Refined welding method for heat-sensitive devices
By identifying the gap information of thermally sensitive devices on the circuit board and optimizing laser welding parameters, the impact of welding temperature on the performance of thermally sensitive devices is solved, and more efficient welding effects are achieved, ensuring the reliability of thermally sensitive devices.
Patent Information
- Application Number
- CN202410827328.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-25
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-06-25
AI Technical Summary
During the manufacturing process of electronic products, thermally sensitive devices are not intelligent during welding due to their unintelligent temperature control, which affects the reliability and performance of the entire electronic product.
By collecting device images of thermally sensitive devices on the circuit board, identifying device gap information, optimizing preheating and welding parameters of laser welding, obtaining the optimal laser preheating and welding parameters, and performing refined welding to reduce welding temperature to thermally sensitive devices The impact of
Effectively reduce the impact of welding temperature on the performance of thermally sensitive devices, improve welding effect, and ensure the performance and reliability of thermally sensitive devices on the circuit board.
Smart Images

Figure CN118371863B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of intelligent welding, and in particular to a refined welding method for heat-sensitive devices. Background Art
[0002] In the manufacturing process of modern electronic products, the welding of components on circuit boards is a crucial process. As the functions of electronic products become increasingly complex, the types and numbers of components integrated on circuit boards are also increasing. Among them, heat-sensitive devices (such as thermistors, temperature sensors, etc.) are increasingly widely used in circuit boards. However, these heat-sensitive devices may be affected in performance due to uneven heating or excessive heating during the welding process, which in turn affects the reliability and performance of the entire electronic product.
[0003] Laser welding technology has unique advantages in the welding of circuit board components, such as high precision, high efficiency and non-contact. However, when welding heat-sensitive devices, there are still technical problems such as unintelligent temperature control, which affects the performance of heat-sensitive devices and the welding effect. Summary of the invention
[0004] The present application provides a refined welding method for heat-sensitive devices, which is used to solve the technical problem in the prior art that circuit board welding affects the performance of heat-sensitive devices on the circuit board and the welding effect.
[0005] In a first aspect of the present application, a refined welding method for a heat-sensitive device is provided, the method comprising: collecting a device image of a heat-sensitive device on a circuit board to be welded from a preset direction to obtain a device image set;
[0006] Perform device fixed gap identification according to the device image set to obtain a device gap information array;
[0007] According to the device gap information array, fine optimization of laser preheating parameters in laser welding is performed to obtain optimal laser preheating parameters, and optimal pad temperature information after laser preheating is predicted, wherein the optimization is performed with the purpose of reducing the influence of preheating temperature on the performance of heat-sensitive devices and improving preheating effect;
[0008] Based on the optimal pad temperature information and device gap information array, fine optimization of laser welding parameters in laser welding is performed to obtain optimal laser welding parameters, wherein the optimization is performed with the purpose of reducing the influence of welding temperature on the performance of heat-sensitive devices and improving welding effect;
[0009] The optimal laser preheating parameters and the optimal laser welding parameters are used to perform fine welding of the heat sensitive device.
[0010] In a second aspect, the present application provides an electronic device comprising: a processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the steps of the method in the above-mentioned first aspect.
[0011] In a third aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the steps of the method in the first aspect when executed.
[0012] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0013] The refined welding method for heat-sensitive devices provided in the present application collects a set of device images of heat-sensitive devices on a circuit board to be welded from a preset direction, identifies and obtains a device gap information array, optimizes laser preheating parameters, obtains optimal laser preheating parameters, and predicts optimal pad temperature information after laser preheating. Based on the optimal pad temperature information and the device gap information array, the laser welding parameters are optimized to obtain optimal laser welding parameters. The optimal laser preheating parameters and the optimal laser welding parameters are used to perform refined welding of heat-sensitive devices. The present application optimizes laser welding preheating parameters and welding parameters by taking into account the gap between the heat-sensitive device and the circuit board through the above method, which is more intelligent and accurate, can effectively reduce the influence of welding temperature on the performance of heat-sensitive devices, and at the same time improve the preheating and welding effects of laser welding, thereby ensuring the performance and reliability of heat-sensitive devices on the circuit board. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0015] Figure 1 A schematic structural diagram of a refined welding method for a heat-sensitive device provided in an embodiment of the present application;
[0016] Figure 2 A schematic diagram of a process for obtaining a device image set in a refined welding method for a heat-sensitive device provided in an embodiment of the present application;
[0017] Figure 3 A structural schematic diagram of an electronic device is provided for the present application.
[0018] Explanation of the reference numerals: electronic device 300 , memory 301 , processor 302 , communication interface 303 , bus architecture 304 . DETAILED DESCRIPTION
[0019] The present application provides a refined welding method for heat-sensitive devices, which is used to solve the technical problem in the prior art that circuit board welding affects the performance of heat-sensitive devices on the circuit board and the welding effect.
[0020] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.
[0021] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Embodiment 1
[0022] like Figure 1 As shown, the present application provides a refined welding method for heat-sensitive devices, the method comprising:
[0023] S10: collecting device images of heat-sensitive devices on the circuit board to be welded from a preset direction to obtain a device image set;
[0024] In the embodiment of the present application, the circuit board to be welded refers to a circuit board that is currently being prepared for welding operations, on which heat-sensitive devices that require fine welding are installed. Heat-sensitive devices are easily affected by high temperatures during the welding process, and therefore special welding methods and parameters are required to protect their performance.
[0025] The device image of the heat-sensitive device on the circuit board to be soldered is collected from a preset direction. The preset direction refers to determining a fixed direction or angle before image collection so as to uniformly collect the images of the heat-sensitive devices on the circuit board to analyze the installation status of the heat-sensitive devices on the circuit board.
[0026] And device images of multiple heat-sensitive devices are collected from preset directions to obtain a device image set to analyze the installation conditions of heat-sensitive devices in multiple positions. The installation conditions of heat-sensitive devices can affect the degree of influence of welding temperature on heat-sensitive devices during laser welding.
[0027] like Figure 2 As shown, in the embodiment of the present application, step S10 includes:
[0028] The direction parallel to the circuit board to be welded is used as the preset direction;
[0029] Based on the preset direction, device images of heat-sensitive devices on the circuit board are collected from multiple positions to obtain a device image set.
[0030] In an embodiment of the present application, a device image is collected to identify the gap between the heat-sensitive device and the circuit board before it is installed on the circuit board for welding. The size of the gap will affect the degree to which the performance of the heat-sensitive device is affected by the welding temperature during the welding process. For example, the larger the gap, the smaller the degree of influence.
[0031] Therefore, the direction parallel to the circuit board to be welded is used as the preset direction. For example, in actual operation, the circuit board can be placed on a fixed platform so that the shooting direction of the camera device is parallel to the surface of the circuit board to collect images.
[0032] Based on the preset direction, device images of the heat-sensitive device on the circuit board are collected from multiple positions. Each device image includes a side image of the heat-sensitive device and a surface image of the circuit board. The gap between the heat-sensitive device and the circuit board can be identified. The heat-sensitive device on the circuit board is photographed from different angles or positions to obtain comprehensive image data of the heat-sensitive device at multiple angles. A device image set is obtained, which is helpful to analyze the comprehensiveness and completeness of the gap between the heat-sensitive device and the circuit board.
[0033] S20: performing device fixed gap identification according to the device image set to obtain a device gap information array;
[0034] In the embodiment of the present application, the fixed gap identification of the thermal sensitive device is to identify the gap between the thermal sensitive device and the circuit board from the device image set through image processing technology, so as to serve as the basis for analyzing the influence of welding temperature on the performance of the thermal sensitive device.
[0035] Device fixed gaps are identified based on multiple device images in a device image set to obtain device gap information observed at multiple positions. A device gap information array is constructed by combining the position coordinates or angles of the multiple positions to analyze the impact of welding temperature on the performance of heat-sensitive devices from different directions.
[0036] Exemplarily, image processing and machine learning algorithms (such as convolutional neural networks) are used for recognition.
[0037] Step S20 in the method provided in the embodiment of the present application includes:
[0038] Based on fixed data of heat-sensitive devices on multiple circuit boards, a set of sample device images is collected, and the size of the gap between the heat-sensitive device and the circuit board in each sample device image is marked to obtain a set of sample gap information;
[0039] Using the sample device image set and the sample gap information set, based on a convolutional neural network, training to obtain a device gap identifier;
[0040] Using the device gap identifier to identify the device image set, and obtain a device gap information set;
[0041] Based on the multiple positions and in combination with the device gap information set, a device gap information array is constructed.
[0042] In the embodiment of the present application, machine learning is used to identify the fixed gap of the device.
[0043] Based on the fixed data of heat-sensitive devices on multiple circuit boards, sample data collection for machine learning is performed, and images of heat-sensitive devices are collected from multiple circuit boards according to the above-mentioned preset directions to form a set of sample images. These images should clearly show the relative position and gap size between the heat-sensitive devices and the circuit boards.
[0044] Furthermore, each sample device image is marked, and the actual gap size between the thermal sensitive device and the circuit board is recorded to form a sample gap information set.
[0045] Convolutional neural network is a deep learning model that is particularly suitable for image recognition and processing. It includes convolutional layers, pooling layers and fully connected layers, and achieves target recognition by learning features in images.
[0046] The sample device image set is used as input, and the sample gap information set is used as label data. They are input into CNN for training. By repeatedly iteratively optimizing the model parameters, CNN can accurately identify the gap between the heat-sensitive device and the circuit board. After the training is completed, a device gap identifier that can automatically identify the gap size is obtained.
[0047] Furthermore, the trained device gap identifier is used to identify new device images, automatically identify the gap between the heat-sensitive device and the circuit board, and the previously collected device image set is input into the device gap identifier, which outputs the gap size in each image to form a device gap information set.
[0048] Based on the angles of the position coordinates of multiple positions relative to the thermally sensitive device, the gap information of the corresponding device image recognition is summarized and sorted to form a comprehensive device gap information array.
[0049] Through the above steps, the sample images and gap information collected are trained using a convolutional neural network to obtain a recognizer that can automatically identify the gap size. The recognizer is used to recognize the actual device image, obtain comprehensive gap information, and construct a device gap information array. These data provide an important basis for subsequent refined welding, help improve welding accuracy, and protect the performance of heat-sensitive devices.
[0050] S30: performing fine optimization of laser preheating parameters in laser welding according to the device gap information array, obtaining optimal laser preheating parameters, and predicting optimal pad temperature information after laser preheating, wherein the optimization is performed for the purpose of reducing the influence of preheating temperature on the performance of heat-sensitive devices and improving preheating effect;
[0051] In the embodiment of the present application, the device gap information array is an array constructed based on the gap data between the identified thermally sensitive device and the circuit board. Based on the device gap information array, the laser preheating parameters in laser welding are finely optimized.
[0052] Among them, the fine optimization of laser preheating parameters refers to the precise adjustment of various laser preheating parameters (such as laser power, preheating time, etc.) before welding to obtain the best preheating effect, while minimizing the impact of preheating temperature on the performance of heat-sensitive devices to prevent high-temperature damage.
[0053] Among them, during the optimization process, the optimization is carried out with the purpose of reducing the influence of preheating temperature on the performance of heat-sensitive devices and improving the preheating effect. The influence of preheating temperature on the performance of heat-sensitive devices can be predicted by combining the device gap information array and laser prediction parameters, as well as predicting the preheating temperature of the pad, and the laser preheating parameters can be evaluated and optimized.
[0054] The laser welding of the circuit board includes laser pad preheating and laser welding. After the optimization is completed, the optimal laser preheating parameters are obtained, and the optimal pad temperature information after the laser preheating is predicted based on the optimal laser preheating parameters, which serves as the basis for subsequent laser welding parameter optimization.
[0055] Exemplarily, the laser preheating parameters can be optimized using an optimization algorithm, such as a genetic algorithm, particle swarm optimization, etc., to gradually adjust the preheating parameters until the optimal laser preheating parameters with the best preheating effect and the least impact on the performance of heat-sensitive devices are obtained.
[0056] Step S30 in the method provided in the embodiment of the present application includes:
[0057] Get the laser preheating parameter range;
[0058] In order to reduce the impact of preheating temperature on the performance of heat-sensitive devices and improve the preheating effect, a preheating optimization function is constructed as follows:
[0059] ;
[0060] Among them, LPF is the preheating fitness, and are performance weight and warm-up weight, and The sum of is 1, M is the number of device gap information in the device gap information array, is the preheating performance influence coefficient of the heat-sensitive device obtained by combining the gap information of the i-th device with the laser preheating parameter prediction, To predict the pad temperature after preheating according to the laser preheating parameters, It is the standard preheating temperature;
[0061] According to the device gap information array and the preheating optimization function, the laser preheating parameters are optimized within the laser preheating parameter range to obtain the optimal laser preheating parameters, and the optimal pad temperature information after the laser preheating is completed is predicted.
[0062] In the embodiment of the present application, a laser preheating parameter range is obtained. The laser preheating parameter range refers to the adjustable range of various parameters (such as laser power, preheating time, etc.) during laser preheating, which serves as an optimization space for laser prediction parameters.
[0063] The laser preheating parameter range can be constructed based on past laser welding preheating data.
[0064] In order to reduce the influence of preheating temperature on the performance of heat-sensitive devices and improve the preheating effect, a function for evaluating and optimizing the adaptability of laser preheating parameters is constructed to achieve the best preheating effect and the minimum influence on heat-sensitive devices, that is, to construct a preheating optimization function, as shown in the following formula:
[0065] ;
[0066] Among them, LPF is the preheating fitness, and are performance weight and preheating weight, indicating the importance of preheating performance and temperature control. and The sum of is 1, for example, it can be set to 0.5 and 0.5, M is the number of device gap information in the device gap information array, is the preheating performance influence coefficient of the heat-sensitive device obtained by combining the gap information of the i-th device with the laser preheating parameter prediction, To predict the pad temperature after preheating according to the laser preheating parameters, It is the standard preheating temperature, which is used to compare and evaluate the preheating effect. It can be set according to the temperature standard of the pad in the preheating of the PCB laser welding.
[0067] in, It represents the average preheating performance influence coefficient after all gap information is combined with preheating parameters. Through logarithmic function conversion, the optimization process is made more sensitive. The larger the average preheating performance influence coefficient, the smaller the preheating adaptability. It indicates the relative deviation between the predicted temperature and the standard temperature, so as to more accurately evaluate the preheating temperature control effect. The larger the deviation, the smaller the preheating adaptability.
[0068] The device gap information array and the preheating optimization function are used as the basis for optimizing the analysis and evaluating the laser preheating parameters, and the laser preheating parameter range is used as the optimization basis. The laser preheating parameters are optimized within the range to obtain the optimal laser preheating parameters, and the optimal pad temperature information after the laser preheating is completed is predicted.
[0069] The step of "optimizing the laser preheating parameters within the laser preheating parameter range according to the device gap information array and the preheating optimization function" in the method provided in the embodiment of the present application includes:
[0070] Randomly generate a plurality of first laser preheating parameters within the laser preheating parameter range;
[0071] According to the plurality of first laser preheating parameters and the device gap information array, a thermal sensitive device performance impact prediction and a pad temperature prediction are performed to obtain a plurality of first preheating performance impact coefficient sets and a plurality of first pad temperatures, and according to the preheating optimization function, a plurality of first preheating fitness degrees are calculated;
[0072] According to the plurality of first preheating adaptabilities, selecting and obtaining an adjustment direction from a plurality of first laser preheating parameters;
[0073] According to the adjustment direction, the plurality of first laser preheating parameters are adjusted to obtain a plurality of second laser preheating parameters;
[0074] Continue to optimize until convergence, output the laser preheating parameters with the maximum preheating adaptability, obtain the optimal laser preheating parameters, and predict the optimal pad temperature information after the laser preheating is completed.
[0075] In the embodiment of the present application, within the determined laser preheating parameter range, a plurality of different laser preheating parameters are generated using a random generation method, and these laser preheating parameters are used as initial values of the optimization process, that is, as a plurality of first laser preheating parameters. Each first laser preheating parameter includes a different laser preheating power or laser preheating time.
[0076] According to the plurality of first laser preheating parameters and the device gap information array, the performance impact of the heat-sensitive device and the pad temperature are predicted to obtain a plurality of first preheating performance impact coefficient sets and a plurality of first pad temperatures. For example, the performance impact of the heat-sensitive device and the pad temperature are predicted using a pre-trained prediction model for predicting the welding preheating performance impact and the pad temperature. Each first preheating performance impact coefficient set includes a plurality of first preheating performance impact coefficients predicted according to the corresponding first laser preheating parameters and the plurality of device gap information in the device gap information array.
[0077] Based on the predicted multiple first preheating performance influence coefficient sets and multiple first pad temperatures, according to the preheating optimization function, multiple first preheating adaptabilities are obtained by substitution into the calculation.
[0078] According to the multiple first preheating fitnesses, the adjustment direction is obtained by screening in the multiple first laser preheating parameters. For example, the first laser preheating parameter corresponding to the largest first preheating fitness is selected as the adjustment direction. Alternatively, the multiple first laser preheating parameters are arranged in order from large to small according to the multiple first preheating fitnesses, and then the change direction of the laser preheating parameters as the preheating fitness increases is counted as the adjustment direction.
[0079] According to the adjustment direction, the parameters in the plurality of first laser preheating parameters are adjusted to generate new laser preheating parameters, and the plurality of second laser preheating parameters are obtained. Exemplarily, the plurality of first laser preheating parameters are adjusted by using the adjustment step of the laser power and the adjustment step of the preheating time. For example, the adjustment step of the preheating time is 2s.
[0080] If the adjustment direction is the first laser preheating parameter corresponding to the largest first preheating adaptability, the first laser preheating parameter is not adjusted, and only other first laser preheating parameters are adjusted.
[0081] Continue to predict the performance impact of the thermally sensitive device and the pad temperature according to the multiple second laser preheating parameters combined with the device gap information array, calculate the second preheating fitness, and continue to optimize until convergence, that is, when the convergence condition is reached, the optimization process is stopped. Exemplarily, the convergence condition is to reach the optimization number threshold, and the optimization number threshold is, for example, 100 to 150 times.
[0082] After convergence, the laser preheating parameters with the largest preheating adaptability are output as the optimal laser preheating parameters, and the optimal laser preheating parameters are used for laser preheating prediction. The temperature of the pad after preheating is obtained as the optimal pad temperature information, which serves as the basis for subsequent optimization of laser welding parameters.
[0083] Through the above steps, multiple initial parameters are randomly generated within the laser preheating parameter range, and prediction and optimization are performed in combination with the device gap information, and the parameters are gradually adjusted to improve the preheating adaptability, and finally the optimal laser preheating parameters are obtained. This process ensures the precise control of the preheating process, effectively reduces the impact of high temperature on the performance of heat-sensitive devices, and improves the preheating effect.
[0084] The step of "predicting the performance impact of a thermally sensitive device and the pad temperature according to the plurality of first laser preheating parameters and the device gap information array" in the method provided in the embodiment of the present application includes:
[0085] According to the historical data of laser welding of circuit boards, a set of sample device gap information, a set of sample laser preheating parameters, and a set of sample pad temperatures are collected, and the performance of heat-sensitive devices after preheating with different device gap information and laser preheating parameters is tested and identified to obtain a set of sample preheating performance influence coefficients;
[0086] Using the sample device gap information set and the sample laser preheating parameter set as input and the sample preheating performance influence coefficient set as output, training the preheating performance influence prediction branch;
[0087] Using the sample laser preheating parameter set as input and the sample pad temperature set as output, training the pad temperature prediction branch, and combining the preheating performance impact prediction branch to obtain a laser preheating predictor;
[0088] The laser preheating predictor is used to predict the performance impact of thermally sensitive devices and pad temperatures of the multiple first laser preheating parameters and device gap information arrays, and obtain the multiple first preheating performance impact coefficient sets and multiple first pad temperatures.
[0089] In an embodiment of the present application, based on the historical data of laser welding of the circuit board, that is, the data recorded in the past laser welding process, including various welding parameters, results and component information on the circuit board, the gap data between the heat-sensitive device on the circuit board and the circuit board, the various laser preheating parameters used in the historical welding process, and the temperature data of the pad after preheating under different preheating parameters are collected to obtain a sample device gap information set, a sample laser preheating parameter set, and a sample pad temperature set.
[0090] Furthermore, actual preheating experiments are carried out on different device gap information and laser preheating parameter combinations to test and record performance changes of thermally sensitive devices, such as the amplitude of current changes in thermally sensitive devices. The test results are identified as preheating performance influence coefficients to form a sample preheating performance influence coefficient set.
[0091] The sample device gap information set and the sample laser preheating parameter set are used as input, and the sample preheating performance influence coefficient set is used as output to train the preheating performance influence prediction branch. For example, a machine learning or deep learning algorithm, such as a convolutional neural network (CNN) or a regression model, is used for training. Through multiple iterations and parameter adjustments, the model is optimized so that it can accurately predict the preheating performance influence coefficient. After the training is completed, the preheating performance influence prediction branch is obtained.
[0092] And, a pad temperature prediction branch is trained by using a sample laser preheating parameter set as input and a sample pad temperature set as output.
[0093] By combining the trained preheating performance impact prediction branch and pad temperature prediction branch, a laser preheating predictor is obtained, which can simultaneously predict the preheating performance impact and pad temperature.
[0094] A laser preheating predictor is used to predict the performance impact of thermally sensitive devices and pad temperatures of multiple first laser preheating parameters and device gap information arrays, and multiple first preheating performance impact coefficient sets and multiple first pad temperatures are obtained as prediction results of preheating using multiple first laser preheating parameters.
[0095] S40: Based on the optimal pad temperature information and device gap information array, fine optimization of laser welding parameters in laser welding is performed to obtain optimal laser welding parameters, wherein the optimization is performed with the purpose of reducing the influence of welding temperature on the performance of heat-sensitive devices and improving welding effect;
[0096] In the embodiment of the present application, fine optimization of laser welding parameters in laser welding is performed based on the optimal pad temperature information and device gap information array. During the optimization process, optimization is performed with the purpose of reducing the influence of welding temperature on the performance of heat-sensitive devices and improving welding effect, so as to reduce the adverse effect of high welding temperature on the performance of heat-sensitive devices and ensure high welding quality.
[0097] Among them, by combining the optimal pad temperature information and device gap information array with laser welding parameters, the impact of temperature after laser welding on the performance of heat-sensitive devices and welding quality are predicted. Among them, the pad temperature will affect the melting and flattening effect of the solder, and then affect the welding effect, so the optimal pad temperature information is used for prediction.
[0098] In an embodiment of the present application, the optimization of laser welding parameters can be performed using an optimization algorithm, such as a genetic algorithm, particle swarm optimization, etc., to gradually adjust the laser welding parameters until the optimal laser welding parameters with high welding quality and minimal impact on the performance of heat-sensitive devices are obtained.
[0099] Step S40 in the method provided in the embodiment of the present application includes:
[0100] Get the laser welding parameter range;
[0101] In order to reduce the impact of welding temperature on the performance of heat-sensitive devices and improve the welding effect, a welding optimization function is constructed as follows:
[0102] ;
[0103] Among them, LWF is welding fitness, and is the weight, The sum of is 1, M is the number of device gap information in the device gap information array, is the welding performance influence coefficient of the heat-sensitive device obtained by combining the gap information of the i-th device with the laser welding parameters prediction, and Q is the welding quality coefficient after welding according to the laser welding parameters;
[0104] Randomly generate a plurality of first laser welding parameters within the laser welding parameter range, perform welding performance influence prediction and welding quality prediction in combination with the optimal pad temperature information and device gap information array, obtain a plurality of first welding performance influence coefficient sets and a plurality of first welding quality coefficients, and calculate and obtain a plurality of first welding fitnesses;
[0105] According to the plurality of first welding fitnesses, adjusting, updating and optimizing the plurality of first laser welding parameters;
[0106] Continue to optimize until convergence, output the laser welding parameters with the maximum welding fitness, and obtain the optimal laser welding parameters.
[0107] In an embodiment of the present application, the laser welding parameter range suitable for the current circuit board is determined through the laser welding experimental data of the circuit board and past experience. These laser welding parameter ranges are the optimization space for the subsequent optimization process. The laser welding parameter range includes, for example, the parameter range of the laser power.
[0108] In order to reduce the impact of welding temperature on the performance of heat-sensitive devices and improve the welding effect, a welding optimization function is constructed as follows:
[0109] ;
[0110] Among them, LWF is welding fitness, which is used to evaluate the quality of laser welding parameters. The larger it is, the better the laser welding parameters are. and is the weight, The sum of is 1, indicating the importance attached to the influence of welding performance and welding quality, for example are all set to 0.5, which has the same influence on welding performance and welding quality. M is the number of device gap information in the device gap information array, is the welding performance influence coefficient of the heat-sensitive device obtained by combining the i-th device gap information with the laser welding parameter prediction, Q is the welding quality coefficient after welding according to the laser welding parameters, such as the proportion of the area in the solder joint that meets the welding requirements after welding is completed, or the welding quality score, etc. The larger the welding quality coefficient, the better the welding quality.
[0111] Within the laser welding parameter range, a random generation method is used to generate a plurality of different laser welding parameters, and a plurality of first laser welding parameters are obtained as the first generation solutions for optimization. The random generation method can be performed based on a random algorithm of a computer.
[0112] According to multiple first laser welding parameters, combined with the optimal pad temperature information and the device gap information array, the welding performance impact prediction and the welding quality prediction are performed to obtain multiple first welding performance impact coefficient sets and multiple first welding quality coefficients, which are used as the prediction results after laser welding based on the multiple first welding performance parameters and the optimal pad temperature information predicted after preheating in the aforementioned content.
[0113] For example, based on machine learning, a trained prediction model can be used to take a plurality of first laser welding parameters, optimal pad temperature information, and device gap information array as inputs for prediction.
[0114] A plurality of first welding performance influence coefficient sets and a plurality of first welding quality coefficients are input into the above-mentioned welding optimization function, and a plurality of first welding fitnesses are obtained by calculation.
[0115] According to the multiple first welding fitnesses, the multiple first laser welding parameters are adjusted, updated and optimized.
[0116] Exemplarily, based on the steps in step S30 in the aforementioned content, the first laser welding parameter with the largest first welding fitness can be selected as the adjustment direction according to multiple first welding fitnesses. Alternatively, based on other optimization algorithms, an update direction can be selected in the initial solution to perform update optimization to gradually improve the fitness of the laser welding parameters.
[0117] Furthermore, the optimization is continued, and the prediction of welding performance influence, welding quality prediction, fitness calculation and laser welding parameter adjustment are repeatedly performed to gradually optimize the laser welding parameters until the convergence conditions are reached. The convergence conditions, for example, reach a preset optimization times threshold, stop the optimization, output the laser welding parameters with the maximum welding fitness, and obtain the optimal laser welding parameters.
[0118] Through the above steps, multiple initial parameters are randomly generated within the laser welding parameter range, and the optimal pad temperature information and device gap information are combined for prediction and optimization, and the parameters are gradually adjusted to improve the welding adaptability, and finally the optimal laser welding parameters are obtained. This process ensures the high quality and high reliability of the welding process, minimizes the impact of high welding temperature on the performance of heat-sensitive devices, and improves the welding quality and reliability of heat-sensitive devices.
[0119] The step of "combining the optimal pad temperature information and the device gap information array to predict the welding performance impact and welding quality" in the method provided in the embodiment of the present application includes:
[0120] According to the historical laser welding data of the circuit board, a sample pad temperature information set, a sample laser welding parameter set, a sample device gap information set, a sample welding performance influence coefficient set and a sample welding quality coefficient set are collected;
[0121] Using the sample pad temperature information set, the sample laser welding parameter set, the sample device gap information set and the sample welding performance influence coefficient set, training the welding performance influence prediction branch;
[0122] The sample pad temperature information set, the sample laser welding parameter set and the sample welding quality coefficient set are used to train a welding quality prediction branch, and the laser welding predictor is obtained by combining the welding performance impact prediction branch;
[0123] The laser welding predictor is used to predict the optimal pad temperature information, multiple first laser welding parameters and device gap information array to obtain multiple first welding performance influence coefficient sets and multiple first welding quality coefficients.
[0124] In an embodiment of the present application, all data recorded during the laser welding process in the past, including various welding parameters, temperature changes, welding effects, etc., are used as historical laser welding data of the circuit board, and as the basis for training a machine learning model for laser welding prediction, thereby providing real and rich sample data for model training and improving prediction accuracy.
[0125] According to the historical laser welding data of the circuit board, the sample pad temperature information set, sample laser welding parameter set, and sample device gap information set are collected. And according to the temperature of the welding process, the experimental test of the performance impact of the heat-sensitive device is carried out, such as testing the change amplitude of the voltage and current in the heat-sensitive device, and according to the degree of the performance impact of the heat-sensitive device in the test results, the sample welding performance impact coefficient set is obtained.
[0126] In addition, welding quality test and evaluation are performed based on historical welding results, such as testing the area ratio of the weld that meets the welding quality requirements after welding is completed, or performing welding quality scoring and marking it as a sample welding quality coefficient to obtain a set of sample welding quality coefficients.
[0127] Among them, the pad temperature, laser welding parameters and device gap will affect the degree of influence of temperature on the performance of heat-sensitive devices during welding. Therefore, a sample pad temperature information set, a sample laser welding parameter set, and a sample device gap information set are used as input data, and a sample welding performance influence coefficient set is used as output data. Exemplarily, an appropriate machine learning or deep learning algorithm (such as a convolutional neural network or a regression model) is selected, such as a feedforward neural network, to construct a welding performance influence prediction branch, and supervised training is performed, and the model is optimized through iteration and parameter adjustment until the training is completed to obtain a welding performance influence prediction branch. Among them, exemplarily, the training of the welding performance influence prediction branch is completed when the accuracy meets the requirement of 90%.
[0128] Furthermore, the pad temperature and laser welding parameters will affect the melting and welding of the solder during the welding process. Therefore, the sample pad temperature information set and the sample laser welding parameter set are used as input data, and the sample welding quality coefficient set is used as output data. An appropriate machine learning or deep learning algorithm (such as a convolutional neural network or a regression model) is selected. For example, a feedforward neural network is used to construct a welding quality prediction branch, and supervised training is performed. The model is optimized through iteration and parameter adjustment until the training is completed, and a welding quality prediction branch is obtained.
[0129] By combining the welding performance impact prediction branch and the welding quality prediction branch, a laser welding predictor for predicting the impact of laser welding parameters on the performance of thermosensitive devices and welding quality after welding is obtained.
[0130] The laser welding predictor is used to predict multiple first laser welding parameters and device gap information arrays, wherein the optimal pad temperature information, multiple first laser welding parameters and device gap information arrays are used as input data, combined and input welding performance impact prediction branch predictions are performed to obtain multiple first welding performance impact coefficient sets.
[0131] Furthermore, the optimal pad temperature information and a plurality of first laser welding parameters are combined and input into a welding quality prediction branch prediction to obtain a plurality of first welding quality coefficients.
[0132] In the embodiment of the present application, the historical laser welding data of the circuit board is used to train the welding performance impact prediction branch and the welding quality prediction branch to form a comprehensive laser welding predictor. Then, the predictor is used to predict the performance impact of the thermosensitive device and the welding quality of the initially generated welding parameters to obtain detailed welding performance impact and welding quality data. This process ensures that the optimization of welding parameters is based on an accurate prediction model, which improves the accuracy and reliability of laser welding parameter optimization.
[0133] S50: Using the optimal laser preheating parameters and the optimal laser welding parameters to perform fine welding of the heat sensitive device.
[0134] Based on the optimization results of the above content, the optimal laser preheating parameters and the optimal laser welding parameters are used to carry out refined welding of heat-sensitive devices, which can ensure the welding quality while minimizing the impact of high welding temperature on the performance of heat-sensitive devices and ensuring the quality of circuit boards.
[0135] In summary, the embodiments of the present application have at least the following technical effects:
[0136] The refined welding method for heat-sensitive devices provided in the present application collects a set of device images of heat-sensitive devices on a circuit board to be welded from a preset direction, identifies and obtains a device gap information array, optimizes laser preheating parameters, obtains optimal laser preheating parameters, and predicts optimal pad temperature information after laser preheating. Based on the optimal pad temperature information and the device gap information array, the laser welding parameters are optimized to obtain optimal laser welding parameters. The optimal laser preheating parameters and the optimal laser welding parameters are used to perform refined welding of heat-sensitive devices. The present application optimizes laser welding preheating parameters and welding parameters by taking into account the gap between the heat-sensitive device and the circuit board through the above method, which is more intelligent and accurate, can effectively reduce the influence of welding temperature on the performance of heat-sensitive devices, and at the same time improve the preheating and welding effects of laser welding, thereby ensuring the performance and reliability of heat-sensitive devices on the circuit board. Embodiment 2
[0137] Based on the same inventive concept as the refined welding method for heat-sensitive devices in the aforementioned embodiment, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method in embodiment one is implemented.
[0138] Through the detailed description of the refined welding method for heat sensitive devices in the above description, those skilled in the art can clearly know the refined welding method for heat sensitive devices in this embodiment, so for the sake of brevity of the description, it will not be described in detail here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.
[0139] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
[0140] Reference below Figure 3 To describe the electronic device of the embodiment of the present application:
[0141] The electronic device 300 includes: a processor 302, a communication interface 303, and a memory 301. Optionally, the electronic device 300 may also include a bus architecture 304. The communication interface 303, the processor 302, and the memory 301 may be interconnected via the bus architecture 304; the bus architecture 304 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus. The bus architecture 304 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 3 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.
[0142] The processor 302 may be a CPU, a microprocessor, an ASIC, or one or more integrated circuits for controlling the execution of the program of the present application.
[0143] The communication interface 303 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), wired access network, etc.
[0144] The memory 301 may be a ROM or other type of static storage device that can store static information and instructions, a RAM or other type of dynamic storage device that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may exist independently and be connected to the processor via a bus architecture 304. The memory may also be integrated with the processor, wherein the memory 301 is used to store computer-executable instructions for executing the solution of the present application, and the execution is controlled by the processor 302. The processor 302 is used to execute the computer-executable instructions stored in the memory 301.
[0145] An embodiment of the present application further provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to implement the refined welding method for heat-sensitive devices in the above-mentioned embodiment when executed by a processor.
[0146] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.
[0147] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0148] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
[0149] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.
Claims
1. A refined welding method for heat-sensitive devices, characterized in that: The method comprises: Collect device images of heat-sensitive devices on a circuit board to be welded from a preset direction to obtain a device image set; Perform device fixed gap identification according to the device image set to obtain a device gap information array; According to the device gap information array, fine optimization of laser preheating parameters in laser welding is performed to obtain optimal laser preheating parameters, and optimal pad temperature information after laser preheating is predicted, wherein the optimization is performed with the purpose of reducing the influence of preheating temperature on the performance of heat-sensitive devices and improving preheating effect; Based on the optimal pad temperature information and device gap information array, fine optimization of laser welding parameters in laser welding is performed to obtain optimal laser welding parameters, wherein the optimization is performed with the purpose of reducing the influence of welding temperature on the performance of heat-sensitive devices and improving welding effect; Using the optimal laser preheating parameters and the optimal laser welding parameters to perform fine welding of the heat sensitive device; The device image of the heat-sensitive device on the circuit board to be welded is collected from a preset direction to obtain a device image set, including: The direction parallel to the circuit board to be welded is used as the preset direction; Based on the preset direction, collecting device images of heat-sensitive devices on the circuit board from multiple positions to obtain a device image set; Performing device fixed gap identification according to the device image set to obtain a device gap information array includes: Based on fixed data of heat-sensitive devices on multiple circuit boards, a set of sample device images is collected, and the size of the gap between the heat-sensitive device and the circuit board in each sample device image is marked to obtain a set of sample gap information; Using the sample device image set and the sample gap information set, based on a convolutional neural network, training to obtain a device gap identifier; Using the device gap identifier to identify the device image set, and obtain a device gap information set; Based on the multiple positions and in combination with the device gap information set, a device gap information array is constructed.
2. The refined welding method for heat sensitive devices according to claim 1, characterized in that: According to the device gap information array, fine optimization of laser preheating parameters in laser welding is performed to obtain optimal laser preheating parameters, including: Get the laser preheating parameter range; In order to reduce the impact of preheating temperature on the performance of heat-sensitive devices and improve the preheating effect, a preheating optimization function is constructed as follows: ; Among them, LPF is the preheating fitness, and are performance weight and warm-up weight, and The sum of is 1, M is the number of device gap information in the device gap information array, is the preheating performance influence coefficient of the heat-sensitive device obtained by combining the gap information of the i-th device with the laser preheating parameter prediction, To predict the pad temperature after preheating according to the laser preheating parameters, It is the standard preheating temperature; According to the device gap information array and the preheating optimization function, the laser preheating parameters are optimized within the laser preheating parameter range to obtain the optimal laser preheating parameters, and the optimal pad temperature information after the laser preheating is completed is predicted.
3. The refined welding method for heat sensitive devices according to claim 2, characterized in that: Optimizing laser preheating parameters within the laser preheating parameter range according to the device gap information array and the preheating optimization function includes: Randomly generate a plurality of first laser preheating parameters within the laser preheating parameter range; According to the plurality of first laser preheating parameters and the device gap information array, a thermal sensitive device performance impact prediction and a pad temperature prediction are performed to obtain a plurality of first preheating performance impact coefficient sets and a plurality of first pad temperatures, and according to the preheating optimization function, a plurality of first preheating fitness degrees are calculated; According to the plurality of first preheating adaptabilities, selecting and obtaining an adjustment direction from a plurality of first laser preheating parameters; According to the adjustment direction, the plurality of first laser preheating parameters are adjusted to obtain a plurality of second laser preheating parameters; Continue to optimize until convergence, output the laser preheating parameters with the maximum preheating adaptability, obtain the optimal laser preheating parameters, and predict the optimal pad temperature information after the laser preheating is completed.
4. The refined welding method for heat sensitive devices according to claim 3, characterized in that: According to the plurality of first laser preheating parameters and the device gap information array, a prediction of the performance impact of a thermally sensitive device and a prediction of a pad temperature are performed, including: According to the historical data of laser welding of circuit boards, a set of sample device gap information, a set of sample laser preheating parameters, and a set of sample pad temperatures are collected, and the performance of heat-sensitive devices after preheating with different device gap information and laser preheating parameters is tested and identified to obtain a set of sample preheating performance influence coefficients; Using the sample device gap information set and the sample laser preheating parameter set as input and the sample preheating performance influence coefficient set as output, training the preheating performance influence prediction branch; Using the sample laser preheating parameter set as input and the sample pad temperature set as output, training the pad temperature prediction branch, and combining the preheating performance impact prediction branch to obtain a laser preheating predictor; The laser preheating predictor is used to predict the performance impact of thermally sensitive devices and pad temperatures of the multiple first laser preheating parameters and device gap information arrays, and obtain the multiple first preheating performance impact coefficient sets and multiple first pad temperatures.
5. The refined welding method for heat sensitive devices according to claim 1, characterized in that: Based on the optimal pad temperature information and device gap information array, fine optimization of laser welding parameters in laser welding is performed to obtain optimal laser welding parameters, including: Get the laser welding parameter range; In order to reduce the impact of welding temperature on the performance of heat-sensitive devices and improve the welding effect, a welding optimization function is constructed as follows: ; Among them, LWF is welding fitness, and is the weight, The sum of is 1, M is the number of device gap information in the device gap information array, is the welding performance influence coefficient of the heat-sensitive device obtained by combining the gap information of the i-th device with the laser welding parameters prediction, and Q is the welding quality coefficient after welding according to the laser welding parameters; Randomly generate a plurality of first laser welding parameters within the laser welding parameter range, perform welding performance influence prediction and welding quality prediction in combination with the optimal pad temperature information and device gap information array, obtain a plurality of first welding performance influence coefficient sets and a plurality of first welding quality coefficients, and calculate and obtain a plurality of first welding fitnesses; According to the plurality of first welding fitnesses, adjusting, updating and optimizing the plurality of first laser welding parameters; Continue to optimize until convergence, output the laser welding parameters with the maximum welding fitness, and obtain the optimal laser welding parameters.
6. The refined welding method for heat sensitive devices according to claim 5, characterized in that: A plurality of first laser welding parameters are randomly generated within the laser welding parameter range, and welding performance influence prediction and welding quality prediction are performed in combination with the optimal pad temperature information and device gap information array, including: According to the historical laser welding data of the circuit board, a sample pad temperature information set, a sample laser welding parameter set, a sample device gap information set, a sample welding performance influence coefficient set and a sample welding quality coefficient set are collected; Using the sample pad temperature information set, the sample laser welding parameter set, the sample device gap information set and the sample welding performance influence coefficient set, training the welding performance influence prediction branch; The sample pad temperature information set, the sample laser welding parameter set and the sample welding quality coefficient set are used to train a welding quality prediction branch, and the laser welding predictor is obtained by combining the welding performance impact prediction branch; The laser welding predictor is used to predict the optimal pad temperature information, multiple first laser welding parameters and device gap information array to obtain multiple first welding performance influence coefficient sets and multiple first welding quality coefficients.
7. An electronic device, characterized in that: The electronic device comprises: At least one processor; and a memory in communication with the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the refined welding method for heat-sensitive devices according to any one of claims 1 to 6.
8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the refined welding method for heat-sensitive devices according to any one of claims 1 to 6 when executed.
