Parameter optimization devices, parameter optimization methods, and computer program products
By using a parameter optimization device and a multi-task Bayesian optimization algorithm, parameter determination and film thickness information acquisition are executed alternately, solving the problem of low efficiency in control parameter optimization in coating machines. This achieves efficient and accurate control parameter optimization and film thickness uniformity, reducing substrate consumption and environmental impact.
Patent Information
- Application Number
- CN202510258153.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2024-03-22
- Filing Date
- 2025-03-05
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2045-03-05
AI Technical Summary
Existing technologies require extensive and repeated measurements and adjustments when optimizing control parameters in coating machines, resulting in low efficiency.
A parameter optimization device is employed, which alternately performs parameter determination, film thickness information acquisition, and estimation model processing. Combined with a multi-task Bayesian optimization algorithm, the control parameters are optimized. The model accuracy is improved by learning from the process information included in the input of the estimation model.
It achieves efficient and precise control parameter optimization, reduces the number and time of film thickness measurement, improves the uniformity of coating film and production efficiency, reduces the consumption of substrate and processing liquid, and reduces environmental impact.
Smart Images

Figure CN120688026B_ABST
Abstract
Description
Technical Field
[0001] The subject matter disclosed in this specification relates to a parameter optimization device, a parameter optimization method, and a computer program. Background Technology
[0002] In the manufacturing process of flat panel displays, a device called a coating machine is used. A coating machine is a substrate processing device that uses a pump to spray a processing liquid from a slit nozzle and coat the entire substrate being transported with the processing liquid. In such coating machines, with the increasing demand for higher product quality in recent years, it is required that the processing liquid be coated in a manner that results in a uniform film thickness across the entire substrate. For example, in Patent Document 1, optimization is achieved by repeatedly measuring the spray characteristics of the processing liquid during spraying, thereby adjusting the parameters used to control the pump.
[0003] Patent Document 1: Japanese Patent Application Publication No. 2020-040046
[0004] Patent Document 2: International Publication No. 2019 / 244474
[0005] Non-patent literature 1: K. Swersky et al., Multi-Task Bayesian Optimization, [online], [retrieved June 1, 2018], NIPS, 2013, URL <https: / / papers.nips.cc / paper / 5086-multi-task-bayesian-optimization.pdf>
[0006] However, in the case of Patent Document 1, since the control process and ejection characteristics are measured repeatedly, the time required to optimize the control parameters may be long. Summary of the Invention
[0007] The purpose of this invention is to provide a technique for effectively optimizing control parameters.
[0008] To address the aforementioned problems, a first approach is a parameter optimization apparatus that optimizes control parameters for controlling a coating apparatus that sprays a processing liquid onto a substrate to form a coating film. The parameter optimization apparatus comprises: a parameter determination unit that determines control parameters to be evaluated next using a plurality of film thickness information values each associated with a specific control parameter; a film thickness information acquisition unit that acquires film thickness information based on a measurement of the film thickness of the coating film, wherein the coating film is formed by controlling the coating apparatus using the control parameters determined by the parameter determination unit; a film thickness information estimation unit that estimates film thickness information corresponding to the control parameters determined by the parameter determination unit using an estimation model that takes control parameters and process information as input and outputs film thickness information; and a cycle control unit that controls the parameter determination unit, the film thickness information acquisition unit, and the film thickness information estimation unit to alternately execute a first process and a second process, wherein the first process includes the processing of the parameter determination unit and the processing of the film thickness information acquisition unit, and the second process includes the processing of the parameter determination unit and the processing of the film thickness information estimation unit.
[0009] In the second approach, the parameter optimization device of the first approach also includes a learning unit, which uses the film thickness information obtained by the film thickness information acquisition unit to learn the estimation model.
[0010] In the third method, in the parameter optimization device of the first or second method, the second process is a process of repeatedly performing the processing of the parameter determination unit and the processing of the film thickness information estimation unit. When the number of times the film thickness information is estimated by the film thickness information estimation unit in the second process has reached a predetermined number, the cycle control unit executes the first process.
[0011] In the fourth method, in any of the parameter optimization devices of the first to third methods, the parameter determination unit calculates an evaluation value based on the film thickness information and determines the control parameters to be evaluated next based on the evaluation value.
[0012] In the fifth method, in the parameter optimization device of the fourth method, the parameter determination unit uses a multi-task Bayesian optimization algorithm to determine the control parameters to be evaluated next.
[0013] A sixth approach, a parameter optimization method, optimizes parameters for controlling a coating apparatus that sprays a processing liquid onto a substrate to form a coating film. The parameter optimization method includes: a) a parameter determination step, using a plurality of film thickness information points each associated with a specific parameter to determine control parameters to be evaluated; b) a film thickness information acquisition step, acquiring film thickness information based on a measurement of the film thickness of the coating film, wherein the coating film is formed by controlling the coating apparatus using the control parameters determined in the parameter determination step; and c) a film thickness information estimation step, using an estimation model that takes the control parameters and process information as inputs and outputs film thickness information to estimate film thickness information corresponding to the control parameters determined in the parameter determination step. The parameter optimization method alternately executes a first process and a second process, the first process including the parameter determination step and the film thickness information acquisition step, and the second process including the parameter determination step and the film thickness information estimation step.
[0014] The seventh method is a computer program product comprising a computer program, wherein the computer program, when executed by a processor, implements the steps of the parameter optimization method of the sixth method.
[0015] According to the first to seventh methods, since the search for control parameters via the estimation model and the search for control parameters based on the film thickness measurement results requiring time are performed alternately, it is possible to both reduce film thickness measurements and efficiently optimize control parameters. Furthermore, since the input to the estimation model includes process information, film thickness information obtained from different processes can be used as teacher data. Therefore, by inputting process information, the output accuracy of the estimation model can be improved. Thus, high-precision optimization of control parameters is possible.
[0016] The parameter optimization device of the second method can improve the accuracy of the estimation model by learning by using the film thickness information of the coating film measured during the optimization process.
[0017] The optimization time for parameters can be further shortened by using a third-party parameter optimization device.
[0018] According to the parameter optimization device of the fifth method, even if there is a difference between the distribution of evaluation values based on the film thickness information of the measurement results and the distribution of evaluation values based on the film thickness information of the estimation model, Bayesian optimization search can be performed more effectively. Attached Figure Description
[0019] Figure 1 This is a schematic diagram showing the overall structure of the coating apparatus according to the embodiment.
[0020] Figure 2 It means Figure 1A diagram showing the structure of the treatment fluid supply mechanism in the coating apparatus.
[0021] Figure 3 It is a block diagram showing the structure of the control unit.
[0022] Figure 4 It is a diagram that shows the structure of the control unit together with the flow of data.
[0023] Figure 5 This is a diagram illustrating an example of the process of optimizing control parameters using the control unit.
[0024] Explanation of reference numerals in the attached figures:
[0025] 1: Coating device
[0026] 9: Control Unit (Parameter Optimization Device)
[0027] 913: Parameter Determination Section
[0028] 914: Film Thickness Information Acquisition Department
[0029] 915: Film thickness information estimation department
[0030] 916: Academic Department
[0031] 918: Cyclic Control Department
[0032] 931: Computer Programs
[0033] Y: Presumed Model Detailed Implementation
[0034] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. Furthermore, the constituent elements described in these embodiments are merely illustrative, and the scope of the present invention is not limited thereto. In the accompanying drawings, the dimensions or quantities of various parts may be exaggerated or simplified as needed for ease of understanding.
[0035] <1. Implementation Method>
[0036] Figure 1 This is a schematic diagram showing the overall structure of the coating apparatus 1 according to the embodiment. The coating apparatus 1 is a substrate processing apparatus that forms a coating film on the substrate S by spraying a processing liquid onto the upper surface Sf of the substrate S. As will be described later, the coating apparatus 1 functions as a parameter optimization device for optimizing the parameters used to control the spraying of the processing liquid.
[0037] The substrate S is, for example, a glass substrate for a liquid crystal display device. Alternatively, the substrate S can be various substrates to be processed for electronic devices such as semiconductor wafers, photomasks, plasma displays, magneto-optical discs, organic EL (Electroluminescence) glass substrates, solar cells, silicon substrates, other flexible substrates, and printed circuit boards. The coating apparatus 1 is, for example, a slot coater.
[0038] exist Figure 1 In this document, an XYZ coordinate system is defined to illustrate the configuration of the elements of the coating apparatus 1. The transport direction of the substrate S is defined as the "X direction". The direction in which the substrate S travels in the X direction (downstream of the transport direction) is defined as the +X direction, and its opposite direction (upstream of the transport direction) is defined as the -X direction. Furthermore, the direction orthogonal to the X direction is the Y direction, and the direction orthogonal to both the X and Y directions is the Z direction. In the following description, the Z direction is defined as the vertical direction, and the X and Y directions are defined as the horizontal directions. In the Z direction, the +Z direction is the upward direction, and the -Z direction is the downward direction. It should be noted that these directions are not intended to limit the configuration of the coating apparatus.
[0039] The coating apparatus 1, in the +X direction, sequentially comprises an input conveyor 100, an input transfer unit 2, a floating stage unit 3, an output transfer unit 4, and an output conveyor 110. The input conveyor 100, input transfer unit 2, floating stage unit 3, output transfer unit 4, and output conveyor 110 form a transport path for the substrate S to pass through. Additionally, the coating apparatus 1 also includes a substrate transport unit 5, a coating mechanism 7, a processing liquid supply mechanism 8, and a control unit 9.
[0040] The substrate S is transported from the upstream device of the coating apparatus 1 to the input conveyor 100. The input conveyor 100 includes a roller conveyor 101 and a rotary drive mechanism 102. The rotary drive mechanism 102 rotates each roller of the roller conveyor 101. By rotating each roller of the roller conveyor 101, the substrate S is transported downstream (in the +X direction) in a horizontal orientation. "Horizontal orientation" means that the main surface (the surface with the largest area) of the substrate S is parallel to the horizontal plane (XY plane).
[0041] The input transfer unit 2 includes a roller conveyor 21 and a rotation and lifting drive mechanism 22. The rotation and lifting drive mechanism 22 rotates each roller of the roller conveyor 21 and raises and lowers the roller conveyor 21. By rotating the roller conveyor 21, the substrate S is transported downstream in a horizontal position (+X direction). In addition, by raising and lowering the roller conveyor 21, the position of the substrate S in the Z direction is changed. The substrate S is transferred from the input conveyor 100 to the floating worktable unit 3 via the input transfer unit 2.
[0042] like Figure 1 As shown, the floating worktable section 3 is generally flat. The floating worktable section 3 is divided into three parts along the X direction. Facing the +X direction, the floating worktable section 3 has, in sequence, an inlet floating worktable 31, a coating worktable 32, and an outlet floating worktable 33. The upper surfaces of the inlet floating worktable 31, the coating worktable 32, and the outlet floating worktable 33 are located on the same plane. The floating worktable section 3 also has a lifting pin drive mechanism 34, a floating control mechanism 35, and a lifting drive mechanism 36. The lifting pin drive mechanism 34 causes the plurality of lifting pins disposed on the inlet floating worktable 31 to rise and fall. The floating control mechanism 35 supplies compressed air to the inlet floating worktable 31, the coating worktable 32, and the outlet floating worktable 33 to raise and fall the substrate S. The lifting drive mechanism 36 causes the outlet floating worktable 33 to rise and fall.
[0043] A plurality of ejector holes are arranged in a matrix on the upper surfaces of the inlet floating stage 31 and the outlet floating stage 33, and compressed air supplied from the floating control mechanism 35 is ejected from these ejector holes. When compressed air is ejected from each ejector hole, the substrate S floats upward relative to the floating stage 3. The lower surface Sb of the substrate S then separates from the upper surface of the floating stage 3, and the substrate S is supported in a horizontal position. The distance (floatation amount) between the lower surface Sb of the substrate S and the upper surface of the floating stage 3 when the substrate S is floating is, for example, 10 μm or more and 500 μm or less.
[0044] The upper surface of the coating stage 32 is provided with an ejector hole for discharging compressed air supplied from the levitation control mechanism 35 and an suction hole for drawing in air. The ejector hole and suction hole are alternately arranged in the X and Y directions. The levitation control mechanism 35 controls the amount of compressed air ejected from the ejector hole and the amount of air drawn from the suction hole. This precisely controls the levitation amount of the substrate S relative to the coating stage 32, so that the position of the upper surface Sf of the substrate S above the coating stage 32 in the Z direction is a predetermined value. Furthermore, based on the detection results of the sensor 61 or sensor 62 (described later), the control unit 9 calculates the levitation amount of the substrate S relative to the coating stage 32. Preferably, the levitation amount of the substrate S relative to the coating stage 32 can be adjusted with high precision by airflow control.
[0045] The substrate S, which is moved into the floating worktable section 3, is propelled in the +X direction by the roller conveyor 21 and transported to the inlet floating worktable 31. The inlet floating worktable 31, the coating worktable 32, and the outlet floating worktable 33 support the substrate S in a floating state. For example, the structure described in Japanese Patent No. 5346643 can be used as the floating worktable section 3.
[0046] The substrate transport unit 5 is positioned below the floating worktable unit 3. The substrate transport unit 5 includes a chuck mechanism 51 and a suction and travel control mechanism 52. The chuck mechanism 51 has a suction pad (not shown) disposed on the suction member. The chuck mechanism 51 abuts the suction pad against the periphery of the lower surface Sb of the substrate S, thereby supporting the substrate S from below. The suction and travel control mechanism 52 applies negative pressure to the suction pad, thereby adsorbing the substrate S onto the suction pad. Furthermore, the suction and travel control mechanism 52 causes the substrate transport unit 5 to reciprocate in the X direction.
[0047] The chuck mechanism 51 holds the substrate S with the lower surface Sb of the substrate S positioned above the upper surface of the floating stage 3. While the peripheral portion of the substrate S is held by the chuck mechanism 51, the buoyancy applied by the floating stage 3 maintains a horizontal posture.
[0048] like Figure 1 As shown, the coating apparatus 1 has a sensor 61 for measuring the thickness of the substrate. The sensor 61 is positioned near the roller conveyor 21. The sensor 61 detects the position of the upper surface Sf of the substrate S held by the chuck mechanism 51 in the Z direction. Furthermore, by positioning the chuck (not shown) in a state where the substrate S is not held directly below the sensor 61, the sensor 61 can detect the position of the adsorption surface of the upper surface of the adsorption member in the vertical Z direction.
[0049] The chuck mechanism 51 holds the substrate S loaded into the floating stage 3 and moves the substrate S in the +X direction. As a result, the substrate S is transported from above the inlet floating stage 31 to above the outlet floating stage 33 via the coating stage 32. Then, the substrate S is moved from the outlet floating stage 33 to the output transfer section 4.
[0050] The output transfer unit 4 moves the substrate S from a position above the outlet floating worktable 33 toward the output conveyor 110. The output transfer unit 4 includes a roller conveyor 41 and a rotation and lifting drive mechanism 42. The rotation and lifting drive mechanism 42 rotates the roller conveyor 41 and moves it up and down in the Z direction. By rotating each roller of the roller conveyor 41, the substrate S moves in the +X direction. Furthermore, by moving the roller conveyor 41 up and down, the substrate S is displaced in the Z direction.
[0051] The output conveyor 110 includes a roller conveyor 111 and a rotary drive mechanism 112. The output conveyor 110 transports the substrate S in the +X direction by rotating the rollers of the roller conveyor 111, thus moving the substrate S out of the coating apparatus 1. Furthermore, the input conveyor 100 and the output conveyor 110 are part of the coating apparatus 1. However, the input conveyor 100 and the output conveyor 110 can also be incorporated into devices different from the coating apparatus 1.
[0052] The coating mechanism 7 applies a coating liquid to the upper surface Sf of the substrate S. The coating mechanism 7 is positioned above the transport path of the substrate S. The coating mechanism 7 has a nozzle 71. The nozzle 71 is a slit nozzle with a slit-shaped outlet on its lower surface. The nozzle 71 is connected to a positioning mechanism (not shown). The positioning mechanism positions the nozzle 71 at a coating position above the coating worktable 32. Figure 1 The position shown by the solid line in the middle moves between the maintenance position described later. The treatment fluid supply mechanism 8 is connected to the nozzle 71. The treatment fluid supply mechanism 8 supplies treatment fluid to the nozzle 71, thereby spraying the treatment fluid from the spray outlet disposed on the lower surface of the nozzle 71.
[0053] Figure 2 It means Figure 1 The diagram shows the structure of the treatment fluid supply mechanism 8 in the coating apparatus 1. The treatment fluid supply mechanism 8 includes a pump 81, piping 82, a treatment fluid replenishment unit 83, piping 84, an on / off valve 85, a pressure sensor 86, and a drive unit 87. The pump 81 is a delivery source for conveying the treatment fluid to the nozzle 71, delivering the treatment fluid through volume changes. For example, the pump 81 can be a bellows-type pump as described in Japanese Patent Application Publication No. 10-61558. Figure 2 As shown, the pump 81 has a flexible tube 811 that expands and contracts freely in the radial direction. One end of the flexible tube 811 is connected to the treatment fluid replenishment unit 83 via a piping 82. The other end of the flexible tube 811 is connected to the nozzle 71 via a piping 84.
[0054] Pump 81 has a bellows 812 that can freely and elastically deform in the axial direction. Bellows 812 has a small bellows section 813, a large bellows section 814, a pump chamber 815, and a working disc section 816. Pump chamber 815 is disposed between flexible pipe 811 and bellows 812. An incompressible medium is sealed within pump chamber 815. Working disc section 816 is connected to drive unit 87.
[0055] The processing fluid replenishment unit 83 includes a storage tank 831 for storing processing fluid. The storage tank 831 is connected to the pump 81 via a piping 82. An on / off valve 833 is installed on the piping 82. The on / off valve 833 opens and closes according to instructions from the control unit 9. If the on / off valve 833 is open, processing fluid can be replenished from the storage tank 831 to the flexible pipe 811 of the pump 81. Conversely, if the on / off valve 833 is closed, the replenishment of processing fluid from the storage tank 831 to the flexible pipe 811 of the pump 81 can be restricted.
[0056] Pipe 84 is connected to the output side of pump 81. An on / off valve 85 is provided on pipe 84. The on / off valve 85 opens and closes pipe 84 according to instructions from control unit 9. By opening and closing the on / off valve 85, switching can be performed between discharging treatment fluid to nozzle 71 and stopping the discharging of treatment fluid to nozzle 71. A pressure sensor 86 is provided on pipe 84. The pressure sensor 86 detects the pressure (discharge pressure) applied to the treatment fluid discharged to nozzle 71 and outputs a signal indicating the detected pressure value to control unit 9.
[0057] like Figure 1 and Figure 2 As shown, a sensor 62 is disposed on the nozzle 71 from which processing liquid is supplied from the processing liquid supply mechanism 8. The sensor 62 detects the height of the substrate S in the Z direction non-contactly. The sensor 62 is electrically connected to the control unit 9. Based on the detection result of the sensor 62, the control unit 9 measures the distance (separation distance) between the floating substrate S and the upper surface of the coating stage 32. Then, the control unit 9 adjusts the coating position of the nozzle 71 by means of a positioning mechanism according to the measured separation distance. Furthermore, as the sensor 62, for example, an optical sensor or an ultrasonic sensor can be used.
[0058] Furthermore, the substrate S removed from the output conveyor 110 is dried by a drying device, thereby forming a coating film. Then, as... Figure 1 As shown, the substrate S with the coated film is transported to the film thickness gauge AP1 as needed to measure the film thickness. The film thickness gauge AP1 can be, for example, a spectroscopic ellipsometer or an X-ray reflectance measuring device.
[0059] The coating mechanism 7 includes a nozzle cleaning standby unit 72. The nozzle cleaning standby unit 72 performs prescribed maintenance on the nozzles 71 positioned in the maintenance position. The nozzle cleaning standby unit 72 includes a roller 721, a cleaning section 722, and a roller groove 723. The nozzle cleaning standby unit 72 cleans the nozzles 71 and forms a liquid accumulation, thereby adjusting the nozzle 71's outlet position to a state suitable for coating processing.
[0060] Figure 3This is a block diagram showing the structure of the control unit 9. The control unit 9 controls the operation of each component within the coating apparatus 1. The control unit 9 can use a computer. The control unit 9 has a processor 91 and a memory 93. The processor may be, for example, a CPU (Central Processing Unit). The memory 93 may be a temporary storage device such as RAM (Random Access Memory). In addition, the memory 93 may also be a non-temporary storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The memory 93 is connected to the processor 91 via bus wiring.
[0061] The control unit 9 includes a display device 95 for displaying various information and an input device 97 for receiving user commands. The display device 95 and the input device 97 are connected to the processor 91 via bus wiring. The display device 95 is, for example, a liquid crystal display (LCD). The input device 97 includes, for example, a mouse or keyboard. Furthermore, the display device 95 has a touch panel, which also enables it to function as an input device.
[0062] The memory 93 stores the computer program 931. The computer program 931 is provided to the control unit 9 via the recording medium M. That is, the computer program 931 is readable and recordable on the recording medium M by the computer, i.e., the control unit 9. Specifically, the recording medium M is a USB (Universal Serial Bus) memory, a DVD (Digital Versatile Disc) or other optical discs, magnetic disks, etc.
[0063] The processor 91 executes the computer program 931 and functions as the ejection control unit 910, parameter determination unit 913, film thickness information acquisition unit 914, film thickness information estimation unit 915, learning unit 916, and cycle control unit 918.
[0064] Figure 4 This diagram illustrates the structure of the control unit along with the flow of data. The ejection control unit 910 controls the operation (transfer operation) of the pump 81 that delivers the processing liquid to the nozzle 71 according to preset control parameters. In the coating apparatus 1, the processing liquid ejected from the nozzle 71 is coated onto the upper surface Sf of the substrate S with a uniform film thickness. Therefore, before starting production of the substrate S (or mass production), the control parameters closely related to the ejection pressure waveform can be optimized in advance to make the ejection pressure waveform an ideal shape.
[0065] The control parameters are setpoints used for pump control, such as various parameters that define the movement of the working disc section 816 (e.g., acceleration time, constant speed, time to maintain constant speed, deceleration time, etc.). The control parameters are optimized via the parameter determination unit 913, film thickness information acquisition unit 914, film thickness information estimation unit 915, learning unit 916, and cycle control unit 918. In other words, the parameter determination unit 913, film thickness information acquisition unit 914, film thickness information estimation unit 915, learning unit 916, and cycle control unit 918 constitute a parameter optimization device for optimizing control parameters.
[0066] The parameter determination unit 913 uses a plurality of acquired film thickness information, each associated with a specific control parameter, to determine the control parameter to be evaluated next. For example... Figure 4 As shown, the acquired film thickness information used by the parameter determination unit 913 is stored in the database DB1 in a state corresponding to specific control parameters. The database DB1 is a function implemented by the memory 93. The film thickness information is information related to the film thickness of the coating film formed on the substrate S, and is a film thickness distribution (film thickness profile) in one direction. Furthermore, the film thickness information is not limited to the film thickness distribution. As film thickness information, for example, indicators representing the uniformity of film thickness (mean and variance, etc.) can also be used.
[0067] The film thickness information acquisition unit 914 acquires the film thickness information TH1, which is the measurement result of the coating film formed on the substrate S by controlling the coating apparatus 1 using the control parameter P1 determined by the parameter determination unit 913. Specifically, the ejection control unit 910 performs ejection processing based on the control parameter P1 to form a coating film on the substrate S, and then measures the film thickness of the formed coating film using the film thickness meter AP1. The film thickness information acquisition unit 914 uses the measured film thickness to acquire the film thickness information TH1. The film thickness information acquisition unit 914 stores the acquired film thickness information TH1 in a state associated with the control conditions at which the film thickness information TH1 was obtained, i.e., the control parameter P1 (i.e., the corresponding control parameter P1).
[0068] The film thickness information estimation unit 915 uses an estimation model Y to estimate film thickness information TH2 corresponding to the control parameter P2 determined by the parameter determination unit 913. The estimation model Y is a learned model that takes the control parameter and process information as input and outputs film thickness information. The process information is information related to the process conditions for forming the coating film and is different from the control parameter. As process information, for example, the model name of the coating apparatus, the size of the substrate for forming the coating film or the type of substrate (surface material, etc.), and information related to the processing liquid (physical property information, etc.) can be used. In addition, the process information used as input can be one type of information or multiple types of information. The film thickness information estimation unit 915 stores the estimated film thickness information TH2 in the database DB1 in a state that corresponds to the control parameter P2.
[0069] The parameter determination unit 913 uses the film thickness information TH1 stored in the database DB1 by the film thickness information acquisition unit 914 or the film thickness information TH2 stored in the database DB1 by the film thickness information estimation unit 915 to determine the control parameters to be evaluated next.
[0070] Learning Unit 916 uses machine learning with teacher data, taking acquired control parameters and process information as input and film thickness information as output, to learn a prediction model Y. As the prediction model Y, for example, by using an RNN (Recurring Neural Network), one-dimensional data, i.e., the film thickness distribution, can be predicted as film thickness information. Furthermore, when film thickness information is used as an indicator of uniformity, random forest regression can be used as the prediction model Y, for example.
[0071] like Figure 4 As shown, the teacher data (learning dataset) used for machine learning by the learning unit 916 is stored in the database DB2. Database DB2 is a function implemented through the memory 93. Film thickness information is stored in database DB2 in a state associated with the corresponding control parameters and corresponding process information. Film thickness information acquisition unit 914 stores the acquired film thickness information TH1 in database DB2 in a state associated with the corresponding control parameters and process information. Additionally, as... Figure 4 As shown, database DB2 also stores film thickness information provided by database DB3, which is different from coating device 1, via a network (not shown).
[0072] The loop control unit 918 controls the parameter determination unit 913, the film thickness information acquisition unit 914, and the film thickness information estimation unit 915 to alternately execute the first process and the second process. Here, the first process includes the determination of control parameters by the parameter determination unit 913 and the acquisition of film thickness information by the film thickness information acquisition unit 914. The second process includes the determination of control parameters by the parameter determination unit 913 and the estimation of film thickness information by the film thickness information estimation unit 915. As will be described later, in the second process, the loop control unit 918 controls the repeated execution of the processes performed by the parameter determination unit 913 and the film thickness information estimation unit 915. Further details will be provided in [reference needed]. Figure 5 Please provide an explanation.
[0073] <Parameter Optimization Processing>
[0074] Figure 5 This is a diagram illustrating an example of the parameter optimization process performed by the control unit 9. Figure 5 In the process, the parameter determination step S10 is equivalent to the control parameter determination process performed by the parameter determination unit 913. Furthermore, the estimation step S3 is equivalent to the film thickness information estimation process performed by the film thickness information estimation unit 915, and the acquisition step S5 is equivalent to the film thickness information estimation process performed by the film thickness information acquisition unit 914. That is, the first process is equivalent to the parameter determination step S10 and the acquisition step S5, and the second process is equivalent to the parameter determination step S10 and the estimation step S3.
[0075] In addition, Figure 5 In the process, a pre-set upper limit N of the number of attempts is defined. est Estimated upper limit of N est This refers to the number of times the estimated step S3 is repeated in a single second process. Furthermore, the cycle control unit 918 counts the number of times the estimated step S3 is executed, thereby recording the estimated execution count C. est Used as a variable.
[0076] In addition, Figure 5 In the process, a pre-set upper limit N for the number of searches is implemented. all Search limit N all This refers to the number of times the control parameters are searched during an optimization process (i.e., the number of times the parameter determination unit 913 determines the control parameters). The loop control unit 918 counts the number of times the search is performed, thereby setting the search execution count C. ser Used as a variable.
[0077] If optimization processing begins, the parameter determination unit 913 first determines the initial control parameters (initial parameter determination step S1). The initial control parameters can be values already used in the coating apparatus. Alternatively, the initial control parameters can be randomly determined values.
[0078] Next, the loop control unit 918 determines the estimated number of executions C. est Has the estimated maximum number of attempts N been reached? est (Decision Process S2). In decision process S2, the decision is made to determine the estimated number of executions C. est The estimated maximum number of times N was not reached est In this case, the cycle control unit 918 causes the film thickness information estimation unit 915 to perform estimation processing. That is, the film thickness information estimation unit 915 inputs the control parameters determined by the initial parameter determination process S1 (or, the parameter determination process S10 described later) into the estimation model Y, thereby estimating the film thickness information (estimation process S3). In addition, the cycle control unit 918 performs estimation execution C times. est Increase by 1 (incremental process S4). If the incremental process S4 ends, the control unit 9 enters the evaluation value calculation process S9.
[0079] In the determination process S2, the estimated number of executions C is determined. est The estimated maximum number of times N is reached est In this case, the film thickness information acquisition unit 914 performs an acquisition process (acquisition step S5). Specifically, the ejection control unit 910 uses the control parameter P1 determined in the initial parameter determination step S1 (or the parameter determination step S10 described later) to control the pump 81, thereby forming a coating film on the substrate S. Then, the film thickness is measured using the film thickness gauge AP1. The film thickness information acquisition unit 914 acquires the film thickness measurement result, i.e., the film thickness information. After the acquisition step S5, the cycle control unit 918 estimates the number of executions C. est Set to 0 (reset process S6).
[0080] After the reset step S6, the loop control unit 918 determines whether to update the estimated model Y (determination step S7). Regarding whether to update the estimated model Y, for example, if the film thickness information TH1 obtained through the acquisition step S5—that is, the film thickness information TH1 not used for relearning the estimated model Y—is more than a predetermined amount, then the estimated model Y can be updated. If it is determined in the determination step S7 that the estimated model Y needs to be updated, the loop control unit 918 causes the learning unit 916 to update the estimated model Y (model update step S8). That is, the film thickness information added to the database DB1 through the acquisition step S5 is added and used to relearn the estimated model Y. After the model update step S8 is completed, the control unit 9 enters the evaluation value calculation step S9. Alternatively, if it is determined in the determination step S7 that it should not be updated, the control unit 9 skips the model update step S8 and enters the evaluation value calculation step S9.
[0081] In the evaluation value calculation step S9, the parameter determination unit 913 calculates the evaluation value of the film thickness information TH2 estimated by the estimation step S3 or the film thickness information TH1 obtained by the acquisition step S5. For example, as an evaluation value for obtaining a uniform film thickness, an index representing the magnitude of the deviation in film thickness distribution can be used. In this case, the error or deviation of the error relative to the specified target film thickness can also be used as the evaluation value.
[0082] Next, the parameter determination unit 913 determines the control parameters to be evaluated next (parameter determination step S10) based on the evaluation values obtained through the evaluation value calculation step S9. Specifically, methods such as reinforcement learning (RL), Bayesian optimization, or particle swarm optimization can be used. According to these methods, the control parameters can be updated at each trial to achieve the best evaluation value.
[0083] Alternatively, as a parameter search method, multi-task Bayesian optimization as described in Non-Patent Document 1 or Patent Document 2 (International Publication No. 2019 / 244474) can also be used. The film thickness information TH1 obtained by the acquisition step S5 may be affected by noise components generated from actual measurements. Therefore, the distribution of evaluation values may differ between the film thickness information TH2 estimated by the estimation step S3 and the film thickness information TH1 obtained by the acquisition step S5. Therefore, by using one as the object observation data and the other as the reference observation data, multi-task Bayesian optimization can be used to more effectively search for Bayesian optimization even if the distribution of evaluation values differs between the two.
[0084] After parameter determination step S10, the loop control unit 918 will perform the search execution count C. serIncrement by 1 (incremental process S11). Then, the loop control unit 918 determines the number of search executions C. ser Has the maximum number of searches (N) been reached? all (Decision step S12). If the decision step S12 determines that the number of search executions is C. ser The maximum number of searches (N) has not been reached. all If the result is not found, the control unit 9 returns to the decision process S2 to continue processing. On the other hand, if the decision process S12 determines that the number of search executions is C... ser Reaching the maximum number of searches N all Then the control unit 9 ends the processing. As described above, the optimized control parameters are stored in the memory 93 and used in subsequent coating processes of the coating apparatus 1.
[0085] As described above, the control unit 9 of this embodiment alternately searches for control parameters via the estimation model Y and for control parameters based on film thickness measurements requiring time. This reduces the number of coating film formations and film thickness measurements, thus shortening the time required to adjust the control parameters. Furthermore, since the number of coating film formations is reduced, the consumption of the substrate S and processing liquid is reduced, thereby mitigating environmental impact. Additionally, the input to the estimation model includes process information, allowing film thickness information obtained from different processes to be used as teacher data. Therefore, by inputting process information, the output accuracy of the estimation model can be improved. Thus, since the control parameter search can be performed with high precision, optimization can be performed efficiently.
[0086] Furthermore, by using the film thickness information measured during the optimization of control parameters to relearn the estimation model Y, it is possible to optimize and improve the output accuracy of the estimation model Y.
[0087] Furthermore, in the second process, the subsequent control parameter determination process performed by the parameter determination unit and the film thickness information TH2 estimation performed by the film thickness information estimation unit 915 are executed multiple times until the upper limit number of estimations N is reached. est This allows for a further reduction in the time required for parameter optimization.
[0088] <2. Variations>
[0089] The implementation methods have been described above, but the present invention is not limited to the above content and can be modified in various ways.
[0090] For example, in Figure 5 In the illustrated process, in the second processing step, the parameter determination step S10 and the estimation step S3 are repeatedly performed to estimate the upper limit N times. estIn contrast, in the first process, the parameter determination step S10 and the acquisition step S5 are executed only once. However, in the first process, the loop control unit 918 can also control the process so that the parameter determination step S10 and the acquisition step S5 are executed more than twice.
[0091] Alternatively, the film thickness information obtained in the estimation step S3 and the acquisition step S5 can be set to be the same as the evaluation value obtained in the evaluation value calculation step S9. In this case, the estimation model Y can also be learned so that the estimation model Y outputs an evaluation value based on the control parameters. In addition, the film thickness information acquisition unit 914 can also calculate the evaluation value based on the film thickness distribution.
[0092] In addition, Figure 5 In the process shown, in the judgment step S12, the control parameters are repeatedly generated until the number of search executions C is reached. ser Reaching the maximum number of searches N all Instead, the loop control unit 918 can also perform control to repeatedly generate control parameters until the evaluation value calculated in the evaluation value calculation step S9 exceeds the specified benchmark value.
[0093] In addition, in the above embodiment, the parameter optimization device is implemented by the control unit 9 of the coating device 1, but it can also be configured as a device different from the coating device 1.
[0094] While the invention has been described in detail, the foregoing description is exemplary in all respects and the invention is not limited thereto. It is understood that many modifications not illustrated can be contemplated without departing from the scope of the invention. The structures described in the above embodiments and modifications can be appropriately combined or omitted as long as they do not contradict each other.
Claims
1. A parameter optimization apparatus for optimizing control parameters used to control a coating apparatus, wherein the coating apparatus sprays a processing liquid onto a substrate to form a coating film, wherein... The parameter optimization device has the following features: The parameter determination unit uses multiple film thickness information that are associated with specific control parameters to determine the control parameters to be evaluated next. The film thickness information acquisition unit acquires film thickness information based on the measurement results of the film thickness of the coated film, wherein the coated film is formed by controlling the coating device using control parameters determined by the parameter determination unit; The film thickness information estimation unit uses an estimation model that takes control parameters and process information as input and outputs film thickness information to estimate the film thickness information corresponding to the control parameters determined by the parameter determination unit. as well as The loop control unit controls the parameter determination unit, the film thickness information acquisition unit, and the film thickness information estimation unit to alternately execute a first process and a second process. The first process includes the processing of the parameter determination unit and the processing of the film thickness information acquisition unit, and the second process includes the processing of the parameter determination unit and the processing of the film thickness information estimation unit.
2. The parameter optimization device according to claim 1, wherein, It also has a learning unit that uses the film thickness information obtained by the film thickness information acquisition unit to learn the estimation model.
3. The parameter optimization device according to claim 1, wherein, The second process involves repeatedly performing the parameter determination unit's processing and the film thickness information estimation unit's processing multiple times. If the number of times the film thickness information is estimated by the film thickness information estimation unit in the second process has reached a predetermined number, the cycle control unit executes the first process.
4. The parameter optimization apparatus according to any one of claims 1 to 3, wherein, The parameter determination unit calculates an evaluation value based on the film thickness information and determines the control parameters to be evaluated next based on the evaluation value.
5. The parameter optimization device according to claim 4, wherein, The parameter determination unit uses a multi-task Bayesian optimization algorithm to determine the control parameters to be evaluated next.
6. A parameter optimization method for optimizing parameters used to control a coating apparatus, the coating apparatus spraying a processing liquid onto a substrate to form a coating film, wherein, The parameter optimization method includes: a) The parameter determination process uses multiple film thickness information, each associated with a specific parameter, to determine the control parameters to be evaluated next; b) A film thickness information acquisition step, acquiring film thickness information based on the measurement results of the film thickness of the coated film, wherein the coated film is formed by controlling the coating apparatus using control parameters determined in the parameter determination step; and c) Film thickness estimation step: Using an estimation model that takes control parameters and process information as input and outputs film thickness information, the film thickness information corresponding to the control parameters determined by the parameter determination step is estimated. The parameter optimization method alternately performs a first process and a second process. The first process includes the parameter determination process and the film thickness information acquisition process, and the second process includes the parameter determination process and the film thickness information estimation process.
7. A computer program product comprising a computer program, wherein, When the computer program is executed by the processor, it implements the steps of the parameter optimization method of claim 6.
Citation Information
Patent Citations
JP1978046643A
Chemicals supplying device
JP1998061558A
Substrate treatment device, substrate treatment method, and computer program for substrate treatment
JP2020040046A
Parameter search method, parameter search device, and parameter search program
WO2019244474A1
Method for predicting sulfur cycle coupling denitrifying phosphorus removal process performance in two stages
CN116861341A
