Wave soldering intelligent optimization control system and control method

Abstract

The invention discloses a wave-soldering intelligent optimization control system and method. The system comprises a wave-soldering module, an optical detection module and an intelligent optimization control module. The wave soldering module is used for welding the PCBA according to the welding parameters and outputting product identification information; the optical detection module detects the welding quality according to the identification information and outputs corresponding soldering tin quality data; the intelligent optimization control module receives the welding parameters, the product identification information and the corresponding soldering tin quality data, and selectively controls the wave soldering module according to whether the quality data meets a preset target or not; and if not, starting a parameter optimization process, performing statistical analysis based on the associated data, determining an optimal welding parameter combination, and updating operation parameters of the wave-soldering module. According to the method, the influence of the welding parameters on the soldering tin quality is accurately quantified through multi-dimensional statistical analysis, the dependence on personnel experience is reduced, subjective errors are avoided, and the efficiency and accuracy of parameter optimization are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of electronic manufacturing technology, and in particular to an intelligent optimization control system and control method for wave soldering. Background Technology

[0002] Wave soldering is a critical step in the electronic assembly process, and its soldering quality directly determines the reliability and performance of the final product. Currently, the mainstream wave soldering process parameter optimization methods in the industry heavily rely on engineers' personal experience and offline experiments. However, after long-term operation, it has been found that this method has many drawbacks: Firstly, the initial process parameters are usually derived from the historical settings of similar models. Engineers then manually fine-tune these parameters based on their personal experience and search for relatively optimal parameter combinations through offline experimental design. This method makes it difficult to fully consider the interactive effects between multiple parameters, and the optimization process is blind and inefficient.

[0003] Secondly, the entire offline DOE process, from experimental design and execution to data analysis, usually takes tens of hours or even longer. During this period, the production line needs to stop normal production, resulting in huge time costs and wasted capacity. In addition, the experiment itself also requires a lot of consumables, leading to high economic costs.

[0004] In addition, existing methods cannot monitor and adjust in real time, resulting in high quality risks. Based on the above shortcomings, existing wave soldering process parameter optimization methods cannot meet the needs of enterprises. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art by providing an intelligent optimization control system and control method for wave soldering, which can automatically and in real time optimize wave soldering parameters, thereby enabling products to achieve the set solder quality and achieve a high product yield.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: an intelligent optimization control system for wave soldering, comprising: The wave soldering module is configured to perform soldering operations on PCBAs according to soldering parameters and output product identification information for each PCBA. An optical inspection module, which is communicatively connected to the wave soldering module, is configured to receive the product identification information, perform quality inspection on the soldered PCBA, and output solder quality data associated with the product identification information. The intelligent optimization control module is communicatively connected to the wave soldering module and the optical detection module, and is configured as follows: Receive soldering parameters and product identification information from the wave soldering module, and solder quality data associated with the product identification information from the optical inspection module; The wave soldering module is selectively controlled based on whether the solder quality data meets a preset target. When the conditions are met, the wave soldering module is controlled to maintain the current soldering parameters. If the conditions are not met, the parameter optimization process is initiated. Based on the associated welding parameters and solder quality data, statistical analysis is performed to determine the optimal combination of welding parameters, and the wave soldering module is controlled to update to the optimal combination of welding parameters.

[0007] Furthermore, the intelligent optimization control module includes a data receiving unit, a DOE experimental design unit, a multi-dimensional statistical analysis unit, and a parameter optimization unit. The data receiving unit is communicatively connected to the wave soldering module and the optical inspection module, and is configured to receive product identification information, welding parameters during welding, and related solder quality data in real time. The DOE experimental design unit is connected to the data receiving unit and is configured to generate a multi-factor orthogonal array based on the current soldering parameters in response to solder quality data falling below a preset quality target. The multidimensional statistical analysis unit is connected to the DOE experimental design unit and the data receiving unit, and is configured to perform significant factor screening, quantitative modeling and model verification on the experimental solder quality data corresponding to the experiment based on the multi-factor orthogonal array. The parameter optimization unit is connected to the multi-dimensional statistical analysis unit and is configured to calculate the optimal combination of welding parameters based on the significant factors and quantitative relationships verified by the model, and output the result to the wave soldering module to control its update of the operating welding parameters.

[0008] Furthermore, the multidimensional statistical analysis unit includes a range analysis module, a variance analysis module, a regression analysis module, and a residual analysis module; The range analysis module is configured to calculate the range of solder quality data corresponding to wave soldering parameters based on the solder quality data after the multi-factor orthogonal array experiment in order to screen candidate parameters, and transmit the results to the variance analysis module. The analysis of variance module is configured to perform significance tests on candidate parameters, screen significant factors, and transmit them to the regression analysis module; The regression analysis module is configured to construct a regression equation with significant factors as independent variables and solder quality data as dependent variables, and then transmit the regression equation to the residual analysis module. The residual analysis module is configured to calculate the residual by combining the actual observed values ​​of solder quality inspected by the optical detection module with the predicted values ​​of the regression equation. If there is no systematic deviation in the residual, the effective regression equation is output to the parameter optimization unit. If there is a deviation, the multi-dimensional statistical analysis unit is triggered to re-execute the analysis, and the statistical processing flow including range analysis, variance analysis and regression analysis is re-executed.

[0009] A control method for an intelligent optimization control system for wave soldering includes the following steps: S1: Perform soldering operations on the PCBA using the current soldering parameters, and output product identification information for each PCBA; S2: Receive the product identification information, perform quality inspection on the soldered PCBA, and output solder quality data associated with the product identification information; S3: Receive soldering parameters and product identification information from the wave soldering module, and solder quality data associated with the product identification information from the optical inspection module; selectively control the wave soldering module based on whether the solder quality data meets a preset target. S31: When the condition is met, control the wave soldering module to maintain the current welding parameters. S32: When the conditions are not met, initiate the parameter optimization process, perform statistical analysis based on the associated welding parameters and solder quality data to determine the optimal welding parameter combination, and control the wave soldering module to update to the optimal welding parameter combination.

[0010] Furthermore, the specific steps of step S32 include the following: It can receive product identification information, welding parameters during welding, and related solder quality data in real time. In response to the solder quality data not meeting the preset quality target, a multi-factor orthogonal array is generated based on the current soldering parameters; For the experimental solder quality data corresponding to the experiments conducted based on the multi-factor orthogonal array, significant factor screening, quantitative modeling, and model validation were completed. Based on the significant factors and quantitative relationships validated by the model, the optimal combination of welding parameters is calculated and output to the wave soldering module.

[0011] Furthermore, the specific steps for screening significant factors, quantitative modeling, and model validation of the experimental solder quality data corresponding to the experiments conducted based on the aforementioned multi-factor orthogonal array are as follows: Based on the solder quality data after the multi-factor orthogonal array experiment, the range of solder quality data corresponding to wave soldering parameters is calculated to screen candidate parameters. The candidate parameters are subjected to a significance test to screen significant factors; A regression equation was constructed using the aforementioned significant factors as independent variables and solder quality data as dependent variables. The residuals are calculated by combining the actual observed values ​​of solder quality inspected by the optical inspection module with the predicted values ​​of the regression equation. If there is no systematic bias in the residuals, a valid regression equation is output. If there is a systematic bias in the residuals, re-execution is triggered.

[0012] Furthermore, based on the solder quality data after the multi-factor orthogonal array experiment, the specific steps for calculating the range of solder quality data corresponding to the wave soldering parameters to screen candidate parameters are as follows: Based on range analysis, according to the formula R=x max -x min , where x max x represents the maximum quality data corresponding to a certain welding parameter. min To minimize the value, the influence of different levels of various welding parameters on solder quality indicators was calculated, and candidate welding parameters with a significant impact on solder quality were preliminarily screened.

[0013] Furthermore, the candidate parameters are subjected to a significance test, and the specific steps for screening significant factors are as follows: Based on the principles of analysis of variance, total variance is decomposed into between-group variance and within-group variance; the calculation logic of within-group variance follows the same formula as the total variance. Consistent, in which the solder quality data within each group of experiments are taken as the analysis object, xi is a single solder quality data within the group, μ is the group mean of the solder quality data in the group, N is the number of data in the group, and the experimental data are tested for significance by calculating the ratio of the mean square between groups to the mean square within groups, and the significant factor is retained. Furthermore, the specific steps for constructing a regression equation using the aforementioned significant factor as the independent variable and solder quality data as the dependent variable are as follows: Using the significant factors retained by the analysis of variance as independent variables and the solder quality index as the dependent variable, a quadratic regression model was established to obtain the regression equation. , .

[0014] Furthermore, the residuals are calculated by combining the actual observed values ​​of solder quality inspected by the optical inspection module with the predicted values ​​of the regression equation. If the residuals have no systematic bias, a valid regression equation is output. If the residuals have systematic bias, the specific steps for triggering re-execution are as follows: Based on formula Where ei is the residual and yi is the measured value of the experimental solder quality data.

[0015] For the regression model's predicted values, calculate the residuals between the predicted values ​​and the experimentally measured values. If the residuals are randomly distributed and have no systematic bias, the model is considered to be a good fit. If the residuals have bias, return to the multidimensional statistical analysis unit to re-execute the statistical analysis.

[0016] Due to the application of the above technical solution, the present invention has the following advantages compared with the prior art: This invention utilizes multi-dimensional statistical analysis methods, including the DOE experimental design unit in the intelligent optimization control module and the range analysis, variance analysis, regression analysis, and residual analysis modules in the multi-dimensional statistical analysis unit, to accurately quantify the impact of each welding parameter on solder quality. It no longer relies on the subjective experience and judgment of operators, but instead uses objective statistical data to screen significant factors, effectively avoiding experience errors and making parameter optimization more scientific and accurate, reducing the product defect rate by approximately 80%.

[0017] Secondly, the intelligent optimization control module is directly integrated into the wave soldering module, forming a tight data interaction link with the wave soldering module and the optical inspection module. It does not rely on external independent data processing devices, avoiding delays and losses in data transmission between multiple devices, greatly improving the timeliness of data flow, providing a solid foundation for real-time parameter optimization, increasing maintenance response speed by 60%, and reducing ineffective energy consumption from 50-60% of traditional equipment to below 5%.

[0018] In addition, the reduction of precise parameters improves the physical soldering quality, avoids overload damage to the solder pot, thereby increasing the equipment life by 45-50%, directly saving material and labor rework costs, and better meeting actual usage needs. Attached Figure Description

[0019] The technical solution of the present invention will be further described below with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the intelligent optimization control system for wave soldering in this invention; Figure 2 This is a flowchart illustrating the control method of the intelligent optimization control system for wave soldering in this invention. Figure 3 This is a schematic diagram of how the DOE experimental design unit in this invention generates a two-level multi-factor orthogonal array based on the current operating range of the wave soldering module's welding parameters. Figure 4 This is a schematic diagram of 13 groups in the two-level multi-factor orthogonal table in Embodiment 1 of the present invention; Figure 5 This is a flowchart of the control method of the wave soldering intelligent optimization control system in Embodiment 1 of the present invention; The module includes: wave soldering module 1, optical inspection module 2, intelligent optimization control module 3, data storage and traceability unit 4, data receiving unit 30, DOE experimental design unit 31, multi-dimensional statistical analysis unit 32, and parameter optimization unit 33. Detailed Implementation

[0020] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0021] For ease of understanding, the specific processes in the embodiments of this application are described below. Please refer to [link / reference]. Figure 1 An intelligent optimization control system for wave soldering, as described in this application embodiment, includes a wave soldering module 1, an optical inspection module 2, and an intelligent optimization control module 3.

[0022] Wave soldering module 1 is used to solder PCBAs according to set soldering parameters and output product identification information. The wave soldering module is equipped with an identification scanning module and a communication interface. The scanning module is used to obtain the unique product identification information of the PCBA, such as a QR code, and the communication interface is used to realize data interaction and parameter command transmission with the optical inspection module. Soldering parameters include at least one of the following: solder pot conveyor speed, solder pot temperature, flux flow rate, preheating temperature, and wave frequency. Optimal soldering quality is ensured by subsequently adjusting these soldering parameters.

[0023] The optical inspection module 2 is communicatively connected to the wave soldering module 1 and is configured to receive product identification information output by the wave soldering module 1. After performing quality inspection on the soldered PCBA, it outputs solder quality data associated with the product identification information, which is subsequently output to the intelligent optimization control module 3. The main functions of the optical inspection module 2 include identifying defects such as cold solder joints, bridging, and insufficient solder, and calculating the DPMO / DPPM (defects per million pieces) of each PCBA as solder quality data using image algorithms.

[0024] The intelligent optimization control module 3 is built into the wave soldering module 1. The intelligent optimization control module 3 is communicatively connected to the wave soldering module 1 and the optical detection module 2, and is configured to receive soldering parameters and product identification information from the wave soldering module 1, and solder quality data associated with the product identification information from the optical detection module 2; and selectively control the wave soldering module according to whether the solder quality data meets the preset target.

[0025] When the conditions are met, the wave soldering module 1 is controlled to maintain the current soldering parameters; when the conditions are not met, the parameter optimization process is initiated, and statistical analysis is performed based on the associated soldering parameters and solder quality data to determine the optimal combination of soldering parameters, and the wave soldering module 1 is controlled to update to the optimal combination of soldering parameters.

[0026] In addition, this control system also includes a data storage and traceability unit 4, which is communicatively connected to the intelligent optimization control module 3. It is used to store historical welding parameters, historical solder quality data and optimization process data, and can perform full-process data traceability according to product identification information.

[0027] Furthermore, the intelligent optimization control module 3 includes a data receiving unit 30, a DOE experimental design unit 31, a multi-dimensional statistical analysis unit 32, and a parameter optimization unit 33. These units are connected sequentially to form a serial data processing and analysis link. The data receiving unit 30 is communicatively connected to the wave soldering module 1 and the optical inspection module 2, receiving product identification information, soldering parameters, and solder quality data in real time. The DOE experimental design unit is connected to the data receiving unit 30, and in response to solder quality data falling below a preset target, generates a multi-factor orthogonal array based on the current soldering parameters. The multi-dimensional statistical analysis unit 32 is connected to the DOE experimental design unit 31 and the data receiving unit 30, and is used to complete significant factor screening, quantitative modeling, and model verification for the experimental solder quality data corresponding to the experiment based on the multi-factor orthogonal array. The parameter optimization unit 33 is connected to the multi-dimensional statistical analysis unit 32, establishes a model based on the screened significant factors, calculates the optimal combination of soldering parameters, and outputs it to the wave soldering module to control its updating of operating soldering parameters.

[0028] Furthermore, the multi-dimensional statistical analysis unit includes a range analysis module, a variance analysis module, a regression analysis module, and a residual analysis module. The range analysis module calculates the range of solder quality data corresponding to wave soldering parameters based on the solder quality data after the multi-factor orthogonal array experiment to screen candidate parameters and transmits the data to the variance analysis module. The variance analysis module performs significance testing on the candidate parameters, screens significant factors, and transmits the data to the regression analysis module. The regression analysis module constructs a regression equation with significant factors as independent variables and solder quality data as dependent variables and transmits the equation to the residual analysis module. The residual analysis module calculates the residuals by combining the actual observed values ​​of solder quality checked by the optical inspection module with the predicted values ​​of the regression equation. If there is no systematic bias in the residuals, the effective regression equation is output to the parameter optimization module; if there is a bias, each module is triggered to re-execute the analysis.

[0029] Furthermore, this invention also discloses a control method for an intelligent optimization control system for wave soldering, comprising the following steps: S1: Perform soldering operations on the PCBA using the current soldering parameters and output product identification information for each PCBA. Specifically, the wave soldering module scans the product identification information of the PCBA and simultaneously sends it to the optical inspection module. This identification information is a QR code or other information number. The module then solders the PCBA according to the currently set initial soldering parameters, including solder pot temperature, conveyor speed, flux flow rate, preheating temperature, and wave frequency.

[0030] S2: Receive the product identification information, perform quality inspection on the soldered PCBA, and output solder quality data associated with the product identification information.

[0031] The quality inspection here involves: image acquisition and defect identification of the soldered PCBA, calculation of the PMO value of each PCBA, and the optical inspection module transmitting the product identification and corresponding DPMO data back to the built-in data receiving unit of the wave soldering module to form a correlation record from product identification information to soldering parameters and solder quality data.

[0032] S3: Receive soldering parameters and product identification information from the wave soldering module, and solder quality data associated with the product identification information from the optical inspection module; selectively control the wave soldering module based on whether the solder quality data meets a preset target; S31: When the condition is met, control the wave soldering module to maintain the current welding parameters. S32: When the conditions are not met, initiate the parameter optimization process, perform statistical analysis based on the associated welding parameters and solder quality data to determine the optimal welding parameter combination, and control the wave soldering module to update to the optimal welding parameter combination.

[0033] Furthermore, the specific steps of step S32 are as follows: S321. Receive product identification information, welding parameters during welding, and related solder quality data; S322. In response to the solder quality data not meeting the preset quality target, a multi-factor orthogonal array is generated based on the current soldering parameters. (See reference...) Figure 3 ; S323. For the experimental solder quality data corresponding to the experiment based on the multi-factor orthogonal array, complete the screening of significant factors, quantitative modeling, and model validation; the specific steps are as follows: Based on the solder quality data after the multi-factor orthogonal array experiment, calculate the range of solder quality data corresponding to the wave soldering parameters to screen candidate parameters; perform significance testing on the candidate parameters to screen significant factors; construct a regression equation with the significant factors as independent variables and solder quality data as dependent variables; calculate the residuals by combining the actual observed values ​​of solder quality checked by the optical detection module with the predicted values ​​of the regression equation; if the residuals have no systematic bias, output the effective regression equation; if the residuals have systematic bias, trigger re-execution.

[0034] S324. Based on the significant factors and quantitative relationships validated by the model, the optimal combination of welding parameters is calculated and output to the wave soldering module.

[0035] Furthermore, in step S323, the steps based on the solder quality data after the multi-factor orthogonal array experiment are as follows: the intelligent optimization control module sends an experimental command to the wave soldering module, the wave soldering module reads the parameter combination corresponding to each group of experiments in the orthogonal array, and executes at least 19 sets of comparative experiments in sequence according to the orthogonal array sequence. Each set of experiments continuously solders ≥100 PCBAs of the same model. During the execution of each set of experiments, the optical detection module synchronously collects the solder quality data of each PCBA and transmits it back to the intelligent optimization control module in real time, forming a one-to-one correspondence between experimental data and parameter combinations.

[0036] Furthermore, the specific steps for calculating the range of solder quality data corresponding to welding parameters to screen candidate parameters are as follows: Based on range analysis, according to the formula R=x max -x min , where x max x represents the maximum quality data corresponding to a certain welding parameter. min To minimize the value, the influence of different levels of various welding parameters on solder quality indicators was calculated, and candidate welding parameters with a significant impact on solder quality were preliminarily screened.

[0037] Furthermore, in step S323, the candidate parameters are subjected to a significance test, and the specific steps for screening significant factors are as follows: Based on the principles of analysis of variance, total variance is decomposed into between-group variance and within-group variance; the calculation logic of within-group variance follows the same formula as the total variance. Consistency is achieved by taking the solder quality data within each experimental group as the analysis object, where xi is a single solder quality data point within the group, μ is the group mean of the solder quality data in that group, and N is the number of data points within that group. The experimental data are tested for significance by calculating the ratio of the mean square between groups to the mean square within groups, and significant factors are retained. Furthermore, in step S323, the specific steps for constructing a regression equation using the significant factor as the independent variable and the solder quality data as the dependent variable are as follows: Using the significant factors retained by the analysis of variance as independent variables and the solder quality index as the dependent variable, a quadratic regression model was established to obtain the regression equation. , .

[0038] Furthermore, in step S323, the residual is calculated by combining the actual observed value of solder quality checked by the optical inspection module with the predicted value of the regression equation; if the residual has no systematic deviation, the effective regression equation is output; if the residual has a systematic deviation, the specific steps for triggering re-execution are as follows. Based on formula Where ei is the residual and yi is the measured value of the experimental solder quality data.

[0039] For the regression model's predicted values, calculate the residuals between the predicted values ​​and the experimentally measured values. If the residuals are randomly distributed and have no systematic bias, the model is considered to be a good fit. If the residuals have bias, return to the multidimensional statistical analysis unit to re-execute the statistical analysis.

[0040] The following is an example illustrating a specific use case: Example

[0041] See Figure 5 A control method for an intelligent optimization control system for wave soldering includes the following steps: S1. Perform soldering operations on the PCBA using the current soldering parameters and output product identification information for each PCBA. The initial soldering parameters include: wave soldering chain feed 1300mm / min, solder pot temperature 260℃, preheating 1 temperature 130℃, preheating 2 temperature 150℃, preheating 3 temperature 180℃, upper preheating 3 temperature 90℃, wave soldering 1 frequency 16.0Hz, wave soldering 2 frequency 28.0Hz, DPMO 200, where DPMO is the number of defects per million pieces.

[0042] S2. Receive the product identification information, perform quality inspection on the soldered PCBA, and output solder quality data associated with the product identification information.

[0043] S3. Receive welding parameters and product identification information from the wave soldering module, and solder quality data associated with the product identification information from the optical inspection module; selectively control the wave soldering module based on whether the solder quality data meets a preset target.

[0044] If DPMO is less than 200, the solder quality is normal. Continue the current operation, and the wave soldering module will continue to run with the current soldering parameters.

[0045] If DPMO is greater than 200, the solder quality is determined to be abnormal, triggering the parameter optimization process, which includes the following steps: S41. Generate a multi-factor orthogonal table based on the current welding parameters. (See below) Figure 4 The intelligent optimization control module sends experimental instructions to the wave soldering module. The wave soldering module execution unit reads the parameter combination corresponding to each group of experiments in the orthogonal array and executes 19 sets of comparative experiments in sequence according to the orthogonal array. Each set of experiments continuously solders 100 PCBAs of the same model. During the execution of each set of experiments, the optical detection module synchronously collects the solder quality data of each PCBA and transmits it back to the intelligent optimization control module in real time, forming a one-to-one correspondence between experimental data and parameter combinations.

[0046] S42. Based on the solder quality data after the multi-factor orthogonal array experiment, calculate the range of solder quality data corresponding to the wave soldering parameters to screen candidate parameters; according to the formula R=x max -x min , where x max x represents the maximum quality data corresponding to a certain welding parameter. min To minimize the value, the influence of different levels of various welding parameters on solder quality indicators was calculated, and candidate welding parameters with a significant impact on solder quality were preliminarily screened.

[0047] Among them, the range of each welding parameter is as follows: transport speed = 27.95, solder pot temperature = 27.08, preheating 1 temperature = 16.80, preheating 2 temperature = 18.21, preheating 3 temperature = 22.93, upper preheating 3 temperature = 22.21, peak 1 frequency = 18.61, peak 2 frequency = 15.97.

[0048] Screening results: The range of the preheating temperature 3 (29.33) was the largest among all parameters, indicating that it had the most significant impact on DPMO, and was selected as a candidate factor for the next step of analysis of variance.

[0049] S43. Perform a significance test on the candidate parameters and screen significant factors; check whether the preheating temperature 3 is a "significant factor" to eliminate random errors.

[0050] Based on the principles of analysis of variance, total variance is decomposed into between-group variance and within-group variance; the calculation logic of within-group variance is the same as that of the overall variance formula. Consistency was maintained, with solder quality data within each experimental group as the analysis object. xi represents a single solder quality data point within the group, μ represents the group mean of solder quality data in that group, and N represents the number of data points within that group. The significance of the experimental data was tested by calculating the ratio of the between-group mean square to the within-group mean square. A screening threshold of P-value ≤ 0.05 was set, eliminating non-significant factors with P-value > 0.05 and retaining significant factors with P-value ≤ 0.05.

[0051] Using the formula for total variance The calculated P-value is 0.032, which is less than 0.05. Therefore, the preheating temperature 3 is set as a significant factor, and other welding parameters are deleted.

[0052] S44. Construct a regression equation with the significant factor as the independent variable and the solder quality data as the dependent variable.

[0053] Using the significant factors retained by the analysis of variance as independent variables and the solder quality index as the dependent variable, a quadratic regression model was established to obtain the regression equation. , ; Specifically, pairwise data for preheating temperature 3 (X5) and DPMO (y) were extracted from 19 experimental groups. The coefficients were solved using the least squares method: b0 = 303.5073, b5 = -1.0473. The final regression equation is: = 303.5073 - 1.0473 X5, calculate the coefficient of determination R2=0.88. The closer R2 is to 1, the higher the simulation fit.

[0054] S45. Combine the actual observed values ​​of solder quality inspected by the optical inspection module with the predicted values ​​of the regression equation to calculate the residuals; if there is no systematic bias in the residuals, output the valid regression equation; if there is a systematic bias in the residuals, trigger re-execution.

[0055] First, residual verification is performed based on the formula. Where ei is the residual and yi is the measured value of the experimental solder quality data. For the predicted values ​​of the regression model, calculate the residuals between the predicted values ​​of the regression model and the experimentally measured values.

[0056] Specifically, taking the fifth group of experiments as an example: X5 = 194℃, measured DPMOy i =102.1, substituting into the regression equation, we get: ; Calculate the residual e i =102.1-100.33=1.77; Residual percentage: All models meet the condition that "absolute residual value ≤ 5%" and the residuals are randomly distributed, thus the regression model is considered to be a good fit.

[0057] S324. Based on the significant factors and quantitative relationships validated by the model, the optimal combination of welding parameters is calculated and output to the wave soldering module to control its updating of operating welding parameters. Specifically, based on the regression equation =303.5073 - 1.0473 X5, according to the minimum preheating temperature required to solve for DPMO: Target Substituting ≤200 into the regression equation above, we get X5≥98.8. This result indicates that a preheating temperature of not less than 98.8℃ is sufficient to meet the quality target. Considering that in wave soldering, appropriately increasing the preheating temperature is generally beneficial for flux activation and reducing PCBA thermal shock, thereby improving the stability of the process window, within the range of X5≥98.8℃ and X5≤194℃, a temperature close to the upper limit, 194℃, is selected as the optimal setting value. This satisfies the quality requirement of "DPMO≤200" while remaining within the allowable range of equipment parameters. In this case, the wave soldering module maintains the above parameters.

[0058] In summary, the multi-dimensional statistical analysis methods employed by this control system, including range analysis, variance analysis, regression analysis, and residual analysis modules, can accurately quantify the impact of each welding parameter on solder quality. This eliminates reliance on the subjective experience of operators, effectively avoiding experience-based errors and reducing the product defect rate by approximately 80%, thus meeting the corresponding usage requirements.

[0059] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A wave soldering intelligent optimization control system, characterized in that, include: The wave soldering module is configured to perform soldering operations on PCBAs according to soldering parameters and output product identification information for each PCBA. An optical inspection module, which is communicatively connected to the wave soldering module, is configured to receive the product identification information, perform quality inspection on the soldered PCBA, and output solder quality data associated with the product identification information. The intelligent optimization control module is communicatively connected to the wave soldering module and the optical detection module, and is configured as follows: Receive soldering parameters and product identification information from the wave soldering module, and solder quality data associated with the product identification information from the optical inspection module; The wave soldering module is selectively controlled based on whether the solder quality data meets a preset target. When the conditions are met, the wave soldering module is controlled to maintain the current soldering parameters. If the conditions are not met, the parameter optimization process is initiated. Based on the associated welding parameters and solder quality data, statistical analysis is performed to determine the optimal combination of welding parameters, and the wave soldering module is controlled to update to the optimal combination of welding parameters.

2. The intelligent optimization control system for wave soldering as described in claim 1, characterized in that: The intelligent optimization control module includes a data receiving unit, a DOE experimental design unit, a multi-dimensional statistical analysis unit, and a parameter optimization unit. The data receiving unit is communicatively connected to the wave soldering module and the optical inspection module, and is configured to receive product identification information, welding parameters during welding, and related solder quality data in real time. The DOE experimental design unit is connected to the data receiving unit and is configured to generate a multi-factor orthogonal array based on the current soldering parameters in response to solder quality data falling below a preset quality target. The multidimensional statistical analysis unit is connected to the DOE experimental design unit and the data receiving unit, and is configured to perform significant factor screening, quantitative modeling and model verification on the experimental solder quality data corresponding to the experiment based on the multi-factor orthogonal array. The parameter optimization unit is connected to the multi-dimensional statistical analysis unit and is configured to calculate the optimal combination of welding parameters based on the significant factors and quantitative relationships verified by the model, and output the result to the wave soldering module to control its update of the operating welding parameters.

3. The intelligent optimization control system for wave soldering as described in claim 2, characterized in that: The multidimensional statistical analysis unit includes a range analysis module, a variance analysis module, a regression analysis module, and a residual analysis module; The range analysis module is configured to calculate the range of solder quality data corresponding to wave soldering parameters based on the solder quality data after the multi-factor orthogonal array experiment in order to screen candidate parameters, and transmit the results to the variance analysis module. The analysis of variance module is configured to perform significance tests on candidate parameters, screen significant factors, and transmit them to the regression analysis module; The regression analysis module is configured to construct a regression equation with significant factors as independent variables and solder quality data as dependent variables, and then transmit the regression equation to the residual analysis module. The residual analysis module is configured to calculate the residual by combining the actual observed values ​​of solder quality inspected by the optical detection module with the predicted values ​​of the regression equation. If there is no systematic deviation in the residual, the effective regression equation is output to the parameter optimization unit. If there is a deviation, the multi-dimensional statistical analysis unit is triggered to re-execute the analysis, and the statistical processing flow including range analysis, variance analysis and regression analysis is re-executed.

4. A control method for an intelligent optimization control system for wave soldering, characterized in that, Includes the following steps: S1: Perform soldering operations on the PCBA using the current soldering parameters, and output product identification information for each PCBA; S2: Receive the product identification information, perform quality inspection on the soldered PCBA, and output solder quality data associated with the product identification information; S3: Receive soldering parameters and product identification information from the wave soldering module, and solder quality data associated with the product identification information from the optical inspection module; selectively control the wave soldering module based on whether the solder quality data meets a preset target. S31: When the condition is met, control the wave soldering module to maintain the current welding parameters. S32: When the conditions are not met, initiate the parameter optimization process, perform statistical analysis based on the associated welding parameters and solder quality data to determine the optimal welding parameter combination, and control the wave soldering module to update to the optimal welding parameter combination.

5. The control method of the intelligent optimization control system for wave soldering as described in claim 4, characterized in that, The specific steps of step S32 include the following: It can receive product identification information, welding parameters during welding, and related solder quality data in real time. In response to the solder quality data not meeting the preset quality target, a multi-factor orthogonal array is generated based on the current soldering parameters; For the experimental solder quality data corresponding to the experiments conducted based on the multi-factor orthogonal array, significant factor screening, quantitative modeling, and model validation were completed. Based on the significant factors and quantitative relationships validated by the model, the optimal combination of welding parameters is calculated and output to the wave soldering module.

6. The control method of the intelligent optimization control system for wave soldering as described in claim 5, characterized in that: The specific steps for screening significant factors, quantitative modeling, and model validation of the experimental solder quality data corresponding to the experiments conducted based on the aforementioned multi-factor orthogonal array are as follows: Based on the solder quality data after the multi-factor orthogonal array experiment, the range of solder quality data corresponding to wave soldering parameters is calculated to screen candidate parameters. The candidate parameters are subjected to a significance test to screen significant factors; A regression equation was constructed using the aforementioned significant factors as independent variables and solder quality data as dependent variables. The residuals are calculated by combining the actual observed values ​​of solder quality inspected by the optical inspection module with the predicted values ​​of the regression equation; if there is no systematic bias in the residuals, the effective regression equation is output. If there is a systematic bias in the residual, a re-execution is triggered.

7. The control method of the intelligent optimization control system for wave soldering according to claim 6, characterized in that, Based on the solder quality data after the multi-factor orthogonal array experiment, the specific steps for calculating the range of solder quality data corresponding to wave soldering parameters to screen candidate parameters are as follows: Based on range analysis, according to the formula R=x max -x min , where x max x represents the maximum quality data corresponding to a certain welding parameter. min To minimize the value, the influence of different levels of various welding parameters on solder quality indicators was calculated, and candidate welding parameters with a significant impact on solder quality were preliminarily screened.

8. The control method of the intelligent optimization control system for wave soldering according to claim 6, characterized in that, The specific steps for performing a significance test on the candidate parameters and screening for significant factors are as follows: Based on the principles of analysis of variance, total variance is decomposed into between-group variance and within-group variance; the calculation logic of within-group variance follows the same formula as the total variance. Consistency is achieved by taking the solder quality data within each experimental group as the analysis object, where xi is a single solder quality data point within the group, μ is the group mean of the solder quality data in that group, and N is the number of data points within that group. The experimental data are tested for significance by calculating the ratio of the mean square between groups to the mean square within groups, and significant factors are retained.

9. The control method of the intelligent optimization control system for wave soldering according to claim 6, characterized in that, The specific steps for constructing a regression equation using the aforementioned significant factor as the independent variable and solder quality data as the dependent variable are as follows: Using the significant factors retained by the analysis of variance as independent variables and the solder quality index as the dependent variable, a quadratic regression model was established to obtain the regression equation. , .

10. The control method of the intelligent optimization control system for wave soldering according to claim 6, characterized in that, The residuals are calculated by combining the actual observed values ​​of solder quality inspected by the optical inspection module with the predicted values ​​of the regression equation. If the residuals have no systematic bias, a valid regression equation is output. If the residuals have systematic bias, the specific steps for triggering re-execution are as follows: Based on formula Where ei is the residual and yi is the measured value of the experimental solder quality data. For the regression model's predicted values, calculate the residuals between the predicted values ​​and the experimentally measured values. If the residuals are randomly distributed and have no systematic bias, the model is considered to be a good fit. If the residuals have bias, return to the multidimensional statistical analysis unit to re-execute the statistical analysis.

Images