Monitoring control method for welding process parameters

Abstract

The invention relates to the technical field of data processing, in particular to a welding process parameter monitoring control method which comprises the steps that welding data of welding equipment at the welding moment are collected in real time; welding parameters at the welding moment are calculated; the amplitude characteristic value at the welding moment is calculated; calculating a change difference value at the welding moment; determining a welding monitoring value at a welding moment; calculating an abnormal deviation value at the welding moment; and according to the abnormal deviation values of all the historical moments and the welding moments, whether the welding data at the welding moments are abnormal or not is judged through a K-means clustering algorithm, so that monitoring control over the welding process parameters is achieved. According to the method, the problem that the K-means clustering algorithm is difficult to distinguish between normal and abnormal small-amplitude change welding data can be effectively solved, and the accuracy and reliability of a welding process parameter monitoring system are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, to a method for monitoring and controlling welding process parameters. Background Art

[0002] Welding plays an important role in modern manufacturing, such as machinery, automobiles, aerospace and construction. Nowadays, industrial production has increasingly stringent requirements on welding quality and efficiency, and accurate monitoring and effective control of welding process parameters have become the key. The welding process involves many parameters such as current, voltage, speed, etc., which interact with each other and have a profound impact on welding quality. For example, too large a current can easily lead to deep penetration of the weld and overheating of the metal, making the weld structure coarse, reducing mechanical properties, and aggravating deformation; too small a current may cause problems such as insufficient penetration and incomplete penetration; voltage fluctuations affect arc stability and weld formation, and improper settings will cause defects such as uneven weld width and undercut; speed is related to heat input, too fast will cause the weld to cool too quickly, which is prone to cracks and pores, and too slow will increase the heat-affected zone and weaken the joint performance. Due to the limitations of the welding equipment itself, it is difficult to ensure that its output welding process parameters are always in the best state in actual welding operations, resulting in large fluctuations in welding quality, high scrap rate, increased costs and resource waste, and threats to product quality and safety.

[0003] The patent application document with the publication number CN118917223A discloses a multi-parameter weld fatigue life prediction method that takes into account the influence of manufacturing process, aiming to solve the problem of difficulty in predicting weld fatigue life under complex process conditions caused by too many influencing factors in the welding manufacturing process, and determine the optimal welding process combination on this basis. Specifically, the patent application document combines the feature analysis method and the conventional data-driven model to propose a life prediction method. The method uses feature component analysis to reduce the impact of changes in weld fatigue performance caused by different weld process parameters on the model, enhances the generalization ability of the model under complex process conditions, and then uses existing multi-condition data to assist in life prediction under target conditions, reducing data requirements; further, the method constructs a hyperparameter optimization index based on K-means clustering, optimizes the process characteristics through a parameter optimization algorithm, thereby enhancing the optimization effect between complex conditions and ensuring prediction accuracy.

[0004] However, the above patent application documents are aimed at the problem of difficulty in predicting weld fatigue life under complex process conditions caused by too many influencing factors in the welding manufacturing process, and do not solve the problem that when using the K-means clustering algorithm to monitor welding process parameters, there may be normal welding data with small changes and abnormal welding data with small changes in welding. The K-means clustering algorithm is difficult to accurately distinguish and accurately judge these two types of data with completely different properties but similar changes. As a result, the abnormal data that may really affect the welding quality cannot be accurately identified in the welding process parameter monitoring link, which reduces the accuracy and reliability of the entire welding process parameter monitoring system, resulting in the failure to timely detect welding equipment abnormalities and the inability to perform equipment maintenance in advance. At the same time, the performance and safety of subsequent related welding products cannot be guaranteed. Summary of the invention

[0005] In order to solve the problem that the K-means clustering algorithm cannot accurately distinguish between normal small changes and abnormal small changes in welding data, which reduces the accuracy and reliability of the welding process parameter monitoring system, resulting in failure to timely detect welding equipment abnormalities and thus failure to ensure the performance and safety of subsequent welding products, the present invention proposes a monitoring and control method for welding process parameters, which includes the following steps: Collect welding data of welding equipment at the time of welding in real time, the welding data includes current, voltage and speed; calculate welding parameters at the time of welding based on the welding data at the time of welding , is the current at the time of welding, is the voltage at the time of welding, is the speed at the time of welding, is a normalized function; taking the welding moment as the center, multiple moments including the welding moment are recorded as welding time periods; based on the difference in welding parameters between the welding moment and the other moments in the welding time period, and the sorting interval between the welding moment and the other moments in the welding time period after sorting by size, the amplitude characteristic value of the welding moment is calculated , is the number of moments in the welding period, The first The welding parameters at each moment, The welding time and the welding period The welding data at the moment The sorting interval of group data, is the absolute value symbol; based on the difference in amplitude characteristic values ​​between the welding moment and the rest of the moments in the welding period, the change difference value at the welding moment is calculated; the voltage at the welding moment is weighted by the change difference value as a weight to obtain the welding monitoring value at the welding moment; the maximum value and the minimum value of the welding monitoring values ​​at all moments in the welding period are respectively recorded as the upper limit monitoring value and the lower limit monitoring value at the welding moment; based on the difference in the upper limit monitoring value and the lower limit monitoring value between the welding moment and the rest of the moments in the welding period, the abnormal offset value at the welding moment is calculated; according to the abnormal offset values ​​of all historical moments and the welding moment, the K-means clustering algorithm is used to determine whether the welding data at the welding moment is abnormal, so as to realize the monitoring and control of the welding process parameters.

[0006] The beneficial effects are: by integrating the amplitude characteristic value, change difference value and abnormal offset value, the problem that the K-means clustering algorithm cannot accurately distinguish between normal small changes and abnormal small changes is effectively avoided, and the accuracy and reliability of monitoring are improved; it can more accurately capture abnormal fluctuations at the current welding moment, timely discover equipment abnormalities, avoid quality risks, and ensure the stability of the welding process; by accurately monitoring welding data, ensure the stability of the welding process, thereby improving the performance and safety of welding products and avoiding quality problems caused by abnormal fluctuations; it can timely discover problems with welding equipment, so as to take preventive maintenance measures, reduce the occurrence of failures, and improve production continuity. Among them, the calculation of welding parameters reflects the welding moment more comprehensively by integrating the influence of current, voltage and speed; normalization eliminates the dimensional differences between different parameters, improves the consistency and comparability of data, helps to balance the fluctuations of various parameters, reduce the impact of abnormal fluctuations, and improve stability. The calculation of the amplitude characteristic value can sensitively identify abnormal fluctuations at the current welding moment by considering the differences in welding parameters and the sorting intervals; the dimension differences are eliminated after normalization, making the stability monitoring of the welding process more accurate and consistent; and the comprehensiveness and accuracy of the welding process control are improved by considering the sorting differences of multiple groups of sequences.

[0007] Furthermore, the multiple moments include several historical moments and several future moments of welding; based on the welding data at the welding moment, a linear regression algorithm is used to perform prediction to obtain the welding data at several future moments, and then obtain the welding parameters at several future moments.

[0008] Furthermore, the change difference value satisfies the following relationship: ; In the formula, is the difference value of welding time, and are the number of historical moments and future moments of welding moments within the welding period, The number of historical moments of welding time in the welding period The amplitude characteristic value at the moment, is the number of future moments of welding time in the welding period The amplitude eigenvalue at a certain moment.

[0009] The beneficial effects are: being able to detect changes in the amplitude characteristics of the current welding moment and identify potential abnormal fluctuations; timely discover drastic changes in the current welding moment, provide early warning, and avoid quality problems; help adjust welding parameters to ensure a stable welding process and reduce defects; and more accurately reflect short-term fluctuations by combining changes in adjacent moments.

[0010] Furthermore, the welding monitoring value satisfies the following relationship: ; In the formula, is the welding monitoring value at the welding time, is the difference value of welding time.

[0011] Furthermore, the abnormal offset value satisfies the following relationship: ; In the formula, is the abnormal deviation value at the welding moment, and They are the upper and lower limit monitoring values ​​at the welding moment, and They are respectively The upper and lower monitoring limits at a given moment.

[0012] The beneficial effects are: by accurately calculating the difference between the upper and lower limit monitoring values, the sensitivity and accuracy of abnormality detection are enhanced; abnormal fluctuations that deviate from the normal range at the current welding moment can be detected in time, and the problem can be located quickly; data support is provided for parameter adjustment during the welding process to ensure process stability and quality consistency; and process fluctuations are fully reflected, which helps to monitor and adjust the welding process in real time and improve control capabilities.

[0013] Further, the determining whether the welding data at the welding moment is abnormal includes: Using the K-means clustering algorithm, the abnormal offset values ​​of all historical moments and welding moments are clustered into two categories to obtain two clusters; in response to the Euclidean distance of any data point between the two clusters being less than the preset abnormal threshold, the welding data at the welding moment is considered normal; in response to the Euclidean distance of any data point between the two clusters being not less than the preset abnormal threshold, the cluster with the least data points in the two clusters is recorded as the abnormal cluster, the welding data at the welding moment is considered abnormal, and an early warning prompt is issued.

[0014] The present invention has the following beneficial effects: By calculating the welding parameters and the amplitude characteristic values, the multi-faceted correlations and differences between the welding moment and the other moments in the welding period are fully considered, and the fluctuation characteristics of the welding data can be captured, so as to accurately distinguish between normal and abnormal small changes; by calculating the change difference value and weighting it to obtain the welding monitoring value, and then determining the upper limit monitoring value and the lower limit monitoring value, and finally calculating the abnormal offset value and using the K-means clustering algorithm, not only can the abnormal welding data be accurately identified, but also the abnormal conditions of the welding equipment can be discovered in time, thereby providing reliable guarantees for the performance and safety of subsequent welding products; it helps to improve the quality stability of welding production, reduce the scrap rate, and reduce the risk of safety accidents caused by welding quality problems. In summary, the welding process parameter monitoring and control method of the present invention can effectively solve the problem that the K-means clustering algorithm is difficult to distinguish between normal and abnormal small changes in welding data, and greatly improves the accuracy and reliability of the welding process parameter monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals are the same or corresponding parts, wherein: Figure 1 The present invention is a flow chart of the steps of a method for monitoring and controlling welding process parameters according to an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0017] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0018] See also Figure 1 , which shows a flow chart of a method for monitoring and controlling welding process parameters provided by an embodiment of the present invention, the method comprising the following steps: S1: Real-time collection of welding data of welding equipment at the time of welding.

[0019] The welding data of the welding equipment is collected through the digital screen display of the welding equipment itself or the corresponding sensors, including: current: welding current is the main process parameter of arc welding, which affects the welding quality and speed; voltage: welding voltage and welding current are the basis for achieving good weld formation; speed: the adjustment of welding speed affects the formation of the weld and the size of the heat-affected zone. The welding moment refers to the current moment in the welding operation process.

[0020] S2: Calculate welding parameters at the welding time.

[0021] It should be noted that the current, voltage and speed in the welding parameters are the main energy parameters that determine the weld size and quality of the final welding product, and there is a mutual influence between them. Current: The increase of welding current will lead to an increase in the penetration depth and residual height of the weld, while the weld width will not change much or increase slightly. This is because after the current increases, the arc force and heat input on the workpiece will increase, and the heat source position will move downward, resulting in an increase in penetration depth. The penetration depth is nearly proportional to the welding current; at the same time, after the current increases, the amount of wire melting increases. Since the weld width does not change much, the residual height increases. Voltage: The increase of welding voltage will lead to an increase in arc power, an increase in workpiece heat input, and at the same time, the arc length will be elongated, and the distribution radius will increase, resulting in a slight decrease in penetration depth and an increase in weld width; the residual height decreases, because the weld width increases, but the amount of wire melting decreases slightly. Speed: The increase in welding speed will lead to a decrease in energy, a decrease in penetration depth and weld width, and a decrease in residual height; the amount of wire metal deposited on a unit length weld is inversely proportional to the welding speed, and the weld width is nearly inversely proportional to the square root of the welding speed. These three parameters interact with each other and jointly determine the heat input and weld formation during welding. For example, if the welding speed increases, in order to maintain the size and quality of the weld, it may be necessary to increase the welding current and arc voltage to compensate for the reduction in heat input; on the contrary, if the welding current or voltage is too high, it may cause the weld to burn through or undercut, while if it is too low, it may cause defects such as incomplete penetration and slag inclusion. Therefore, in the actual welding process, it is necessary to comprehensively consider the adjustment of these three parameters according to the specific welding conditions and requirements to obtain the best welding effect. Because the data volume of the three parameters is large, in order to facilitate the subsequent data analysis, it is necessary to reduce the data dimension to reduce the subsequent calculation amount, while maintaining the data characteristics so that the reduced-dimensional data can continue to reflect the characteristics of the data.

[0022] Based on the welding data at the welding time, the welding parameters at the welding time are calculated, and the welding parameters satisfy the following relationship: ; In the formula, is the welding parameter at the welding time, is the current at the time of welding, is the voltage at the time of welding, is the speed at the time of welding, is the normalization function.

[0023] S3: Calculate the amplitude characteristic value at the welding moment.

[0024] It should be noted that it is difficult to accurately determine whether the welding data at the moment of welding is normal fluctuation or abnormal data only by relying on the welding data at the moment of welding and the historical moment; because the welding process is affected by multiple factors, such as different positions on the surface of the welding product, changes in ambient temperature and humidity, wear of equipment components, etc., even under similar conditions, the current data may deviate greatly from the historical data, and the nature of the deviation cannot be determined based on this alone. Therefore, the welding data at the future moment is obtained through the prediction algorithm, and the data is combined with the predicted data to make the subsequent calculation more accurate.

[0025] Taking the welding time as the center, multiple times including the welding time are recorded as welding time periods.

[0026] Specifically, the multiple moments include several historical moments and several future moments of the welding moment; According to the welding data at the welding moment, a linear regression algorithm is used to make predictions to obtain welding data at several future moments, and then obtain welding parameters at several future moments.

[0027] It should be noted that because the data of the welding equipment is continuous over time, under normal circumstances, the operation of the welding equipment is stable. Whether it is adjusted during operation or in continuous stable operation, the data generated by the welding equipment in continuous time are relatively coherent, so the difference in continuous time data will not be too large. Therefore, the amplitude characteristic value at the welding moment is calculated for subsequent judgment.

[0028] Based on the difference between the welding parameters at the welding moment and the other moments in the welding period, and the sorting interval between the welding data at the welding moment and the other moments in the welding period after sorting by size, the amplitude characteristic value at the welding moment is calculated, and the amplitude characteristic value satisfies the following relationship: ; In the formula, is the amplitude characteristic value at the welding moment, is the number of moments in the welding period, is the welding parameter at the welding time, The first The welding parameters at each moment, The welding time and the welding period The welding data at the moment The sorting interval of group data, is the normalization function, is the absolute value symbol.

[0029] in, Indicates the amplitude characteristic value at the welding moment. The larger the value, the greater the change of the welding equipment state at the welding moment relative to the other moments in the welding period. It is used to reflect the amplitude performance of the welding machine state in a short time. Reflected in the welding time and welding period The change amplitude of each moment is used to reflect the data difference. When the difference is larger, it means that the difference between the welding moment and the other moments in the welding period in terms of welding parameters is more significant. After comprehensive calculation with the sorting interval, the amplitude characteristic value will increase, indicating that the data fluctuation may be large and more likely to deviate from the normal fluctuation range; for example, if the welding parameters at the welding moment suddenly increase or decrease a lot compared with the other moments, this part of the difference will have a greater positive driving effect on the amplitude characteristic value, increasing the possibility of the data being judged as abnormal fluctuation; It reflects the relative position difference between the welding data at the welding moment and the rest of the welding time period. The larger sorting interval indicates that the relative position changes significantly. Even if the welding parameters are not much different, the existence of the sorting interval will affect the amplitude characteristic value, causing the amplitude characteristic value to rise, indicating that the data has a large change in sorting and there may be abnormal conditions. For example, in the sorting of current data, if the current value at a certain moment is close to the value at other moments, but the sorting has a large jump, the influence of the sorting interval will be reflected in the amplitude characteristic value, which helps to judge the particularity of data fluctuations.

[0030] S4: Calculate the change difference value of the welding time.

[0031] It should be noted that there are many different stages in the operation of welding equipment, so the judgment of data anomalies requires different standards for different stages of the operation process. This can make the standards more accurate, ensure that the data can be accurately detected, and detect abnormalities of welding equipment in time during operation.

[0032] Based on the difference between the amplitude characteristic values ​​at the welding moment and the remaining moments in the welding period, the change difference value at the welding moment is calculated.

[0033] Specifically, the change difference value satisfies the following relationship: ; In the formula, is the difference value of welding time, is the amplitude characteristic value at the welding moment, and are the number of historical moments and future moments of welding moments within the welding period, The number of historical moments of welding time in the welding period The amplitude characteristic value at the moment, is the number of future moments of welding time in the welding period The amplitude characteristic value at each moment is is the absolute value symbol.

[0034] in, and The two parameters directly reflect the degree of difference in the amplitude characteristic value between the welding moment and the rest of the welding period. The greater the difference between the amplitude characteristic value at the welding moment and the amplitude characteristic value at the historical moment or the future moment, the greater the absolute value of the difference. For example, if the amplitude characteristic value at the welding moment suddenly increases, while the amplitude characteristic value during the welding period is relatively stable, the change difference value will increase. This increase in difference means that the data fluctuation at the welding moment has changed significantly compared with the rest of the moments in the welding period, and is more likely to be a signal of abnormal data.

[0035] S5: Determine the welding monitoring value at the welding moment.

[0036] It should be noted that the variation difference value reflects the degree of change in the amplitude characteristic value at the welding moment relative to the rest of the welding period, and its size characterizes the degree of suspicion that the data fluctuation deviates from the normal state. When the variation difference value is large, it means that the possibility of abnormal data fluctuation at the welding moment is high, and voltage is a key parameter in the welding process. Its value itself has an important influence on the welding quality and the operating status of the equipment. By multiplying the variation difference value with the voltage as a weight, the voltage data is given a dynamic importance adjustment during the evaluation.

[0037] The voltage at the welding moment is weighted by using the change difference value as a weight to obtain the welding monitoring value at the welding moment.

[0038] Specifically, the welding monitoring value satisfies the following relationship: ; In the formula, is the welding monitoring value at the welding time, is the difference value of welding time, is the voltage at the time of welding.

[0039] in, The larger the value is, the voltage data will be amplified when calculating the welding monitoring value, thereby highlighting the influence of voltage parameters when the data fluctuates abnormally at the welding moment, so that it can be more accurately judged whether the welding process is abnormal and the degree of the abnormality based on the welding monitoring value in the future, providing a more targeted and effective data basis for the entire welding process parameter monitoring and control system, and timely discovering possible problems with welding equipment and taking corresponding measures.

[0040] S6: Calculate the abnormal deviation value at the welding moment.

[0041] It should be noted that the abnormality of the welding moment data relative to the data at other moments in the welding period is quantified by comparing the upper and lower limit differences of the welding monitoring values ​​at the welding moment and the rest of the welding period. First, the maximum and minimum values ​​of the welding monitoring values ​​at all moments in the welding period are determined as the upper and lower limit monitoring values ​​of the welding moment. These two values ​​define the range of the monitoring values ​​at the welding moment in the welding period; then the sum of the differences between the welding moment and the upper and lower limit monitoring values ​​at other moments in the welding period is calculated to obtain the abnormal offset value. The larger the value, the more the data at the welding moment deviates from the data distribution range of the rest of the moments in the entire welding period, and the more likely it is that there is an abnormal situation.

[0042] The maximum and minimum values ​​of the welding monitoring values ​​at all times in the welding period are recorded as the upper and lower monitoring values ​​of the welding moment; based on the differences between the upper and lower monitoring values ​​of the welding moment and the rest of the moments in the welding period, the abnormal deviation value of the welding moment is calculated.

[0043] Specifically, the abnormal offset value satisfies the following relationship: ; In the formula, is the abnormal deviation value at the welding moment, is the number of moments in the welding period, and They are the upper and lower limit monitoring values ​​at the welding moment, and The welding period is The upper and lower limit monitoring values ​​at each moment, is the absolute value symbol.

[0044] in, The larger it is, the larger the abnormal offset value is, because if the upper limit monitoring value at the welding moment is much higher than that at other moments, and the lower limit monitoring value is much lower than that at other moments, it means that the welding monitoring value range at the welding moment deviates greatly from that at other moments, and the possibility of data abnormality is higher.

[0045] S7: Based on the abnormal offset values ​​of all historical moments and welding moments, the K-means clustering algorithm is used to determine whether the welding data at the welding moment is abnormal, so as to realize the monitoring and control of the welding process parameters.

[0046] Specifically, the step of determining whether the welding data at the welding moment is abnormal includes: Using the K-means clustering algorithm, the abnormal offset values ​​of all historical moments and welding moments are clustered into two categories, and two clusters are obtained; In response to the Euclidean distance of any data point between two clusters being less than a preset abnormal threshold, the welding data at the welding moment is determined to be normal; In response to the Euclidean distance of any data point between two clusters being not less than a preset abnormal threshold, the cluster with the least data points in the two clusters is recorded as an abnormal cluster, the welding data at the welding moment is determined to be abnormal, and an early warning prompt is issued.

[0047] Through the above steps, the monitoring and control of welding process parameters are completed. In the real-time monitoring process, it is determined whether there is any abnormality at the welding moment, and problems are discovered in time, which effectively improves the operating efficiency and safety of welding equipment, reduces downtime, extends equipment life, and ensures the stability of welding equipment, thereby ensuring the performance and safety of subsequent welding products.

[0048] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring and controlling welding process parameters, characterized in that: include: Collect welding data of welding equipment at the time of welding in real time, the welding data includes current, voltage and speed; calculate welding parameters at the time of welding based on the welding data at the time of welding , is the current at the time of welding, is the voltage at the time of welding, is the speed at the time of welding, is the normalization function; Taking the welding moment as the center, multiple moments including the welding moment are recorded as welding time periods; based on the difference in welding parameters between the welding moment and the other moments in the welding time period, and the sorting interval between the welding moment and the other moments in the welding time period after sorting by size, the amplitude characteristic value of the welding moment is calculated. , is the number of moments in the welding period, The first The welding parameters at each moment, The welding time and the welding period The welding data at the moment The sorting interval of group data, is an absolute value symbol; based on the difference between the amplitude characteristic values ​​at the welding moment and the other moments in the welding period, the change difference value at the welding moment is calculated; the voltage at the welding moment is weighted by using the change difference value as a weight to obtain the welding monitoring value at the welding moment; The maximum value and the minimum value of the welding monitoring values ​​at all times in the welding period are recorded as the upper limit monitoring value and the lower limit monitoring value of the welding time; based on the difference between the upper limit monitoring value and the lower limit monitoring value between the welding time and the other times in the welding period, the abnormal deviation value of the welding time is calculated; According to the abnormal deviation values ​​of all historical moments and welding moments, the K-means clustering algorithm is used to judge whether the welding data at the welding moment is abnormal, so as to realize the monitoring and control of welding process parameters.

2. A method for monitoring and controlling welding process parameters according to claim 1, characterized in that: The multiple moments include several historical moments and several future moments of the welding moment; According to the welding data at the welding moment, a linear regression algorithm is used to make predictions to obtain welding data at several future moments, and then obtain welding parameters at several future moments.

3. The method for monitoring and controlling welding process parameters according to claim 1, characterized in that: The change difference value satisfies the following relationship: ； In the formula, is the difference value of welding time, and are the number of historical moments and future moments of welding moments within the welding period, The number of historical moments of welding time in the welding period The amplitude characteristic value at the moment, is the number of future moments of welding time in the welding period The amplitude eigenvalue at a certain moment.

4. The method for monitoring and controlling welding process parameters according to claim 1, characterized in that: The welding monitoring value satisfies the following relationship: ； In the formula, is the welding monitoring value at the welding time, is the difference value of welding time.

5. The method for monitoring and controlling welding process parameters according to claim 1, characterized in that: The abnormal offset value satisfies the following relationship: ； In the formula, is the abnormal deviation value at the welding moment, and They are the upper and lower limit monitoring values ​​at the welding moment, and They are respectively The upper and lower monitoring limits at a given moment.

6. A method for monitoring and controlling welding process parameters according to claim 1, characterized in that: The step of judging whether the welding data at the welding moment is abnormal includes: Using the K-means clustering algorithm, the abnormal offset values ​​of all historical moments and welding moments are clustered into two categories, and two clusters are obtained; In response to the Euclidean distance of any data point between two clusters being less than a preset abnormal threshold, the welding data at the welding moment is determined to be normal; In response to the Euclidean distance of any data point between two clusters being not less than a preset abnormal threshold, the cluster with the least data points in the two clusters is recorded as an abnormal cluster, the welding data at the welding moment is determined to be abnormal, and an early warning prompt is issued.

Images