CN117182370A - Intelligent welding optimization and error source analysis method - Google Patents

Abstract

The invention relates to the field of intelligent manufacturing technology, and in particular to an intelligent welding optimization and error source analysis method, which includes the following steps: collecting welding parameters during the welding process and performing preprocessing; calculating and obtaining the overall performance change value, and combining the welding parameters and the overall performance change Values are input into the pre-trained deep learning model to obtain the sensitivity of each geometric error source and identify key geometric error sources; the welding parameters are dynamically adjusted in real time based on the adjustment strategy, the adjusted performance indicators are evaluated, and based on the evaluation results By optimizing the deep learning model and adjustment strategy, this invention can ensure optimal performance under various production environments and conditions; adapt to changing production needs; can effectively reduce production costs and improve production efficiency, while also improving the quality of the final product , has significant economic benefits.

Description

Intelligent welding optimization and error source analysis method

Technical field

The present invention relates to the field of intelligent manufacturing technology, and in particular to an intelligent welding optimization and error source analysis method.

Background technique

As a key assembly process in the manufacturing industry, welding has a direct impact on product quality, production efficiency, and overall manufacturing costs. With the continuous development of advanced manufacturing technology, the accuracy and efficiency requirements for the welding process are becoming higher and higher.

The welding process involves multiple parameters (such as weld geometry parameters, welding speed, pressure, etc.). These parameters interact with each other, resulting in the optimization of one parameter may affect the performance of other parameters. Traditional welding technology is usually based on preset parameters, which is very difficult to achieve. It is difficult to dynamically adjust based on real-time conditions (such as material properties, equipment status, etc.); usually, the evaluation of the welding process focuses more on a single or a few performance indicators, such as weld quality, while ignoring others such as production speed and equipment utilization and other factors.

Existing welding technologies and methods are often limited to fixed parameter settings and lack adaptive capabilities, which results in the need for manual intervention when facing different production environments and needs, increasing production costs and time; existing methods usually only focus on a few Performance indicators, such as weld quality, are given less attention to other indicators that may be equally or more important, such as production efficiency and equipment utilization; current welding systems often lack real-time data analysis and feedback mechanisms, making even when When there are deviations or problems, it is difficult to make timely adjustments.

Contents of the invention

In view of the problems existing in the above-mentioned existing technologies, the present invention provides an intelligent welding optimization and error source analysis method. First, by applying a deep learning model, the present invention can accurately identify and quantify various key geometric error sources and improve welding quality and accuracy; Secondly, the present invention supports real-time adjustment of welding parameters and equipment scheduling to cope with uncertainties and changes in the production environment, thereby improving equipment utilization and production efficiency. Finally, the present invention has self-optimization capabilities and can continuously adjust based on real-time evaluation results. Deep learning models and scheduling strategies ensure continuous optimization and adaptability; the invention has a high degree of reliability and efficiency in industrial welding applications.

Intelligent welding optimization and error source analysis method includes the following steps:

Collect welding parameters during the welding process and perform preprocessing;

Calculate and obtain the overall performance change value, input the welding parameters and overall performance change value into the pre-trained deep learning model, obtain the sensitivity of each geometric error source, and identify key geometric error sources;

Welding parameters are dynamically adjusted in real time based on the adjustment strategy, the adjusted performance indicators are evaluated, and the deep learning model and adjustment strategy are optimized based on the evaluation results.

Preferably, the welding parameters include: welding seam geometric parameters, welding speed, welding pressure, material properties and equipment status, where,

The weld geometric parameters include: the width, depth and angle of the weld;

The material properties include: material type, thickness and temperature;

The device status includes: current, voltage and operating temperature of the device.

Preferably, the preprocessing includes: data cleaning and normalization.

Preferably, the calculation expression of the overall performance change value is:

Among them, ΔE is the overall performance change value, w _i is the weight of the i-th geometric error source, and Δe _i is the change value of the i-th geometric error source.

Preferably, the structure of the deep learning model is a convolutional neural network. By taking the welding parameters in the historical welding records as input and taking the overall performance change value as the output, the deep learning model is pre-trained for predicting and analyzing each Effect of geometric error sources on overall performance.

Preferably, the welding parameters and overall performance change values are input into the pre-trained deep learning model, and the calculation expression for obtaining the sensitivity of each geometric error source is:

Among them, _Si is the sensitivity value of the i-th geometric error source.

Preferably, the geometric error source whose sensitivity is greater than the first threshold and whose overall performance change value is greater than the second threshold is used as the key geometric error source.

Preferably, the adjustment strategy includes:

When the sensitivity of a certain key geometric error source exceeds the third threshold, the welding parameters are adjusted in real time. The third threshold is determined based on the sensitivity and the overall performance change value; if a certain key geometric error source affects the welding speed, the welding parameters are adjusted by reducing or increasing the sensitivity. Welding speed, the adjustment value depends on the sensitivity; if a key geometric error source affects the welding pressure, the welding pressure will be adjusted according to the sensitivity value; if a key geometric error source affects multiple parameters, the welding pressure will be adjusted according to the weight of the sensitivity. Each parameter is adjusted one by one;

When the sensitivities of at least two key geometric error sources exceed the third threshold, a comprehensive adjustment coefficient is calculated based on the sensitivity of each key geometric error source; the welding parameters are adjusted based on the comprehensive adjustment coefficient;

When the total sensitivity of a specific type of key geometric error sources exceeds the fourth threshold, a comprehensive evaluation of all equipment parameters affected by the key geometric error sources is performed, and the welding parameters are adjusted based on the evaluation results.

Preferably, the performance indicators include: welding accuracy, welding speed, equipment utilization and weld quality. The performance indicators are compared with the overall performance change value to complete the evaluation of the performance indicators.

Preferably, optimizing the deep learning model and adjustment strategy based on the evaluation results includes:

When the evaluation result is that the welding accuracy is poor, fine-tune the parameters related to the key geometric error sources in the deep learning model; in the adjustment strategy, reduce the welding speed and increase the welding pressure;

When the evaluation result is that the welding speed is slow, the deep learning model is retrained and the key geometric error sources related to speed are adjusted; in the adjustment strategy, the welding speed is increased and the welding pressure is reduced;

When the evaluation result is that the equipment utilization is low, analyze the relevant key geometric error sources and other additional factors that affect the equipment utilization, and adjust the deep learning model based on this information; in the adjustment strategy, adjust the equipment scheduling strategy.

Compared with the existing technology, the advantages and beneficial effects of the present invention are:

(1) This invention identifies and adjusts key geometric error sources in real time through a deep learning model, and can provide highly personalized welding parameter settings under various production environments and conditions, thereby ensuring optimal performance;

(2) By integrating multiple performance indicators and overall performance change values, the present invention can not only comprehensively evaluate welding quality, speed and equipment utilization, but also dynamically adjust models and strategies based on real-time feedback to adapt to changing production needs;

(3) Due to the accurate identification and optimization of key geometric error sources, the present invention can effectively reduce production costs and improve production efficiency, while also improving the quality of final products, and has significant economic benefits.

Description of the drawings

Figure 1 is a schematic flow chart of the method of the present invention;

Figure 2 is a schematic diagram of the content of the adjustment strategy in the present invention;

Figure 3 is a schematic block diagram of the structure of performance indicators in the present invention;

Figure 4 is a schematic block diagram of the evaluation result structure in the present invention.

Detailed ways

Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood, however, that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for convenience of explanation, numerous specific details are set forth to provide a comprehensive understanding of the disclosed embodiments. It will be apparent, however, that one or more embodiments may be practiced without these specific details. Furthermore, in the following description, descriptions of well-known structures and techniques are omitted to avoid unnecessarily confusing the concepts of the present disclosure.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the disclosure. The terms "comprising," "comprising," and the like, as used herein, indicate the presence of stated features, steps, operations, and/or components but do not exclude the presence or addition of one or more other features, steps, operations, or components.

All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art, unless otherwise defined. It should be noted that the terms used here should be interpreted to have meanings consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

Where an expression similar to "at least one of A, B, C, etc." is used, it should generally be interpreted in accordance with the meaning that a person skilled in the art generally understands the expression to mean (e.g., "having A, B and C "A system with at least one of" shall include, but is not limited to, systems with A alone, B alone, C alone, A and B, A and C, B and C, and/or systems with A, B, C, etc. ). Where an expression similar to "at least one of A, B or C, etc." is used, it should generally be interpreted in accordance with the meaning that a person skilled in the art generally understands the expression to mean (for example, "having A, B or C "A system with at least one of" shall include, but is not limited to, systems with A alone, B alone, C alone, A and B, A and C, B and C, and/or systems with A, B, C, etc. ).

Several block diagrams and/or flow diagrams are shown in the accompanying drawings. It will be understood that some blocks in the block diagrams and/or flowchart illustrations, or combinations thereof, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus such that the instructions, when executed by the processor, create functions for implementing the functions illustrated in these block diagrams and/or flowcharts. /operating device. The technology of the present disclosure may be implemented in the form of hardware and/or software (including firmware, microcode, etc.). Additionally, the technology of the present disclosure may take the form of a computer program product on a computer-readable storage medium having instructions stored thereon, and the computer program product may be used by or in conjunction with an instruction execution system.

As shown in Figure 1, the intelligent welding optimization and error source analysis method includes the following steps:

Collect welding parameters during the welding process and perform preprocessing;

Calculate and obtain the overall performance change value, input the welding parameters and overall performance change value into the pre-trained deep learning model, obtain the sensitivity of each geometric error source, and identify key geometric error sources;

The present invention dynamically adjusts welding parameters in real time based on the adjustment strategy, evaluates the adjusted performance indicators, and optimizes the deep learning model and adjustment strategy based on the evaluation results. Due to real-time monitoring and adjustment, the inventive solution is expected to significantly improve welding accuracy. Through deep learning models and sensitivity analysis, real-time adjustments can make the welding process smoother, thereby improving efficiency. The solution of the present invention can self-adjust and optimize based on real-time data and evaluation results, and has high adaptability. The invention solves the precision control problem in the complex welding process through data-driven and intelligent adjustment strategies, and has high theoretical and practical value.

Preferably, the welding parameters include: welding seam geometric parameters, welding speed, welding pressure, material properties and equipment status, where,

The welding seam geometric parameters include: the width, depth and angle of the welding seam; these parameters (the width, depth and angle of the welding seam) directly affect the mechanical properties and appearance quality of the welded joint. Improper geometric parameters lead to weld defects, such as slag inclusions, holes, etc.; if welding a storage tank, the width and depth of the weld must be large enough to withstand the internal pressure, and at the same time, the angle needs to be set appropriately to ensure that there is no stress concentration point.

Welding speed, welding pressure, these parameters are related to heat input and metal flow, thereby affecting the geometry and microstructure of the weld; different welding speeds and pressures will affect the size and shape of the heat-affected zone.

The material properties include: material type, thickness and temperature; material type, thickness and temperature have a great impact on the stability of the welding process; different materials have different melting points, thermal conductivity and thermal expansion coefficients, which all need to be adjusted during the welding process. be considered; if welding stainless steel and carbon steel, their different thermal expansion coefficients and melting points need to be considered, and the current and voltage need to be adjusted.

The equipment status includes: current, voltage, and operating temperature of the equipment; including current, voltage, and operating temperature of the equipment; these parameters are related to the stability of the power supply and the working status of the welding gun, thus affecting the controllability and accuracy of the entire welding process. , During the long-term continuous welding process, the operating temperature of the equipment increases. At this time, the current and voltage need to be appropriately reduced, or the machine needs to be shut down for cooling.

The above parameters are all related and affect each other. For example, increasing the current requires increasing the voltage, but this increases heat input, which affects the material's microstructure and mechanical properties.

Since various parameters influence each other, a comprehensive and integrated approach is needed to optimize these parameters to achieve ideal welding results. This is why deep learning models are needed to understand these complex interrelationships.

By carefully selecting and optimizing these welding parameters, the present invention can significantly improve welding quality and reduce defects and unqualified products; appropriate parameter settings can reduce downtime and repair time during the welding process, thereby improving production efficiency; optimizing parameters can reduce material waste and energy consumption, thus reducing production costs.

Preferably, the preprocessing includes: data cleaning and normalization.

Data cleaning includes:

Data integrity: Any missing or incomplete data needs to be identified and processed; missing data occurs due to sensor failure, data transmission errors, or other non-standard operations;

Outlier processing: During the welding process, outliers or outliers will be encountered; these data are generated due to equipment failure or operator error and need to be identified and eliminated;

Consistency: Ensure that all data follows the same units and scale; for example, if one sensor reports current in amperes and another in milliamperes, they need to be unified to a standard unit.

In one embodiment, assume that during the welding process, the current reading suddenly jumps to an abnormally high value, for example, from 20A to 200A. Such outliers can be identified and removed through a data cleaning process.

Normalization processing includes:

Range scaling: Scale the data range of all features to the same interval (usually [0,1] or [-1,1]). This helps machine learning models converge faster.

Mean and Variance Adjustment: Another common normalization technique is Z-Score normalization, which subtracts the mean and divides by the standard deviation, resulting in data with a mean of 0 and a variance of 1.

Weight adjustment: In some cases, some features are more important than others. Normalization can also be used to adjust based on these weights.

In one embodiment, welding is performed in multiple production lines, and the equipment status on each production line is different and will have different current and voltage ranges. Normalization ensures that all data is on the same scale, allowing deep learning models to make more accurate predictions.

It is assumed that material type has a greater influence on welding quality, relative to other parameters such as current and voltage. In this case, the material type can be given a higher weight during the normalization process.

In the present invention, preprocessed data is more easily processed by deep learning models or other machine learning models, thereby improving the accuracy and stability of the model; because the data range and distribution have been standardized, the model is more likely to converge quickly during the training process ;Data cleaning can eliminate errors and anomalies, thereby improving the reliability of model predictions.

Preferably, the calculation expression of the overall performance change value is:

Among them, ΔE is the overall performance change value, w _i is the weight of the i-th geometric error source, and Δe _i is the change value of the i-th geometric error source.

The weight w _i of the i-th geometric error source, which is a coefficient, indicates the degree of influence of the i-th geometric error source on the overall performance change. Weights can be empirically based, based on physical models, or derived through optimization algorithms. Weights range from 0 to 1, and all weights usually sum to 1 to represent a normalized scale.

The change value Δe _i of the i-th geometric error source represents the change amount of the i-th geometric error source (such as weld width, depth, angle, etc.). These changes can be actual measured values, or differences relative to a certain reference value (such as ideal weld geometry parameters).

The minimum value of ΔE represents optimal welding performance because all sources of geometric errors are minimized or weighted to their optimal state.

Since the weights and error values of the present invention are both variables, the above expression provides a flexible way to quantify the comprehensive impact of multiple geometric error sources on the overall performance; the weight w _i can be optimized according to different applications or different welding tasks, so that The model is able to adapt to different scenarios; expressions provide a quantitative way to evaluate the performance of the welding process, which is very useful for further analysis and optimization.

In one embodiment, it is assumed that there are three geometric error sources: the width (Δe ₁ ), depth (Δe ₂ ) and angle (Δe ₃ ) of the weld, and their weights are 0.4, 0.3 and 0.3 respectively.

In an ideal state: Δe ₁ = Δe ₂ = Δe ₃ = 0. At this time, ΔE = 0, which means there is no geometric error and the performance is optimal.

In actual operation: assuming that Δe ₁ =0.1, Δe ₂ =0.2, and Δe ₃ =0.1 are measured, the overall performance change value ΔE=0.4×0.1+0.3×0.2+0.3×0.1=0.16.

Preferably, the structure of the deep learning model is a convolutional neural network. By taking the welding parameters in the historical welding records as input and taking the overall performance change value as the output, the deep learning model is pre-trained for predicting and analyzing each Effect of geometric error sources on overall performance.

The deep learning model selected in this invention is Convolutional Neural Network (CNN). CNN mainly consists of convolutional layers, activation functions, pooling layers and fully connected layers. CNN is usually used to process data with a grid structure (such as images), but here, it is applied to process welding parameters and overall performance variation values.

Input: welding parameters in historical welding records (such as geometric parameters of the weld, welding speed, welding pressure, material properties and equipment status).

Output: Overall performance change value (ΔE).

The model is pre-trained using a large amount of historical welding records. This training process involves the minimization of a loss function, which is usually the difference between the actual output (i.e. the true ΔE) and the model's predicted output.

Once the model is pre-trained, it can be used to predict and analyze the impact of each geometric error source (such as weld width, depth, angle, etc.) on overall performance (ΔE) in real time.

Convolutional neural networks have a high degree of accuracy in complex pattern recognition, which helps accurately predict the impact of each source of geometric error on overall performance; because the model is pre-trained on the data, it automatically learns how to make trade-offs Different sources of geometric errors; once the model is trained, it can be used in a real-time environment to monitor and optimize the welding process.

In one embodiment, assuming a pre-trained CNN model, the inputs include the width, depth and angle of the weld, welding speed, welding pressure, material type, thickness, temperature, current and voltage.

During training, the model has learned, for example, that the depth and width of the weld are the most important sources of geometric error affecting overall performance (ΔE).

During the actual welding process, if the input parameters indicate that the weld depth is too small, the model will predict that ΔE will increase (performance decreases).

Based on this prediction, welding parameters (such as increasing welding pressure or slowing down welding speed) can be adjusted in real time to improve weld depth and thereby optimize overall performance.

In this way, the pre-trained CNN model can not only predict the impact of each geometric error source on the overall performance, but also provide a basis for real-time optimization decisions.

Preferably, the welding parameters and overall performance change values are input into the pre-trained deep learning model, and the calculation expression for obtaining the sensitivity of each geometric error source is:

Among them, _Si is the sensitivity value of the i-th geometric error source.

Sensitivity (S _i ) is a key indicator in the present invention, which is used to quantify the impact of the i-th geometric error source (for example, weld width, depth, etc.) on the overall performance change value (ΔE), aiming to explain the relationship between the model output and Dependencies between inputs.

By calculating the sensitivity of each geometric error source, the present invention is able to quantify the impact of these error sources on the overall performance; a high sensitivity value (S _i ) means that the corresponding geometric error source has a greater impact, so that adjustments can be prioritized; based on sensitivity With this information, the welding process can be optimized and adjusted in real time.

In one embodiment, it is assumed that there are three sources of geometric errors: weld width (Width), weld depth (Depth), and weld angle (Angle).

Computing sensitivity: Using a pre-trained deep learning model, we get:

Width sensitivity (S ₁ ) = 0.5

Depth sensitivity (S ₂ ) = 0.3

Angle sensitivity (S ₃ )=0.2

According to sensitivity analysis, width has the greatest impact on overall performance. Therefore, the optimization strategy should first focus on controlling the weld width.

If real-time monitoring shows deviations in weld width, make adjustments immediately to optimize overall performance.

In this way, sensitivity analysis not only provides the impact of each geometric error source on overall performance, but also provides a basis for real-time optimization and adjustment.

Preferably, the geometric error source whose sensitivity is greater than the first threshold and whose overall performance change value is greater than the second threshold is used as the key geometric error source.

The first threshold and the second threshold are set as a basis for judging the importance of geometric error sources. The first threshold is for sensitivity ( _Si ) and the second threshold is for overall performance change (ΔE). The threshold is generally determined by previous experience, simulation or statistical analysis, or it can be specifically set based on actual needs.

By considering both sensitivity and overall performance variation values, the present invention more fully evaluates the impact of each geometric error source, ensuring that only the truly important factors are selected as critical geometric error sources.

The present invention uses two thresholds to more effectively identify those geometric error sources that have the greatest impact on welding quality and performance; by focusing on key geometric error sources, the complexity and risk caused by over-optimizing non-critical parameters can be reduced; Based on these key geometric error sources, more targeted real-time dynamic adjustments can be made to optimize overall performance.

In one embodiment, assume that in a welding project, there are the following sensitivity and overall performance variation values for several geometric error sources:

Weld width (S ₁ =0.6, ΔE = 0.7)

Weld depth (S ₂ =0.2, ΔE = 0.1)

Weld angle (S ₃ =0.4, ΔE = 0.2)

Suppose the first threshold is 0.5 and the second threshold is 0.6.

In this case, only the "weld width" meets the conditions that the sensitivity is greater than 0.5 (first threshold) and the overall performance change value is greater than 0.6 (second threshold). Therefore, weld width is considered to be a key geometric error source.

Engineers or operators can focus on optimizing and adjusting weld width to achieve the greatest improvement in overall performance. This also reduces undue attention and waste of resources on less important sources of geometric errors.

Preferably, as shown in Figure 2, the adjustment strategy includes:

When the sensitivity of a certain key geometric error source exceeds the third threshold, the welding parameters are adjusted in real time. The third threshold is determined based on the sensitivity and the overall performance change value; if a certain key geometric error source affects the welding speed, the welding parameters are adjusted by reducing or increasing the sensitivity. Welding speed, the adjustment value depends on the sensitivity; if a key geometric error source affects the welding pressure, the welding pressure will be adjusted according to the sensitivity value; if a key geometric error source affects multiple parameters, the welding pressure will be adjusted according to the weight of the sensitivity. Each parameter is adjusted one by one;

The third threshold is determined based on sensitivity and overall performance change values. This means that when the impact of an error source exceeds this value, its impact on overall performance is considered significant and requires immediate real-time adjustments.

Depending on the sensitivity of the geometric error source, the adjustment strategy for the welding parameters can be decided.

If the sensitivity is high, the adjustment range will be larger. Conversely, if the sensitivity is lower, the adjustment will be smaller. The direction of adjustment (such as increasing or decreasing the welding speed) depends on the positive or negative value of the sensitivity or other performance-related factors.

If you know that a specific geometric error source mainly affects a certain parameter (such as welding speed or welding pressure), you can make targeted adjustments instead of blindly adjusting all parameters.

By setting the third threshold and performing real-time monitoring and adjustment, the present invention can ensure that the welding process is always performed under optimal conditions; targeted adjustments are made to key geometric error sources to avoid unnecessary waste of resources; as welding progresses, With real-time optimization of parameters, improvements in welding quality and stability can be expected.

In one embodiment, assuming that there is a key geometric error source - the depth of the weld, the sensitivity is 0.9, which far exceeds the third threshold of 0.7. Since this error source is known to mainly affect welding speed and welding pressure, adjustments need to be made:

If the increased depth of the weld causes the welding speed to be too fast, the welding speed needs to be reduced. The specific reduction will be determined based on the sensitivity value of 0.9; assuming that the increase in the depth of the weld also causes the welding pressure to be too high, the welding pressure also needs to be reduced. The size of the adjustment is also determined based on the sensitivity value.

If the weld depth affects welding speed, welding pressure and other parameters at the same time, it is also necessary to consider how the sensitivity of 0.9 is allocated to each parameter to ensure that the adjustment of each parameter can achieve the best effect.

When the sensitivities of at least two key geometric error sources exceed the third threshold, a comprehensive adjustment coefficient is calculated based on the sensitivity of each key geometric error source; the welding parameters are adjusted based on the comprehensive adjustment coefficient;

When the sensitivities of at least two critical geometric error sources exceed a third threshold, a single adjustment is not sufficient to solve the problem. At this time, a more complex and diversified solution is needed. The comprehensive adjustment coefficient is a comprehensive value that represents the sensitivity of multiple error sources. It is their weighted sum or other composite value, which is used to synthesize the impact of multiple key geometric error sources on welding parameters.

Based on the calculated comprehensive adjustment coefficient, real-time dynamic adjustment of welding parameters can be performed more accurately. Because this coefficient combines the effects of multiple error sources, it provides a multi-dimensional perspective to adjust welding parameters, making the adjustment more comprehensive and accurate.

By considering the influence of multiple key geometric error sources, the present invention provides a more accurate adjustment basis for the comprehensive adjustment coefficient, thereby optimizing the welding process more accurately; in complex situations where multiple key geometric error sources need to be adjusted, Using the comprehensive adjustment coefficient can ensure that the overall performance is more stable; through precise adjustment, limited resources can be effectively utilized and waste avoided.

In one embodiment, it is assumed that there are two key geometric error sources - weld depth and weld width. Their sensitivities are 0.9 and 0.8 respectively, both exceeding the third threshold of 0.7.

Calculate the combined adjustment factor: This factor can be calculated in a number of ways, such as simply taking the average of the two sensitivities, or using a more complex weighted average. Assuming that the weighted average is used and the weights are 0.6 and 0.4 respectively, then the comprehensive adjustment coefficient = 0.9×0.6+0.8×0.4=0.86

Because the comprehensive adjustment coefficient is 0.86, which is a quite high value, it means that the welding parameters need to be adjusted significantly. For example, if the weld depth mainly affects the welding speed, and the weld width mainly affects the welding pressure, then both parameters need to be significantly adjusted according to the comprehensive adjustment coefficient.

Through such multi-dimensional adjustment, not only can a single parameter be more accurately controlled, but the welding process can also be more comprehensively optimized.

When the total sensitivity of a specific type of key geometric error sources exceeds the fourth threshold, a comprehensive evaluation of all equipment parameters affected by the key geometric error sources is performed, and the welding parameters are adjusted based on the evaluation results.

"Total sensitivity to a specific type of critical geometric error sources", which usually means that such error sources (such as weld geometry parameters, material properties, etc.) have a cumulative or combined effect, which is obtained by weighted averaging or other mathematical means of. Total sensitivity is actually a number that reflects the overall impact of these error sources.

After the total sensitivity exceeds the fourth threshold, a comprehensive evaluation of all equipment parameters (such as welding speed, welding pressure, current, voltage, etc.) affected by these error sources needs to be carried out. This assessment will use a multi-metric evaluation approach that considers not only the geometric error sources themselves, but also their interactions with other parameters.

The evaluation results will be used as a basis for adjusting welding parameters. The goal of this stage is, based on the evaluation results, to identify optimal or suboptimal welding parameter settings that minimize resource usage and costs while maintaining or improving welding quality.

In the present invention, when the total sensitivity of multiple key geometric error sources exceeds a threshold, it means that a more global and higher-level optimization is required, not just the adjustment of a single parameter; through comprehensive evaluation and optimization of multiple parameters , resources can be used more effectively; comprehensive evaluation and optimization can identify parameter settings that can stably improve welding quality.

In one embodiment, assume that there is a specific type of critical geometric error source, such as a "weld geometry parameter," which includes the width and depth of the weld. Assuming that the sensitivities of the two are 0.8 and 0.9 respectively, the total sensitivity is 0.8×0.5+0.9×0.5=0.85, which exceeds the fourth threshold of 0.8.

The influence of these geometric parameters on all equipment parameters such as welding speed and welding pressure is evaluated. Suppose the evaluation results show that the welding speed needs to be reduced and the welding pressure needs to be increased; based on the evaluation results, you will reduce the welding speed and increase the welding pressure. For example, if the current welding speed is 4mm/s, you will reduce it to 3.8mm/s; if the current welding pressure is 150MPa, you will increase it to 155MPa.

In this way, when the total sensitivity of a specific type of critical geometric error sources exceeds the fourth threshold, a comprehensive evaluation and parameter adjustment can more comprehensively address the problem.

Preferably, as shown in Figure 3, the performance indicators include: welding accuracy, welding speed, equipment utilization and weld quality. The performance indicators are compared with the overall performance change value to complete the evaluation of the performance indicators.

Performance metrics are key elements in quantifying mission objectives. These indicators directly or indirectly reflect various aspects of the welding process, such as operating efficiency, economy of equipment use, welding quality, etc.

The overall performance change value is a more comprehensive quantitative expression, which is obtained by integrating the change values and weights of multiple geometric error sources. This metric not only quantifies a single aspect of performance, but also integrates multiple aspects to give a more global perspective.

By comparing each performance indicator with the overall performance change value, a more comprehensive and comprehensive evaluation result can be obtained. Doing so can help identify which performance metric or metrics are most consistent with overall performance changes, allowing for more targeted optimization.

Before comparison, each performance indicator and the overall performance change value need to be normalized so that they can be compared under the same dimension. In addition, individual performance indicators have different weights, which also need to be considered when calculating the overall performance change value.

This step allows the system to self-assess and adjust in a more scientific and precise manner. If a certain performance indicator is highly consistent with the overall performance change value, then optimizing this indicator will further improve the overall performance. On the contrary, if a certain indicator is inconsistent with the overall performance change value, then the weight of the indicator needs to be reconsidered or the calculation method of the overall performance change value needs to be revised.

In one embodiment, it is assumed that in a welding task, the performance indicators of welding speed and equipment utilization are highly consistent with the overall performance change value, while the welding accuracy is relatively low. This means that optimizing welding speed and equipment utilization improves overall performance. Therefore, in the subsequent optimization process, special attention can be paid to these two indicators, such as improving equipment utilization by adjusting workflow or optimizing equipment scheduling, or increasing welding speed by changing welding parameters.

Preferably, as shown in Figure 4, optimizing the deep learning model and adjustment strategy based on the evaluation results includes:

When the evaluation result is that the welding accuracy is poor, the parameters related to the key geometric error sources in the deep learning model are fine-tuned; in the adjustment strategy, the welding speed is reduced and the welding pressure is increased; this means optimizing the model weights and activation functions, etc. To improve the sensitivity of the model to geometric errors; at the same time, reduce the welding speed and increase the welding pressure, which is based on physical principles, because slowing down the speed and increasing the pressure can usually improve the quality of the weld.

When the evaluation result is that the welding speed is slow, the deep learning model is retrained and adjusted for the key geometric error sources related to speed; in the adjustment strategy, the welding speed is increased and the welding pressure is reduced; here the deep learning model needs to be retrained, focusing on Adjustment of key geometric error sources related to welding speed. Increasing welding speed and reducing welding pressure can improve overall production efficiency, but it is also necessary to ensure that welding quality is not sacrificed.

When the evaluation result is low equipment utilization, analyze the relevant key geometric error sources and other additional factors that affect equipment utilization, and adjust the deep learning model based on this information; in the adjustment strategy, adjust the equipment scheduling strategy. In this case , in addition to geometric error sources, other factors affecting equipment utilization must also be considered. For example, equipment maintenance time, raw material supply, etc. Make appropriate adjustments to the deep learning model and optimize the equipment scheduling strategy.

The device scheduling policy is initially set to an "initial state", which is typically a policy based on experience or simple rules that governs the allocation and use of devices under normal operating conditions. When the evaluation results show that the equipment utilization is lower than expected, the initial equipment scheduling strategy will be re-examined and adjusted. The following are possible optimization directions:

Dynamic priority reassignment: Re-evaluate the priority of each welding task and allocate equipment according to the new priority.

Flexible time window scheduling: Adjust the original time window to respond more flexibly to changing workloads or other unpredictable factors.

Load rebalancing: Redistribute upcoming tasks to less loaded devices based on real-time device operating status.

Predictive maintenance adjustments: If equipment is about to need maintenance, prioritize high-priority tasks and reschedule low-priority tasks to other equipment.

Multitasking compatibility re-evaluation: Taking into account the multi-tasking capabilities of the device, re-evaluate which tasks can be performed in parallel to improve device utilization.

Energy consumption optimization considerations: When adjusting the strategy, the energy efficiency of the device is also taken into consideration, and tasks are assigned to devices with higher energy efficiency as much as possible.

These adjustments are based on real-time assessments and the output of deep learning models and are designed to increase actual equipment utilization while maintaining or improving welding quality and efficiency. This dynamic adjustment mechanism makes equipment scheduling more flexible and responsive, and can better adapt to changing needs and conditions in the production environment.

The entire optimization process is dynamic and adaptive, and the optimization of models and strategies is based on real-time or near-real-time evaluation results. This enables the entire system to adapt to changing production conditions and needs.

The optimized models and strategies of the present invention should show significant improvements in improving welding accuracy, speeding up welding, and increasing equipment utilization. This will ultimately lead to reduced costs, increased production efficiency and improved product quality.

In one embodiment, the welding accuracy is poor: If the evaluation results show that the welding accuracy is low, after model fine-tuning and parameter adjustment, the new welding task shows higher welding accuracy under the same conditions.

Slow welding speed: Assume that 10 welding tasks were originally completed per hour. After optimization of the model and strategy, 12 welding tasks per hour can now be completed while maintaining or improving welding quality.

Low equipment utilization: The original equipment utilization rate was 60%, which was increased to 75% after optimization, which means less idle time and higher production efficiency.

These examples illustrate that by optimizing deep learning models and adjustment strategies based on evaluation results, various performance indicators of the production process can be significantly improved.

This solution is based on the ideas of deep learning and dynamic optimization, especially convolutional neural networks (CNN), to solve the geometric error problem and performance optimization in the welding industry. The plan covers the following key aspects:

Before model training, the welding data is cleaned and normalized to reduce noise and facilitate model training; a mathematical expression is used to quantify the change in overall performance, which takes into account the changing value of each geometric error source ( Δe _i ) and the corresponding weight (wi ₎ ; use historical welding data to pre-train the CNN to predict the impact of each geometric error source on the overall performance; use the pre-trained model to predict the impact of each geometric error source Quantify influence or "sensitivity"; set thresholds to identify which geometric error sources are key factors through sensitivity and overall performance change values; if the sensitivity of key geometric error sources exceeds the preset threshold, welding will be adjusted in real time based on the sensitivity Parameters; if there are multiple key error sources, a comprehensive adjustment coefficient will be calculated, and the welding parameters will be adjusted according to this coefficient; based on the evaluation results (such as welding accuracy, welding speed and equipment utilization, etc.), dynamically fine-tune the model parameters and adjustment strategies ; Finally, use various performance indicators to compare with the calculated overall performance change value to complete a comprehensive evaluation of performance.

By adjusting welding parameters in real time, the solution of the present invention can effectively reduce various geometric errors, thereby improving welding accuracy; by optimizing welding speed and equipment utilization, the solution helps to improve overall production efficiency; by reducing geometric errors and improving equipment Utilization rate, the solution can significantly reduce production costs; fine welding parameter control and real-time optimization help to improve the quality of the final product; the solution of the present invention is highly adaptable and flexible, and can be dynamically adjusted according to the actual production environment and needs Optimization; by comparing with predetermined performance indicators, the solution can provide a comprehensive and quantitative performance evaluation.

In summary, the solution of the present invention provides a comprehensive, adaptive and highly optimized solution, which can significantly improve the performance of the welding industry in terms of geometric error control and performance optimization.

Those skilled in the art will understand that embodiments of the present application may be provided as methods, systems or computer program products. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment that combines software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each process and/or block in the flowchart illustrations and/or block diagrams, and combinations of processes and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce a use A device for realizing the functions specified in one process or multiple processes of the flowchart and/or one block or multiple blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory that causes a computer or other programmable data processing apparatus to operate in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including the instruction means, the instructions The device implements the functions specified in a process or processes of the flowchart and/or a block or blocks of the block diagram.

These computer program instructions may also be loaded onto a computer or other programmable data processing device, causing a series of operating steps to be performed on the computer or other programmable device to produce computer-implemented processing, thereby executing on the computer or other programmable device. Instructions provide steps for implementing the functions specified in a process or processes of a flowchart diagram and/or a block or blocks of a block diagram.

In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

Memory may include non-volatile memory in computer-readable media, random access memory (RAM), and/or non-volatile memory in the form of read-only memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.

Computer-readable media includes both persistent and non-volatile, removable and non-removable media that can be implemented by any method or technology for storage of information. Information may be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), and read-only memory. (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, Magnetic tape cassettes, tape disk storage or other magnetic storage devices or any other non-transmission medium can be used to store information that can be accessed by a computing device. As defined in this article, computer-readable media does not include transitory media, such as modulated data signals and carrier waves.

It should also be noted that the terms "comprises," "comprises," or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements not only includes those elements, but also includes Other elements are not expressly listed or are inherent to the process, method, article or equipment. Without further limitation, an element qualified by the statement "comprises a..." does not exclude the presence of additional identical elements in the process, method, good, or device that includes the element.

The above are only examples of the present application and are not used to limit the present application. To those skilled in the art, various modifications and variations may be made to this application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application shall be included in the scope of the claims of this application.

Claims (10)

1. The intelligent welding optimization and error source analysis method is characterized by comprising the following steps of:

collecting welding parameters in the welding process, and preprocessing;

calculating to obtain an overall performance change value, inputting welding parameters and the overall performance change value into a pre-trained deep learning model, obtaining the sensitivity of each geometric error source, and identifying key geometric error sources;

And dynamically adjusting welding parameters in real time based on an adjustment strategy, evaluating the adjusted performance indexes, and optimizing the deep learning model and the adjustment strategy based on an evaluation result.

2. The intelligent welding optimization and error source analysis method of claim 1, wherein the welding parameters include: weld geometry, weld speed, weld pressure, material properties, and equipment status, wherein,

the weld geometry parameters include: width, depth, and angle of the weld;

the material properties include: material type, thickness, and temperature;

the device state includes: current, voltage and operating temperature of the device.

3. The intelligent welding optimization and error source analysis method of claim 1, wherein the preprocessing comprises: and (5) data cleaning and normalization processing.

4. The intelligent welding optimization and error source analysis method according to claim 1, wherein the calculation expression of the overall performance change value is:

wherein ΔE is the overall performance change value, w _i Is the weight of the ith geometric error source,Δe _i is the variation value of the ith geometric error source.

5. The intelligent welding optimization and error source analysis method according to claim 1, wherein the deep learning model structure is a convolutional neural network, and the deep learning model is pre-trained by taking the welding parameters in the history welding record as input and the overall performance change value as output, so as to predict and analyze the influence of each geometrical error source on the overall performance.

6. The intelligent welding optimization and error source analysis method according to claim 1, wherein the input of the welding parameters and the overall performance variation values into the pre-trained deep learning model obtains a calculation expression of the sensitivity of each geometrical error source as follows:

wherein S is _i Is the sensitivity value of the ith geometric error source.

7. The intelligent welding optimization and error source analysis method according to claim 1, wherein a geometric error source with sensitivity greater than a first threshold and an overall performance variation value greater than a second threshold is used as the key geometric error source.

8. The intelligent welding optimization and error source analysis method of claim 1, wherein the tuning strategy comprises:

when the sensitivity of a certain key geometric error source exceeds a third threshold, the welding parameters are adjusted, and the third threshold is determined based on the sensitivity and the overall performance change value; if a certain key geometric error source affects the welding speed, adjusting the numerical value according to the sensitivity by reducing or increasing the welding speed; if a certain key geometric error source influences the welding pressure, the welding pressure is adjusted according to the numerical value of the sensitivity; if a certain key geometric error source affects a plurality of parameters, adjusting each parameter one by one according to the weight of the sensitivity;

When the sensitivity of at least two key geometric error sources exceeds a third threshold, calculating a comprehensive adjustment coefficient according to the sensitivity of each key geometric error source; according to the comprehensive adjustment coefficient, adjusting welding parameters;

and when the total sensitivity of the specific type of key geometric error source exceeds a fourth threshold, comprehensively evaluating all equipment parameters influenced by the key geometric error source, and adjusting welding parameters according to an evaluation result.

9. The intelligent welding optimization and error source analysis method of claim 1, wherein the performance metrics comprise: and (3) comparing the performance index with the overall performance change value to finish the evaluation of the performance index, wherein the welding precision, the welding speed, the equipment utilization rate and the weld quality are the same.

10. The intelligent welding optimization and error source analysis method of claim 1, wherein optimizing the deep learning model and the adjustment strategy based on the evaluation result comprises:

when the estimation result is that the welding precision is poor, fine tuning parameters related to a key geometric error source in the deep learning model; in the regulation strategy, the welding speed is reduced, and the welding pressure is increased;

when the evaluation result is that the welding speed is low, retraining the deep learning model, and adjusting the key geometric error sources related to the speed; in the regulation strategy, the welding speed is increased, and the welding pressure is reduced;

When the evaluation result is that the equipment utilization rate is low, analyzing related key geometric error sources and other additional factors influencing the equipment utilization rate, and adjusting the deep learning model based on the information; and in the regulation strategy, regulating the equipment scheduling strategy.