Apparatus and method for optimizing control parameters of a solder paste printing device

Abstract

An electronic device of various embodiments of the present disclosure can include a communication circuit communicably connected to a solder paste printing device configured to print a solder paste on a plurality of substrates based on a plurality of control parameters, and a detection device configured to detect a state of the solder paste printed on the plurality of substrates, one or more memories, and one or more processors. The one or more processors according to various embodiments can obtain a first set of control parameters of the solder paste printing device for printing the solder paste on a first substrate, transmit information indicating the first set of control parameters, obtain first solder paste detection information indicating a state of the solder paste printed on the first substrate, determine a first yield for the first substrate based on the first solder paste detection information, and generate a model for exploring an optimal set of control parameters based on a first pair of data including the first set of control parameters and the first yield.

Description

TECHNICAL FIELD

[0001] The disclosure relates to an apparatus and a method for optimizing control parameters of a solder paste printing device. BACKGROUND

[0002] Before mounting components on a substrate (e.g., a printed circuit substrate), solder paste can be applied to pads of the substrate by a solder paste printing device (e.g., a screen printer). Then, the applied solder paste can be inspected by a solder paste inspection (SPI) device. After the inspection is completed, components can be mounted on the pads of the printed circuit substrate on which the solder paste is applied according to surface mount technology (SMT).

[0003] A plurality of control parameters is used to control an operation of the solder paste printing device to print solder paste on a substrate. The plurality of control parameters, for example, includes a control parameter for adjusting a printing pressure, a control parameter for adjusting a printing speed, and a control parameter for adjusting a separation speed. When the optimal plurality of control parameters is used, the yield (or the good product rate) of the solder paste printed on the substrate can be improved. SUMMARY

[0004] TECHNICAL PROBLEM

[0005] The disclosure provides a technology for optimizing a plurality of control parameters of a solder paste printing device.

[0006] MEANS FOR SOLVING THE PROBLEM

[0007] According to various embodiments of the disclosure, an electronic device can include: communication circuitry communicably connected to a solder paste printing device configured to print solder paste on a plurality of substrates based on a plurality of control parameters, respectively, and a detection device configured to detect a state of the solder paste printed on the plurality of substrates transferred from the solder paste printing device, one or more memories, and one or more processors operatively connected to the communication circuitry and the one or more memories. The one or more processors of the various embodiments are configured to perform operations of obtaining a first control parameter group of the solder paste printing device for printing solder paste on a first substrate, transmitting information indicating the first control parameter group to the solder paste printing device, obtaining first solder paste detection information indicating a state of the solder paste printed on the first substrate from the detection device, determining a first yield for the first substrate based on the first solder paste detection information, and generating a model for exploring an optimal control parameter group based on a first data pair including the first control parameter group and the first yield.

[0008] According to various embodiments of the disclosure, in a method of optimizing control parameters of an electronic device, the electronic device is communicably connected to a solder paste printing apparatus configured to print a solder paste on a plurality of substrates based on a plurality of control parameters and a detection apparatus configured to detect a state of the solder paste printed on the plurality of substrates transferred from the solder paste printing apparatus, the method includes obtaining a first control parameter set of the solder paste printing apparatus for printing the solder paste on a first substrate, transmitting information indicating the first control parameter set to the solder paste printing apparatus, obtaining first solder paste detection information indicating a state of the solder paste printed on the first substrate from the detection apparatus, determining a first yield for the first substrate based on the first solder paste detection information, and generating a model for exploring an optimal control parameter set based on a first data pair including the first control parameter set and the first yield.

[0009] Effects of Invention

[0010] According to various embodiments of the disclosure, a model for exploring an optimal control parameter set is generated and updated based on real-time collected data (e.g., control parameter set and yield), thereby not requiring the model to be generated based on historical data and performing optimization. In addition, since variations and states of a solder paste, a stencil, and a squeegee affecting a solder paste printing process can change over time, historical data cannot reflect a current state of the solder paste printing process. According to various embodiments of the disclosure, a surrogate model for exploring an optimal control parameter set corresponding to a current state of the solder paste printing process can be generated and updated based on real-time collected data.

[0011] The model according to various embodiments of the disclosure only outputs (or explores) a control parameter set for obtaining a yield above a pre-set threshold value. Accordingly, even if optimization of a control parameter set is performed in real time during execution of a solder paste printing process on a plurality of substrates, a yield above a threshold value can be ensured for the solder paste printing process performed in a process in which the corresponding optimization is performed.

[0012] According to various embodiments of the disclosure, only a pre-set number of control parameter sets corresponding to a pre-set number of substrates can be explored, and then an optimal control parameter set is determined from among the explored control parameter sets to be used, thereby being able to prevent an excessive number of substrates from being required when exploring control parameter sets. Depending on the situation, an additional judgment criterion can be used to adjust the number of substrates to be explored.

[0013] According to various embodiments of the present disclosure, since optimization of the control parameter set is performed in real time during the solder paste printing process for a plurality of substrates, yield reduction that can occur due to sudden changes in the solder paste printing process environment can be minimized. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 FIG. 1 is a diagram illustrating an operation process of an electronic device according to an embodiment of the present disclosure.

[0015] Figure 2 FIG. 2 is a block diagram of an electronic device according to an embodiment of the present disclosure.

[0016] Figure 3 FIG. 3 is a diagram for explaining a model according to an embodiment of the present disclosure.

[0017] Figure 4a FIG. 4 is a diagram for explaining a process of exploring an optimal control parameter set according to an embodiment of the present disclosure. Figure 4b

[0018] FIG. 5 is a diagram for explaining a method of considering stability in a process for exploring an optimal control parameter set according to an embodiment of the present disclosure. Figure 5

[0019] FIG. 6 is a flowchart of an operation of an electronic device according to an embodiment of the present disclosure. Figure 6

[0020] FIG. 7 is a flowchart of an operation of an electronic device according to an embodiment of the present disclosure. Figure 7 DETAILED DESCRIPTION

[0021] Embodiments of the present disclosure are exemplified for the purpose of explaining the technical idea of the present disclosure. The scope of protection of the present disclosure is not limited to the embodiments given below or specific descriptions of the embodiments.

[0022] Unless otherwise defined, technical and scientific terms used in the present disclosure have the same meaning as would be understood by one of ordinary skill in the art to which the present disclosure belongs. All terms used in the present disclosure are selected from general knowledge in the art for the purpose of more specifically explaining the present disclosure, and are not intended to limit the scope of protection of the present disclosure.

[0023] As for the expressions "include", "have" and "possess" and the like used in the present disclosure, unless otherwise noted in the sentence or paragraph in which the corresponding expression is used, it should be understood as an open-ended term that can include other embodiments.

[0024] Unless otherwise noted, singular expressions described in the present disclosure can include the meaning of the plural, and the same applies to singular expressions described in the claims.

[0025] ​The expressions "first", "second", and so on used in the present disclosure are used in order to distinguish a plurality of constituent elements, and are not used to limit the order or importance of the corresponding constituent elements.

[0026] The term "unit" used in the present disclosure means a hardware constituent element such as software, a field-programmable gate array (FPGA), and an application specific integrated circuit (ASIC). However, the "unit" is not limited to hardware and software. The "unit" can be configured to be located in an addressable storage medium, and can be configured to operate one or more processors. Accordingly, as an example, the "unit" includes constituent elements such as a software constituent element, an object-oriented software constituent element, a class constituent element, and a task constituent element; and a processor, a function, a property, a procedure, a subroutine, a program code segment, a driver, a firmware, a microcode, a circuit, data, a database, a data constituent, a platform, an array, and a variable. The functions provided within the constituent elements and the "unit" can be combined into a smaller number of constituent elements and "units" or further separated into additional constituent elements and "units".

[0027] The expression "based on" used in the present disclosure is used to describe one or more factors to which a determination, a judgment action, or an operation described in a sentence or a clause including the corresponding expression is influenced, and the expression does not exclude additional factors influencing the determination, the judgment action, or the operation.

[0028] In the present disclosure, when it is mentioned that a certain constituent element is "coupled" or "connected" with other constituent elements, it is understood that the certain constituent element can be directly coupled or connected with the other constituent elements, or can be coupled or connected with the other constituent elements via a new other constituent element as a medium.

[0029] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. In the drawings, the same reference numerals are used for the same or corresponding constituent elements. Also, in the description of the following embodiments, repeated descriptions of the same or corresponding constituent elements can be omitted. However, even if the descriptions of the constituent elements are omitted, it does not mean that the constituent elements are not included in certain embodiments.

[0030] Although process steps, method steps, and algorithms are described in a sequential order, such processes, methods, and algorithms can be configured to work in any suitable order. In other words, any sequential order of process steps, method steps, and algorithms described in various embodiments of the present disclosure need not be performed in that order. Additionally, some process steps, method steps, and algorithms can be performed simultaneously. Furthermore, examples of the processes described by the figures are not exclusive of other variations and modifications that can be applied to it without departing from the disclosure. Also, the examples of the processes described by the figures are not exclusive of other variations and modifications that can be applied to it without departing from the disclosure, and none of the processes described by the figures are essential to one or more embodiments of the present disclosure, nor are the processes described by the figures preferred over other processes.

[0031] Figure 1 is a diagram representing an operation process of the electronic device 100 for optimizing a control parameter set of the solder paste printing apparatus 120 according to an embodiment of the present disclosure. The electronic device 100 of the present disclosure can determine a control parameter set of the solder paste printing apparatus 120 based on information obtained in a solder paste inspection process.

[0032] For one or more substrates, a substrate processing process can be performed sequentially. The substrate processing process can include a process of printing a solder paste on a substrate and a process of mounting and bonding a component on the substrate according to a surface mounting technology. The solder paste printing apparatus 120, the first inspection apparatus 130, the component mounting apparatus 140, the second inspection apparatus 150, an oven 160, and / or the third inspection apparatus 170 can be used in the substrate processing process. In the substrate processing process, the solder paste printing apparatus 120 can print a solder paste on a substrate. The first inspection apparatus 130 can inspect a state of the solder paste printed on the substrate. The component mounting apparatus 140 can mount a component on the substrate on which the solder paste is printed. The second inspection apparatus 150 can inspect a state of the mounted component. The second inspection apparatus 150 can be referred to as a pre-Automated Optical Inspection (pre-AOI) apparatus. The substrate on which the component is mounted can be transported to the oven 160 and undergo a reflow process. In the reflow process, the solder paste is melted and solidified again while the component is bonded to the substrate. The third inspection apparatus 170 can inspect a state of the component on the substrate after the reflow process. The third inspection apparatus 170 can be referred to as a post-Automated Optical Inspection apparatus.

[0033] The solder paste printing apparatus 120 can be configured to print the solder paste on a plurality of substrates based on a plurality of control parameters. The solder paste printing apparatus 120, for example, can be a screen printer. A control parameter set of the solder paste printing apparatus 120 used herein can mean a set of a plurality of control parameters. The plurality of control parameters, for example, can include a control parameter for adjusting a print pressure of the solder paste printing apparatus 120, a control parameter for adjusting a print speed, and a control parameter for adjusting a separation speed. Before printing the solder paste, a stencil mask having an opening is disposed on the substrate. The solder paste printing apparatus 120 prints the solder paste on the substrate while a squeegee is advanced in a prescribed direction on the substrate on which the stencil mask is disposed. The print pressure can mean a pressure applied to the substrate by the squeegee when the solder paste is printed. The print speed can mean a speed at which the squeegee is advanced when printing. The separation speed can mean a speed at which the stencil mask is separated from the substrate after printing. One or more pads can be provided on the substrate, and the pad can mean a pair of electrodes disposed on the substrate at a position at which a component is to be bonded. The solder paste printing apparatus 120 can print the solder paste on one or more pads on the substrate, respectively.

[0034] The control parameter set of the solder paste printing apparatus 120 can correspond to a point on a multi-dimensional parameter space defined by a plurality of control parameters. For example, when two control parameters are set as respective axes, the control parameter set can be expressed as a point in a two-dimensional parameter space (i.e., a plane). For example, when three control parameters are set as respective axes, the control parameter set can be expressed as a point in a three-dimensional parameter space.

[0035] According to various embodiments of the present disclosure, in order to generate a model for exploring optimal control parameters of the solder paste printing apparatus 120, a substrate (hereinafter referred to as "first substrate 111") can be subjected to a substrate processing process. An initial control parameter set (hereinafter referred to as "first control parameter set") of the solder paste printing apparatus 120 for printing the solder paste on the first substrate 111 can be set in advance and stored in the electronic device 100. In an embodiment, the first control parameter set can be set by a user. The electronic device 100 can transmit information indicating the first control parameter set to the solder paste printing apparatus 120. The information for indicating the first control parameter set, for example, can be a signal for controlling the solder paste printing apparatus 120 to perform a solder paste printing process on the first substrate 111 based on the first control parameter set.

[0036] The solder paste printing apparatus 120 according to various embodiments can receive information indicating the first control parameter group and perform a solder paste printing process on the first substrate 111 based on the first control parameter group. After printing the solder paste on the first substrate 111, the first substrate 111 can be transferred from the solder paste printing apparatus 120 to the first inspection apparatus 130. The first inspection apparatus 130 can inspect a state of the solder paste printed on the first substrate 111. The first inspection apparatus 130 can be referred to as a solder paste inspection (SPI) apparatus. The first inspection apparatus 130 can output solder paste inspection information (hereinafter referred to as "first solder paste inspection information") indicating an inspection state of the solder paste printed on the first substrate 111. The state of the solder paste can include at least one selected from a position, a direction, a volume, a height, and an area of the corresponding solder paste. The first solder paste inspection information, for example, can be a volume value of the solder paste printed on the first substrate 111.

[0037] The electronic device 100 can obtain the first solder paste inspection information from the first inspection apparatus 130. The electronic device 100 can determine (calculate) a yield (hereinafter referred to as "first yield") for the first substrate 111 based on the first solder paste inspection information. The first yield can mean a probability that the first substrate 111 on which the solder paste is coated according to the first control parameter group is produced as a good product after completion of the substrate processing process. The first yield, for example, can be determined based on the volume value of the solder paste printed on the first substrate 111. In an embodiment, the yield can be calculated as a probability distribution through a probabilistic modeling method. For example, the first yield can be a probability distribution having values of "mean: 1.96, standard deviation: 0.5". This can mean that the first substrate 111 on which the solder paste is printed has a yield of an average of 1.96 with a standard deviation of 0.5 in probability.

[0038] The electronic device 100 can generate a model based on a data pair (hereinafter referred to as a "first data pair") including a first control parameter set and a first yield. The model can be a model for exploring an optimal control parameter set. The optimal control parameter set can be a control parameter set of the solder paste printing apparatus that maximizes the yield. The first control parameter set can be set as an independent variable (cause) of the model, and the first yield can be set as an explanatory variable (result) of the model to generate the model. For example, the model can be generated on the premise that the first control parameter set and the first yield have a Gaussian distribution. According to another embodiment, a squeeze direction and a squeeze angle can also be added to the explanatory variable. The squeeze direction can mean a direction (e.g., forward, reverse) in which the squeezer of the solder paste printing apparatus 120 moves in order to apply pressure. In an embodiment, each of the squeeze directions can also be separated to generate an additional model. The squeeze angle can mean an angle that the squeezer has with respect to the substrate at the time of printing. The generated model can be stored in one or more memories of the electronic device 100. In an embodiment, the model can be a model that is generated in other apparatuses through the aforementioned process, is transferred to the electronic device 100, and is stored in the memory.

[0039] The electronic device 100 can obtain a second control parameter set of the solder paste printing apparatus 120 for printing a solder paste on a second substrate 113 from the generated model. The second substrate 113 can be, for example, a substrate that is processed in sequence after the first substrate 111 in a substrate processing process. That is, after processing the first substrate 111, a solder paste can be printed on the second substrate 113. In an embodiment, the second substrate 113 can also be a substrate that is processed after a predetermined number of substrates are processed after the first substrate 111, rather than a substrate that is processed immediately after the first substrate 111.

[0040] The electronic device 100 can transmit the obtained information indicating the second control parameter group to the solder paste printing apparatus 120. The information indicating the second control parameter group can be, for example, a signal to control the solder paste printing apparatus 120 to perform the solder paste printing process on the second substrate 113 based on the second control parameter group. The solder paste printing apparatus 120 can receive the information indicating the second control parameter group and perform the solder paste printing process on the second substrate 113 based on the second control parameter group. After the solder paste is printed on the second substrate 113, the second substrate 113 can be transferred from the solder paste printing apparatus 120 to the first inspection apparatus 130. The first inspection apparatus 130 can inspect the state of the solder paste printed on the second substrate 113. The first inspection apparatus 130 can output second solder paste inspection information indicating the inspection state of the solder paste printed on the second substrate 113. The second solder paste inspection information can be, for example, a volume value of the solder paste printed on the second substrate 113. The electronic device 100 can obtain the second solder paste inspection information from the first inspection apparatus 130 and can determine a yield (hereinafter referred to as a "second yield") for the second substrate 113 based on the first solder paste inspection information. The electronic device 100 can update the model based on a first data pair including the first control parameter group and the first yield and a second data pair including the second control parameter group and the second yield. Details of the model will be described later.

[0041] Figure 2 is a block diagram of an electronic device 100 according to an embodiment of the disclosure. The electronic device 100 can include one or more processors 210, one or more memories 220, and / or a communication circuit 230. According to various embodiments, at least one of the constituent elements of the electronic device 100 can be omitted or other constituent elements can be added. According to various embodiments, part of the constituent elements can be implemented by integration with each other, or can be implemented as a single or a plurality of individuals, additionally or alternatively. The one or more processors 210 can be referred to as a processor 210. Unless otherwise explicitly described in context, the expression processor 210 can mean a set of one or more processors. The one or more memories 220 can be referred to as a memory 220. Unless otherwise explicitly described in context, the expression memory 220 can mean a set of one or more memories. In an embodiment, at least part of the constituent elements inside and outside the electronic device 100 can be connected to each other through a bus, a general purpose input / output (GPIO), a serial peripheral interface (SPI), a mobile industry processor interface (MIPI), or the like, and receive and transmit data and / or signals.

[0042] The processor 210 can control at least one constituent element of the electronic device 100 by driving software (e.g., instructions, programs). In addition, the processor 210 can perform various calculations, processes, data generation, processing, etc. operations related to the present disclosure. In addition, the processor 210 can load data, etc. from the memory 220 or store data, etc. in the memory 220.

[0043] The memory 220 can store various data. The data stored in the memory 220 is data obtained, processed, or used by at least one constituent element of the electronic device 100, and can include software (e.g., instructions, programs, etc.). The memory 220 can include volatile and / or non-volatile memory 220. The instructions and programs in the present disclosure are software stored in the memory 220, and can include an operating system for controlling device resources, an application program, and / or middleware for providing various functions to the application program to enable the application program to apply device resources, etc. In an embodiment, the memory 220 can store instructions that cause the processor 210 to perform calculations when executed by the processor 210. In an embodiment, the memory 220 can store the above-described model. In an embodiment, the memory 220 can store the solder paste detection information and the control parameter set of the solder paste printing device 120.

[0044] The communication circuit 230 in various embodiments can perform wireless or wired communication between the electronic device 100 and a server or between the electronic device 100 and other devices. For example, the communication circuit 230 can perform wireless communication based on enhanced Mobile Broadband (eMBB), Ultra Reliable Low-Latency Communications (URLLC), Massive Machine Type Communications (MMTC), Long-Term Evolution (LTE), LTE Advance (LTE-A), New Radio (NR), Universal Mobile Telecommunications System (UMTS), Global System for Mobile communications (GSM), Code Division Multiple Access (CDMA), Wideband CDMA (WCDMA), Wireless Broadband (WiBro), Wireless Fidelity (WiFi), Bluetooth, Near Field Communication (NFC), Global Positioning System (GPS), or Global Navigation Satellite System (GNSS), etc. For example, the communication circuit 230 can perform wired communication based on Universal Serial Bus (USB), High Definition Multimedia Interface (HDMI), Recommended Standard-232 (RS-232), or Plain Old Telephone Service (POTS), etc. In an embodiment, the electronic device 100 can communicate with the solder paste printing apparatus 120 and / or the first inspection apparatus 130 through the communication circuit 230. According to an embodiment, the electronic device 100 can also communicate with the component mounting apparatus 140, the second inspection apparatus 150, the oven 160, and / or the third inspection apparatus 170 through the communication circuit 230.

[0045] In an embodiment, the electronic device 100 can exist in a form combined with at least one of the solder paste printing apparatus 120, the first inspection apparatus 130, the component mounting apparatus 140, the second inspection apparatus 150, the oven 160, and / or the third inspection apparatus 170, and can exist as an additional apparatus. When existing in a form combined with at least one of the above-described apparatuses (for example, the solder paste printing apparatus 120), the electronic device 100 can not pass through the communication circuit 230, but can directly perform reception and transmission of various information with the corresponding apparatus.

[0046] In an embodiment, the electronic device 100 can further include a user interface 240. The user interface 240 can receive a user input and output (display) information to a user. In an embodiment, the user interface 240 can include an input apparatus and / or an output apparatus. The input apparatus can be an apparatus for receiving information input from the outside to be delivered to at least one constituent element of the electronic device 100. For example, the input apparatus can include a mouse, a keyboard, a touchpad, etc. The output apparatus can be an apparatus for providing various information of the electronic device 100 to a user in a visual / audible form. For example, the output apparatus can include a display, a projector, a hologram, a speaker, etc. In an embodiment, the user interface 240 can receive information input by a user for controlling the electronic device 100, information for controlling a substrate processing process, or information related to a substrate.

[0047] In an embodiment, the user interface 240 can receive an initial setting input by a user, that is, a first control parameter group of the solder paste printing apparatus 120 for printing a solder paste on the first substrate 111. The processor 210 can obtain the first control parameter group of the solder paste printing apparatus 120 for printing a solder paste on the first substrate 111, which is initially set by a user, from the memory 220, and transmit information indicating the first control parameter group to the solder paste printing apparatus 120. The processor 210 can obtain first solder paste inspection information indicating a state of a solder paste printed on the first substrate 111 from the first inspection apparatus 130, and determine a first yield for the first substrate 111 based on the first solder paste inspection information. The processor 210 can generate a model for exploring an optimal control parameter group based on a first data pair including the first control parameter group and the first yield. The processor 210 can store the generated model in the memory 220.

[0048] The processor 210 can obtain a second control parameter set of the solder paste printing device 120 for printing a solder paste on a second substrate 113 to be printed with a solder paste after the first substrate 111 from the generated model, and transmit information indicating the second control parameter set to the solder paste printing device 120. The processor 210 can obtain second solder paste detection information indicating a state of a solder paste printed on the second substrate 113 from the first detection device 130, and determine a second yield for the second substrate 113 based on the second solder paste detection information. The processor 210 can update the model based on the first data pair and a second data pair including the second control parameter set and the second yield.

[0049] In an embodiment, the electronic device 100 can be various forms of devices. For example, the electronic device 100 can be a portable communication device, a computer device, a wearable device, or a device combining two or more of the above devices. However, the electronic device 100 of the present disclosure is not limited to the above devices.

[0050] Various embodiments of the electronic device 100 mentioned in the present disclosure can be combined with each other. Each embodiment can be combined according to the number of cases, and the embodiment of the electronic device 100 combined also belongs to the scope of the present disclosure. In addition, the internal / external constituent elements of the above electronic device 100 can be added, changed, replaced, or deleted according to the embodiment. In addition, the internal / external constituent elements of the above electronic device 100 can be implemented by a hardware component.

[0051] Figure 3 is a diagram for explaining a model 300 for various embodiments of the present disclosure. As described above, the model 300 can output a control parameter set 320 for printing a solder paste on a next substrate in a substrate processing process based on a data pair 310 of a control parameter set and a yield related to each of a plurality of substrates explored before. The model 300 can be a model generated by an optimization algorithm for exploring an optimal control parameter set. In an embodiment, the model 300 can output a next control parameter set based on an optimization algorithm. The algorithm that can be applied can be a probability process (Stochastic Process) such as a Monte Carlo method, a Bayesian Optimization algorithm, and / or a likelihood based inference, etc.

[0052] First, the Monte Carlo method can be used to solve most problems involving probabilistic analysis. According to the law of large numbers, the integral of the predicted values ​​of any random variable takes the experimental mean (i.e., the sample mean) of the independent samples of the variable, thus providing an approximation. Using this property, a Markov chain Monte Carlo (MCMC) sampler can be used when the probability distribution of the variable is parameterized. The Monte Carlo method can also utilize the Gibbs sampling algorithm. Gibbs sampling is an MCMC algorithm used to obtain a set of observations from a given multivariate probability distribution when direct sampling is difficult.

[0053] A probabilistic process generally refers to a mathematical object defined by random variables. A probabilistic process can represent the numerical value of a system that changes randomly over time. That is, a probabilistic process can be used as a mathematical model for systems and phenomena that appear to change in arbitrary ways. Probabilistic processes can be interpreted through random variables. Types of probabilistic processes include Markov processes, Gaussian processes, statistical models, Bayesian inference, random walks, martingales, Levy processes, random fields, renewal processes, and branching processes. Bayesian inference, also known as Bayes' inference, is a statistical inference method that updates the hypothesized probabilities using Bayes' theorem after obtaining additional information through experiments. Bayesian inference is primarily used when dynamically analyzing data to adapt it to meet given conditions. In the field of artificial intelligence, it can be used when knowledge learned from previous data is updated with additional data according to conditions, and Bayesian optimization algorithms based on this are applicable.

[0054] The Gaussian process algorithm is an algorithm that uses the covariance and mean of variables in the probability distribution of multiple variables to construct a model 300. Therefore, unlike previous neural network circuits, it has the advantage that once the initial hyperparameters are determined, no additional manual parameter setting is required.

[0055] Likelihood-based inference can include algorithms such as the expectation-maximization algorithm. The expectation-maximization algorithm can be an iterative algorithm that finds parameter estimates with maximum likelihood or maximum a posteriori probability in a probabilistic model 300 that depends on unobserved latent variables. The iteration of the expectation-maximization algorithm may include performing: an expectation step using the current estimates of the parameters to generate an expectation function of the log-expected value of the evaluation; and a maximization step calculating the parameters used to maximize the expected log-expected value. The expectation-maximization algorithm can be used when the maximum expectation of a statistical model 300 cannot be found by directly solving the equations.

[0056] Model 300 can output a set of control parameters for printing solder paste on the next substrate based on a set of control parameters for the solder paste printing apparatus 120 used to print solder paste on the substrate and a data pair of yield data for the substrate. Electronic device 100 can generate model 300 for exploring the optimal set of control parameters based on a first data pair, wherein the first data pair includes a first set of control parameters for the first substrate (i.e., the first substrate 111) for which the substrate processing process has been performed and a first yield data for the first substrate 111. Subsequently, model 300 can be updated according to the sequence of the substrate processing processes, based on data pairs of control parameter sets and yield data for each of the [2nd, 3rd, ..., n-1, n]th substrates. In the above case, model 300 can output a set of control parameters based on the data pairs that enables the yield for the next substrate to be above a preset threshold. The threshold can be preset by the user.

[0057] Figure 4a and Figure 4b These are diagrams illustrating the process of exploring the optimal control parameter set in an embodiment of this disclosure. Figure 4a It is a diagram representing the two-dimensional parameter space 410 defined by two control parameters. Figure 4b This is graph 420, which represents the yield according to the order of substrate processing steps. Figure 4a and Figure 4b This illustration assumes the exploration of optimal control parameters in a two-dimensional parameter space (plane) defined by two control parameters, but it is not limited to this. Clearly, it is equally applicable in a three-dimensional parameter space defined by three control parameters or a multi-dimensional parameter space defined by four or more control parameters. The control parameter set illustrated in this figure can mean including both the first and second control parameters. Furthermore, although this figure illustrates the case of exploring the optimal control parameter set while changing it all at once for convenience, it is of course possible to explore the optimal control parameters separately for each of the multiple control parameters included in the control parameter set.

[0058] Electronic device 100 can optimize the control parameter set of solder paste printing apparatus 120 in real time based on solder paste detection information indicating the state of solder paste printed on the substrate as detected in real time by first detection device 130 (e.g., SPI device). Step t0 is a step related to the first substrate 111 that is processed first in the substrate processing process. The user can set the first control parameter set of solder paste printing apparatus 120 for printing solder paste on the first substrate 111 through user interface 240. For example, the user can input multiple control parameters of solder paste printing apparatus 120 that enable a yield above a preset threshold as the first control parameter set. Electronic device 100 can transmit information indicating the first control parameter set to solder paste printing apparatus 120, and solder paste printing apparatus 120 can print solder paste on the first substrate 111 based on the first control parameter set. Subsequently, the first substrate 111 printed with solder paste can be transferred from the solder paste printing apparatus 120 to the first detection apparatus 130. The first detection apparatus 130 can detect the state of the solder paste printed on the first substrate 111 and generate solder paste detection information indicating the state of the solder paste printed on the first substrate 111. The electronic device 100 can obtain the solder paste detection information indicating the state of the solder paste printed on the first substrate 111 from the first detection apparatus 130. Based on the obtained solder paste detection information, the electronic device 100 can determine a first yield for the first substrate 111. Based on the first set of control parameters and the first yield, the electronic device 100 can generate a model (e.g., model 300) for exploring the optimal set of control parameters.

[0059] Step t1 is a step related to the next substrate after the first substrate 111, namely the second substrate 113. That is, solder paste can be printed on the second substrate 113 after the first substrate 111. The electronic device 100 can obtain a second set of control parameters for the solder paste printing apparatus 120 for printing solder paste on the second substrate 113 from the generated model. According to one embodiment, the model can explore control parameter sets corresponding to unexplored portions in the two-dimensional parameter space (plane), portions where more new data can be obtained, or portions in the two-dimensional parameter space that are far from previously explored portions (e.g., points corresponding to the first set of control parameters). For example, the model can output the control parameter set corresponding to any unexplored point among a plurality of points included in the two-dimensional parameter space as the second set of control parameters based on the first data pair. For example, the model can output the control parameter set corresponding to points in the two-dimensional parameter space that are far from the points corresponding to the first set of control parameters as the second set of control parameters. For example, the model can output the control parameter set corresponding to the point with the greatest uncertainty of the predicted yield among a plurality of points included in the two-dimensional parameter space as the second set of control parameters. The point with the greatest uncertainty in the yield can, for example, mean the point with the largest discrete (or standard deviation) value of the yield calculated as a probability distribution. For instance, the model can output a second set of control parameters corresponding to the point with the largest discrete value of the predicted yield among multiple points included in the two-dimensional parameter space, or it can output a second set of control parameters corresponding to the points among multiple points where the discrete value of the predicted yield is greater than a preset value. According to one embodiment, the model can output a second set of control parameters predicting a second yield, where the second yield is higher than the first yield.

[0060] Electronic device 100 can transmit information indicating a second set of control parameters to solder paste printing apparatus 120, which can then print solder paste on a second substrate 113 based on the second set of control parameters. Afterwards, electronic device 100 can obtain second solder paste detection information indicating the state of the solder paste printed on the second substrate 113 from a first detection device 130, and can determine a second yield for the second substrate 113 based on the second solder paste detection information. Electronic device 100 can update its model based on a first data pair including the first set of control parameters and the first yield, and a second data pair including the second set of control parameters and the second yield. According to one embodiment, electronic device 100 can update its model to output a set of control parameters predicted to yield a rate higher than both the first and second yields. According to another embodiment, electronic device 100 can update its model to output a set of control parameters corresponding to points in the two-dimensional parameter space that are far from the points corresponding to the first and second sets of control parameters, respectively. In this case, electronic device 100 can update its model to output only the set of control parameters predicted to have a yield above a preset threshold.

[0061] Step t2 is a step related to the next substrate after the second substrate 113, namely the third substrate. That is, solder paste can be printed on the third substrate after the second substrate 113. The electronic device 100 can obtain a third set of control parameters for the solder paste printing apparatus 120 for printing solder paste on the third substrate from the updated model. According to one embodiment, the model can explore control parameter sets corresponding to unexplored portions in the two-dimensional parameter space, portions where more new data can be obtained, or portions far from previously explored portions (e.g., points corresponding to the first and second control parameter sets, respectively). According to one embodiment, the model can output the control parameter set predicted to have a third yield as the third control parameter set, wherein the third yield is higher than the first and second yields. Similar to the description in step t2, the electronic device 100 can transmit information indicating the third control parameter set to the solder paste printing apparatus 120, obtain third solder paste detection information indicating the state of the solder paste printed on the third substrate from the first detection device 130, and determine the third yield for the third substrate based on the third solder paste detection information. The electronic device 100 can update the model based on the first data pair, the second data pair, and the third data pair including the third control parameter group and the third yield.

[0062] Step t3 is related to the next substrate after the third substrate, namely the fourth substrate. That is, solder paste can be printed on the fourth substrate after the third substrate. Similar to the description in step t2, a fourth control parameter and a fourth yield for the fourth substrate can be obtained, and the model can be updated based on the fourth control parameter and the fourth yield.

[0063] According to the aforementioned method, the electronic device 100 can update the model until a predetermined number of control parameter sets are explored. For example, the electronic device 100 can update the model only until 20 control parameter sets for 20 substrates are explored. Then, the electronic device 100 can determine the optimal parameter set based on the predetermined number of explored control parameter sets. The process of determining the optimal control parameter set using such previous exploration results can be called an exploitation process. An exploitation process can mean exploring parameters in the parameter space in an existing exploration area in a direction close to the optimum in order to find the optimal solution (optimal control parameter set). The exploitation process can find the optimal solution based on the tendency shown by the results of previous exploration processes. For example, when the electronic device 100 is exploring 20 control parameter sets, it can determine the control parameter set with the highest yield among the 20 control parameter sets as the optimal control parameter set and transmit it to the solder paste inspection device. The aforementioned 20 is an illustrative number; the number of control parameter sets used for exploration can be set by the user to various different numbers.

[0064] In one embodiment, if an environmental change occurs during a substrate processing process based on the optimal set of parameters determined through the aforementioned process, the electronic device 100 can restart a new process to explore the optimal set of control parameters. In one embodiment, at least one sensor or detection device mounted on the substrate processing apparatus can sense environmental changes related to the substrate processing process. The at least one sensor or detection device can transmit information indicating an environmental change in the substrate processing process to the electronic device 100. Upon receiving the corresponding information, the electronic device 100 can restart the process to explore the optimal set of control parameters. For example, when the electronic device 100 senses an environmental change related to the substrate processing process, it can regenerate a model based on data pairs including the control parameter set and yield, and use the regenerated model to re-explore the optimal set of control parameters.

[0065] Figure 5 This is a diagram illustrating a method for considering stability during the process of exploring an optimal set of control parameters according to an embodiment of this disclosure. Figure 5 This is a coordinate graph 500 representing the yield based on the control parameter values. In this disclosure, since a model for exploring the optimal control parameter set is generated and updated based on data pairs including control parameter sets and yields collected during the execution of the substrate processing process, losses may occur when the yield drops below a preset threshold due to control parameter sets tried during the exploration of the optimal control parameter set. To minimize losses during the exploration of the optimal control parameter set, control parameters that predict a yield drop below the preset threshold should not be explored. Therefore, the electronic device 100 can avoid exploring control parameter sets t that are predicted to have a yield below the preset threshold. x .

[0066] Electronic device 100 can generate a model for exploring the optimal set of control parameters based on a first set of control parameters for a solder paste printing apparatus 120 for printing solder paste on a first substrate 111, which includes user-set first data pairs for the first yield of the first substrate 111. Based on the first data pairs, the model can output a second set of control parameters that enables the second yield of the next substrate, i.e., the second substrate 113, to be above a preset threshold. That is, the model can consider whether the set of control parameters to be output meets the preset threshold for yield, and can ensure the stability of the substrate processing process and prevent defective production due to changes in control parameters by not outputting control parameter sets whose predicted yield is less than the preset threshold.

[0067] Figure 6This is an operation flowchart of an electronic device 100 according to an embodiment of the present disclosure. Referring to the operation flowchart 600, in operation 610, the processor 210 of the electronic device 100 can obtain a first set of control parameters for the solder paste printing apparatus 120 for printing solder paste on the first substrate 111. The first set of control parameters may be a combination of multiple control parameters, which can be initially set by the user through the user interface 240.

[0068] In operation 620, the processor 210 may transmit information indicating the first control parameter set to the solder paste printing apparatus 120. The information indicating the first control parameter set may be, for example, a signal for controlling the solder paste printing apparatus 120 to perform a solder paste printing process on the first substrate 111 based on the first control parameter set.

[0069] In operation 630, the processor 210 can obtain first solder paste detection information indicating the state of solder paste printed on the first substrate 111 from the detection device. The detection device may be a first detection device (SPI device) 130 for detecting the state of solder paste printed on the substrate. The first solder paste detection information may, for example, be the volume value of the solder paste printed on the first substrate 111.

[0070] In operation 640, processor 210 may determine a first yield for the first substrate 111 based on the first solder paste detection information. The yield may, for example, be a yield determined based on the volume value of the solder paste printed on the first substrate 111. The yield may be determined as a probability distribution using a probabilistic modeling method.

[0071] In operation 650, processor 210 can generate a model for exploring the optimal set of control parameters based on a first data pair including a first set of control parameters and a first yield. The model can be generated, for example, assuming that the first set of control parameters and the first yield have a Gaussian distribution. After generating the model, operation A can be performed to update the model. The specific operations for updating the model will be discussed later. Figure 7 This will be explained in the description.

[0072] Figure 7 This is an operation flowchart of an electronic device 100 according to an embodiment of the present disclosure. Operation flowchart 700 may include... Figure 6 The operation following operation 650. In operation 710, the processor 210 can obtain from the generated model a second set of control parameters for the solder paste printing apparatus 120 for printing solder paste on the next substrate, namely the second substrate 113, of the first substrate 111.

[0073] The model can explore control parameter sets corresponding to unexplored portions of the multidimensional parameter space, portions from which more new data can be obtained, or portions far removed from previously explored portions. For example, based on a first data pair, the model can output a second control parameter set corresponding to any unexplored point among multiple points included in the two-dimensional parameter space. For example, the model can output a second control parameter set corresponding to a point in the two-dimensional parameter space far removed from the point corresponding to the first control parameter set. For example, the model can output a second control parameter set corresponding to the point among multiple points included in the two-dimensional parameter space with the greatest uncertainty in the predicted yield. According to one embodiment, the model can output a second control parameter set predicting a second yield, where the second yield is higher than the first yield. In all the above cases, the model can output a second control parameter set that enables the second yield to be above a pre-set threshold.

[0074] In operation 720, processor 210 can transmit information indicating the second control parameter group to solder paste printing apparatus 120. In operation 730, processor 210 can obtain second solder paste detection information indicating the state of solder paste printed on the second substrate 113 from the detection device. In operation 740, processor 210 can determine a second yield of the second substrate 113 based on the second solder paste detection information.

[0075] Processor 210 can update the model based on a first data pair and a second data pair including a second set of control parameters and a second yield. According to one embodiment, processor 210 can update the model to output a set of control parameters predicted to have a yield higher than both the first and second yields. According to another embodiment, processor 210 can update the model to output a set of control parameters corresponding to points in the multidimensional parameter space that are far from the points corresponding to the first and second sets of control parameters, respectively. In this case, processor 210 can update the model to output only the set of control parameters predicted to have a yield above a pre-set threshold.

[0076] While the methods described above have been illustrated through specific embodiments, they can also be implemented as computer-readable code in a computer-readable recording medium. Computer-readable recording media include all types of recording devices for storing readable data of a computer system. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage devices, etc. Furthermore, computer-readable recording media can be distributed across a network-connected computer system and can store and execute computer-readable code in a distributed manner. Moreover, those skilled in the art to which this disclosure pertains can readily derive functional programs, code, and code segments for implementing the above embodiments.

Claims

1. An electronic device comprising: communication circuitry communicably connected to solder paste printing apparatus configured to print solder paste on a plurality of substrates based on a plurality of control parameters, and to inspection apparatus configured to inspect a state of solder paste printed on the plurality of substrates transferred from the solder paste printing apparatus; one or more memories; and one or more processors operatively connected to the communication circuitry and the one or more memories, the one or more processors configured to: obtain a first set of control parameters of the solder paste printing apparatus for printing solder paste on a first substrate; transmit information indicating the first set of control parameters to the solder paste printing apparatus; obtain first solder paste inspection information indicating a state of solder paste printed on the first substrate from the inspection apparatus; determine a first yield for the first substrate based on the first solder paste inspection information; generate a model for exploring an optimal set of control parameters based on a first pair of data comprising the first set of control parameters and the first yield; obtain a second set of control parameters of the solder paste printing apparatus for printing solder paste on a second substrate to be printed after the first substrate from the model, wherein the second set of control parameters is a set of control parameters determined based on discrete values of yields corresponding to a plurality of points included in a multi-dimensional parameter space defined by the plurality of control parameters. 2.The electronic device of claim 1, wherein the one or more processors are configured to: transmit information indicating the second set of control parameters to the solder paste printing apparatus; obtain second solder paste inspection information indicating a state of solder paste printed on the second substrate from the inspection apparatus; determine a second yield for the second substrate based on the second solder paste inspection information; update the model based on the first pair of data and a second pair of data comprising the second set of control parameters and the second yield. 3.The electronic device of claim 2, wherein the model outputs, based on the first pair of data, the second set of control parameters capable of making the second yield be above a threshold value set in advance or automatically. 4.The electronic device of claim 2, wherein the model outputs, based on the first pair of data, the second set of control parameters capable of making the second yield be higher than the first yield. 5.The electronic device of claim 2, wherein the one or more processors update the model until a preset number of sets of control parameters are explored. 6.The electronic device of claim 5, wherein the one or more processors are configured to: after exploring the preset number of sets of control parameters, determine a set of control parameters corresponding to a maximum yield among the preset number of sets of control parameters as an optimal set of control parameters; transmit information indicating the optimal set of control parameters to the solder paste printing apparatus. 7.The electronic device of claim 1, wherein The plurality of control parameters include a control parameter for adjusting a printing pressure of the solder paste printing device, a control parameter for adjusting a printing speed, and a control parameter for adjusting a substrate separation speed. 8.The electronic device of claim 2, wherein The first control parameter set and the second control parameter set respectively correspond to one point in a multi-dimensional parameter space defined by a plurality of the control parameters. 9.The electronic device of claim 8, wherein The model outputs, based on the first data pair, the second control parameter set corresponding to any point not yet explored among a plurality of points included in the multi-dimensional parameter space. 10.The electronic device of claim 9, wherein The any point is a point among the plurality of points for which uncertainty of a yield is predicted to be the greatest. 11.The electronic device of claim 2, wherein The first yield and the second yield are determined as probability distributions by a probabilistic modeling method. 12.The electronic device of claim 1, wherein The first control parameter set is set by a user. 13.A method of optimizing control parameters of an electronic device communicably connected to a solder paste printing device configured to print solder paste on a plurality of substrates based on a plurality of control parameters and a detection device configured to detect a state of solder paste printed on the plurality of substrates transferred from the solder paste printing device, the method comprising: obtaining a first control parameter set of the solder paste printing device for printing solder paste on a first substrate; transmitting information indicating the first control parameter set to the solder paste printing device; obtaining first solder paste detection information indicating a state of solder paste printed on the first substrate from the detection device; determining a first yield for the first substrate based on the first solder paste detection information; generating a model for exploring an optimal control parameter set based on a first data pair including the first control parameter set and the first yield; and obtaining a second control parameter set of the solder paste printing device for printing solder paste on a second substrate to be printed after the first substrate from the model, wherein the second control parameter set is a control parameter set determined based on discrete values of yields corresponding to a plurality of points included in a multi-dimensional parameter space defined by a plurality of the control parameters. 14.The method of claim 13, further comprising: transmitting information indicating the second control parameter set to the solder paste printing device; obtaining second solder paste detection information indicating a state of solder paste printed on the second substrate from the detection device; determining a second yield for the second substrate based on the second solder paste detection information; and updating the model based on the first data pair and a second data pair including the second control parameter set and the second yield. 15.The method of claim 14, wherein The model outputs, based on the first data pair, the second control parameter set capable of making the second yield be above a threshold value set in advance. ​ 16. The method of claim 14, wherein, the model outputs, based on the first pair of data, the second set of control parameters that can make the second yield higher than the first yield.

17. The method of claim 13, wherein, the plurality of control parameters comprises a control parameter for adjusting a printing pressure of the solder paste printing device, a control parameter for adjusting a printing speed, and a control parameter for adjusting a substrate separation speed.

18. The method of claim 14, wherein, the first set of control parameters and the second set of control parameters respectively correspond to a point in a multi-dimensional parameter space defined by the plurality of control parameters.

19. The method of claim 18, wherein, the model outputs, based on the first pair of data, the second set of control parameters that corresponds to any point that has not been explored among a plurality of points included in the multi-dimensional parameter space.

20. The method of claim 14, wherein, the first yield and the second yield are determined as probability distributions by a probabilistic modeling method.

Images