-time feedback mechanisms.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method for optimizing welding current applicable to multiple material thicknesses as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for optimizing welding current applicable to multiple material thicknesses according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for optimizing welding current applicable to multiple material thicknesses according to any one of claims 1 to 6 are implemented.

Images