Parameter optimization device, parameter optimization method, and computer program

Abstract

This technology provides efficient optimization of control parameters even when there are errors between the trained model and the device being optimized. [Solution] The slide parameter generation unit 912 determines the slide amount that changes the control parameters and generates slide control parameters by changing the control parameters for verification by the slide amount. The optimal slide amount determination unit 916 determines the optimal slide amount based on the first film thickness information of the coating film when the control parameters for verification are applied to the coating apparatus 1, and the second film thickness information of the coating film when the slide control parameters are applied. The optimization unit 917 optimizes the control parameters applied to the coating apparatus 1 using the basic estimation model Y1 and the optimal slide amount.

Description

[Technical Field]

[0001] The subject matter disclosed herein relates to parameter optimization devices, parameter optimization methods, and computer programs. [Background technology]

[0002] In the manufacturing process of flat panel displays (FPDs), a device called a coater is used. A coater is a substrate processing device that uses a pump to discharge a processing liquid from a slit nozzle, coating the entire substrate being transported with the processing liquid. In recent years, with the increasing demand for higher product quality, it has become important for the processing liquid to form a uniform film thickness across the entire substrate. Conventionally, to achieve a uniform film thickness, the application of the processing liquid to the substrate and the measurement of the coated film thickness were repeatedly performed, and control parameters that control the discharge of the processing liquid were adjusted. In industrial machinery in general, control devices have various control parameters, which are often adjusted before shipment. This control parameter adjustment work is generally performed by engineers. However, there is a strong demand for automation of the adjustment work in order to reduce costs through labor savings and suppress variations. For this reason, many methods for efficiently optimizing control parameters have been proposed. Prior art related to the present invention includes, for example, those described in Patent Documents 1 and 2. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2020-027370 [Patent Document 2] Japanese Patent Publication No. 2022-074880 [Overview of the project] [Problems that the invention aims to solve]

[0004] Patent Document 1 describes a method for optimizing control parameters by repeatedly performing simulations and actual measurements. However, if the error between the trained model and the device being optimized is large, the optimization efficiency decreases, and it can take a considerable amount of time. Possible factors contributing to the error between the trained model and the device include individual differences in the device and differences depending on the type of processing fluid. Patent Document 2 discloses a method for efficiently searching for optimization, but even in this case, there was a concern that the optimization efficiency of the control parameters would decrease significantly if the error between the trained model and the device being optimized was large.

[0005] The object of the present invention is to provide a technique that enables efficient optimization of control parameters even when there is an error between the trained model and the device to be optimized. [Means for solving the problem]

[0006] To solve the above problems, the first embodiment is a parameter optimization apparatus comprising: a test parameter determination unit that determines a test parameter for testing; a slide parameter generation unit that determines a slide amount to change at least one parameter element of the test parameter based on the output of a basic estimation model that takes the control parameter as input and outputs film thickness information, and generates a slide control parameter that changes the test parameter by the slide amount; a film thickness information acquisition unit that acquires first film thickness information of the coated film when the test parameter is applied to the coating apparatus to be optimized, and second film thickness information of the coated film when the slide control parameter is applied, respectively; an optimal slide amount determination unit that determines an optimal slide amount to change at least one parameter element based on the first film thickness information and the second film thickness information; and an optimization unit that optimizes the control parameter to be applied to the coating apparatus to be optimized using the basic estimation model and the optimal slide amount.

[0007] The second embodiment is a parameter optimization apparatus comprising: a slide parameter generation unit that determines a slide amount to change at least one parameter element of a control parameter and generates a slide dataset from a basic dataset which is a set of pairs of the control parameter and film thickness information associated with the control parameter, including a slide control parameter obtained by changing the control parameter by the slide amount; a learning unit that learns a slide estimation model using the slide dataset which takes the slide control parameter as input and outputs the film thickness information; a test parameter determination unit that determines a test control parameter using the basic estimation model which takes the control parameter as input and outputs the film thickness information, and also determines a test slide control parameter using the slide estimation model; a film thickness information acquisition unit that acquires a first film thickness information of the coated film when the test control parameter is applied to the coating apparatus to be optimized, and a second film thickness information of the coated film when the test slide control parameter is applied, respectively; an optimal estimation model determination unit that determines the optimal estimation model from the basic estimation model and the slide estimation model based on the first film thickness information and the second film thickness information; and an optimization unit that optimizes the control parameter to be applied to the coating apparatus to be optimized using the optimal estimation model determined by the optimal estimation model determination unit.

[0008] The third embodiment is a parameter optimization apparatus according to the first or second embodiment, wherein the slide parameter generation unit determines the slide amount based on the evaluation value distribution indicated by a plurality of film thickness information output from the basic estimation model.

[0009] The fourth embodiment is a parameter optimization apparatus according to the third embodiment, wherein the slide parameter generation unit determines the slide amount based on the characteristics of the evaluation value profile on a straight line passing through the minimum value of the evaluation value distribution.

[0010] The fifth aspect is a parameter optimization apparatus according to the fourth aspect, wherein the straight line is parallel to the axis of one parameter element.

[0011] The sixth aspect is a parameter optimization apparatus according to the fourth aspect, wherein the straight line intersects the regression line of the evaluation value distribution.

[0012] The seventh embodiment is a parameter optimization device according to any of the first to sixth embodiments, wherein the control parameters for the test are control parameters optimized using the basic estimation model.

[0013] The eighth aspect is a parameter optimization method, comprising: a test parameter determination step of determining a test parameter; a slide parameter generation step of determining a slide amount to change at least one parameter element of the test parameter based on the output of a basic estimation model that takes the control parameter as input and outputs film thickness information, and generating a slide control parameter that changes the test parameter by the slide amount; a film thickness information acquisition step of acquiring first film thickness information of the coated film when the test parameter is applied to the coating apparatus to be optimized, and second film thickness information of the coated film when the slide control parameter is applied, respectively; an optimal slide amount determination step of determining an optimal slide amount to change at least one parameter element based on the first film thickness information and the second film thickness information; and an optimization step of optimizing the control parameter to be applied to the coating apparatus to be optimized using the basic estimation model and the optimal slide amount.

[0014] The ninth aspect is a parameter optimization method comprising: a slide parameter generation step of determining a slide amount to change at least one parameter element of a control parameter, and generating a slide dataset from a basic dataset which is a set of pairs of the control parameter and film thickness information associated with the control parameter, including a slide control parameter obtained by changing the control parameter by the slide amount; a learning step of learning a slide estimation model that takes the slide control parameter as input and outputs the film thickness information using the slide dataset; a test parameter determination step of determining a test control parameter using the basic estimation model that takes the control parameter as input and outputs the film thickness information, and determining a test slide control parameter using the slide estimation model; a film thickness information acquisition step of acquiring a first film thickness information of the coated film when the test control parameter is applied to a coating apparatus to be optimized, and a second film thickness information of the coated film when the test slide control parameter is applied, respectively; an optimal estimation model determination step of determining the optimal estimation model from the basic estimation model and the slide estimation model based on the first film thickness information and the second film thickness information; and an optimization step of optimizing the control parameter to be applied to the coating apparatus to be optimized using the optimal estimation model determined in the optimal estimation model determination step.

[0015] The tenth aspect is a computer program that can be executed by a computer, which causes the computer to execute the parameter optimization method of the eighth or ninth aspect. [Effects of the Invention]

[0016] According to the first to tenth embodiments, by changing the control parameters applied to the coating apparatus by an optimal slide amount, the film thickness information of the formed coating film can be brought closer to the output of the acquired trained basic estimation model. Therefore, even when using a trained basic estimation model that has errors with the coating apparatus, the control parameters can be efficiently optimized by applying the optimal slide amount.

[0017] According to the parameter optimization apparatus of the fourth embodiment, by obtaining the evaluation value profile of a single parameter element in the axial direction, the change in the evaluation value for that single parameter element can be grasped. Therefore, the amount of slide of a single parameter element that reduces the error between the basic estimation model and the coating apparatus can be appropriately calculated.

[0018] According to the parameter optimization apparatus of the fifth embodiment, the direction intersecting the regression line indicates the direction in which the evaluation value changes sharply. That is, by setting the slide amount in this direction, the change in the evaluation value can be made relatively large. Therefore, by determining the slide amount using the characteristics of the evaluation value profile in this direction as an indicator, it is possible to search for a slide amount that effectively reduces the error between the basic estimation model and the coating apparatus 1. [Brief explanation of the drawing]

[0019] [Figure 1] This diagram schematically shows the overall configuration of the coating apparatus according to the first embodiment. [Figure 2] This figure shows the configuration of the processing liquid supply mechanism included in the coating apparatus shown in Figure 1. [Figure 3] This is a block diagram showing the configuration of the control unit. [Figure 4] This diagram shows the functional blocks and data flow of the control unit. [Figure 5] This is a flowchart showing the parameter optimization method according to the first embodiment. [Figure 6] Figure 5 is a flowchart detailing the process for determining the optimal slide amount. [Figure 7A] This figure shows the distribution of evaluation values ​​in the parameter space. [Figure 7B] This figure shows the linear evaluation value profile as shown in Figure 7A. [Figure 7C] This figure shows the positions of the two control parameters generated by the slide parameter generation unit in the parameter space. [Figure 8]Figure 5 is a detailed flowchart of the optimization process. [Figure 9] This is a block diagram showing the functions of the control unit according to the second embodiment. [Figure 10] This is a flowchart showing the parameter optimization method of the second embodiment. [Figure 11] Figure 10 is a flowchart detailing the process for determining the optimal estimation model. [Figure 12] This figure shows the distribution of evaluation values ​​for film thickness information output by the basic estimation model and the slide estimation model. [Figure 13] Figure 11 is a flowchart showing the details of the optimization process. [Figure 14A] This figure shows the distribution of evaluation values ​​in the parameter space. [Figure 14B] This figure shows the linear evaluation value profile as shown in Figure 14A. [Figure 14C] This figure shows the positions of the two control parameters generated by the slide parameter generation unit in the parameter space. [Modes for carrying out the invention]

[0020] Embodiments of the present invention will be described below with reference to the attached drawings. Note that in the drawings, the dimensions and number of parts may be exaggerated or simplified for ease of understanding.

[0021] <1. First Embodiment> Figure 1 is a schematic diagram showing the overall configuration of the coating apparatus 1 according to the first embodiment. The coating apparatus 1 is a substrate processing apparatus that forms a coating film on a substrate S by discharging a processing liquid onto the upper surface Sf of the substrate S. As will be described later, the control unit 9 of the coating apparatus 1 has the function of a parameter optimization device that optimizes the control parameters for controlling the discharge of the processing liquid.

[0022] The substrate S is, for example, a glass substrate for a liquid crystal display device. The substrate S may also be a semiconductor wafer, a glass substrate for a photomask, a glass substrate for a plasma display, a substrate for a magnetic / optical disk (glass or ceramic substrate), a glass substrate for an organic EL display, a glass substrate or silicon substrate for a solar cell, or other various substrates for electronic devices such as flexible substrates and printed circuit boards. The coating apparatus 1 is, for example, a slit coater.

[0023] In Figure 1, an XYZ coordinate system is defined to show the arrangement of each element of the coating apparatus 1. The transport direction Dt of the substrate S is the "X direction". In the X direction, the direction in which the substrate S moves (downstream of the transport direction Dt) is defined as the +X direction, and the opposite direction (upstream of the transport direction Dt) is defined as the -X direction. The direction perpendicular to the X direction is the Y direction, and the direction perpendicular to both the X and Y directions is defined as the Z direction. In the following explanation, 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 defined as the upward direction, and the -Z direction is defined as the downward direction. Note that these directions are not intended to limit the arrangement direction of the coating apparatus.

[0024] The coating apparatus 1 comprises, in order in the +X direction, 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 through which the substrate S passes. The coating apparatus 1 further comprises a substrate transport unit 5, a coating mechanism 7, a processing liquid supply mechanism 8, and a control unit 9.

[0025] The substrate S is transported from a device located upstream of the coating apparatus 1 to the input conveyor 100. The input conveyor 100 comprises a roller conveyor 101 and a rotary drive mechanism 102. The rotary drive mechanism 102 rotates each roller of the roller conveyor 101. The rotation of each roller of the roller conveyor 101 transports the substrate S downstream (+X direction) in a horizontal position. "Horizontal position" refers to a state in which the main surface (the surface with the largest area) of the substrate S is parallel to the horizontal plane (XY plane).

[0026] The input transfer unit 2 includes a roller conveyor 21 and a rotation / lifting drive mechanism 22. The rotation / lifting drive mechanism 22 raises and lowers the roller conveyor 21. The rotation of each roller constituting the roller conveyor 21 causes the substrate S to be transported downstream (+X direction) in a horizontal position. In addition, the raising and lowering of the roller conveyor 21 changes the position of the substrate S in the Z direction. The substrate S is transferred from the input conveyor 100 to the levitation stage unit 3 via the input transfer unit 2.

[0027] As shown in Figure 1, the levitation stage section 3 has a substantially flat plate-like structure. The levitation stage section 3 is divided into three regions along the X direction. The levitation stage section 3 comprises, in order toward the +X direction, an inlet levitation stage 31, a coating stage 32, and an outlet levitation stage 33. The upper surfaces of the inlet levitation stage 31, the coating stage 32, and the outlet levitation stage 33 are on the same plane. The levitation stage section 3 further comprises a lift pin drive mechanism 34, a levitation control mechanism 35, and a lifting drive mechanism 36. The lift pin drive mechanism 34 raises and lowers a plurality of lift pins located on the inlet levitation stage 31. The levitation control mechanism 35 supplies compressed air for levitating the substrate S to the inlet levitation stage 31, the coating stage 32, and the outlet levitation stage 33, respectively. The lifting drive mechanism 36 raises and lowers the outlet levitation stage 33.

[0028] Multiple ejection holes for ejecting compressed air supplied from the levitation control mechanism 35 are arranged in a matrix on the upper surfaces of the inlet levitation stage 31 and the outlet levitation stage 33. When compressed air is ejected from each ejection hole, the substrate S levitates above the levitation stage section 3. As a result, the lower surface Sb of the substrate S separates from the upper surface of the levitation stage section 3, and the substrate S is supported in a horizontal position. When the substrate S is levitated, the distance between the lower surface Sb of the substrate S and the upper surface of the levitation stage section 3 (levitation amount) is, for example, in the range of 10 μm to 500 μm.

[0029] The upper surface of the coating stage 32 is provided with ejection holes for ejecting compressed air supplied by the levitation control mechanism 35 and suction holes for drawing in gas. The ejection holes and suction holes are arranged alternately in both the X and Y directions. The levitation control mechanism 35 controls the amount of compressed air ejected from the ejection holes and the amount of air drawn in from the suction holes. By controlling the ejection and suction amounts, the amount of levitation of the substrate S relative to the coating stage 32 is precisely controlled so that the position of the upper surface Sf of the substrate S passing above the coating stage 32 in the Z direction reaches the target value. The amount of levitation of the substrate S relative to the coating stage 32 is calculated by the control unit 9 based on the detection results of the sensor 61 or sensor 62, which will be described later. Furthermore, the amount of levitation of the substrate S relative to the coating stage 32 can be adjusted with high precision by airflow control.

[0030] The substrate S, which has been brought into the levitation stage section 3, receives a thrust force in the +X direction from the roller conveyor 21 and is transported onto the inlet levitation stage 31. The inlet levitation stage 31, coating stage 32, and outlet levitation stage 33 support the substrate S in a levitated state. For example, the configuration described in Japanese Patent No. 5346643 can be applied to the levitation stage section 3.

[0031] The substrate transport unit 5 is located below the levitation stage unit 3. The substrate transport unit 5 comprises a chuck mechanism 51 and a suction / travel control mechanism 52. The chuck mechanism 51 is equipped with a suction pad (not shown) formed on a suction member. The chuck mechanism 51 supports the substrate S from below by bringing the suction pad into contact with the peripheral edge of the lower surface Sb of the substrate S. The suction / travel control mechanism 52 suctions the substrate S to the suction pad by applying a negative pressure less than atmospheric pressure to the suction pad. The suction / travel control mechanism 52 also moves the substrate transport unit 5 back and forth in the X direction.

[0032] The chuck mechanism 51 holds the substrate S at a position where its lower surface Sb is higher than the upper surface of the levitation stage 3. With its peripheral edge held by the chuck mechanism 51, the substrate S maintains a horizontal position due to the buoyancy acting from the levitation stage 3.

[0033] As shown in Figure 1, the coating apparatus 1 is equipped with a sensor 61 for measuring the plate thickness. The sensor 61 is positioned adjacent to the roller conveyor 21. The sensor 61 detects the position in the Z direction of the upper surface Sf of the substrate S held by the chuck mechanism 51. Furthermore, by positioning a chuck (not shown) that does not hold the substrate S directly below the sensor 61, the sensor 61 can detect the position in the vertical Z direction of the suction surface, which is the upper surface of the suction member.

[0034] The chuck mechanism 51 moves in the +X direction while holding the substrate S that has been brought into the levitation stage section 3. As a result, the substrate S is transported from above the inlet levitation stage 31, through above the coating stage 32, to above the outlet levitation stage 33. Then, the substrate S is transported from the outlet levitation stage 33 to the output transfer section 4.

[0035] The output transfer unit 4 moves the substrate S from above the exit floating stage 33 to the output conveyor 110. The output transfer unit 4 includes a roller conveyor 41 and a rotation / lifting drive mechanism 42. The rotation / lifting drive mechanism 42 rotates the roller conveyor 41 and also raises and lowers the roller conveyor 41 along the Z direction. The rotation of each roller of the roller conveyor 41 moves the substrate S in the +X direction. In addition, the raising and lowering of the roller conveyor 41 changes the position of the substrate S in the Z direction.

[0036] 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 the rotation of each roller of the roller conveyor 111, and transports the substrate S to the outside of the coating apparatus 1. In this embodiment, the input conveyor 100 and the output conveyor 110 are configured as part of the coating apparatus 1, but they can also be incorporated into a separate apparatus from the coating apparatus 1.

[0037] The coating mechanism 7 applies the processing 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 discharge port formed on its lower surface. The nozzle 71 is connected to a positioning mechanism (not shown). The positioning mechanism moves the nozzle 71 between a coating position above the coating stage 32 (indicated by a solid line in Figure 1) and a maintenance position, which will be described later. The processing liquid supply mechanism 8 is connected to the nozzle 71. The processing liquid is supplied from the processing liquid supply mechanism 8 and discharged from the discharge port located on the lower surface of the nozzle 71.

[0038] Figure 2 shows the configuration of the processing liquid supply mechanism 8 provided in the coating apparatus 1 shown in Figure 1. The processing liquid supply mechanism 8 comprises a pump 81, piping 82, a processing liquid replenishment unit 83, piping 84, an on-off valve 85, a pressure sensor 86, and a drive unit 87. The pump 81 is the source of the processing liquid and supplies the processing liquid by volume change. As the pump 81, for example, a bellows-type pump as described in Japanese Patent Application Publication No. 10-61558 can be used. As shown in Figure 2, the pump 81 has a flexible tube 811 that can elastically expand and contract in the radial direction. One end of the flexible tube 811 is connected to the processing liquid replenishment unit 83 via piping 82. The other end of the flexible tube 811 is connected to the nozzle 71 via piping 84.

[0039] The pump 81 has a bellows 812 that is elastically deformable in the axial direction. The bellows 812 has a small bellows section 813, a large bellows section 814, a pump chamber 815, and an operating disk section 816. The pump chamber 815 is located between the flexible tube 811 and the bellows 812. An incompressible medium is sealed inside the pump chamber 815. The operating disk section 816 is connected to the drive unit 87.

[0040] The processing liquid replenishment unit 83 has a storage tank 831 for storing processing liquid. The storage tank 831 is connected to the pump 81 via piping 82. A shut-off valve 833 is provided in piping 82. The shut-off valve 833 opens and closes in response to commands from the control unit 9. When the shut-off valve 833 is open, processing liquid can be replenished from the storage tank 831 to the flexible tube 811 of the pump 81. When the shut-off valve 833 is closed, the replenishment of processing liquid from the storage tank 831 to the flexible tube 811 of the pump 81 is cut off.

[0041] The piping 84 is connected to the output side of the pump 81. The on-off valve 85 is located on the piping 84. The on-off valve 85 opens and closes the piping 84 in response to a command from the control unit 9. By opening and closing the on-off valve 85, the supply of processing liquid to the nozzle 71 is switched between supplying and stopping the liquid supply. The pressure sensor 86 is located on the piping 84. The pressure sensor 86 detects the pressure (discharge pressure) applied to the processing liquid supplied to the nozzle 71 and outputs a signal indicating the detected pressure value to the control unit 9.

[0042] As shown in Figures 1 and 2, a sensor 62 is positioned on the nozzle 71 to which the processing liquid is supplied from the processing liquid supply mechanism 8. The sensor 62 non-contactually detects the height of the substrate S in the Z direction. 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, based on the measured separation distance, the control unit 9 adjusts the coating position of the nozzle 71 by the positioning mechanism. For example, an optical sensor or an ultrasonic sensor can be used as the sensor 62.

[0043] The substrate S, which is discharged from the output conveyor 110, is dried in a drying device or the like to form a coating film. Then, as shown in Figure 1, the substrate S with the coated film is transported to the film thickness measuring instrument AP1 as needed, and the film thickness of the coating film is measured. For example, a spectroscopic ellipsometer or an X-ray reflectance measuring device can be used as the film thickness measuring instrument AP1.

[0044] The coating mechanism 7 includes a nozzle cleaning standby unit 72. The nozzle cleaning standby unit 72 performs predetermined maintenance on the nozzle 71 positioned at the maintenance location. The nozzle cleaning standby unit 72 includes a roller 721, a cleaning section 722, and a roller butt 723. The nozzle cleaning standby unit 72 prepares the nozzle 71's discharge port for coating by cleaning and forming a liquid reservoir.

[0045] Figure 3 is a block diagram showing the configuration of the control unit 9. The control unit 9 controls the operation of each component in the coating apparatus 1. A computer can be used as the control unit 9. The control unit 9 comprises a processor 91 and a memory 93. The processor 91 has, for example, a CPU (Central Processing Unit). The memory 93 has a transient storage device such as RAM (Random Access Memory). The memory 93 may also have a non-transient storage device such as an HDD (Hard Disk Drive) or SSD (Solid State Drive). The memory 93 is connected to the processor 91 via bus wiring.

[0046] 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. The input device 97 is, for example, a mouse or keyboard. The display device 95 may also function as an input device if it has a touch panel.

[0047] 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 recorded on the recording medium M in a readable format by the control unit 9, which is a computer. The recording medium M is a removable media such as a USB (Universal Serial Bus) memory, an optical disc such as a DVD (Digital Versatile Disc), or a magnetic disc.

[0048] Figure 4 is a diagram showing the functional blocks and data flow of the control unit 9. The ejection control unit 910, verification parameter determination unit 911, slide parameter generation unit 912, film thickness information acquisition unit 913, learning unit 914, estimation unit 915, optimal slide amount determination unit 916, and optimization unit 918 shown in Figure 4 are functional blocks realized by the processor 91 executing the computer program 931.

[0049] The discharge control unit 910 controls the operation (supply operation) of the pump 81 that supplies the processing liquid to the nozzle 71 based on preset control parameters. Here, in order to coat the processing liquid discharged from the nozzle 71 onto the upper surface Sf of the substrate S with a uniform film thickness, control parameters closely related to the discharge pressure waveform are optimized in advance so that the discharge pressure waveform has an ideal shape before production (or mass production) of the substrate S begins. The control parameters are set values ​​for pump control and include, for example, various parameters that define the movement of the operating disk unit 816 (e.g., acceleration time, steady speed, time to maintain steady speed, deceleration time, etc.).

[0050] The test parameter determination unit 911 determines the control parameters for the test. The slide parameter generation unit 912 determines the slide amount ΔS that changes at least one parameter element of the control parameters for the test. The slide parameter generation unit 912 also generates slide control parameters obtained by changing the control parameters for the test by the slide amount ΔS.

[0051] The film thickness information acquisition unit 913 acquires first film thickness information when the control parameters for verification are applied to the coating apparatus 1, and second film thickness information when the slide control parameters for verification are applied. The optimal slide amount determination unit 916 determines the optimal slide amount ΔS based on the evaluation values ​​indicated by the first and second film thickness information. opt To decide.

[0052] The learning unit 914 uses the basic dataset DB1 to learn a basic estimation model Y1 that takes control parameters as input and outputs film thickness information. The estimation unit 915 uses the basic estimation model Y1 to estimate film thickness information for the control parameters. The optimization unit 918 optimizes the basic estimation model Y1 and the optimal slide amount ΔS based on a predetermined algorithm. opt The control parameters applied to the coating device 1 are optimized using this method.

[0053] <Parameter Optimization Method> Figure 5 is a flowchart illustrating the parameter optimization method according to the first embodiment. The parameter optimization method includes an optimal slide amount determination process S1 and an optimization process S2 performed thereafter. These steps will be described in order below.

[0054] <Optimal slide amount determination process S1> Figure 6 is a flowchart detailing the optimal slide amount determination process S1 shown in Figure 5. As shown in Figure 6, the control unit 9 uses the basic data set DB1 to optimize the control parameters. The basic data set DB1 is a set of pairs of control parameters and film thickness information. The film thickness information is information about the film thickness of the coating film formed on the substrate S, for example, the film thickness distribution (film thickness profile) in one direction. Note that the film thickness information is not limited to the film thickness distribution, and may be an evaluation value that evaluates the film thickness distribution. For example, an index that represents the magnitude of the variation in the film thickness distribution may be used as an evaluation value that indicates uniformity. Specifically, the average and variation of the error relative to a predetermined target film thickness can be used as an index.

[0055] The data pairs included in the basic dataset DB1 include data obtained from a coating apparatus other than the coating apparatus 1 being optimized, or data obtained from the coating apparatus 1 being optimized, but under different process conditions (e.g., type of coating solution, type of substrate S, etc.). The basic dataset DB1 may be stored in memory 93, or it may be stored on a device other than the control unit 9, such as a server. The control unit 9 may also access the basic dataset DB1 via a network.

[0056] In the optimal slide amount determination process S1, the learning unit 914 first learns a basic estimation model Y1 using the basic dataset DB1 as training data, taking control parameters as input and film thickness information as output (Figure 6: First learning process S11). By using, for example, an RNN (Recurrent Neural Network) as the estimation model used for learning, it becomes possible to estimate the one-dimensional data of the film thickness distribution as film thickness information. Furthermore, if the film thickness information is used as an evaluation value indicating uniformity, for example, random forest regression can be used as the estimation model. Note that it is not mandatory for the learning of the basic estimation model Y1 to be performed by the control unit 9; a basic estimation model Y1 learned by another device may be used.

[0057] An error may occur between the output (film thickness information) of the basic estimation model Y1, which has been trained on the basic dataset DB1, and the actual output of the coating apparatus 1. This is due to individual differences in the apparatus and differences in process conditions such as the type of processing liquid and substrate S. Such errors can potentially be reduced by changing (sliding) the values ​​of some parameter elements of the control parameters. From this perspective, the optimal slide amount determination process S1 determines the optimal amount to change the parameter elements (optimal slide amount).

[0058] The test parameter determination unit 911 determines N ini pieces(N inigenerates a random control parameter (hereinafter referred to as "random control parameter") of an integer of 1 or more (Figure 6: Random control parameter generation step S12). Then, the test parameter determination unit 911 uses the basic estimation model Y1 to determine N ini Among the random control parameters of the individual, the control parameter P for which the film thickness information becomes the minimum evaluation value min_A is determined (Figure 6: Minimum evaluation value parameter determination step S13). Specifically, the estimation unit 915 estimates the film thickness information for each random control parameter using the basic estimation model Y1. Then, the test parameter determination unit 911 determines the random control parameter with the minimum evaluation value among the estimated N ini pieces of film thickness information as the control parameter P min_A . Note that the minimum evaluation value means that it is the optimal control parameter for the best uniformity of the film thickness distribution.

[0059] Note that it is not essential that the control parameter for testing is determined using the basic estimation model Y1. The control parameter for testing may be randomly selected from, for example, the basic dataset DB1.

[0060] The slide parameter generation unit 912 determines the slide amount based on the evaluation value distribution E ini shown by the N _A pieces of film thickness information obtained in the process of the minimum evaluation value parameter determination step S13. The evaluation value distribution E _A is the evaluation value distribution shown by a plurality of film thickness information output from the basic estimation model Y1. The slide amount is the amount that changes at least one parameter element among the control parameters. The method for determining the slide amount will be described while referring to FIGS. 7A, 7B, and 7C.

[0061] FIG. 7A is a diagram showing the evaluation value distribution E _A in the parameter space. Here, the case where the control parameter is composed of two parameter elements x1 and x2 will be described, but the case where the number of parameter elements is three or more can be performed in the same procedure. The evaluation value distribution E shown in FIG. 7A _AIn this diagram, groups of evaluation values ​​with small values ​​are shown with dark hatching, and groups of evaluation values ​​with large values ​​are shown with light hatching. The control parameter P for the minimum evaluation value. min_A This is the evaluation value distribution E _A It is located almost in the center.

[0062] This evaluation value distribution E _A In this case, the slide parameter generation unit 912 generates the control parameter P of the minimum evaluation value. min_A The straight line L1 (parallel line) that passes through the curve and is parallel to the axis of parameter element x1 is calculated (Figure 6: Straight line calculation process S14).

[0063] The slide parameter generation unit 912 determines the slide amount based on the evaluation value profile on this straight line L1. Figure 7B shows the evaluation value profile on the straight line L1 shown in Figure 7A. In Figure 7B, the horizontal axis represents x1 and the vertical axis represents the evaluation value. As shown in Figure 7B, the evaluation value profile increases monotonically in both the positive and negative directions of the x1 axis, with the point where the minimum evaluation value is taken as the lower end. The slide parameter generation unit 912 uses the standard deviation σ, which is one of the characteristics of this evaluation value profile, as one of its properties. _eval The slide parameter generation unit 912 then calculates the calculated standard deviation σ. _eval The values ​​ΔS1 and ΔS2 obtained as positive and negative integer multiples of σ are defined as the slide amount (Figure 6: Slide amount determination step S15). For example, the standard deviation σ _eval The value obtained by multiplying by 3 is ΔS1 (=3σ). _eval ) and the standard deviation σ _eval The value obtained by multiplying by -3 is ΔS²(=-3σ). _eval ) and are determined as the slide amount.

[0064] In the slide amount determination step S15, a positive slide amount ΔS1 that increases the value of the parameter element and a negative slide amount ΔS2 that decreases the value of the parameter element are determined. This effectively broadens the search range for an appropriate slide amount. Note that the slide amounts determined in the slide amount determination step S15 may be one or three or more.

[0065] Also, the standard deviation σ of the evaluation value profile _eval By determining the slide amounts ΔS1 and ΔS2 based on this, an effective slide amount can be searched using the change characteristics of the evaluation value profile as an indicator. Note that the slide amount is the standard deviation σ _eval It is not mandatory to make a decision based on this. For example, the slide amount may be determined based on a default value. Specifically, the slide amount is determined by the control parameter P of the minimum evaluation value. min_A The value obtained by multiplying x1 by a decimal number less than or equal to 1 (e.g., ±0.1, ±0.2, ±0.3, etc.) is also acceptable.

[0066] The slide parameter generation unit 912 determines the slide amounts ΔS1 and ΔS2, and then generates control parameters Pmin_A The control parameter P is obtained by changing the parameter element x1 by the slide amounts ΔS1 and ΔS2. min_B (=P min_A +ΔS1),P min_C (=P min_A The slide control parameters (+ΔS2) are calculated (Figure 6: Slide parameter calculation process S16).

[0067] Figure 7C shows two control parameters P generated by the slide parameter generation unit 912. min_B ,P min_C This figure shows the position of the parameter space. As shown in Figure 7C, the two generated control parameters P min_B ,P min_C The position is along the straight line L1, and the control parameter P min_A These are the positions obtained by sliding it in the positive and negative directions, respectively.

[0068] Control parameter P min_A ,P min_B ,P min_C Once determined, the discharge control unit 910 controls the control parameter P min_A ,P min_B ,P min_C The coating film is formed using the applied coating (Figure 6: Coating film formation process S17). Then, the film thickness information acquisition unit 913 obtains the control parameter P from the film thickness measuring instrument AP1.min_A Film thickness information T min _ A (First film thickness information), and control parameter P min_B ,P min_C Film thickness information T min _ B ,T min_C (Second film thickness information) is acquired for each (Figure 6: Film thickness information acquisition process S18).

[0069] The optimal slide amount determination unit 916 uses the obtained film thickness information T min_A ,T min_B ,T min_C Based on the evaluation value shown, the optimal slide amount ΔS opt The optimal slide amount is determined (Figure 6: Optimal slide amount determination process S19). Specifically, the estimation unit 915 uses the basic estimation model Y1 to determine the control parameters Pmin_A,P min_B, P min_C Film thickness information P min_A_est ,P min_B_est ,P min_C_est The optimal slide amount determination unit 916 then estimates the actual film thickness information T. min_A ,T min_B ,T min_C The evaluation value shown and the estimated film thickness information P min_A_est ,P min_B_est ,P min_C_est The error is calculated for each. Then, the optimal slide amount determination unit 916 determines the slide amount of the control parameter with the smallest error as the optimal slide amount ΔS opt For example, control parameter P min_B If the error is minimized, the optimal slide amount ΔS opt The sliding amount ΔS1 is defined as the sliding amount. Also, the control parameter P min_A If the error is minimized, the optimal slide amount ΔS opt This is considered to be zero. The determined optimal slide amount ΔS opt This is stored in memory 93. This completes the optimal slide amount determination process S1.

[0070] <Optimization process S2> Next, the optimization process S2 shown in Figure 5 will be explained. Figure 8 is a detailed flowchart of the optimization process S2 shown in Figure 5. In the optimization process S2, as will be described later, the optimization unit 918 determines the optimal slide amount ΔS determined in the optimal slide amount determination process S1. opt Using the basic estimation model Y1, the control parameters are optimized.

[0071] Furthermore, when the optimization unit 918 executes the optimization process S2, it sets an estimated upper limit of N in advance. est and the upper limit of the number of searches N all This is set. Estimated upper limit number N est This indicates the number of times the estimation process S23, described later, is repeated. The optimization unit 918 also counts the number of times the estimation process S23 has been executed, so the estimated execution count C is... est Use this as a variable. Also, the search limit is N. all This is the number of times the control parameters should be searched in a single optimization process (i.e., the number of times the optimization unit 918 determines the control parameters). The optimization unit 918 counts the number of times the search has been performed, so the number of search executions C ser Use it as a variable.

[0072] When the optimization process S2 is started, the optimization unit 918 determines the initial control parameters to be evaluated (Figure 8: Initial parameter determination step S21). The initial control parameters may be those already used in the coating apparatus 1, or they may be randomly determined values. In addition, the control parameters P determined in the minimum evaluation value parameter determination step S13 shown in Figure 6 may also be used. min_A This may be used.

[0073] Next, the optimization unit 918 calculates the estimated number of executions C. est The estimated upper limit of number of times N est It is determined whether or not the target has been reached (Figure 8: Determination step S22). In determination step S22, the estimated number of executions C est The estimated upper limit of number of times N estIf it is determined that the target has not been reached, the optimization unit 918 causes the estimation unit 915 to perform the estimation process. That is, the estimation unit 915 uses the basic estimation model Y1 to estimate the film thickness information for the control parameters determined in the initial parameter determination step S21 (or the evaluation parameter determination step S29 described later) (Figure 8: Estimation step S23). The optimization unit 918 also determines the number of estimation executions C. est Add 1 to it (Figure 8: Increment process S24). Then the optimization unit 918 proceeds to the evaluation parameter determination process S29.

[0074] In the determination step S22, the estimated number of executions C est The estimated upper limit of number of times N est If it is determined that the target has been reached, the film thickness information acquisition unit 913 acquires the film thickness information (Figure 8: Film thickness information acquisition step S25). Specifically, the control parameters determined in the initial parameter determination step S21 (or the evaluation parameter determination step S29 described later) are applied to the coating apparatus 1. However, here, the value of the parameter element x1 of the control parameters is set to the optimal slide amount ΔS described above. opt The slide control parameters, which have been changed by only a small amount, are applied to the coating apparatus 1. More specifically, the discharge control unit 910 controls the pump 81 based on the slide control parameters to form a coating film on the substrate S, and furthermore, the thickness of the coating film is measured by the thickness measuring instrument AP1. Then, the thickness information acquisition unit 913 acquires the thickness information, which is the measurement result of the thickness. After the thickness information acquisition step S25, the optimization unit 918 estimates the number of executions C est Set it to 0 (Figure 8: Reset process S26).

[0075] After the reset step S26, the optimization unit 918 determines whether or not to update the basic estimation model Y1 (Figure 8: update determination step S27). The criterion for deciding whether or not to update the basic estimation model Y1 is, for example, whether or not the number of film thickness data points acquired in the film thickness information acquisition step S25 that have not yet been used to train the basic estimation model Y1 exceeds a predetermined number. In this case, if the number of unused film thickness data points exceeds a predetermined number, the basic estimation model Y1 is updated.

[0076] In the update determination step S27, when it is determined to update the basic estimation model Y1, the learning unit 914 updates the basic estimation model Y1 (Fig. 8: model update step S28). That is, using the dataset of the control parameters and the film thickness information added in the film thickness information acquisition step S25 as teacher data, the basic estimation model Y1 is relearned. In the model update step S28, the control parameters of the dataset used for relearning are the original control parameters before the optimal slide amount ΔS opt is applied. Thereby, it is possible to prevent the basic estimation model Y1 from being affected by the optimal slide amount ΔS opt .

[0077] When the model update step S28 is completed, the optimization unit 918 executes the evaluation parameter determination step S29. Also, when it is determined not to update in the update determination step S27, the optimization unit 918 skips the model update step S28 and executes the evaluation parameter determination step S29.

[0078] In the evaluation parameter determination step S29, the optimization unit 918 determines the control parameter to be evaluated next based on the evaluation value indicated by the film thickness information estimated in the estimation step S23 or the film thickness information obtained in the film thickness information acquisition step S25 (Fig. 8: evaluation parameter determination step S29). As an algorithm for determining the control parameter for evaluation, for example, methods such as reinforcement learning (RL: Reinforcement Learning), Bayesian optimization, or particle swarm optimization are used. According to these methods, the control parameter can be updated for each trial so that the evaluation value becomes the best.

[0079] After the evaluation parameter determination step S29, the optimization unit 918 adds 1 to the exploration execution count C ser . Then, the optimization unit 918 determines whether the exploration execution count C ser has reached the exploration upper limit count N all (Fig. 8: determination step S31). In the determination step S31, whether the exploration execution count C ser has reached the exploration upper limit count N allIf it is determined that the limit has not been reached, the optimization unit 918 returns to the determination step S22 and continues the process. On the other hand, in the determination step S31, if the number of search executions C ser has reached the upper limit of the number of searches N all it is determined that the optimization unit 918 ends the optimization process.

[0080] As described above, the optimal control parameters obtained by optimization are stored in the memory 93 and used in the subsequent coating process in the coating device 1. However, in the coating device 1, the optimal control parameters are not used as they are, but the slide control parameters obtained by changing the optimal control parameters by the optimal slide amount ΔS opt are used.

[0081] <Effect> According to the parameter optimization method of the present embodiment, by applying the slide control parameters obtained by changing the control parameters by the optimal slide amount ΔS opt to the coating device 1, the film thickness information of the formed coating film can be made closer to the output of the basic estimation model Y1. Therefore, even when using the learned basic estimation model Y1 with an error from the coating device 1, by applying the optimal slide amount ΔS opt the optimization of the control parameters can be efficiently performed.

[0082] In addition, the evaluation value profile of the straight line L1 parallel to the axis of x1 which is a parameter element of the control parameter represents the change in the evaluation value when only x1 is changed and other parameter elements are fixed to the optimal values. Therefore, when the error between the output of the basic estimation model Y1 and the output of the coating device 1 is caused by the parameter element x1, an effective optimal slide amount can be searched by calculating the slide amount for that parameter element.

[0083] <2. Second Embodiment> Next, the second embodiment will be described. In the following description, elements having the same functions as the elements already described may be denoted by the same reference numerals or reference numerals with added alphabetic characters, and detailed description may be omitted.

[0084] Figure 9 is a block diagram showing the functions of the control unit 9a according to the second embodiment. Each functional block shown in Figure 9 is a functional block realized by the processor 91 executing the computer program 931. The control unit 9a has an optimal estimation model determination unit 917 instead of an optimal slide amount determination unit 916 (Figure 4). The optimal estimation model determination unit 917 performs the process of determining the optimal estimation model for estimating film thickness information. The optimization unit 918 also processes the optimal estimation model Y determined by the optimal estimation model determination unit. _opt Use this to perform optimization of the control parameters.

[0085] Figure 10 is a flowchart showing the parameter optimization method of the second embodiment. The parameter optimization method of this embodiment includes an optimal estimation model determination process S1a and an optimization process S2. In this embodiment, the optimal estimation model determined by the optimal estimation model determination process S1a is used in the optimization process S2a.

[0086] Figure 11 is a flowchart detailing the optimal estimation model determination process S1a shown in Figure 10. In this optimal estimation model determination process S1a, as in the optimal slide amount determination process S1 of the first embodiment (Figure 6), the slide amounts ΔS1 and ΔS2 are determined by executing the first learning process S11 to the slide amount determination process S15.

[0087] The slide parameter generation unit 912 (Figure 9) creates two slide datasets DB2 and DB3 from the basic dataset DB1, which include slide control parameters obtained by changing the control parameters by the slide amounts ΔS1 and ΔS2 (Figure 11: Slide dataset creation process S16a). For example, if the basic dataset DB1 has a pair of control parameter P and film thickness information T, which is represented as [P,T], then the pair in slide dataset DB2 is represented as [P+ΔS1,T] and the pair in slide dataset DB3 is represented as [P+ΔS2,T].

[0088] The learning unit 914 uses two slide datasets DB2 and DB3 to train two slide estimation models Y2 and Y3, respectively, which take slide control parameters as input and output film thickness information (Figure 11: Learning process S16b).

[0089] After the learning process S16b, the test parameter determination unit 911 performs the same process as the random control parameter generation process S12, N ini pieces(N ini Random control parameters (where is an integer greater than or equal to 1) are generated (Figure 11: Random control parameter generation step S16c). Note that in random control parameter generation step S16c, the random control parameters generated in random control parameter generation step S12 may be used again. Then, the test parameter determination unit 911 uses two slide estimation models Y2 and Y3 to determine N ini From among the random control parameters, the control parameter P that yields the minimum evaluation value for the film thickness information estimated by each model is selected. min_B ,P min_C These are determined (Figure 11: Minimum evaluation value parameter determination process S16d).

[0090] Figure 12 shows the distribution of evaluation values ​​E for film thickness information output by the basic estimation model Y1 and slide estimation models Y2 and Y3. _A ,E _B ,E _C This figure shows the evaluation value distribution E shown by the outputs of the two slide estimation models Y2 and Y3, as shown in Figure 12. _B ,E _C This is the evaluation value distribution E shown by the output of the basic estimation model Y1. _A The distribution obtained by shifting by sliding amounts ΔS1 and ΔS2 respectively is as follows.

[0091] The slide estimation model Y2 is a model trained using the slide dataset DB2, which is a set of pairs of slide control parameters and film thickness information. Therefore, the control parameter P min_B This becomes a control parameter corresponding to the slide control parameter. Control parameter P min_C The same applies to the slide control parameters.

[0092] After the minimum evaluation value parameter determination step S16d, the coating film formation process S17 and the film thickness information acquisition step S18 are executed in the same manner as the optimal slide amount determination process S1 (Figure 6). The film thickness information acquisition step S18 determines the control parameter P min_A Film thickness information T min _ A (First film thickness information), and control parameter P min_B ,P min_C Film thickness information T min _ B ,T min_C (Second film thickness information) is acquired for each.

[0093] The optimal estimation model determination unit 917 uses the obtained film thickness information T min_A ,T min_B ,T min_C Based on the evaluation values ​​shown, the optimal estimation model is determined (optimal estimation model determination step S19a). Specifically, the film thickness information T based on actual measurements is used. min_A ,T min_B ,T min_C The evaluation values ​​shown and the control parameters P estimated using each model. min_A ,P min_B ,P min_C Film thickness information T min_A_est ,T min_B_est ,T min_C_est The errors of each model are compared. The estimation model with the smallest error (i.e., the estimation model that outputs an estimation result close to the actual measurement result) is then identified as the optimal estimation model Y. _opt It will be decided.

[0094] Figure 13 is a flowchart detailing the optimization process S2a shown in Figure 11. The basic flow of the optimization process S2a is the same as the flow of the optimization process S2 in the first embodiment (Figure 8). However, in the optimization process S2a, in the estimation step S23, the estimation unit 915 uses the optimal estimation model Y instead of the basic estimation model Y1. _opt The film thickness information is estimated using this method. In the optimization process S2a, the film thickness information acquisition process S25 is applied directly to the coating apparatus 1 without changing the control parameters.

[0095] In the second embodiment, an optimal estimation model Y outputs film thickness information that has a small error with the film thickness information based on actual measurements of the coated film formed by the coating apparatus 1. _opt This is determined. Therefore, the optimal estimation model Y _opt This allows for efficient optimization of the control parameters applied to the coating apparatus 1.

[0096] <3. Variant Example> Although embodiments have been described above, the present invention is not limited to those described above, and various modifications are possible.

[0097] For example, the method for determining the slide amounts ΔS1 and ΔS2 is not limited to the methods shown in Figures 7A and 7C. Figure 14A shows the evaluation value distribution E in the parameter space. _A This figure shows the result. In this modified example, the slide parameter generation unit 912 generates the evaluation value distribution E in the linear calculation step S14. _A The regression line L2 is calculated for this. Then, the slide parameter generation unit 912 generates the control parameter P for the minimum evaluation value. min_A We calculate the line L1a that passes through the point and intersects (in this case, is orthogonal to) the regression line L2.

[0098] Next, in the slide parameter generation unit 912, in the slide amount determination step S15, the slide amount is determined based on the evaluation value profile on the straight line L1a. Figure 14B is a diagram showing the evaluation value profile on the straight line L1a shown in Figure 14A. In Figure 14B, the horizontal axis shows the parameter elements (x1 and x2), and the vertical axis shows the evaluation value. The slide parameter generation unit 912 determines the standard deviation σ shown in this evaluation value profile. _eval The slide parameter generation unit 912 calculates the standard deviation σ. _eval The values ​​ΔS1 and ΔS2 obtained by multiplying by positive and negative integers are defined as the sliding amounts.

[0099] Figure 14C shows two control parameters P generated by the slide parameter generation unit 912. min_B ,P min_CThis figure shows the position of the parameter space. As shown in Figure 14C, the two generated control parameters P min_B ,P min_C The position of the control parameter P min_A This represents the position obtained by sliding along the line L1a in the positive and negative directions, respectively. Furthermore, the amount of sliding determined by this method is the amount by which multiple parameter elements (x1, x2) are changed.

[0100] The direction in which the regression line L2 intersects indicates the direction in which the evaluation value changes sharply. In other words, by setting the slide amount in this direction, the change in the evaluation value can be made relatively larger. For this reason, the characteristics of the evaluation value profile in this direction (e.g., standard deviation σ) _eval By using this as an indicator to determine the slide amounts ΔS1 and ΔS2, it is possible to search for a slide amount that effectively reduces the error between the basic estimation model Y1 and the coating device 1.

[0101] Although this invention has been described in detail, the above description is illustrative in all respects, and the invention is not limited thereto. It is understood that countless variations not illustrated can be conceived without falling outside the scope of this invention. The components described in each of the above embodiments and variations can be combined or omitted as appropriate, as long as they do not contradict each other. [Explanation of Symbols]

[0102] 1: Coating device 911: Test parameter determination unit 912: Slide parameter generation unit 913: Film Thickness Information Acquisition Unit 914: Learning Department 916: Optimal slide amount determination unit 917: Optimal Estimation Model Determination Unit 918: Optimization Department 931: Computer program DB1: Basic dataset DB2, DB3: Slide dataset DB3: Slide Dataset Y1: Basic estimation model Y2, Y3: Slide estimation model Y _opt :Optimal estimation model

Claims

1. A parameter optimization device, A test parameter determination unit that determines the control parameters for the test, A slide parameter generation unit determines a slide amount to change at least one parameter element of the control parameter for verification based on the output of a basic estimation model that takes control parameters as input and outputs film thickness information, and generates a slide control parameter obtained by changing the control parameter for verification by the slide amount, A film thickness information acquisition unit acquires, respectively, first film thickness information of the coated film when the control parameters for verification are applied to the coating apparatus to be optimized, and second film thickness information of the coated film when the slide control parameters are applied. An optimal slide amount determination unit that determines the optimal slide amount for changing the at least one parameter element based on the first film thickness information and the second film thickness information, An optimization unit that optimizes the control parameters to be applied to the coating apparatus to be optimized using the basic estimation model and the optimal slide amount, A parameter optimization device equipped with the following features.

2. A parameter optimization device, A slide parameter generation unit determines a slide amount that changes at least one parameter element of the control parameter, and generates a slide dataset from a basic dataset which is a set of pairs of the control parameter and film thickness information associated with the control parameter, including a slide control parameter obtained by changing the control parameter by the slide amount. A learning unit that learns a slide estimation model using the slide dataset, taking the slide control parameters as input and outputting the film thickness information, A test parameter determination unit that determines the control parameters for verification using a basic estimation model that takes the control parameters as input and the film thickness information as output, and also determines the slide control parameters for verification using the slide estimation model, A film thickness information acquisition unit acquires, respectively, first film thickness information of the coated film when the verification control parameters are applied to the coating apparatus to be optimized, and second film thickness information of the coated film when the verification slide control parameters are applied. An optimal estimation model determination unit determines the optimal estimation model from among the basic estimation model and the slide estimation model based on the first film thickness information and the second film thickness information, An optimization unit that optimizes the control parameters to be applied to the coating apparatus to be optimized, using the optimal estimation model determined by the optimal estimation model determination unit, A parameter optimization device equipped with the following features.

3. A parameter optimization device according to claim 1 or claim 2, The slide parameter generation unit is a parameter optimization device that determines the slide amount based on the evaluation value distribution shown by multiple film thickness information output from the basic estimation model.

4. A parameter optimization device according to claim 3, The slide parameter generation unit is a parameter optimization device that determines the slide amount based on the characteristics of the evaluation value profile on a straight line passing through the minimum value of the evaluation value distribution.

5. A parameter optimization device according to claim 4, A parameter optimization device in which the aforementioned straight line is parallel to the axis of one parameter element.

6. A parameter optimization device according to claim 4, The aforementioned straight line intersects the regression line of the evaluation value distribution, in a parameter optimization device.

7. A parameter optimization device according to claim 1 or claim 2, The parameter optimization device wherein the control parameters for the aforementioned test are control parameters optimized using the aforementioned basic estimation model.

8. A parameter optimization method, A test parameter determination step in which control parameters for the test are determined, A slide parameter generation step involves determining a slide amount to change at least one parameter element of the verification control parameter based on the output of a basic estimation model that takes control parameters as input and outputs film thickness information, and generating a slide control parameter obtained by changing the verification control parameter by the slide amount, A film thickness information acquisition step is performed to acquire first film thickness information of the coated film when the control parameters for verification are applied to the coating apparatus to be optimized, and second film thickness information of the coated film when the slide control parameters are applied, respectively. An optimal slide amount determination step, which determines the optimal slide amount for changing the at least one parameter element based on the first film thickness information and the second film thickness information, An optimization step to optimize the control parameters to be applied to the coating apparatus to be optimized, using the basic estimation model and the optimal slide amount, A parameter optimization method that includes this.

9. A parameter optimization method, A slide parameter generation step involves determining a slide amount that changes at least one parameter element of the control parameter, and generating a slide dataset from a basic dataset which is a set of pairs of the control parameter and film thickness information associated with the control parameter, including a slide control parameter obtained by changing the control parameter by the slide amount. A learning process involves training a slide estimation model that takes the slide control parameters as input and outputs the film thickness information using the slide dataset. A test parameter determination step involves determining the control parameters for verification using a basic estimation model that takes the control parameters as input and the film thickness information as output, and determining the slide control parameters for verification using the slide estimation model, A film thickness information acquisition step is performed to acquire first film thickness information of the coated film when the verification control parameters are applied to the coating apparatus to be optimized, and second film thickness information of the coated film when the verification slide control parameters are applied, respectively. An optimal estimation model determination step in which the optimal estimation model is determined from among the basic estimation model and the slide estimation model based on the first film thickness information and the second film thickness information, An optimization step to optimize the control parameters to be applied to the coating apparatus to be optimized, using the optimal estimation model determined by the optimal estimation model determination step, A parameter optimization method comprising the following features.

10. A computer program that can be executed by a computer, A computer program that causes the computer to execute the parameter optimization method described in claim 8 or 9.

Images