WO2022050149A1

WO2022050149A1 - Soldering system

Info

Publication number: WO2022050149A1
Application number: PCT/JP2021/031180
Authority: WO
Inventors: 徹也川添; 典也南; 浩儀山下; 俊介佐々木
Original assignee: 三菱電機株式会社
Priority date: 2020-09-04
Filing date: 2021-08-25
Publication date: 2022-03-10
Also published as: CN115996809A; JPWO2022050149A1

Abstract

A soldering system (1) is provided with: a flux coating machine (3) for applying flux to a substrate (10); a preheater (4) for preheating the substrate (10); a solder bath (5) for storing molten solder; a solder bath heater (14) for melting the solder inside the solder bath (5); a jet flow nozzle (13) for spraying the molten solder in the solder bath (5) toward the substrate (10); a transport mechanism (2) for sequentially transporting the substrate (10) above the flux coating machine (3), the preheater (4), and the solder bath (5); and a temperature measurement device (17) arranged above the jet flow nozzle (13).

Description

Soldering system

This disclosure relates to a soldering system.

A soldering system equipped with a means for automatically checking the quality of soldering is known. For example, in the system of Patent Document 1, the temperature distribution of the substrate immediately after soldering is detected by a non-contact temperature sensor and image processing is performed, and this processing signal is compared and calculated with the optimum temperature data recorded in advance by a computer. Detects the presence or absence of abnormal soldering temperature. The system controls the transfer speed of the substrate, the temperature of the preheater, and the temperature of the molten solder based on the detection results.

Japanese Unexamined Patent Publication No. 7-142852

In Patent Document 1, it is not possible to sequentially grasp the change in the contact state between the substrate and the jetted molten solder. As a result, the soldering quality of the board is not stable.

Therefore, an object of the present disclosure is to provide a soldering system capable of soldering a substrate with stable quality.

The soldering system of the present disclosure includes a flux coating machine that applies flux to a substrate, a preheater that preheats the substrate, a solder bath that stores molten solder, a solder tank heater that melts the solder in the solder tank, and a substrate. A jet nozzle that ejects molten solder in the solder bath toward the solder tank, a transport mechanism that sequentially transports the substrate above the flux coating machine, above the preheater, and above the solder tank, and temperature measurement located above the jet nozzle. Equipped with a device.

According to the soldering system of the present disclosure, since the temperature measuring device arranged above the jet nozzle is provided, the substrate can be soldered with stable quality.

It is a schematic diagram which shows the soldering system 1 of Embodiment 1. FIG. It is a figure which shows the structure of the control device 9. It is a figure which shows the structure of the 1st learning apparatus 22. (A) is a diagram showing an example of temperature distribution data on the surface of the substrate 10. (B) is a diagram showing an example of target temperature distribution data on the surface of the substrate 10. (C) is a diagram showing an example of temperature difference distribution data on the surface of the substrate 10. It is a figure which shows the identification image in a support vector machine. It is a figure which shows the structure of the 2nd learning apparatus 23. It is a figure which shows an example of the structure of a neural network. It is a figure which shows the structure of the 1st inference device 24. It is a figure which shows the structure of the 2nd inference device 25. It is a flowchart which shows the learning procedure by the 1st learning apparatus 22 and the 2nd learning apparatus 23 in Embodiment 1. FIG. It is a flowchart which shows the inference procedure by the 1st inference apparatus 24 and the 2nd inference apparatus 25 in Embodiment 1. FIG. It is a flowchart which shows the learning procedure by the 1st learning apparatus 22 and the 2nd learning apparatus 23 in Embodiment 2. It is a flowchart which shows the inference procedure by the 1st inference apparatus 24 and the 2nd inference apparatus 25 in Embodiment 2. It is a flowchart showing the procedure of inference and relearning by the first inference device 24 and the second inference device 25 in Embodiment 3. It is a figure which shows the arrangement of the infrared camera 7 in Embodiment 4. It is a figure which shows the time change of the surface temperature of the substrate 10. (A) is a figure showing a state when a substrate 10 comes into contact with molten solder. (B) is a diagram showing a state immediately after the substrate 10 is separated from the molten solder. It is a schematic diagram which shows the soldering system 1b of Embodiment 6. It is a figure which shows the structure of the soldering system 1a of Embodiment 7. It is a figure which shows the hardware composition of the control device 9.

The embodiments will be described with reference to the drawings.
Embodiment 1.
FIG. 1 is a schematic view showing the soldering system 1 of the first embodiment.

The soldering system 1 includes a transfer mechanism 2, a flux coating machine 3, a preheater 4, a solder tank 5, a fluxer control unit 18, a jet motor 15, a solder tank heater 14, a temperature measuring device 17, a control device 9, and a soldering inspection device. 16 is provided.

The flux coating machine 3 applies the flux 11 to the lower surface (soldered surface) of the substrate 10 on which the electronic component 201 to be soldered is mounted. As a method of applying the flux 11, there are a spray type, a foaming type, and a dipping type.

The fluxer control unit 18 controls the coating amount of the flux 11. The amount of the flux 11 applied is determined by the flow rate of the flux liquid, the pressure of the compressed air, and the moving speed of the two nozzles. The coating amount of the flux 11 varies depending on the degree of clogging of the nozzle in the flux coating machine 3, the variation in the operation of the flux coating machine 3, the variation in the displacement of the exhaust fan in the flux coating machine 3, and the like. If the amount of the flux 11 applied is large, the time required for the solvent to evaporate in the preheating step of the preheater 4 becomes long, so that the temperature of the substrate 10 does not rise sufficiently. When the amount of the flux 11 applied is small, the solvent is immediately volatilized in the preheating step of the preheater 4, so that the temperature of the substrate 10 becomes too high.

The preheater 4 preheats the substrate 10. The purpose of preheating is to volatilize the solvent of the flux 11 and to heat the substrate 10 before soldering. As a result, the effect of removing the oxide film of the flux 11 can be exhibited, so that the ratio of successful soldering is improved. The heating method of the preheater 4 includes infrared rays, far infrared rays, and hot air. In some cases, only the lower surface (soldered surface) of the substrate 10 is heated, and in other cases, the upper surface (non-soldered surface, component surface) of the substrate 10 is also heated.

The solder bath 5 stores molten solder.
The jet nozzle 13 jets the molten solder 12 stored in the solder bath 5. As a result, the lower surface of the substrate 10 comes into contact with the molten solder 12, and soldering is performed. The jet nozzle 13 includes a primary nozzle 13a and a secondary nozzle 13b. The primary nozzle 13a forms a rough wave and supplies molten solder to every corner of the soldered surface of the substrate 10. The secondary nozzle 13b forms a well-ordered wave so that an appropriate amount of solder adheres to the substrate 10.

The solder bath heater 14 melts the solder.
The jet motor 15 rotates the impeller and feeds the molten solder 12 to the jet nozzle 13. As a result, the molten solder 12 is jetted from the jet nozzle 13.

The control device 9 controls each component of the soldering system 1. For example, the control device 9 controls the temperature of the solder bath heater 14, the rotation speed of the jet motor 15, the angle of the jet nozzle 13, and the like.

The temperature measuring device 17 is arranged above the secondary nozzle 13b constituting the jet nozzle 13. The temperature measuring device 17 measures the surface temperature of the substrate 10. The temperature measuring device 17 includes an infrared camera. The temperature measuring device 17 photographs the substrate 10 with an infrared camera. The temperature measuring device 17 measures the temperature distribution on the surface of the substrate 10 based on the captured infrared image. The temperature measuring device 17 outputs the measured information on the temperature distribution on the surface of the substrate 10 to the control device 9. The infrared camera may be provided with a swing mechanism capable of changing the direction of the optical axis of the infrared camera. When a large component is mounted on the substrate 10, the optical axis is not obstructed by the large component when the temperature of the surface of the substrate 10 cannot be measured due to the obstruction by the large component in the normal orientation of the optical axis. The temperature of the surface of the substrate 10 can be measured by changing the direction of.

Since the temperature measuring device 17 is arranged above the secondary nozzle 13b, the temperature measuring device 17 can measure the temperature of the substrate 10 immediately after the substrate 10 is separated from the jetted molten solder. This stabilizes the quality of soldering to the substrate 10.

The control device 9 is connected to the soldering inspection device 16.
The soldering inspection device 16 inspects the quality of soldering of the substrate 10. Specifically, the soldering inspection device 16 has an image analysis function. The soldering inspection device 16 analyzes the soldering state based on the captured image of the substrate 10 and the like. Specifically, the soldering inspection device 16 detects the presence / absence of an image corresponding to the "icicle-shaped" solder adhering to the terminal of the electronic component 201 and the presence / absence of an image corresponding to the solder bridge. Further, the soldering inspection device 16 determines the area to which the solder is attached, and evaluates the excess or deficiency of this area with respect to the predetermined area. Further, the soldering inspection device 16 evaluates the gloss of the solder and evaluates whether or not the gloss standard is satisfied. The soldering inspection device 16 evaluates a plurality of items as described above. The soldering inspection device 16 inspects the quality of soldering based on the evaluation of each item.

The control device 9 obtains a control parameter for making the temperature distribution on the surface of the substrate 10 within the allowable range. The control device 9 controls each part of the soldering system 1 based on the obtained control parameters.

FIG. 2 is a diagram showing the configuration of the control device 9. The control device 9 includes an identification and setting unit 21, a storage device 27, and a drive control unit 26.

The identification and setting unit 21 includes a first learning device 22, a second learning device 23, a first inference device 24, and a second inference device 25.

The storage device 27 includes a target temperature distribution storage unit 76, an allowable temperature difference distribution storage unit 73, a first learning data storage unit 71, a second learning data storage unit 74, and a first trained model storage unit. 72 and a second trained model storage unit 75 are provided.

The target temperature distribution storage unit 76 stores the target temperature distribution data on the surface of the substrate 10. It is desirable that the temperature distribution on the surface of the substrate 10 on which the molten solder is sprayed is uniform. The surface temperature distribution measured from above the substrate 10 is not uniform because the electronic component 201 is mounted on the substrate 10. Therefore, the target temperature distribution storage unit 76 stores in advance the target temperature distribution data on the surface of the substrate 10 in consideration of the mounted electronic component 201. The target temperature distribution data on the surface of the substrate 10 has been obtained in advance by an experiment. Specifically, the target temperature distribution data on the surface of the substrate 10 is created by statistically processing a plurality of temperature distribution data that did not result in soldering defects.

The permissible temperature difference distribution storage unit 73 stores the permissible temperature difference distribution data. The permissible temperature difference distribution data determines the permissible range of the temperature difference at each position of the temperature difference distribution data on the surface of the substrate.

FIG. 3 is a diagram showing the configuration of the first learning device 22.
The first learning device 22 includes a temperature distribution data creation unit 51, a temperature difference distribution data creation unit 52, a feature amount extraction unit 53, a first learning data creation unit 54, and a first trained model generation unit. It is equipped with 55.

The temperature distribution data creation unit 51 acquires information on the temperature distribution on the surface of the substrate 10 from the temperature measuring device 17. The temperature distribution data creation unit 51 divides the surface of the substrate 10 into, for example, a two-dimensional matrix having a region of 1000 × 1000. The temperature distribution data creation unit 51 sets the coordinates of 1 to 1000 in the vertical direction and 1 to 1000 in the horizontal direction with respect to the matrix. By this processing, it is possible to process the temperature of each position on the surface of the substrate 10 in association with the coordinates. The temperature distribution data creation unit 51 specifies the temperature at each position of the matrix from the acquired temperature distribution information, and creates the temperature distribution data on the surface of the substrate 10.

The temperature difference distribution data creation unit 52 is the temperature of the surface of the substrate 10 representing the difference between the temperature distribution data on the surface of the substrate 10 and the target temperature distribution data on the surface of the substrate 10 stored in the target temperature distribution storage unit 76. Create difference distribution data.

FIG. 4A is a diagram showing an example of temperature distribution data on the surface of the substrate 10. FIG. 4B is a diagram showing an example of target temperature distribution data on the surface of the substrate 10. FIG. 4C is a diagram showing an example of temperature difference distribution data on the surface of the substrate 10.

In FIGS. 4A to 4C, the range surrounded by the outer frame represents the range of the surface of the substrate 10. The lines in the frame represent isotherms. For example, isotherms are shown every 5 ° C.

The feature amount extraction unit 53 extracts the feature amount of the temperature difference distribution data on the surface of the substrate 10. The feature quantity is obtained, for example, the absolute value of the temperature, the density of the isotherm, the shape of the isotherm, the amount of change of the isotherm with respect to the temperature distribution on the surface of the substrate 10 measured last time, or the time before the set time. This is data such as the amount of change in the isotherm with respect to the temperature distribution on the surface of the substrate 10. The set time can be, for example, one hour. The feature amount may be data obtained by performing principal component analysis or independent component analysis of these data.

The first learning data creation unit 54 receives the feature amount of the temperature difference distribution data on the surface of the substrate 10 output from the feature amount extraction unit 53, and the soldering quality of the substrate 10 output from the soldering inspection device 16 is good or bad. Receive the inspection result of.

The first learning data creation unit 54 is a first set including a feature amount of temperature difference distribution data on the surface of the substrate 10 and a label (correct answer) of whether or not the soldering of the substrate 10 is good or bad for a plurality of substrates 10. The learning data is created and stored in the first learning data storage unit 71.

The first learning data storage unit 71 stores the first learning data.
The first trained model generation unit 55 uses the first training data stored in the first training data storage unit 71 to solder the substrate 10 from the feature amount of the temperature difference distribution data on the surface of the substrate 10. Generate a first trained model that estimates the quality of the soldering. The first trained model generation unit 55 stores the generated first trained model in the first trained model storage unit 72.

The first trained model storage unit 72 stores the first trained model.
The first trained model generation unit 55 generates the first trained model by so-called supervised learning according to, for example, a support vector machine. Here, in supervised learning, by giving the first learning data consisting of a set of input and result (label) data to the first trained model generation unit 55, the features of the first learning data can be obtained. A method of learning and inferring results from input.

FIG. 5 is a diagram showing an identification image in the support vector machine. In FIG. 5, for convenience of explanation, it is represented in two dimensions.

The black circles represent the feature amounts of the temperature difference distribution data of the substrate 10 included in the first learning data, which is determined to be defective in soldering. The triangular mark represents the feature amount of the temperature difference distribution data on the surface of the substrate 10 which is included in the first learning data and is judged to be good for soldering.

In the support vector machine, based on the first learning data in the learning phase, the feature amount of the temperature difference distribution data on the surface of the substrate 10 determined to be defective in soldering and the substrate 10 determined to be good in soldering are used. We construct a discriminant surface that distinguishes from the features of the temperature difference distribution data on the surface of the solder.

In the inference phase, the support vector machine identifies which of the two spaces divided by the identification surface belongs to the feature amount of the temperature difference distribution data on the surface of the substrate 10 to be identified. The quality of soldering of the substrate 10 is determined. The asterisk represents an example of the feature amount vector of the temperature difference distribution data of the substrate 10 to be identified. In this example, the substrate 10 to be identified is determined to be defective.

FIG. 6 is a diagram showing the configuration of the second learning device 23.
The second learning device 23 includes a data acquisition unit 61, a control parameter setting unit 62, a simulation unit 63, a second learning data creation unit 64, and a second trained model generation unit 65.

The data acquisition unit 61 acquires temperature difference distribution data on the surface of the substrate 10 when soldering is defective from the first learning device 22.

The control parameter setting unit 62 sets the control parameters. In the learning phase, the control parameter setting unit 62 corrects the control parameter in a plurality of steps from the standard value when the soldering of the substrate 10 becomes defective. The control parameter setting unit 62 raises the temperature of the solder bath heater 14 by 5 ° C, 10 ° C, and 15 ° C higher than the standard value when the control parameter is the temperature of the solder tank heater 14. The control parameter setting unit 62 makes the rotation speed of the jet motor 15 10%, 20%, and 30% faster than the standard value when the control parameter is the rotation speed of the jet motor 15. The control parameter setting unit 62 increases the angle of the jet nozzle 13 by 10 °, 20 °, and 30 ° when the control parameter is the angle of the jet nozzle 13.

The simulation unit 63 obtains the temperature distribution data on the surface of the substrate 10 after the set time has elapsed after the control parameters have been corrected by simulation, based on the modified control parameters output from the control parameter setting unit 62. .. The simulation unit 63 includes the "temperature distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters have been corrected" obtained by the simulation and the "board 10" stored in the target temperature distribution storage unit 76. The "temperature difference distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters have been corrected" indicating the difference from the "target temperature distribution data on the surface of the substrate 10" is output.

The second learning data creation unit 64 includes the temperature difference distribution data on the surface of the substrate 10 when the control parameter is a standard value, the correction amount of the control parameter, and the time lapse set after the control parameter is corrected. A second learning data including a plurality of sets of temperature difference distribution data (correct answer) on the surface of the substrate 10 is created and stored in the second learning data storage unit 74.

The second learning data storage unit 74 stores the second learning data.
The second trained model generation unit 65 uses the second training data stored in the second training data storage unit 74, and the temperature difference distribution data on the surface of the substrate 10 when the control parameter is a standard value. A second trained model is generated that estimates the temperature difference distribution data on the surface of the substrate 10 after the set time has elapsed after the control parameters have been modified. The second trained model generation unit 65 stores the generated second trained model in the second trained model storage unit 75.

The second trained model storage unit 75 stores the second trained model.
As the learning algorithm used by the second trained model generation unit 65, known algorithms such as supervised learning, unsupervised learning, and reinforcement learning can be used. As an example, a case where a neural network is applied will be described.

The second trained model generation unit 65 executes so-called supervised learning according to, for example, a neural network model. Here, in supervised learning, by giving a set of data of input and result (label) to the second trained model generation unit 65, the features in those learning data are learned, and the result is obtained from the input. A method of inferring.

FIG. 7 is a diagram showing an example of the configuration of a neural network.
A neural network is composed of an input layer composed of a plurality of neurons, an intermediate layer (hidden layer) composed of a plurality of neurons, and an output layer composed of a plurality of neurons. The intermediate layer may be one layer or two or more layers.

For example, in the case of a three-layer neural network as shown in FIG. 7, when a plurality of inputs are input to the input layer (X1 to Xn), the value is multiplied by the weight W1 to the intermediate layer (Y1 to Ym). It is input, and the result is further multiplied by the weight W2 and output from the output layer (Z1 to Zs). This output result depends on the values of the weights W1 and W2.

In the present embodiment, the neural network outputs the temperature difference distribution data on the surface of the substrate 10 after the set time elapses after the control parameters are corrected by so-called supervised learning according to the first learning data. learn.

That is, in the neural network, the temperature difference distribution data on the surface of the substrate 10 when the control parameter is a standard value and the correction amount of the control parameter are input to the input layer, and the result output from the output layer is the control parameter. The learning is executed by adjusting the weights W1 and W2 so as to approach the temperature difference distribution data (correct answer) on the surface of the substrate 10 after the set time has elapsed after the correction.

FIG. 8 is a diagram showing the configuration of the first inference device 24.
The first inference device 24 includes a control parameter setting unit 36, a temperature distribution data creation unit 31, a temperature difference distribution data creation unit 32, a feature amount extraction unit 33, a first estimation unit 34, and a data output unit. 35 and.

The control parameter setting unit 62 sets the control parameter to a standard value.
Similar to the temperature distribution data creating unit 51 in the first learning device 22, the temperature distribution data creating unit 31 acquires information on the temperature distribution on the surface of the substrate 10 from the temperature measuring device 17, and the temperature on the surface of the substrate 10 Create distribution data.

The temperature difference distribution data creating unit 32, like the temperature difference distribution data creating unit 52 in the first learning device 22, has the temperature distribution data on the surface of the substrate 10 and the substrate stored in the target temperature distribution storage unit 76. The temperature difference distribution data on the surface of the substrate 10 representing the difference from the target temperature distribution data on the surface of 10 is created.

The feature amount extraction unit 33 extracts the feature amount of the temperature difference distribution data on the surface of the substrate 10 in the same manner as the feature amount extraction unit 53 in the first learning device 22.

The first estimation unit 34 reads out the first trained model that estimates the quality of soldering of the substrate 10 from the feature amount of the temperature difference distribution data on the surface of the substrate 10 from the first trained model storage unit 72. The first estimation unit 34 inputs the feature amount of the temperature difference distribution data on the surface of the substrate 10 output from the feature amount extraction unit 33 into the first trained model, thereby determining whether the soldering of the substrate 10 is good or bad. Get the data to represent.

The data output unit 35 is the substrate 10 when the control parameter generated by the temperature difference distribution data creation unit 32 is a standard value when the soldering of the substrate 10 is estimated to be defective by the first estimation unit 34. Output the surface temperature difference distribution data.

FIG. 9 is a diagram showing the configuration of the second inference device 25.
The second inference device 25 includes a data acquisition unit 41, a second estimation unit 42, and a control parameter setting unit 43.

The control parameter setting unit 43 sets the control parameter to a standard value or a value corrected from the standard value by a correction amount.

The data acquisition unit 41 acquires the temperature difference distribution data on the surface of the substrate 10 when the control parameter output from the data output unit 35 of the first inference device 24 is a standard value.

The second estimation unit 42 corrects the control parameters from the second learned model storage unit 75 based on the temperature difference distribution data on the surface of the substrate 10 when the control parameters are standard values and the correction amount of the control parameters. A second trained model for estimating the temperature difference distribution data on the surface of the substrate 10 after the lapse of a set time is read out.

The second estimation unit 42 is modified by the temperature difference distribution data on the surface of the substrate 10 when the control parameter output from the data acquisition unit 41 to the second trained model is a standard value, and the control parameter setting unit 43. By inputting the correction amount of the control parameter, the temperature difference distribution data on the surface of the substrate 10 after the set time elapses after the control parameter is corrected is obtained.

FIG. 10 is a flowchart showing a learning procedure by the first learning device 22 and the second learning device 23 in the first embodiment.

In step S101, a drive switch (not shown) is set to ON. As a result, the substrate 10 is transported to the upper part of the solder tank 5 by the transport mechanism 2.

In step S102, the control parameter setting unit 62 sets the control parameter to a standard value.

In step S103, the temperature distribution data creation unit 51 creates temperature distribution data on the surface of the substrate 10.

In step S104, the temperature difference distribution data creating unit 52 represents the difference between the temperature distribution data on the surface of the substrate 10 and the target temperature distribution data on the surface of the substrate 10 stored in the target temperature distribution storage unit 76. Create temperature difference distribution data on the surface of.

In step S105, the feature amount extraction unit 53 extracts the feature amount of the temperature difference distribution data on the surface of the substrate 10.

In step S106, the soldering inspection device 16 inspects the quality of soldering of the substrate 10.

In step S107, the first learning data creating unit 54 features the feature amount of the temperature difference distribution data on the surface of the substrate 10 obtained in step S105, and the label of the quality of soldering of the substrate 10 obtained in step S106. The set with the correct answer) is added to the first learning data in the first learning storage unit 71.

When the end instruction is input in step S108, the process proceeds to step S109. When the end instruction is not input, the process returns to step S103, and the process of steps S103 to S107 is repeated for the next board 10 to be conveyed.

In step S109, a drive switch (not shown) is set to off. As a result, the transfer of the substrate 10 by the transfer mechanism 2 is stopped.

In step S110, the first trained model generation unit 55 uses the first training data stored in the first training data storage unit 71 from the feature amount of the temperature difference distribution data on the surface of the substrate 10. A first trained model for estimating the quality of soldering of the substrate 10 is generated. The first trained model generation unit 55 stores the generated first trained model in the first trained model storage unit 72.

In step S111, the data acquisition unit 61 of the second learning device 23 acquires one temperature difference distribution data on the surface of the substrate 10 when soldering is defective from the first learning device 22.

In step S112, the control parameter setting unit 62 corrects the control parameters from the standard values by ΔP, 2 × ΔP, 3 × ΔP, ... And n × ΔP.

In step S113, the simulation unit 63 has the temperature of the surface of the substrate 10 after the time set after the control parameters have been corrected by the simulation based on the modified control parameters output from the control parameter setting unit 62. Obtain distribution data. The simulation unit 63 includes the temperature distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters have been modified, and the target temperature distribution data on the surface of the substrate 10 stored in the target temperature distribution storage unit 76. The temperature difference distribution data on the surface of the substrate 10 after the set time has elapsed after the control parameter representing the difference between the two has been corrected is output. For the modified n control parameters, the temperature difference distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters are modified is output.

In step S114, the second learning data creation unit 64 determines the temperature difference distribution data on the surface of the substrate 10 when the control parameter acquired in step S111 is a standard value, the correction amount of the control parameter in step S112, and step S113. The set with the temperature difference distribution data (correct answer) on the surface of the substrate 10 after the lapse of time set after the control parameters obtained in the above are corrected is used as the second learning data in the second learning data storage unit 74. to add. The above set is added to the second training data for the modified n control parameters.

When the end instruction is input in step S115, the process proceeds to step S116. When the end instruction is not input, the process returns to step S111, and the process of steps S111 to S114 is repeated for another substrate 10 whose soldering is defective.

In step S116, the second trained model generation unit 65 uses the second training data stored in the second training data storage unit 74 to display the surface of the substrate 10 when the control parameter is a standard value. From the temperature difference distribution data and the correction amount of the control parameter, a second trained model for estimating the temperature difference distribution data on the surface of the substrate 10 after the set time elapses after the control parameter is corrected is generated. The second trained model generation unit 65 stores the generated second trained model in the second trained model storage unit 75.

FIG. 11 is a flowchart showing the inference procedure by the first inference device 24 and the second inference device 25 in the first embodiment.

In step S201, a drive switch (not shown) is set to ON. As a result, the substrate 10 is transported to the upper part of the solder tank 5 by the transport mechanism 2.

In step S202, the control parameter setting unit 36 sets the control parameter to a standard value.

In step S203, the temperature distribution data creation unit 31 creates temperature distribution data on the surface of the substrate 10.

In step S204, the temperature difference distribution data creating unit 32 represents the difference between the temperature distribution data on the surface of the substrate 10 and the target temperature distribution data on the surface of the substrate 10 stored in the target temperature distribution storage unit 76. Create temperature difference distribution data on the surface of.

In step S205, the feature amount extraction unit 33 extracts the feature amount of the temperature difference distribution data on the surface of the substrate 10.

In step S206, the first estimation unit 34 estimates the quality of soldering of the substrate 10 from the feature amount of the temperature difference distribution data on the surface of the substrate 10 from the first learned model storage unit 72. Read the model.

In step S207, the first estimation unit 34 solders the substrate 10 by inputting the feature amount of the temperature difference distribution data on the surface of the substrate 10 output from the feature amount extraction unit 33 into the first trained model. Obtain data indicating the quality of the soldering.

If the soldering of the substrate 10 is defective in step S208, the process proceeds to step S209. If the soldering of the substrate 10 is good, the process ends.

In step S209, the data acquisition unit 41 acquires the temperature difference distribution data on the surface of the substrate 10 when the control parameter created by the temperature difference distribution data creation unit 32 in step S204 is a standard value.

In step S210, the second estimation unit 42 obtains from the second learned model storage unit 75, the temperature difference distribution data on the surface of the substrate 10 when the control parameter is a standard value, and the correction amount of the control parameter. A second trained model that estimates the temperature difference distribution data on the surface of the substrate 10 after the set time has elapsed after the control parameters have been modified is read out.

In step S211 the second estimation unit 42 is set to K = 1.
In step S212, the second estimation unit 42 sets the correction amount of the control parameter to K × ΔP.

In step S213, the second estimation unit 42 corrects the temperature difference distribution data on the surface of the substrate 10 and the control parameters when the control parameters output from the data acquisition unit 41 to the second trained model are standard values. By inputting the quantity, the temperature difference distribution data of the substrate after the set time elapses after the control parameters are modified is obtained.

In step S214, in the second estimation unit 42, the temperature difference at each position of the temperature difference distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters are corrected is the allowable temperature difference distribution data storage unit. When the temperature difference is equal to or less than the temperature difference at each position of the allowable temperature difference distribution data in 73, it is determined that the temperature difference distribution data on the surface of the substrate 10 is within the allowable range. In the second estimation unit 42, the temperature difference at each position of the temperature difference distribution data on the surface of the substrate 10 after the lapse of time set after the control parameter is corrected is the temperature difference at each position of the allowable temperature difference distribution data. When it exceeds, it is determined that the temperature difference distribution data on the surface of the substrate 10 is out of the allowable range. When the temperature difference distribution data on the surface of the substrate 10 is within the permissible range, the process proceeds to step S216, and when it is out of the permissible range, the process proceeds to step S215.

In step S215, the second estimation unit 42 increments K by 1. After that, the process returns to step S211.

In step S216, the control parameter setting unit 43 corrects the control parameter by the amount of correction. The corrected control parameter is output to the drive control unit 26. The drive control unit 26 drives the soldering system 1 based on the modified control parameters. This causes, for example, dissolution of the accumulated dross or movement of the accumulated dross position. Then, it is uniformly ejected from the jet nozzle 13, and the temperature distribution on the surface of the substrate 10 falls within the allowable temperature distribution.

Embodiment 2.
FIG. 12 is a flowchart showing the learning procedure by the first learning device 22 and the second learning device 23 in the second embodiment. Since steps S101 to S111 in the flowchart of FIG. 12 are the same as steps S101 to S111 in the flowchart of FIG. 10, the description will not be repeated.

In step S312, the control parameter setting unit 62 sets K = 1.
In step S313, the control parameter setting unit 62 corrects the control parameter from the standard value by K × ΔP.

In step S314, the simulation unit 63 determines the temperature of the surface of the substrate 10 after the time set after the control parameters have been corrected by simulation based on the modified control parameters output from the control parameter setting unit 62. Obtain distribution data. The simulation unit 63 includes the temperature distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters have been modified, and the target temperature distribution data on the surface of the substrate 10 stored in the target temperature distribution storage unit 76. The temperature difference distribution data on the surface of the substrate 10 after the set time has elapsed after the control parameter representing the difference between the two has been corrected is output.

In step S315, in the second learning data creation unit 64, the temperature difference at each position of the temperature difference distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters are corrected is the allowable temperature difference distribution data. When the temperature difference is equal to or less than the temperature difference at each position of the allowable temperature difference distribution data in the storage unit 73, it is determined that the temperature difference distribution data on the surface of the substrate 10 is within the allowable range. In the second learning data creation unit 64, the temperature difference at each position of the temperature difference distribution data on the surface of the substrate 10 after the lapse of time set after the control parameters are corrected is the temperature difference at each position of the allowable temperature difference distribution data. When the temperature difference is exceeded, it is determined that the temperature difference distribution data on the surface of the substrate 10 is out of the allowable range. When the temperature difference distribution data on the surface of the substrate 10 is within the permissible range, the process proceeds to step S317, and when it is out of the permissible range, the process proceeds to step S316.

In step S316, the control parameter setting unit 62 increments K by 1. After that, the process returns to step S313.

In step S317, the second learning data creation unit 64 determines the temperature difference distribution data on the surface of the substrate 10 when the control parameter acquired in step S111 is a standard value, and the correction amount of the control parameter set in step S313. The set with the correct answer) is added to the second learning data in the second learning data storage unit 74.

When the end instruction is input in step S318, the process proceeds to step S319. When the end instruction is not input, the process returns to step S111, and the processes of steps S111 and S312 to S317 are repeated for another substrate 10 whose soldering is defective.

In step S319, the second trained model generation unit 65 uses the second training data stored in the second training data storage unit 74 to display the surface of the substrate 10 when the control parameter is a standard value. From the temperature difference distribution data, a second trained model for estimating the correction amount of the control parameter is generated. The second trained model generation unit 65 stores the generated second trained model in the second trained model storage unit 75.

FIG. 13 is a flowchart showing the inference procedure by the first inference device 24 and the second inference device 25 in the second embodiment. Since steps S201 to S209 in the flowchart of FIG. 13 are the same as steps S101 to S209 in the flowchart of FIG. 11, the description will not be repeated.

In step S410, the second estimation unit 42 estimates the correction amount of the control parameter from the temperature difference distribution data on the surface of the substrate 10 when the control parameter is a standard value from the second learned model storage unit 75. Read out the trained model of 2.

In step S411, the second estimation unit 42 inputs the temperature difference distribution data on the surface of the substrate 10 when the control parameter output from the data acquisition unit 41 is a standard value into the second trained model. Obtain the correction amount of the control parameter.

In step S412, the control parameter setting unit 43 corrects the control parameter by the amount of correction.

Embodiment 3.
FIG. 14 is a flowchart showing the procedure of inference and re-learning by the first inference device 24 and the second inference device 25 in the third embodiment. Since steps S201 to S209 and S410 to S412 in the flowchart of FIG. 14 are the same as steps S201 to S209 and S410 to S412 in the flowchart of FIG. 13, the description will not be repeated.

Step S501 and step S502 are executed after step S207 and before step S208.

In step S501, the first learning data creation unit 54 features the feature amount of the temperature difference distribution data on the surface of the substrate 10 obtained in step S205 and the label (correct answer) of the soldering quality of the substrate obtained in step S207. ) Is added to the first learning data in the first learning data storage unit 71.

In step S502, the first trained model generation unit 55 uses the first training data stored in the first training data storage unit 71 from the feature amount of the temperature difference distribution data on the surface of the substrate 10. The first trained model for estimating the quality of soldering of the substrate 10 is updated. The first trained model generation unit 55 stores the updated first trained model in the first trained model storage unit 72.

By step S501 and step S502, as the operation of the soldering system 1 is continued, the accumulated amount of the first learning data increases, and as the accumulated amount of the first learning data increases, the first learned data is completed. The estimation accuracy of the model is high.

After step S412, steps S503 and S504 are executed.
In step S503, the second learning data creation unit 64 determines the temperature difference distribution data on the surface of the substrate 10 when the control parameter acquired in step S204 is a standard value, and the correction amount of the control parameter obtained in step S411. The set with the correct answer) is added to the second learning data in the second learning data storage unit 74.

In step S504, the second trained model generation unit 65 uses the second training data stored in the second training data storage unit 74 to display the surface of the substrate 10 when the control parameter is a standard value. From the temperature difference distribution data, the second trained model that estimates the correction amount of the control parameter is updated. The second trained model generation unit 65 stores the updated second trained model in the second trained model storage unit 75.

By step S503 and step S504, the accumulated amount of the second learning data increases as the operation of the soldering system 1 is continued, and the second learned data is accumulated as the accumulated amount of the second learning data increases. The estimation accuracy of the model is high.

Embodiment 4.
In this embodiment, the temperature measuring device 17 includes an infrared camera 7.

FIG. 15 is a diagram showing the arrangement of the infrared camera 7 in the fourth embodiment.
As shown in FIG. 15, the transfer mechanism 2 conveys the substrate 10 in a direction different from the horizontal direction by a certain angle (5 ± 1 °).

The infrared camera 7 is arranged so that the direction of the optical axis K of the infrared camera 7 is perpendicular to the surface of the substrate 10 to be conveyed. That is, the direction of the optical axis of the infrared camera 7 is a direction different from the horizontal direction by 90 ° − (5 ± 1) °.

In the horizontal direction, the infrared camera 7 is located between the position PA at 220 [mm] in the horizontal direction upstream from the rear end PN of the secondary nozzle 13b and the position PB at 300 [mm] in the horizontal direction downstream. Is installed. The reason why the installation range of the infrared camera 7 is widened in this way is that the infrared camera 7 may not be installed at the position of the rear end PN of the secondary nozzle 13b due to the difference in the structure of the soldering system 1. Is.

By installing the infrared camera 7 as described above, it is possible to exclude the distortion of the thermal image taken by the infrared camera 7 and perform accurate measurement. Further, by installing the infrared camera 7, it is possible to reduce the area behind the parts mounted on the substrate 10, so that it becomes easy to secure the area required for temperature measurement.

Embodiment 5.
FIG. 16 is a diagram showing a time change of the surface temperature of the substrate 10.

The solid line represents the temperature change on the surface of the substrate 10 when the contact time between the substrate 10 and the molten solder is long. At time t1, the substrate 10 is separated from the molten solder.

At time t2, where the broken line represents the temperature change on the surface of the substrate 10 when the contact time between the substrate 10 and the melting is short, the substrate 10 is separated from the jet melting.

After the substrate 10 is separated from the molten solder, the temperature difference on the surface of the substrate 10 is larger due to the difference in the contact time between the substrate 10 and the molten solder. By measuring the temperature of the substrate 10 when the substrate 10 is separated from the molten solder, the difference in the amount of heat input of the substrate 10 can be clearly detected.

FIG. 17A is a diagram showing a state when the substrate 10 comes into contact with the molten solder.
FIG. 17B is a diagram showing a state immediately after the substrate 10 is separated from the molten solder.

In the present embodiment, the temperature measuring device 17 measures the temperature of the upper surface of the substrate 10 after the substrate 10 is separated from the molten solder jetted from the jet nozzle 13.

Embodiment 6.
FIG. 18 is a schematic view showing the soldering system 1b of the sixth embodiment.

A hole 151 is formed in the exterior 91 of the soldering system 1. The infrared camera 7 included in the temperature measuring device 17 is arranged outside the exterior 91 of the soldering system 1b, and the lens of the infrared camera 7 is arranged in the portion of the hole 151. As a result, it is possible to prevent the solvent of the volatilized flux from adhering to the infrared camera 7, so that accurate temperature measurement can be performed.

Since the infrared camera 7 is located outside the soldering system 1b, the influence of the temperature change of the infrared camera 7 at the time of maintenance or the like can be reduced. In a general infrared camera, if the temperature of the camera body changes, the measured temperature also changes, but by adopting this configuration, more accurate measurement can be performed.

Embodiment 7.
FIG. 19 is a diagram showing the configuration of the soldering system 1a according to the seventh embodiment.

The soldering system 1a of the seventh embodiment includes a protective window 98.
The protective window 98 is provided to protect the lens of the infrared camera 7. By doing so, it is possible to prevent the flux from adhering to the infrared camera 7 even during long-term operation, so that the infrared rays are not blocked and accurate temperature measurement becomes possible.

Embodiment 8.
FIG. 20 is a diagram showing a hardware configuration of the control device 9.

The control device 9 can configure the corresponding operation with the hardware or software of the digital circuit. When the function of the control device 9 is realized by using software, the control device 9 includes, for example, as shown in FIG. 20, a processor 1002 and a memory 1003 connected by a bus 1001. The memory 1003 includes a ROM (Read Only Memory) and a RAM (Random Access Memory). The program executed by the processor 1002 and the data necessary for executing the program are stored in the ROM. The data created during program execution is stored in RAM.

Modification example.
The present disclosure includes the following modifications.

(1) Position of the infrared camera By installing the infrared camera included in the temperature measuring device 17 away from directly under the exhaust duct (not shown) provided in the soldering system 1, flux vapor is generated in the lens of the infrared camera. It may be prevented from adhering.

(2) Temperature distribution data creation The temperature distribution data creation unit divided the surface of the substrate 10 into a 1000 × 1000 two-dimensional matrix. The division of the surface of the substrate 10 is not limited to this. The temperature distribution data creating unit may divide the surface of the substrate 10 according to the size of the substrate 10, the analysis accuracy of the temperature distribution, the control accuracy of the soldering system 1, and the like. For example, the temperature distribution data creation unit may divide the surface of the substrate 10 into 50 × 50, 100 × 800, 7000 × 3000, and the like.

(3) Soldering inspection The quality of soldering may be evaluated by an operator.

(4) Simulation The method for determining the simulation result is not limited to that described in the above embodiment. For example, the temperature distribution obtained by the simulation may be input to the support vector machine to identify whether the group belongs to the defective soldering group or the normal soldering group.

When the accuracy of the first trained model reaches a certain value or more, the re-learning of the first trained model may be stopped. When the accuracy of the second trained model reaches a certain value or more, the retraining of the second trained model may be stopped. By stopping the re-learning, the processing load on the control device 9 can be reduced.

(5) Input of the first trained model In the above embodiment, the input of the first trained model is a feature amount of the temperature difference distribution data, but the input is not limited thereto. The input of the first trained model may be data representing the temperature distribution of the substrate. For example, the input of the first trained model may be the temperature distribution data created by the temperature distribution data creation unit or the temperature difference distribution data created by the temperature difference distribution data creation unit.

(6) Input of the second trained model In the above embodiment, the input of the second trained model is the temperature difference distribution data, but the input is not limited to this. The input of the second trained model may be data representing the temperature distribution of the substrate. For example, the input of the second trained model may be the temperature distribution data created by the temperature distribution data creation unit.

(7) Temperature measuring device As the measuring method of the temperature measuring device, a laser method, an ultrasonic method, or an electromagnetic wave method can be used. The temperature measuring device may include a radiation thermometer instead of an infrared camera.

(8) Learning In the present embodiment, the case where supervised learning is applied to the learning algorithms used by the first trained model generation unit and the second trained model generation unit has been described, but the present invention is not limited to this. do not have. As for the learning algorithm, it is also possible to apply reinforcement learning, unsupervised learning, semi-supervised learning, etc. in addition to supervised learning.

The first trained model generation unit and the second trained model generation unit are the first trained model using the first training data and the second training data created in the plurality of soldering systems. A second trained model may be generated. The first trained model generation unit and the second trained model generation unit may acquire the first training data and the second training data from a plurality of soldering systems used in the same area. , The first training data and the second training data collected from a plurality of soldering systems operating independently in different areas may be used.

It is also possible to add or remove the soldering system that collects the first trained model and the second trained model to the target on the way. Furthermore, the first trained model and the second trained model generated for one soldering system are applied to another soldering system, and the first trained model and the second trained model are obtained by retraining. You may try to update.

As the learning algorithm used in the first trained model generation unit and the second trained model generation unit, deep learning that learns the extraction of the feature amount itself can also be used, and other known ones are known. Machine learning may be performed according to methods such as genetic programming, functional logic programming, or support vector machines.

The first learning device, the first inference device, the second learning device, and the second inference device may be built in the soldering system. Further, the first learning device, the first inference device, the second learning device, and the second inference device may exist on the cloud server.

The first inference device and the second inference device acquire the first trained model and the second trained model from the outside such as other soldering systems, and execute inference based on these. May be good.

(9) Combination of Embodiments Each of the above embodiments may be arbitrarily combined and implemented.

The embodiments disclosed this time should be considered to be exemplary in all respects and not restrictive. The scope of the present disclosure is shown by the scope of claims rather than the above description, and is intended to include all modifications within the meaning and scope of the claims.

1,1a, 1b soldering system, 2 transfer mechanism, 3 flux coating machine, 4 preheater, 5 solder tank, 7 infrared camera, 9 control device, 10 board, 11 flux, 13 jet nozzle, 13a primary nozzle, 13b 2 Next nozzle, 14 solder bath heater, 15 jet motor, 16 soldering inspection device, 17 temperature measuring device, 18 fluxer control unit, 21 identification and setting unit, 22 first learning device, 23 second learning device, 24th 1 inference device, 25 2nd inference device, 26 drive control unit, 27 storage device, 31,51 temperature distribution data creation unit, 32,52 temperature difference distribution data creation unit, 33,53 feature quantity extraction unit, 34th 1 estimation unit, 35 data output unit, 36, 43, 62 control parameter setting unit, 41, 61 data acquisition unit, 42 second estimation unit, 54 first training data creation unit, 55 first trained model Generation unit, 63 simulation unit, 64 second training data creation unit, 65 second trained model generation unit, 71 first training data storage unit, 72 first trained model storage unit, 73 allowable temperature difference distribution. Storage unit, 74 second learning data storage unit, 75 second learned model storage unit, 76 target temperature distribution storage unit, 91 exterior, 98 protective window, 151 hole, 201 electronic parts, 1001 bus, 1002 processor, 1003 Memory, K optical axis.

Claims

A flux coating machine that applies flux to the substrate,
A preheater that preheats the substrate and
A solder bath for storing molten solder and
A solder bath heater that melts the solder in the solder tank, and
A jet nozzle that ejects the molten solder in the solder bath toward the substrate, and a jet nozzle.
A transport mechanism for sequentially transporting the substrate above the flux coating machine, above the preheater, and above the solder bath.
A soldering system comprising a temperature measuring device arranged above the jet nozzle.
By the transport mechanism, the substrate is transported in a direction different from the horizontal direction by a certain angle.
The temperature measuring device includes an infrared camera.
The soldering system according to claim 1, wherein the infrared camera is arranged so that the direction of the optical axis of the infrared camera is perpendicular to the surface of the substrate.
The soldering system according to claim 1 or 2, wherein the temperature measuring device measures the temperature of the upper surface of the substrate after the substrate is separated from the molten solder jetted from the jet nozzle.
Holes are formed in the exterior of the soldering system.
The soldering system according to claim 2, wherein the infrared camera is arranged in a portion of the hole outside the exterior of the soldering system.
The soldering system according to claim 2, wherein the infrared camera includes a protective window for protecting the lens of the infrared camera.
The soldering system according to claim 2, wherein the infrared camera includes a swing mechanism capable of changing the direction of the optical axis of the infrared camera.
A first storage unit that stores the first trained model that estimates the quality of soldering of the board from the data representing the temperature distribution of the board, and
Data representing the temperature distribution of the substrate conveyed by the transfer mechanism is generated based on the temperature measured by the temperature measuring device, and the first trained model stored in the first storage unit is used. Further, any one of claims 1 to 6, further comprising a first inference device for estimating the quality of soldering of the substrate conveyed by the transfer mechanism from the generated data representing the temperature distribution of the substrate. The soldering system according to item 1.
A drive control unit that controls the soldering system based on control parameters,
A second storage unit that stores a second trained model that estimates a correction amount of the control parameter from data representing the temperature distribution of the substrate when the control parameter is a standard value, and a second storage unit.
When the control parameter is a standard value, data representing the temperature distribution of the substrate conveyed by the transfer mechanism is acquired from the temperature measured by the temperature measuring device, and the data is stored in the second storage unit. One of claims 1 to 6, further comprising a second inference device that estimates a correction amount of a control parameter from the acquired data representing the temperature distribution of the substrate using the trained model of 2. Soldering system as described in section.
A drive control unit that controls the soldering system based on control parameters,
From the data representing the temperature distribution of the substrate when the control parameter is a standard value and the amount of modification of the control parameter, the data representing the temperature distribution of the substrate after the set time elapses after the control parameter is modified is estimated. A second storage unit that stores the second trained model,
When the control parameter is a standard value, data representing the temperature distribution of the substrate conveyed by the transfer mechanism is acquired from the temperature measured by the temperature measuring device, and a correction amount of the control parameter is generated.
Using the second trained model stored in the second storage unit, the control parameter is obtained from the acquired data representing the temperature distribution of the substrate and the correction amount of the generated control parameter. The soldering system according to any one of claims 1 to 6, further comprising a second inference device for estimating data representing the temperature distribution of the substrate after the set time has elapsed after the modification.
A first set including a plurality of sets of data representing the temperature distribution of the substrate conveyed by the transfer mechanism and data indicating the quality of soldering of the substrate conveyed by the transfer mechanism indicating the inspection result by the soldering inspection device. The first learning data storage unit that stores the learning data of
Claims 1 to further include a first learning device that generates a first trained model that estimates the quality of soldering of the board from data representing the temperature distribution of the board using the first learning data. 6. The soldering system according to any one of 6.
A drive control unit that controls the soldering system based on control parameters,
The control parameter so that the data representing the temperature distribution of the substrate conveyed by the transfer mechanism when the control parameter is a standard value and the temperature distribution of the substrate conveyed by the transfer mechanism obtained by simulation are within the allowable range. A second training data storage unit that stores a second training data including a plurality of sets of correction amounts of
A second learning device that uses the second learning data to generate a second trained model that estimates the amount of modification of the control parameter from the data representing the temperature distribution of the substrate when the control parameter is a standard value. The soldering system according to any one of claims 1 to 6, further comprising.
A drive control unit that controls the soldering system based on control parameters,
Data representing the temperature distribution of the substrate conveyed by the transfer mechanism when the control parameter is a standard value, the correction amount of the control parameter, and the setting after the control parameter obtained by simulation is corrected by the correction amount. A second learning data storage unit that stores a second learning data including a plurality of sets of data representing the temperature distribution of the substrate conveyed by the transfer mechanism after the lapse of time, and a second learning data storage unit.
After the set time elapses after the control parameter is corrected from the data representing the temperature distribution of the substrate when the control parameter is a standard value and the correction amount of the control parameter using the second learning data. The soldering system according to any one of claims 1 to 6, further comprising a second learning device that generates a second trained model that estimates data representing the temperature distribution of the substrate.
The data representing the temperature distribution of the substrate is a feature amount of the temperature difference distribution data representing the difference between the temperature at the plurality of positions of the substrate measured by the temperature measuring device and the target temperature at the plurality of positions of the substrate. Item 2. The soldering system according to any one of Items 7 to 12.