Reflow process prediction system and prediction method

Abstract

The application discloses a reflow soldering process prediction system and a prediction method. The reflow soldering process prediction method comprises the following steps: obtaining device characteristics of a reflow soldered electronic device, reflow soldering parameters and average reflow soldering temperature time sequences as a first training set; training a first prediction module with the first training set, so that the first prediction module can predict the average reflow soldering temperature time sequences and the reflow soldering parameters according to device characteristics of a to-be-reflow-soldered electronic device; obtaining time sequence characteristics of the first training set; obtaining element characteristics and element positions of the reflow soldered electronic device, and taking the time sequence characteristics, the element characteristics and the element positions as a second training set; training a second prediction module with the second training set, so that the second prediction module can predict element temperature time sequences according to the time sequence characteristics, the element characteristics and the element positions; and combining the first prediction module and the second prediction module into a prediction module.

Description

Technical Field

[0001] This invention relates to the field of reflow soldering production, and in particular to a machine learning prediction system and method for the reflow soldering process. Background Technology

[0002] When reflow soldering electronic devices, the reflow soldering temperature sequence must conform to the reflow soldering temperature specification to become qualified products. Currently, the common practice is to first reflow soldering the device to be reflow soldered using historical reflow soldering parameters from similar devices, obtaining the reflow soldering temperature sequence during this process. Finally, the reflow soldering parameters are adjusted based on the degree of difference between the reflow soldering temperature sequence and the reflow soldering temperature specification until the reflow soldering temperature sequence conforms to the specification.

[0003] Currently, although there is software on the market that can predict reflow soldering temperature timing, this software cannot predict the reflow soldering temperature timing of multiple components at the same time, and is not suitable for actual production lines. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a reflow soldering process prediction system and prediction method to address the shortcomings of the prior art.

[0005] To address the aforementioned technical problems, one technical solution adopted by the present invention is to provide a reflow soldering process prediction system. The reflow soldering process prediction system includes a processing circuit configured to perform the following steps: inputting device features and component features of the electronic device to be reflow soldered into a prediction module to generate reflow soldering parameters for the electronic device to be reflow soldered. The prediction module is generated through a training process, which includes: obtaining the device characteristics, reflow parameters, and average reflow temperature timing of the reflow soldering electronic device as a first training set; training the first prediction module with the first training set to predict the average reflow temperature timing and reflow parameters of the reflow soldering electronic device based on the device characteristics of the reflow soldering electronic device, wherein during the process of the first prediction module predicting the average reflow temperature timing of the reflow soldering electronic device, the processing circuit obtains the timing characteristics of the first training set; obtaining the component characteristics and component positions of the reflow soldering electronic device, and using the timing characteristics, component characteristics, and component positions of the reflow soldering electronic device as a second training set; training the second prediction module with the second training set to predict the component temperature timing of the reflow soldering electronic device based on the timing characteristics, component characteristics, and component positions of the reflow soldering electronic device; and merging the trained first prediction module and second prediction module into a prediction module.

[0006] Preferably, the reflow soldering process prediction system further includes a temperature sensing device for sensing the temperature of the reflow soldering electronic device using the reflow soldering parameters of the reflow soldering electronic device to obtain multiple regional reflow soldering temperature timings of the reflow soldering electronic device; the processing circuit is further configured to calculate the average reflow soldering temperature timing of the reflow soldering electronic device based on the multiple regional reflow soldering temperature timings.

[0007] Preferably, the device features of the reflow soldered electronic device include a circuit board size and a circuit board material.

[0008] Preferably, the component features of the reflow soldered electronic device include multiple component dimensions, multiple component materials, and multiple component surrounding densities.

[0009] Preferably, the reflow soldering parameters of the reflow soldering electronic device include the temperature of a reflow oven zone and the speed of a reflow oven track during the reflow soldering process via the reflow soldering electronic device.

[0010] To address the aforementioned technical problems, another technical solution adopted by the present invention is to provide a reflow soldering process prediction method, comprising: obtaining a device characteristic, a reflow soldering parameter, and an average reflow soldering temperature timing of a reflow soldering electronic device as a first training set; training a first prediction module with the first training set to predict an average reflow soldering temperature timing and a reflow soldering parameter of the electronic device to be reflow soldered based on a device characteristic of the electronic device to be reflow soldered; during the process of the first prediction module predicting the average reflow soldering temperature timing of the electronic device to be reflow soldered, a processing circuit obtains a timing feature of the first training set; obtaining a component characteristic of the reflow soldering electronic device... The system identifies a component location and uses the timing characteristics, the component characteristics of the reflow soldering electronic device, and the component location of the reflow soldering electronic device as a second training set. A second prediction module is trained using the second training set to predict the temperature timing of a component in the reflow soldering electronic device based on the timing characteristics, the component characteristics of the reflow soldering electronic device, and the component location of the reflow soldering electronic device. The trained first prediction module and second prediction module are then merged into a single prediction module. The device characteristics and component characteristics of the reflow soldering electronic device are input into the prediction module to generate the reflow soldering parameters of the reflow soldering electronic device.

[0011] Preferably, the reflow soldering process prediction method further includes: performing temperature sensing on the reflow soldering device using the reflow soldering parameters of the reflow soldering device; obtaining multiple regional reflow soldering temperature timing sequences of the reflow soldering device; and calculating the average reflow soldering temperature timing sequence of the reflow soldering device based on the multiple regional reflow soldering temperature timing sequences.

[0012] Preferably, the device features of the reflow soldered electronic device include a circuit board size and a circuit board material.

[0013] Preferably, the component features of the reflow soldered electronic device include multiple component dimensions, multiple component materials, and multiple component surrounding densities.

[0014] Preferably, the reflow soldering parameters of the reflow soldering electronic device include the temperature of a reflow oven zone and the speed of a reflow oven track during the reflow soldering process via the reflow soldering electronic device.

[0015] To further understand the features and technical content of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are for reference and illustration only and are not intended to limit the present invention. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of an embodiment of the reflow soldering process prediction system of the present invention.

[0017] Figure 2 This is a schematic diagram of an embodiment of the reflow soldering process prediction method of the present invention.

[0018] Figure 3 A schematic diagram of an embodiment for obtaining the average reflow soldering temperature timing of a reflow soldering process prediction system via a reflow soldering electronics device.

[0019] Figure 4 This is a schematic diagram of the reflow soldering temperature sequence for multiple areas of an electronic device.

[0020] Figure 5 This is a schematic diagram of the average reflow soldering temperature timing for reflow soldering electronic devices.

[0021] Figure 6 A schematic diagram of an embodiment in which the first prediction module predicts the average reflow soldering temperature timing and reflow soldering parameters of an electronic device to be reflow soldered.

[0022] Figure 7 A schematic diagram of an embodiment for predicting the component temperature timing of an electronic device to be reflow soldered by a second prediction module.

[0023] Figure 8 A schematic diagram of an embodiment for predicting reflow soldering parameters of an electronic device to be reflow soldered by a prediction module. Detailed Implementation

[0024] The following are specific examples.

[0025] Figure 1 This is a schematic diagram of an embodiment of the reflow soldering process prediction system of the present invention. See also... Figure 1 The reflow soldering process prediction system includes a temperature sensing device 1, a processing circuit 2, and a storage device 3, with the processing circuit 2 connected to the temperature sensing device 1 and the storage device 3.

[0026] The processing circuit 2 is, for example, a central processing unit (CPU), or other programmable general-purpose or special-purpose microcontroller (MCU), microprocessor, digital signal processor (DSP), programmable controller, application-specific integrated circuit (ASIC), graphics processing unit (GPU), image signal processor (ISP), image processing unit (IPU), arithmetic logic unit (ALU), complex programmable logic device (CPLD), field-programmable gate array (FPGA), or other similar elements or combinations thereof.

[0027] Storage device 3 is, for example, any form of fixed or removable random access memory (RAM), read-only memory (ROM), flash memory, hard disk drive (HDD), solid state drive (SSD), or similar element or combination of the above elements.

[0028] Storage device 3 stores device features D1, component features D2, component positions D3, reflow soldering parameters D4, component temperature timing D5, a first prediction module M1, and a second prediction module M2 of the reflow soldering electronic device. Specifically, the reflow soldering electronic device includes a circuit board and multiple components disposed on the circuit board. Device feature D1 includes the circuit board size and material of the reflow soldering electronic device. Component feature D2 includes the dimensions, materials, and surrounding density of the multiple components of the reflow soldering electronic device. Component position D3 includes multiple relative coordinates between the center points of the multiple components of the reflow soldering electronic device and the center point of the circuit board. Reflow soldering parameters D4 include the reflow oven zone temperature and reflow oven track speed during the reflow soldering process. Component temperature timing D5 includes the reflow soldering temperature timing of multiple zones during the reflow soldering process of the multiple components of the reflow soldering electronic device. The first prediction module M1 and the second prediction module M2 are, for example, recurrent neural networks (RNN), long short-term memory (LSTM) neural networks, or bidirectional long short-term memory (BiLSTM) neural networks.

[0029] Figure 2 This is a schematic diagram of an embodiment of the reflow soldering process prediction method of the present invention. Figure 2 The reflow soldering process prediction method is derived from Figure 1 The reflow soldering process prediction system is used to perform this. See also... Figure 2 , Figure 3 , Figures 5 to 8 In step S201, the temperature sensing device 1 generates a component temperature timing D5 via the reflow soldering electronics, and the component temperature timing D5 includes the reflow soldering temperature timing of multiple regions of multiple components via the reflow soldering electronics. In step S202, the processing circuit 2 calculates the average reflow soldering temperature timing D6 via the reflow soldering electronics based on the multiple region reflow soldering temperature timings.

[0030] Figure 3 A schematic diagram of an embodiment for obtaining the average reflow soldering temperature timing D6 via the reflow soldering electronics in a reflow soldering process prediction system. See also Figure 3The reflow soldering electronic device 4 includes a first element E1, a second element E2, a third element E3, and a fourth element E4, and a temperature sensing device 1 is connected to the first element E1, the second element E2, the third element E3, and the fourth element E4. Since the first element E1, the second element E2, the third element E3, and the fourth element E4 have different specific heats, when the first element E1, the second element E2, the third element E3, and the fourth element E4 are subjected to the reflow soldering process, the temperature sensing device 1 generates different first-zone reflow soldering temperature timings RT1, RT2, RT3, and RT4. When the processing circuit 2 obtains the reflow soldering temperature timing RT1 of the first region, the reflow soldering temperature timing RT2 of the second region, the reflow soldering temperature timing RT3 of the third region, and the reflow soldering temperature timing RT4 of the fourth region, the processing circuit 2 is configured to calculate the average reflow soldering temperature timing D6 of the reflow soldering electronics 4 based on the reflow soldering temperature timing RT1 of the first region, the reflow soldering temperature timing RT2 of the second region, the reflow soldering temperature timing RT3 of the third region, and the reflow soldering temperature timing RT4 of the fourth region.

[0031] Figure 4 This is a schematic diagram of an embodiment of the reflow soldering temperature timing for multiple components in reflow soldering electronic device 4. See also... Figure 4 Each of the reflow soldering temperature timing sequences RT1 (first zone), RT2 (second zone), RT3 (third zone), and RT4 (fourth zone) contains multiple time points and corresponding zone reflow soldering temperatures for each of the multiple time points. The time before the first time T1 is the preheating zone of the reflow soldering process; the time after the first time T1 and the time before the second time T2 is the heat absorption zone of the reflow soldering process; the time after the second time T2 and the time before the third time T3 is the reflow soldering zone of the reflow soldering process; and the time after the third time T3 is the cooling zone of the reflow soldering process.

[0032] Figure 5 This is a schematic diagram of an embodiment of the average reflow soldering temperature timing D6 of the reflow soldering electronics device 4. See also... Figure 5 The average reflow soldering temperature timing sequence D6 includes multiple time points and multiple average reflow soldering temperatures corresponding to those time points. The average reflow soldering temperature is... Figure 4 The average value of the reflow soldering temperatures of the four regions: the first region reflow soldering temperature sequence RT1, the second region reflow soldering temperature sequence RT2, the third region reflow soldering temperature sequence RT3, and the fourth region reflow soldering temperature sequence RT4.

[0033] See also Figure 2In step S203, the processing circuit 2 obtains the device characteristics D1, reflow soldering parameters D4, and average reflow soldering temperature timing D6 of the reflow soldering electronic device 4 as the first training set.

[0034] In step S204, the processing circuit 2 trains the first prediction module M1 with the first training set to predict the average reflow soldering temperature timing D8 and reflow soldering parameters D9 of the electronic device to be reflow soldered based on the device characteristics D7 of the electronic device to be reflow soldered.

[0035] Regarding the training of the first prediction module M1, for example, the first prediction module M1 is a neural network module, the device feature D1 is the input of the first prediction module M1, and the predicted reflow soldering parameters D9 and the predicted average reflow soldering temperature timing D8 of the electronic device to be reflowed are the outputs of the first prediction module M1. The weights of the first prediction module M1 are continuously adjusted according to the error between the output of the first prediction module M1 and the reflow soldering parameters D4 and the average reflow soldering temperature timing D6, until the error between the output of the first prediction module M1 and the reflow soldering parameters D4 and the average reflow soldering temperature timing D6 falls within the tolerable error range.

[0036] Specifically, the device feature D7 of the electronic device to be reflow soldered includes the circuit board size and material of the electronic device to be reflow soldered, while the reflow soldering parameters D9 of the electronic device to be reflow soldered include the reflow oven zone temperature and the reflow oven track speed during the reflow soldering process. The reflow oven contains multiple different zones, and the electronic device to be reflow soldered will sequentially enter multiple zones within the reflow oven during the reflow soldering process. The reflow oven zone temperature includes the temperature settings of the multiple different zones within the reflow oven.

[0037] Figure 6 This is a schematic diagram of an embodiment for predicting the average reflow soldering temperature timing D8 and reflow soldering parameters D9 of an electronic device to be reflow soldered by a first prediction module. After the processing circuit 2 trains the first prediction module M1 with a first training set, the device feature D7 of the electronic device to be reflow soldered is input into the trained first prediction module M1. The trained first prediction module M1 outputs the average reflow soldering temperature timing D8 and reflow soldering parameters D9 of the electronic device to be reflow soldered based on the device feature D7.

[0038] See also Figure 2 In step S205, during the process of the first prediction module M1 predicting the average reflow soldering temperature timing D8 of the electronic device to be reflow soldered, the processing circuit 2 obtains the timing features D10 of the first training set.

[0039] In step S206, the processing circuit 2 obtains the component characteristics D2 and component positions D3 of the reflow soldering electronic device 4, and uses the timing characteristics D10, the component characteristics D2 and component positions D3 of the reflow soldering electronic device 4 as the second training set.

[0040] In step S207, the processing circuit 2 trains the second prediction module M2 with the second training set to predict the component temperature timing D13 of the electronic device to be reflowed based on the timing feature D10, the component feature D11 of the electronic device to be reflowed, and the component position D12.

[0041] Regarding the training of the second prediction module M2, for example, the second prediction module M2 is a neural network module. The timing features D10 of the first training set, the component features D2 of the reflow soldering electronic device 4, and the component position D3 are the inputs of the second prediction module M2, and the predicted component temperature timing D13 of the electronic device to be reflow soldered is the output of the second prediction module M2. The weights of the second prediction module M2 are continuously adjusted according to the error between the output of the second prediction module M2 and the component temperature timing D5 until the error between the output of the second prediction module M2 and the component temperature timing D5 falls within the tolerable error range.

[0042] Figure 7 This is a schematic diagram of an embodiment for predicting the component temperature timing D13 of an electronic device to be reflow soldered by the second prediction module M2. After the processing circuit 2 trains the second prediction module M2 with the second training set, the timing features D10 of the first training set, the component features D11 and component positions D12 of the electronic device to be reflow soldered are input into the trained second prediction module M2. The trained second prediction module M2 outputs the component temperature timing D13 of the electronic device to be reflow soldered based on the component features D11 and component positions D12.

[0043] Specifically, the electronic device to be reflow soldered includes a circuit board and multiple components disposed on the circuit board. Component features D11 include the dimensions, materials, and surrounding density of the multiple components in the electronic device to be reflow soldered. Component positions D12 include multiple relative coordinates between the center points of the multiple components in the electronic device to be reflow soldered and the center point of the circuit board. Component temperature timing D13 includes the reflow soldering temperature timing of multiple regions of the multiple components in the electronic device to be reflow soldered during the reflow soldering process.

[0044] In step S208, the processing circuit 2 merges the trained first prediction module M1 and the second prediction module M2 into a prediction module. In step S209, the processing circuit 2 inputs the device feature D7 and component feature D11 of the electronic device to be reflow soldered into the prediction module to generate the reflow soldering parameters D9 of the electronic device to be reflow soldered.

[0045] Figure 8 This is a schematic diagram of an embodiment whereby a prediction module predicts reflow soldering parameters D9 for an electronic device to be reflow soldered. Device features D7 and component features D11 of the electronic device to be reflow soldered are input to the prediction module M3, and the prediction module M3 outputs the reflow soldering parameters D9 based on these features. Specifically, the reflow soldering parameters D9 are the reflow oven zone temperature setting and the reflow oven track speed during the reflow soldering process.

[0046] One of the beneficial effects of this invention is that the reflow soldering process prediction system and method provided by this invention can immediately obtain the reflow soldering parameters of the electronic device to be reflow soldered simply by inputting the device characteristics and component characteristics of the electronic device to be reflow soldered into the prediction module. Furthermore, during the reflow soldering process using these parameters, the component temperature sequence on the electronic device to be reflow soldered also conforms to the reflow soldering temperature specifications. In this way, the time required for the electronic device to become a qualified product can be significantly shortened.

[0047] The content disclosed above is only a preferred and feasible embodiment of the present invention, and is not intended to limit the scope of protection of the claims of the present invention. Therefore, all equivalent technical changes made based on the content of the present invention specification and drawings are included within the scope of protection of the claims of the present invention.

Claims

1. A reflow soldering process prediction system, characterized in that, The reflow soldering process prediction system includes: A processing circuit is configured to perform the following steps: A device feature and a component feature of an electronic device to be reflow soldered are input into a prediction module to generate a reflow soldering parameter for the electronic device to be reflow soldered. The prediction module is generated by executing a training process, which includes: A device characteristic, a reflow soldering parameter, and an average reflow soldering temperature timing of a reflow soldering electronic device are obtained as a first training set. The first prediction module is trained using the first training set to predict an average reflow soldering temperature timing and the reflow soldering parameters of the electronic device to be reflow soldered based on the device characteristics. During the process of the first prediction module predicting the average reflow soldering temperature timing, the processing circuit acquires a timing feature of the first training set. Obtain a component feature and a component location of the reflow soldering electronic device, and use the timing feature, the component feature of the reflow soldering electronic device, and the component location of the reflow soldering electronic device as a second training set; A second prediction module is trained using the second training set to predict the temperature timing of a component in the reflow soldering electronic device based on the timing features, the component features of the reflow soldering electronic device, and the position of a component in the reflow soldering electronic device; and The trained first prediction module and the second prediction module are combined to form the prediction module.

2. The reflow soldering process prediction system according to claim 1, characterized in that, It also includes a temperature sensing device for sensing the temperature of the reflow soldering electronics using the reflow soldering parameters of the reflow soldering electronics to obtain multiple regional reflow soldering temperature timings of the reflow soldering electronics; the processing circuit is further configured to calculate the average reflow soldering temperature timing of the reflow soldering electronics based on the multiple regional reflow soldering temperature timings.

3. The reflow soldering process prediction system according to claim 1, characterized in that, The device features of the reflow soldered electronic device include a circuit board size and a circuit board material.

4. The reflow soldering process prediction system according to claim 1, characterized in that, The component features of the reflow soldered electronic device include multiple component dimensions, multiple component materials, and multiple component surrounding densities.

5. The reflow soldering process prediction system according to claim 1, characterized in that, The reflow soldering parameters of the reflow soldering electronic device include the temperature of a reflow oven zone and the speed of a reflow oven track during the reflow soldering process via the reflow soldering electronic device.

6. A method for predicting reflow soldering processes, characterized in that, The reflow soldering process prediction method includes: A device characteristic, a reflow soldering parameter, and an average reflow soldering temperature timing of a reflow soldered electronic device are obtained as a first training set. A first prediction module is trained using the first training set to predict an average reflow temperature timing and a reflow parameter of an electronic device to be reflowed based on a device characteristic of the electronic device to be reflowed. During the process of the first prediction module predicting the average reflow soldering temperature timing of the electronic device to be reflow soldered, a processing circuit obtains a timing feature of the first training set. Obtain a component feature and a component location of the reflow soldering electronic device, and use the timing feature, the component feature of the reflow soldering electronic device, and the component location of the reflow soldering electronic device as a second training set; A second prediction module is trained using the second training set to predict the temperature timing of a component in the reflow soldering electronic device based on the timing features, a component feature of the reflow soldering electronic device, and the position of a component in the reflow soldering electronic device; and The trained first prediction module and the second prediction module are merged into one prediction module; and The device features and component features of the electronic device to be reflow soldered are input into the prediction module to generate the reflow soldering parameters of the electronic device to be reflow soldered.

7. The reflow soldering process prediction method according to claim 6, characterized in that, Also includes: Temperature sensing is performed on the reflow soldering device using the reflow soldering parameters of the reflow soldering device; the reflow soldering temperature timing of multiple regions of the reflow soldering device is obtained. And calculate the average reflow soldering temperature timing of the reflow soldering electronics based on multiple reflow soldering temperature timings of the aforementioned regions.

8. The reflow soldering process prediction method according to claim 6, characterized in that, The device features of the reflow soldered electronic device include a circuit board size and a circuit board material.

9. The reflow soldering process prediction method according to claim 6, characterized in that, The component features of the reflow soldered electronic device include multiple component dimensions, multiple component materials, and multiple component surrounding densities.

10. The reflow soldering process prediction method according to claim 6, characterized in that, The reflow soldering parameters of the reflow soldering electronic device include the temperature of a reflow oven zone and the speed of a reflow oven track during the reflow soldering process via the reflow soldering electronic device.

Images