A method for monitoring and identifying the welding process in a welding furnace

Abstract

This application relates to the field of intelligent monitoring technology for welding furnaces, and discloses a method for monitoring and identifying the welding process in a welding furnace. The method includes: collecting welding furnace data and welding data, wherein the welding data includes real-time welding data and historical welding data; obtaining a welding instantaneous quality factor based on the real-time welding data, and obtaining an effective welding quality factor based on the historical welding data and the welding instantaneous quality factor; obtaining an equipment degradation factor based on the welding furnace data; evaluating the welding process quality based on the equipment degradation factor and the effective welding quality factor; and monitoring and managing the welding furnace based on the quality evaluation results. This invention, by acquiring real-time welding data and simultaneously adjusting the welding instantaneous quality factor obtained from the real-time welding data using historical welding data, considers the thermal inertia of the welding furnace, sensor delay, and transmission lag factors. Through the obtained effective welding quality factor, it can accurately judge the actual welding quality.

Description

Technical Field

[0001] This application relates to the field of intelligent monitoring technology for welding furnaces, and in particular to a method for monitoring and identifying the welding process in a welding furnace. Background Technology

[0002] Soldering ovens are core equipment in surface mount technology (SMT) and power device packaging. Precise control of their process window directly determines the solder joint quality and product reliability.

[0003] Traditional welding quality assurance mainly relies on offline sampling inspections to determine whether the process meets the peak temperature, liquidus time, and other indicators recommended by the solder paste supplier; or on automated optical inspection (AOI) or X-ray inspection to detect defects that have occurred. However, the above methods cannot achieve real-time monitoring of the process, and there is a lag in the judgment of quality.

[0004] Existing technologies include schemes for real-time monitoring of key parameters during the welding process. However, welding furnaces exhibit significant thermal inertia, sensor delay, and transmission lag. Most existing monitoring methods rely on static threshold judgments, neglecting dynamic responses over time. When furnace temperature or chain speed fluctuates instantaneously, static models cannot accurately predict the timing and extent of its impact on actual welding quality. Furthermore, welding quality depends not only on the temperature profile but also on the coupled influence of multiple factors such as oxygen concentration, humidity, vibration, pressure (for pressure welding furnaces), and equipment aging. Existing technologies typically set upper and lower limits for each parameter independently, failing to quantitatively assess their interaction effects and overall risks. Therefore, the fundamental problem this invention aims to solve is how to integrate multi-source sensor information and consider thermal inertia and delay dynamics to achieve accurate monitoring of the welding process. Summary of the Invention

[0005] To achieve accurate monitoring of the welding process by fusing information from multiple sensors and considering thermal inertia and delay dynamics, this application provides a welding process monitoring and identification method for welding furnaces, employing the following technical solution:

[0006] A method for monitoring and identifying the welding process in a welding furnace, comprising:

[0007] Collect welding furnace data and welding data, including real-time welding data and historical welding data;

[0008] The instantaneous welding quality factor is obtained based on real-time welding data, and the effective welding quality factor is obtained based on historical welding data and the instantaneous welding quality factor.

[0009] Based on welding furnace data, equipment degradation factors are obtained. The welding process is evaluated based on the equipment degradation factors and the effective welding quality factors. The welding furnace is then monitored and managed based on the evaluation results.

[0010] Optionally, the real-time welding data includes several process parameters;

[0011] The process of obtaining the instantaneous welding quality factor includes:

[0012] Compare a single process parameter with its corresponding boundary to obtain the single-parameter compatibility.

[0013] The heat loss term is obtained based on temperature-related process parameters;

[0014] Obtain oxidation parameters based on environmentally relevant process parameters;

[0015] The stability term is obtained based on vibration-related process parameters;

[0016] The instantaneous welding quality factor is obtained based on the weighted cumulative value of single-parameter compatibility, heat loss term, oxidation term, and stability term.

[0017] Optionally, the process of obtaining the effective welding quality factor includes:

[0018] Acquire the sampling period, sensor response constant, and sampling delay of welding data;

[0019] The forgetting factor is set according to the sampling period and the sensor response constant;

[0020] The effective welding quality factor for the previous cycle is obtained based on historical welding data.

[0021] The instantaneous quality factor of welding is adjusted by sampling delay to obtain the adjustment term;

[0022] The effective welding quality factor is obtained by using the forgetting factor, the effective welding quality factor in the previous period, and adjustment terms.

[0023] Optionally, the welding furnace data includes the time since the last maintenance and the maintenance cycle;

[0024] The process of obtaining the equipment degradation factor includes:

[0025] The maximum degradation rate and equipment aging constant were obtained by fitting historical data.

[0026] The equipment degradation factor is defined based on the time since the last maintenance, the maintenance cycle, the maximum degree of degradation, and the equipment aging constant.

[0027] Optionally, temperature-related process parameters include peak temperature and time above the liquidus line;

[0028] The process of obtaining the heat loss term includes:

[0029] The peak temperature and the time above the liquidus line are normalized respectively. The normalized peak temperature is coupled with the liquidus line to obtain the quadratic damage index.

[0030] A thermal damage attenuation coefficient is set, and the quadratic damage index is penalized based on the thermal damage attenuation coefficient to obtain the thermal loss term.

[0031] Optionally, environmentally relevant process parameters include oxygen concentration, furnace relative humidity, and flux activation level;

[0032] The process of obtaining oxidation terms includes:

[0033] Calculate the ratio of oxygen concentration to critical oxygen concentration to obtain the oxygen concentration penalty term;

[0034] The humidity penalty is obtained based on the relative humidity of the furnace.

[0035] The flux volatile pollution penalty item is obtained based on the degree of flux activation.

[0036] The sum of the oxygen concentration penalty, humidity penalty, and flux volatile pollution penalty is obtained and then subjected to exponential decay to obtain the oxidation term.

[0037] Optionally, vibration-related process parameters include vibration intensity, coefficient of variation of chain speed sequence, and pressure fluctuation rate;

[0038] The process of obtaining stability terms includes:

[0039] By setting sensitivity coefficients, the coefficients of variation of vibration intensity, chain speed sequence and pressure fluctuation rate are multiplied by the corresponding sensitivity coefficients and summed. After exponential decay, the stability term is obtained.

[0040] Optionally, the process of conducting a quality assessment includes:

[0041] The product of the equipment degradation factor and the effective welding quality factor is used as the comprehensive score, and multi-level early warnings are implemented based on the range of the comprehensive score values.

[0042] The monitoring and management process includes:

[0043] The contribution of each process parameter to the overall score is calculated based on the single-parameter compatibility, and the parameters are managed according to their contribution to the overall score.

[0044] Optionally, the temperature-related process parameter acquisition process includes:

[0045] Acquire raw infrared thermal image data, preprocess the raw infrared thermal image data, and obtain a three-dimensional temperature matrix;

[0046] The pixel peak temperature and the time above the liquidus line are extracted from the three-dimensional matrix, and the pixel confidence score is calculated. When the pixel confidence score meets the preset conditions, the pixel is regarded as a confidence pixel.

[0047] Obtain the set of peak temperatures and the set of times above the liquidus line for all confidence pixels in each region of interest, and then take the risk quantiles of the set of peak temperatures and the set of times above the liquidus line, respectively.

[0048] The global peak temperature and global time above the liquidus are obtained by using the risk set quantiles of the peak temperature set and the time above the liquidus in all regions of interest, respectively.

[0049] Calculate the corresponding non-uniformity index based on the peak temperature set and the time above the liquidus line set of the confidence pixels, and calculate the penalty increment for the peak temperature and the time above the liquidus line based on the non-uniformity index.

[0050] The sum of the global peak temperature and the penalty increment of the peak temperature is used as the collected peak temperature;

[0051] The sum of the global time above the liquidus line and the penalty increment for the time above the liquidus line is used as the collected time above the liquidus line.

[0052] Optionally, the process for acquiring vibration-related process parameters includes:

[0053] The vibration intensity is obtained by acquiring triaxial acceleration signals through a triaxial accelerometer, taking the square root of the sum of the squares of the triaxial acceleration signals;

[0054] The chain speed measurement sequence is acquired in real time within the sliding time window, and the ratio of the standard deviation to the mean of the chain speed measurement is used as the coefficient of variation of the chain speed sequence.

[0055] Pressure signal sequences are acquired in real time within a sliding time window, and the ratio of the standard deviation to the mean of the pressure signal is used as the pressure volatility.

[0056] In summary, this application includes at least one of the following beneficial technical effects:

[0057] This invention acquires real-time welding data and adjusts the instantaneous welding quality factor obtained from the real-time welding data using historical welding data. This takes into account the thermal inertia of the welding furnace, sensor delay, and transmission lag. By acquiring the effective welding quality factor, the actual welding quality can be accurately judged. At the same time, by taking into account the equipment degradation factors of the welding furnace itself, the intelligent management process of the welding furnace can be achieved through the quality evaluation results. Attached Figure Description

[0058] Figure 1 This is a flowchart of the welding process monitoring and identification method for welding furnaces in this invention.

[0059] Figure 2 This is a flowchart illustrating the process of obtaining the instantaneous welding quality factor in this invention.

[0060] Figure 3 This is a flowchart illustrating the process of obtaining the effective welding quality factor in this invention.

[0061] Figure 4 This is a flowchart illustrating the process of obtaining the equipment degradation factor in this invention. Detailed Implementation

[0062] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0063] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0064] Please see Figure 1 This application discloses a welding process monitoring and identification method for a welding furnace, comprising:

[0065] 01: Collect welding furnace data and welding data, including real-time welding data and historical welding data;

[0066] 02: Obtain the instantaneous welding quality factor based on real-time welding data, and obtain the effective welding quality factor based on historical welding data and the instantaneous welding quality factor;

[0067] 03: Obtain equipment degradation factors based on welding furnace data, evaluate the welding process based on equipment degradation factors and effective welding quality factors, and monitor and manage the welding furnace based on the quality evaluation results.

[0068] The welding furnaces in this embodiment include, but are not limited to, reflow ovens, vacuum eutectic furnaces, hot press furnaces, wave soldering furnaces, and laser welding furnaces. The welding data collected will differ for different welding furnaces.

[0069] This embodiment acquires real-time welding data and adjusts the instantaneous welding quality factor obtained from the real-time welding data using historical welding data. This takes into account the thermal inertia of the welding furnace, sensor delay, and transmission lag. By acquiring the effective welding quality factor, the actual welding quality can be accurately judged. At the same time, by taking into account the equipment degradation factor of the welding furnace itself, the welding furnace can be intelligently managed through the quality evaluation results.

[0070] In one example, taking a hot press furnace as an example, the real-time welding data includes the following process parameters: peak temperature, time above the liquidus line, oxygen concentration, relative humidity of the furnace, flux activation degree, vibration intensity, coefficient of variation of chain speed sequence, and pressure fluctuation rate.

[0071] Please see Figure 2 The process of obtaining the instantaneous welding quality factor includes:

[0072] 021: Compare a single process parameter with its corresponding boundary to obtain the single-parameter compatibility; where the corresponding boundary is determined according to the process range set for the process parameter, and the category of the process parameter is indexed by j. The numerical values ​​of the process parameters are represented, with corresponding boundaries as follows: Obtain single-parameter compatibility ,when Value higher than or Values ​​below At that time, the single-parameter compatibility will decrease. This refers to the width parameter when the current process parameters exceed the upper limit. This refers to the width parameter when the current process parameters are below the lower limit. , Units and Similarly, the process parameters are set based on the numerical characteristics of empirical data; an exponential decay model is used for... The numerical deviation is penalized within a set range, while the normalization of single-parameter compatibility is satisfied. Therefore, single-parameter compatibility reflects the degree to which a single process parameter deviates from its corresponding boundary.

[0073] 022: Obtain the heat loss term based on temperature-related process parameters. In this example, the temperature-related process parameters are the peak temperature and the time above the liquidus line. The time above the liquidus line represents the duration during which the temperature of the solder alloy in the solder paste is higher than its liquidus line temperature from the start of melting to the end of solidification.

[0074] Peak temperature Time above the liquidus line Normalization was performed separately, and the normalized peak temperature was... , To achieve the highest ideal peak temperature for welding strength, Temperature deviation The feature scale, in this embodiment The temperature was set to 10℃; the normalized time above the liquidus was... , For the time above the ideal liquidus line. For time deviation The feature scale, in this embodiment Set to 20 seconds. and All settings are based on laboratory data or empirical data.

[0075] The normalized peak temperature is coupled with the liquidus line to obtain the quadratic damage index. , To separately control the contribution weights of peak temperature deviation, time deviation above the liquidus line, and their interaction to thermal damage, and to satisfy... ,and ; The values ​​are based on fitting of welding process physical and experimental data. The table below is a comparison table of recommended weight values ​​for different solder types.

[0076] Solder type Sn63Pb37 0.63 0.24 0.13 SAC305 0.55 0.36 0.09 SnAgCu high-temperature solder 0.47 0.47 0.06

[0077] As shown in the table above, for the Sn63Pb37 solder, the peak temperature has a significant impact on its welding thermal damage, while for the SnAgCu high-temperature solder, the peak temperature and the time above the liquidus line have a similar degree of influence.

[0078] Then, the thermal damage attenuation coefficient was set. The setting process is based on empirical data corresponding to different solder types. For Sn63Pb37 solder, its thermal damage sensitivity is high, and its corresponding thermal damage attenuation coefficient is 4.0; for SAC305 solder, its thermal damage sensitivity is moderate, and its corresponding thermal damage attenuation coefficient is 2.85; for SnAgCu high-temperature solder, it has high thermal stability, and its corresponding thermal damage attenuation coefficient is 1.90. The thermal damage attenuation coefficient is used to penalize the secondary damage index to obtain the heat loss term. .

[0079] 023: Obtain the oxidation parameters based on environmentally relevant process parameters; in this example, the environmentally relevant process parameters are oxygen concentration, furnace relative humidity, and flux activation level; the process for obtaining the oxidation parameters is as follows:

[0080] Calculate oxygen concentration With critical oxygen concentration The ratio yields the oxygen concentration penalty term. Critical oxygen concentration Set according to the requirements of the welding furnace;

[0081] The humidity penalty is obtained based on the relative humidity H in the furnace. ,in, , .

[0082] The flux volatile pollution penalty item is obtained based on the flux activation level. ; This is an empirical constant; in this embodiment, it is taken as 1.5. The ideal activation level is determined based on the type of flux.

[0083] The sum of the oxygen concentration penalty, humidity penalty, and flux volatile pollution penalty is obtained, and then subjected to exponential decay to obtain the oxidation term. .

[0084] 024: Obtain the stability term based on vibration-related process parameters; in this embodiment, vibration-related process parameters include vibration intensity, coefficient of variation of chain speed sequence, and pressure fluctuation rate;

[0085] The process of obtaining stability terms includes:

[0086] Set sensitivity coefficient The vibration intensity was respectively Coefficient of variation of chain velocity sequence and pressure volatility Sum the squares of the results by multiplying them by their respective sensitivity coefficients. After exponential decay, the stability term is obtained. .

[0087] It should be noted that the sensitivity coefficient Under ideal conditions for other parameters of the welding furnace, chain speed fluctuations of different amplitudes are artificially introduced, the weld quality (shear force, void ratio) is measured, the relationship with the fluctuation amount is fitted, and the result is obtained by reverse calculation.

[0088] 025: Obtain the instantaneous welding quality factor based on the weighted cumulative value of single-parameter compatibility, heat loss term, oxidation term, and stability term. , The weight of the j-th process parameter, The weights of the process parameters are set based on empirical data.

[0089] Please see Figure 3 In one embodiment, the process of obtaining the effective welding quality factor includes:

[0090] 026: Sampling period for acquiring welding data Sensor response constant And sampling delay d; set the forgetting factor according to the sampling period and sensor response constant. , It is the forgetting factor of the first-order inertial filter and the sensor response constant. The sampling delay d is set according to the test data.

[0091] 027: Obtain the effective welding quality factor y(t-1) for the previous cycle based on historical welding data;

[0092] 028: Adjust the instantaneous welding quality factor by sampling delay to obtain the adjustment term v(td);

[0093] 029: The effective welding quality factor is obtained through the forgetting factor, the effective welding quality factor in the previous period, and adjustment items. It should be noted that when there is no historical welding data, y(0)=1, that is, the initial condition is set to be defect-free.

[0094] The above calculation process of the effective welding quality factor can take into account the thermal inertia of the welding furnace, sensor delay, and transmission lag, thus enabling a more accurate judgment of the actual welding quality.

[0095] In one embodiment, the welding furnace data includes the time since the last maintenance, the maintenance cycle, and the equipment aging constant. ;

[0096] Please see Figure 4 The process of obtaining the equipment degradation factor includes:

[0097] 031: Obtain the maximum degradation amplitude by fitting historical data. and equipment aging constant ;

[0098] 032: Based on the time T since the last maintenance and the maintenance cycle Maximum degradation range and equipment aging constant Define equipment degradation factor .

[0099] It should be noted that the maximum degradation range ∈[0,1), which quantifies the maximum performance loss of a welding furnace caused by long-term use (such as heating wire aging, fan efficiency reduction, filter clogging, seal aging, etc.); the fitting process collects welding quality data of the same furnace at different usage times (number of days since the last maintenance) (with control process parameters unchanged), divides the actual score by the model-predicted y(t) to obtain the measured value of M(t), and estimates the maximum degradation range through formula fitting. and equipment aging constant .

[0100] By acquiring equipment degradation factors, we can more comprehensively assess the risks in the welding process.

[0101] In one embodiment, the process of conducting a quality assessment includes:

[0102] 033: The product of the equipment degradation factor and the effective welding quality factor is used as the comprehensive score. Multi-level early warning is implemented based on the range of the comprehensive score value. The higher the comprehensive score value, the closer the process parameters and welding furnace status are to the ideal state. Conversely, the lower the comprehensive score value, the greater the risk of the welding process. Through the multi-level early warning method, management personnel can be provided with reference handling measures to promptly investigate and repair welding furnaces with high risk levels.

[0103] In one embodiment, the monitoring and management process includes:

[0104] 034: Calculate the contribution of each process parameter to the overall score based on the single-parameter compatibility. Management is carried out based on the deduction contribution of process parameters. Process parameters with greater deduction contributions are selected for targeted problem investigation, which can improve the efficiency of welding furnace maintenance and problem handling.

[0105] In one example, a temperature-related process parameter acquisition procedure is given, including:

[0106] Acquire raw infrared thermal image data, preprocess the raw infrared thermal image data, and obtain a three-dimensional temperature matrix. The preprocessing process includes spatial Gaussian filtering and temporal median filtering, which can eliminate shot noise and local fluctuations in emissivity and suppress random impulse noise.

[0107] Extracting pixel peak temperature from a 3D matrix and the time above the pixel liquid phase line Calculate pixel confidence , The emissivity estimate is given by [variable name] and [variable name]. , Based on on-site calibration, this embodiment... Take 0.9, Take 0.05; ; The highest temperature allowed by the process. Temperature decay scale parameter, in this embodiment Set the temperature to 10℃. When the pixel confidence level meets the preset condition (C(x,y)>0.5), the pixel is taken as the confidence pixel.

[0108] Obtain the set of peak temperatures for all confidence pixels in each region of interest. and time sets above the liquidus Risk quantiles were taken for the peak temperature set and the time set above the liquidus line, respectively. In this embodiment, the 95th quantile was taken for the peak temperature set. For the time set above the liquidus line, taking the 90th quantile yields... .

[0109] The global peak temperature is obtained by combining the set of peak temperatures across all regions of interest with the risk quantiles of the time above the liquidus line. and time above the global liquidus , where m is the number of pixels.

[0110] The corresponding non-uniformity indexes are calculated based on the peak temperature set and the time set above the liquidus line of the confidence pixels, respectively. The non-uniformity index corresponding to the peak temperature set is... For all The ratio of the standard deviation to the mean, an inhomogeneity index corresponding to the time set above the liquidus line. For all The ratio of the standard deviation to the mean; calculate the penalty increment for peak temperature and time above the liquidus line based on the inhomogeneity index; penalty increment for peak temperature. The penalty increment for time above the liquidus line ;

[0111] The sum of the global peak temperature and the penalty increment of the peak temperature is used as the collected peak temperature; that is... ;

[0112] The sum of the global time above the liquidus line and the penalty increment for the time above the liquidus line is used as the collected time above the liquidus line, i.e. .

[0113] The method of collecting peak temperature and time above the liquidus line, as described above, solves the problem of missed detection of local hot spots compared to the existing technology that relies on a small number of temperature measurement points. The use of quantiles for regional statistics can amplify the impact of the worst local thermal effect and improve the model's sensitivity to quality risks. The confidence assessment process ensures the quality of the input data. By using weighted fusion and introducing a non-uniformity penalty term, the model can more realistically reflect the actual thermal damage risk.

[0114] In one example, the process of acquiring vibration-related process parameters includes:

[0115] Triaxial acceleration signals are acquired using a triaxial accelerometer. The vibration intensity is obtained by taking the square root of the average of the sum of the squares of the three-axis acceleration signals; that is... .

[0116] The chain speed measurement sequence is acquired in real time within the sliding time window, and the ratio of the standard deviation to the mean of the chain speed measurement is used as the coefficient of variation of the chain speed sequence.

[0117] Pressure signal sequences are acquired in real time within a sliding time window, and the ratio of the standard deviation to the mean of the pressure signal is used as the pressure volatility.

[0118] By collecting the aforementioned vibration-related process parameters, the impact of welding furnace mechanical vibration, conveyor chain speed fluctuations, and pressure instability on weld formation can be quantified, and the robustness of the quality evaluation process can be improved in conjunction with other monitoring parameters.

[0119] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for monitoring and identifying the welding process in a welding furnace, characterized in that, include: Collect welding furnace data and welding data, including real-time welding data and historical welding data; The instantaneous welding quality factor is obtained based on real-time welding data, and the effective welding quality factor is obtained based on historical welding data and the instantaneous welding quality factor. Based on welding furnace data, equipment degradation factors are obtained. The welding process is evaluated based on the equipment degradation factors and the effective welding quality factors. The welding furnace is then monitored and managed based on the evaluation results.

2. The welding process monitoring and identification method for a welding furnace according to claim 1, characterized in that, The real-time welding data includes several process parameters; The process of obtaining the instantaneous welding quality factor includes: Compare a single process parameter with its corresponding boundary to obtain the single-parameter compatibility. The heat loss term is obtained based on temperature-related process parameters; Obtain oxidation parameters based on environmentally relevant process parameters; The stability term is obtained based on vibration-related process parameters; The instantaneous welding quality factor is obtained based on the weighted cumulative value of single-parameter compatibility, heat loss term, oxidation term, and stability term.

3. The welding process monitoring and identification method for a welding furnace according to claim 1, characterized in that, The process of obtaining the effective welding quality factor includes: Acquire the sampling period, sensor response constant, and sampling delay of welding data; The forgetting factor is set according to the sampling period and the sensor response constant; The effective welding quality factor for the previous cycle is obtained based on historical welding data. The instantaneous quality factor of welding is adjusted by sampling delay to obtain the adjustment term; The effective welding quality factor is obtained by using the forgetting factor, the effective welding quality factor in the previous period, and adjustment terms.

4. The welding process monitoring and identification method for a welding furnace according to claim 1, characterized in that, The welding furnace data includes the time since the last maintenance and the maintenance cycle. The process of obtaining the equipment degradation factor includes: The maximum degradation rate and equipment aging constant were obtained by fitting historical data. The equipment degradation factor is defined based on the time since the last maintenance, the maintenance cycle, the maximum degree of degradation, and the equipment aging constant.

5. The welding process monitoring and identification method for a welding furnace according to claim 2, characterized in that, Temperature-related process parameters include peak temperature and time above the liquidus line; The process of obtaining the heat loss term includes: The peak temperature and the time above the liquidus line are normalized respectively. The normalized peak temperature is coupled with the liquidus line to obtain the quadratic damage index. A thermal damage attenuation coefficient is set, and the quadratic damage index is penalized based on the thermal damage attenuation coefficient to obtain the thermal loss term.

6. The welding process monitoring and identification method for a welding furnace according to claim 2, characterized in that, Environmentally relevant process parameters include oxygen concentration, furnace relative humidity, and flux activation level; The process of obtaining oxidation terms includes: Calculate the ratio of oxygen concentration to critical oxygen concentration to obtain the oxygen concentration penalty term; The humidity penalty is obtained based on the relative humidity of the furnace. The flux volatile pollution penalty item is obtained based on the degree of flux activation. The sum of the oxygen concentration penalty, humidity penalty, and flux volatile pollution penalty is obtained and then subjected to exponential decay to obtain the oxidation term.

7. The welding process monitoring and identification method for a welding furnace according to claim 2, characterized in that, Vibration-related process parameters include vibration intensity, coefficient of variation of chain speed sequence, and pressure fluctuation rate; The process of obtaining stability terms includes: By setting sensitivity coefficients, the coefficients of variation of vibration intensity, chain speed sequence and pressure fluctuation rate are multiplied by the corresponding sensitivity coefficients and summed. After exponential decay, the stability term is obtained.

8. The welding process monitoring and identification method for a welding furnace according to claim 2, characterized in that, The process of conducting quality evaluation includes: The product of the equipment degradation factor and the effective welding quality factor is used as the comprehensive score, and multi-level early warnings are implemented based on the range of the comprehensive score values. The monitoring and management process includes: The contribution of each process parameter to the overall score is calculated based on the single-parameter compatibility, and the parameters are managed according to their contribution to the overall score.

9. The welding process monitoring and identification method for a welding furnace according to claim 5, characterized in that, The process of acquiring temperature-related process parameters includes: Acquire raw infrared thermal image data, preprocess the raw infrared thermal image data, and obtain a three-dimensional temperature matrix; The pixel peak temperature and the time above the liquidus line are extracted from the three-dimensional matrix, and the pixel confidence score is calculated. When the pixel confidence score meets the preset conditions, the pixel is regarded as a confidence pixel. Obtain the set of peak temperatures and the set of times above the liquidus line for all confidence pixels in each region of interest, and then take the risk quantiles of the set of peak temperatures and the set of times above the liquidus line, respectively. The global peak temperature and global time above the liquidus are obtained by using the risk set quantiles of the peak temperature set and the time above the liquidus in all regions of interest, respectively. Calculate the corresponding non-uniformity index based on the peak temperature set and the time above the liquidus line set of the confidence pixels, and calculate the penalty increment for the peak temperature and the time above the liquidus line based on the non-uniformity index. The sum of the global peak temperature and the penalty increment of the peak temperature is used as the collected peak temperature; The sum of the global time above the liquidus line and the penalty increment for the time above the liquidus line is used as the collected time above the liquidus line.

10. The welding process monitoring and identification method for a welding furnace according to claim 7, characterized in that, The process of acquiring vibration-related process parameters includes: The vibration intensity is obtained by acquiring triaxial acceleration signals through a triaxial accelerometer, taking the square root of the sum of the squares of the triaxial acceleration signals; The chain speed measurement sequence is acquired in real time within the sliding time window, and the ratio of the standard deviation to the mean of the chain speed measurement is used as the coefficient of variation of the chain speed sequence. Pressure signal sequences are acquired in real time within a sliding time window, and the ratio of the standard deviation to the mean of the pressure signal is used as the pressure volatility.

Images