Film thickness estimation device, film thickness estimation method and computer program

Abstract

This invention provides a technique for accurately estimating the thickness of a coated film formed on a substrate. A first encoder unit (201) takes film thickness data (x1) as input and outputs a first latent variable (z1). A decoder unit (203) takes the first latent variable (z1) as input and outputs reconstructed film thickness data (x1'). A second encoder unit (210) takes process parameters (x2) as input and outputs a second latent variable (z2). In a first learning process (ST1), the autoencoder (200) learns in a manner that minimizes the error between the film thickness data (x1) input to the autoencoder (200) and the output film thickness data (x1'). In the second learning process (ST2), the second encoder unit (210) learns in a way that minimizes the error between the first latent variable (z1) obtained by inputting film thickness data (x1) into the first encoder unit (201) and the second latent variable (z2) obtained by inputting process parameters (x2) into the second encoder unit (210).

Description

[Technical Field]

[0001] The subject matter disclosed in this specification relates to a technique for estimating the thickness of a coating film applied to a substrate. [Previous 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 dispense a processing liquid from a slit nozzle and coats the entire substrate being transported with the processing liquid. With the increasing demand for higher quality products 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, the dispensing characteristics of the processing liquid are repeatedly measured, and the parameters used to control the pump are adjusted to achieve optimization. [Prior Art Documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2020-040046 [Summary of the Invention]

[0004] [Problem to be Solved by the Invention] However, in the prior art, due to the repeated control processing and emission characteristic measurements, the time required for optimizing the control parameters may become excessively long. Therefore, it is considered to minimize the number of actual measurements by alternately performing optimization using a film thickness estimation model and optimization based on actual measurements that require time. However, in the aforementioned system, a film thickness estimation model needs to be prepared, and its accuracy significantly affects the accuracy of the optimization.

[0005] The object of this invention is to provide a technique for accurately estimating the thickness of a coating film formed on a substrate. [Means for Solving the Problem]

[0006] To solve the aforementioned problem, a first embodiment is a film thickness estimation device that estimates the film thickness of a coated film applied to a substrate. The device includes: a learning unit that learns a base model; and a film thickness estimation unit that uses a sub-model based on the learned base model obtained by the learning unit to estimate film thickness data according to process parameters related to the coating process. The base model has an autoencoder and a second encoder unit. The autoencoder has: a first encoder unit that takes the film thickness data as input and outputs a first latent variable compressed into a low dimension; and a decoder unit that takes the first latent variable output from the first encoder unit as input and outputs reconstructed film thickness data. The second encoder unit takes the process parameters as input and outputs a dimension. The sub-model includes a second latent variable consistent with the first latent variable, and a second latent variable. The sub-model comprises: a second encoder unit; and a decoder unit connected to the output side of the second encoder unit. The learning unit performs a first learning process and a second learning process. The first learning process enables the autoencoder to learn in a manner that minimizes the error between the film thickness data input to the autoencoder and the film thickness data output based on the input. The second learning process enables the second encoder unit to learn in a manner that minimizes the error between the first latent variable obtained by inputting the film thickness data to the first encoder unit and the second latent variable obtained by inputting the process parameters corresponding to the film thickness data input to the first encoder unit to the second encoder unit.

[0007] The second form is a film thickness estimation device according to the first form, wherein the process parameters include control parameters for controlling the coating apparatus, and the coating apparatus coats the substrate with a treatment liquid.

[0008] The third form is a film thickness estimation device based on the second form, wherein the process parameters further include device information inherent to the coating device.

[0009] The fourth form is a film thickness estimation method for estimating the film thickness of a coated film applied to a substrate. The method includes: a) a learning step, where a base model is trained; and b) an estimation step, where a sub-model based on the trained base model obtained from the learning step is used to estimate film thickness data according to process parameters related to the coating process. The base model has an autoencoder and a second encoder unit. The autoencoder has: a first encoder unit that takes the film thickness data as input and outputs a first latent variable compressed into a low-dimensional form; and a decoder unit that takes the first latent variable output from the first encoder unit as input and outputs reconstructed film thickness data. The second encoder unit takes the process parameters as input and outputs a low-dimensional form. The sub-model includes: a second encoder unit; and a decoder unit connected to the output side of the second encoder unit. The learning step includes: a1) a first learning step, in which the autoencoder learns in a manner that minimizes the error between the film thickness data input to the autoencoder and the film thickness data output based on the input; and a2) a second learning step, in which the second encoder unit learns in a manner that minimizes the error between the first latent variable obtained by inputting the film thickness data to the first encoder unit and the second latent variable obtained by inputting the process parameters corresponding to the film thickness data input to the first encoder unit to the second encoder unit.

[0010] The fifth aspect is a computer program that causes the computer to execute the film thickness estimation method according to the fourth aspect. [Effects of the Invention]

[0011] According to the first to fifth forms, by using the latent variables of the autoencoder as hints to enable the second encoder unit to learn, the learning accuracy of the autoencoder can be effectively utilized while the second encoder unit learns. Therefore, by using a sub-model that combines the second encoder unit and the decoder unit, film thickness data can be estimated with high accuracy based on process parameters.

[0012] According to the film thickness estimation device of the second embodiment, since the film thickness data can be estimated by taking into account the control information of the coating device, the accuracy of film thickness estimation can be improved.

[0013] According to the film thickness estimation device of the third form, since the film thickness data can be estimated by taking into account the characteristics of the coating device, the accuracy of film thickness estimation can be improved.

Implementation Method

[0015] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. Furthermore, in the drawings, for ease of understanding, the dimensions or quantities of the various parts are sometimes exaggerated or simplified in the illustration.

[0016] <1. Embodiment> FIG1 is a schematic diagram showing the overall structure of a coating apparatus 1 including the film thickness estimation device of the embodiment. The coating apparatus 1 is a substrate processing apparatus that forms a coating film on a substrate S by dispensing a processing liquid onto the upper surface Sf of a substrate S. As described later, the control unit 9 of the coating apparatus 1 functions as a film thickness estimation device for estimating the film thickness of the coating film.

[0017] The substrate S is, for example, a glass substrate for a liquid crystal display device. Alternatively, the substrate S can also be various substrates to be processed for electronic devices, such as semiconductor wafers, glass substrates for light masks, glass substrates for plasma displays, glass or ceramic substrates for magnetic disks / optical disks, glass substrates for organic electroluminescence (EL), glass substrates or silicon substrates for solar cells, other flexible substrates, and printed substrates. The coating apparatus 1 is, for example, a slot coater.

[0018] In Figure 1, an XYZ coordinate system is defined to illustrate the arrangement of the components of the coating apparatus 1. The transport direction Dt of the substrate S is designated as the "X direction". In the X direction, the direction in which the substrate S travels (towards the downstream of the transport direction Dt) is the +X direction, and its opposite direction (towards the upstream of the transport direction Dt) is 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 as the horizontal directions. In the Z direction, the +Z direction is defined as the upward direction, and the -Z direction as the downward direction. Moreover, these directions are not intended to limit the arrangement of the coating apparatus.

[0019] The coating apparatus 1, facing the +X direction, includes, in sequence, an input conveyor 100, an input transfer unit 2, a floating platform unit 3, an output transfer unit 4, and an output conveyor 110. The input conveyor 100, input transfer unit 2, floating platform unit 3, output transfer unit 4, and output conveyor 110 form a transport path for the substrate S to pass through. In addition, 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.

[0020] The substrate S is conveyed from an apparatus located upstream 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 conveyed 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).

[0021] The input transfer unit 2 includes a roller conveyor 21 and a rotation / lifting drive mechanism 22. The rotation / lifting drive mechanism 22 rotates each roller of the roller conveyor 21 and lifts the roller conveyor 21. By rotating the roller conveyor 21, the substrate S is transported downstream (in the +X direction) in a horizontal position. In addition, by lifting 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 platform unit 3 via the input transfer unit 2.

[0022] As shown in Figure 1, the floating stage section 3 is generally flat. The floating stage section 3 is divided into three parts along the X direction. The floating stage section 3 includes, in sequence towards the +X direction, an inlet floating stage 31, a coating stage 32, and an outlet floating stage 33. The upper surfaces of the inlet floating stage 31, the coating stage 32, and the outlet floating stage 33 are located on the same plane. The floating stage section 3 also includes a lifting pin drive mechanism 34, a floating control mechanism 35, and a lifting drive mechanism 36. The lifting pin drive mechanism 34 causes multiple lifting pins disposed on the inlet floating stage 31 to rise and fall. The floating control mechanism 35 supplies compressed air for lifting the substrate S to the inlet floating stage 31, the coating stage 32, and the outlet floating stage 33. The lifting drive mechanism 36 causes the outlet floating stage 33 to rise and fall.

[0023] Multiple ejection holes, which eject compressed air supplied from the levitation control mechanism 35, are arranged in a matrix on the upper surface of the inlet levitation platform 31 and the upper surface of the outlet levitation platform 33. When compressed air is ejected from each ejection hole, the substrate S floats upward relative to the levitation platform 3. As a result, the lower surface Sb of the substrate S separates from the upper surface of the levitation platform 3, and the substrate S is supported in a horizontal position. The distance (levitation amount) between the lower surface Sb of the substrate S and the upper surface of the levitation platform 3 when the substrate S is levitated is, for example, 10 μm or more and 500 μm or less.

[0024] On the upper surface of the coating stage 32, there are ejection holes for spraying compressed air supplied from the levitation control mechanism 35 and suction holes for drawing in air. The ejection holes and suction holes are arranged alternately in 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 from the suction holes. As a result, the levitation amount of the substrate S relative to the coating stage 32 is precisely controlled such that the position of the upper surface Sf of the substrate S passing above the coating stage 32 in the Z direction is a predetermined value. Furthermore, the levitation amount 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 described later. In addition, the levitation amount of the substrate S relative to the coating stage 32 is preferably adjustable with high precision by airflow control.

[0025] The substrate S, which is fed into the floating platform section 3, is given a pushing force in the +X direction by the roller conveyor 21, and is thus transported to the inlet floating platform 31. The inlet floating platform 31, the coating platform 32, and the outlet floating platform 33 support the substrate S in a floating state. As the floating platform section 3, for example, the structure described in Japanese Patent No. 5346643 can be used.

[0026] The substrate transport unit 5 is disposed below the floating platform unit 3. The substrate transport unit 5 includes a suction cup mechanism 51 and an adsorption / transfer control mechanism 52. The suction cup mechanism 51 includes an adsorption pad (not shown) disposed on the adsorption member. The suction cup mechanism 51 supports the substrate S from below by abutting the adsorption pad against the periphery of the lower surface Sb of the substrate S. The adsorption / transfer control mechanism 52 adsorbs the substrate S onto the adsorption pad by applying a negative pressure. In addition, the adsorption / transfer control mechanism 52 causes the substrate transport unit 5 to reciprocate in the X direction.

[0027] The suction cup mechanism 51 holds the substrate S in a position where the lower surface Sb of the substrate S is higher than the upper surface of the lifting platform 3. With the peripheral portion held by the suction cup mechanism 51, the substrate S maintains a horizontal posture by the buoyancy provided by the lifting platform 3.

[0028] As shown in FIG. 1, the coating apparatus 1 includes a sensor 61 for measuring plate thickness. The sensor 61 is disposed near the roller conveyor 21. The sensor 61 detects the position of the upper surface Sf of the substrate S held by the suction cup mechanism 51 in the Z direction. In addition, since the suction cup (not shown) that does not hold the substrate S is located directly below the sensor 61, the sensor 61 can detect the position of the upper surface of the adsorption member, i.e., the adsorption surface, in the vertical direction Z.

[0029] The suction cup mechanism 51 holds the substrate S that has been loaded into the floating stage 3 and moves it 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 above the coating stage 32. Then, the substrate S moves from the outlet floating stage 33 to the output transfer section 4.

[0030] The output transfer unit 4 moves the substrate S from a position above the outlet lifting platform 33 toward 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 drives the roller conveyor 41 to rotate and moves the roller conveyor 41 up and down in the Z direction. By rotating each roller of the roller conveyor 41, the substrate S moves in the +X direction. In addition, by lifting the roller conveyor 41, the substrate S is displaced in the Z direction.

[0031] 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, thereby moving the substrate S outside the coating apparatus 1. In addition, the input conveyor 100 and the output conveyor 110 are part of the coating apparatus 1. The input conveyor 100 and the output conveyor 110 may also be incorporated into a device different from the coating apparatus 1.

[0032] The coating mechanism 7 coats the upper surface Sf of the substrate S with a treatment liquid. The coating mechanism 7 is disposed 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 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 (the position indicated by the solid line in FIG. 1) and a maintenance position described later. The treatment liquid supply mechanism 8 is connected to the nozzle 71. The treatment liquid is supplied to the nozzle 71 by the treatment liquid supply mechanism 8, thereby discharging the treatment liquid from the discharge port disposed on the lower surface of the nozzle 71.

[0033] FIG2 is a diagram showing the structure of the treatment liquid supply mechanism 8 included in the coating apparatus 1 shown in FIG1. ​​The treatment liquid supply mechanism 8 includes: a pump 81, a pipe 82, a treatment liquid replenishment unit 83, a pipe 84, an on / off valve 85, a pressure sensor 86, and a drive unit 87. The pump 81 is a supply source for supplying treatment liquid to the nozzle 71, and supplies treatment liquid by means of volume change. The pump 81 can be, for example, a bellows type pump as described in Japanese Patent Application Publication No. 10-61558. As shown in FIG2, 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 treatment liquid replenishment unit 83 via the pipe 82. The other end of the flexible tube 811 is connected to the nozzle 71 via the pipe 84.

[0034] 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 a working disc section 816. The pump chamber 815 is disposed between the flexible pipe 811 and the bellows 812. An incompressible medium is sealed in the pump chamber 815. The working disc section 816 is connected to the drive section 87.

[0035] 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 pipe 82, and an on / off valve 833 is inserted into the pipe 82. The on / off valve 833 is opened and closed according to instructions from the control unit 9. When 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. In addition, when 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 is restricted.

[0036] Pipeline 84 is connected to the output side of pump 81. An on / off valve 85 is installed on pipeline 84. The on / off valve 85 opens and closes pipeline 84 according to instructions from control unit 9. The opening and closing of the on / off valve 85 switches the supply and stop of the processing fluid to nozzle 71. A pressure sensor 86 is installed on pipeline 84. The pressure sensor 86 detects the pressure (discharge pressure) applied to the processing fluid supplied to nozzle 71 and outputs a signal indicating the detected pressure value to control unit 9.

[0037] As shown in Figures 1 and 2, a sensor 62 is disposed in 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 in a non-contact manner. 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 based on the positioning mechanism. Furthermore, as the sensor 62, for example, an optical sensor or an ultrasonic sensor can be used.

[0038] Furthermore, the substrate S removed from the output conveyor 110 is dried using a drying device or the like, thereby forming a coated film. Moreover, as shown in FIG1, the substrate S with the coated film formed is transported to a film thickness measuring device AP1 as needed to measure the film thickness. The film thickness measuring device AP1 can be, for example, a spectrophotometer, an X-ray reflectance measuring device, etc.

[0039] 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 bar 723. The nozzle cleaning standby unit 72 cleans the nozzles 71 and forms liquid accumulation, thereby adjusting the nozzle outlet of the nozzles 71 to a state suitable for coating processing.

[0040] FIG3 is a block diagram showing the structure of the control unit 9 included in the coating apparatus 1 shown in FIG1. ​​The control unit 9 controls the operation of each component within the coating apparatus 1. The control unit 9 is a computer including a processor 91 and a memory 93. The processor 91 has, for example, a central processing unit (CPU). The memory 93 has a temporary storage device such as random access memory (RAM). The memory 93 sometimes also has a non-temporary storage device such as a hard disk drive (HDD) or a solid state drive (SSD). The memory 93 is connected to the processor 91 via bus wiring.

[0041] The control unit 9 includes a display device 95 for displaying various messages and an input device 97 for accepting 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 a keyboard. In addition, the display device 95 can also function as an input device by having a touch panel.

[0042] The computer program 931 is stored in the memory 93. The computer program 931 is provided to the control unit 9 via a non-transitory recording medium M. That is, the computer program 931 is recorded in the recording medium M in a form that can be read by the computer's control unit 9. Specifically, the recording medium M is a Universal Serial Bus (USB) memory, a digital versatile disc (DVD), or a magnetic disk.

[0043] The processor 91 executes the computer program 931, thereby functioning as the output control unit 910, the film thickness acquisition unit 911, the learning unit 913, the film thickness estimation unit 915, and the optimization unit 917.

[0044] The discharge control unit 910 controls the operation (discharge operation) of the pump 81 that delivers the processing liquid to the nozzle 71 based on preset control parameters. In the coating apparatus 1, 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, the control parameters closely related to the discharge pressure waveform are optimized in advance before the production (or mass production) of the substrate S begins, so that the discharge pressure waveform has an ideal shape.

[0045] The control parameters are set values ​​for pump control, such as various parameters that specify the movement of the working disc 816 (e.g., acceleration time, steady-state speed, time to maintain steady-state speed, deceleration time, etc.).

[0046] The film thickness acquisition unit 911 acquires film thickness data. The film thickness data is information representing the thickness of the coating film formed on the substrate S. The film thickness data is data obtained through measurement based on the film thickness measuring device AP1. The film thickness measuring device AP1, for example, scans the coating film formed on the substrate S in one direction to measure the film thickness. In this case, the distribution of the film thickness in one direction (film thickness profile) is measured as film thickness data. When the film thickness measuring device AP1 scans the coating film at a certain speed, the film thickness data is acquired as time-series data.

[0047] The learning unit 913 has the function of learning the basic model Md1. The basic model Md1 is a learning model incorporated into the computer program 931. The film thickness estimation unit 915 uses the sub-model Md2 to estimate film thickness data based on process parameters. The sub-model Md2 is a model based on the basic model Md1 learned by the learning unit 913, and is a model that takes process parameters as input and outputs film thickness data. The structure of the basic model Md1 and the sub-model Md2 will be described later. The control unit 9, which has the learning unit 913 and the film thickness estimation unit 915, functions as a film thickness estimation device for estimating the film thickness of the coating film formed on the substrate S.

[0048] Process parameters are data related to the coating process used to form the coated film, and are different from film thickness data. Process parameters are, for example, non-time-series data. Process parameters include, for example, control parameters. In addition, process parameters may also include device information inherent to the coating apparatus 1. Device information includes, for example, model name, manufacturing number, equipment type or model, device structure information, firmware or software version, or maintenance history, etc. In addition, process parameters may also include information other than the device information, such as the temperature or humidity of the environment in which the coating process is performed, information related to the processing liquid (composition, etc.), or the type of substrate, etc.

[0049] The optimization unit 917 optimizes the control parameters. That is, the optimization unit 917 brings the control parameters close to their optimal values ​​so that a coating film of ideal thickness can be formed on the substrate S. Specifically, the optimization unit 917 searches for the optimal solution that maximizes or minimizes a predefined objective function (evaluation benchmark) according to an optimization algorithm. The input (variable) of the objective function is the control parameters, and the output of the objective function is, for example, the deviation of the film thickness (standard deviation) when the objective is to achieve uniformity of film thickness. The optimization algorithm may employ nonlinear programming, genetic algorithms, Bayesian optimization, etc.

[0050] The optimization unit 917 selects control parameters and acquires film thickness data corresponding to the control parameters. The film thickness data may be data obtained by actually measuring the coating film formed on the substrate S using the film thickness measuring device AP1, or it may be data estimated by the film thickness estimation unit 915. By using the film thickness data estimated by the film thickness estimation unit 915, the actual measurement of film thickness data can be reduced. Therefore, the optimization of control parameters can be performed efficiently.

[0051] Figure 4 is a flowchart illustrating the film thickness estimation method of an embodiment. The film thickness estimation method shown in Figure 4 is a process executed by the processor 91 of the control unit 9 according to the process specified by the computer program 931.

[0052] The film thickness estimation method includes a learning step S1 and an estimation step S2. The learning step S1 is a step of learning the base model Md1. The learning step S1 is performed by the learning unit 913. The estimation step S2 is a step of estimating film thickness data based on given process parameters. The estimation step S2 is performed by the film thickness estimation unit 915.

[0053] <Learning Steps> Figure 5 is a diagram showing the structure of the base model Md1 used in learning step S1 shown in Figure 4. The base model Md1 has an autoencoder 200. The autoencoder 200 is a neural network that compresses and reconstructs the input film thickness data.

[0054] The auto encoder 200 includes a first encoder section 201 and a decoder section 203. The first encoder section 201 takes film thickness data as input and compresses it into a low-dimensional first latent variable (latest vector). The decoder section 203 takes the first latent variable as input and outputs the reconstructed film thickness data.

[0055] The structures of the first encoder section 201 and the decoder section 203 of the auto encoder 200 are designed to maintain symmetry. For example, the first encoder section 201 is configured with a one-dimensional convolutional layer, and the data size of the latent variable is converted by adjusting the stride or the number of filters of the convolution. In addition, the decoder section 203 is configured with a one-dimensional deconvolutional layer, and performs recovery to the same data size as the original film thickness data x1.

[0056] The basic model Md1 also has a second encoder unit 210. The second encoder unit 210 takes process parameters as input and a second latent variable as output. The dimension of the second latent variable is the same as that of the first latent variable. The second encoder unit 210 may employ a multi-layer (e.g., two-layer) fully bonded layer.

[0057] In the learning step S1, the learning unit 913 generates a dataset of film thickness data x1 and process parameters x2, and enables the base model Md1 to learn. Specifically, firstly, the learning unit 913 learns in a way that minimizes the error between the film thickness data (x1) input to the autoencoder 200 and the reconstructed film thickness data (x1') output based on the input. Hereinafter, this learning is referred to as "first learning process ST1". Through the first learning process ST1, parameters such as weights in the first encoder unit 201 and the decoder unit 203 are optimized. By performing the first learning process ST1, the first encoder unit 201 can output a first latent variable z1, which can be used by the decoder unit 203 to reconstruct film thickness data x1' that is close to the original film thickness data x1 with good accuracy.

[0058] Furthermore, the learning unit 913 performs learning in a manner that minimizes the error between the first latent variable z1 obtained by inputting film thickness data x1 into the first encoder unit 201 and the second latent variable z2 obtained by inputting process parameter x2 into the second encoder unit 210. Hereinafter, this learning is referred to as "second learning process ST2". Through the second learning process ST2, the parameters of the second encoder unit 210 are optimized. That is, by performing the second learning process ST2, the second encoder unit 210 can output a second latent variable z2 that is close to the first latent variable z1 used to reconstruct the original film thickness data x1.

[0059] In addition, during the learning step S1, the second learning process ST2 may also be executed after the first learning process ST1 has completely finished. Alternatively, the first learning process ST1 and the second learning process ST2 may be executed alternately.

[0060] Furthermore, the rectified linear unit (ReLU) function can be used as the activation function in the learning step S1 throughout the model. Additionally, the mean absolute error (MAE) can be used as the loss function. In the loss function of the first learning process ST1, the difference between the film thickness data x1' (output value) and the film thickness data x1 (positive solution value) is considered. Furthermore, in the loss function of the second learning process ST2, the difference between the first latent variable z1 when the process parameter x2 is input and the first latent variable z1 when the film thickness data x1 is input is considered. Furthermore, batch normalization is applied to the convolutional and deconvolutional layers in the autoencoder 200 and the fully coupled layer of the second encoder section 210.

[0061] <Estimation Step> Next, estimation step S2 will be explained. Figure 6 is a diagram showing the structure of the sub-model Md2 used in estimation step S2 shown in Figure 4. Sub-model Md2 has a second encoder section 210 and a decoder section 203 obtained in learning step S1. The decoder section 203 is connected to the output side of the second encoder section 210.

[0062] In the estimation step S2, the film thickness estimation unit 915 calculates the film thickness data by inputting the process parameters of the film thickness to be estimated into the sub-model Md2. Specifically, the second encoder unit 210 of the sub-model Md2 outputs a second latent variable z2 based on the input process parameters. Then, the decoder unit 203 outputs film thickness data x1' corresponding to the input second latent variable z2.

[0063] <Summary> As described above, the film thickness estimation device implemented by the control unit 9 includes a learning unit 913 and a film thickness estimation unit 915. The learning unit 913 enables the base model Md1 to learn. The film thickness estimation unit 915 uses a sub-model Md2 based on the learned base model Md1 obtained by the learning unit 913 to estimate film thickness data based on process parameters related to the coating process. The base model Md1 has an auto encoder 200 and a second encoder unit 210. The auto encoder 200 has a first encoder unit 201 and a decoder unit 203. The first encoder unit 201 takes film thickness data x1 as input and outputs a first latent variable z1 compressed into a low dimension. The decoder unit 203 takes the first latent variable z1 output from the first encoder unit 201 as input and outputs reconstructed film thickness data x1'. The second encoder unit 210 takes process parameter x2 as input and outputs a second latent variable z2 with the same dimension as the first latent variable z1. Sub-model Md2 has a second encoder unit 210 and a decoder unit 203 connected to the output side of the second encoder unit 210. The learning unit 913 performs a first learning process ST1 and a second learning process ST2. In the first learning process ST1, the auto encoder 200 learns in a manner that minimizes the error between the film thickness data x1 input to the auto encoder 200 and the film thickness data x1' output based on the input. In the second learning process ST2, the second encoder unit 210 learns in a manner that minimizes the error between a first latent variable z1 obtained by inputting the film thickness data x1 into the first encoder unit 201 and a second latent variable z2 obtained by inputting the process parameter x2 corresponding to the film thickness data x1 input to the first encoder unit 201 into the second encoder unit 210.

[0064] According to the structure described above, since the second encoder unit 210 learns by using the first latent variable z1 in the auto encoder 200 as a cue, the second encoder unit 210 can learn while effectively utilizing the learning accuracy of the auto encoder 200. Therefore, by using the sub-model Md2 that combines the second encoder unit 210 and the decoder unit 203, film thickness data can be estimated with high accuracy based on process parameters.

[0065] In addition, process parameter x2 includes control parameters for controlling the coating apparatus 1, which coats the substrate S with a processing liquid. According to this structure, since the film thickness data can be estimated by taking into account the control information of the coating apparatus 1, the accuracy of the film thickness estimation can be improved.

[0066] In addition, the process parameter x2 also includes device information inherent to the coating apparatus 1. According to the above structure, since the film thickness data can be estimated taking into account the characteristics of the coating apparatus 1, the accuracy of the film thickness estimation can be improved.

[0067] Although the invention has been described in detail, the description is illustrative in all respects and the invention is not limited thereto. It should be understood that numerous variations not illustrated are conceivable without departing from the scope of the invention. The structures described in the various embodiments and variations may be suitably combined or omitted as long as they do not contradict each other. [Simplified Explanation of the Diagram]

[0014] Figure 1 is a schematic diagram showing the overall structure of a coating apparatus including the film thickness estimation device of the embodiment. Figure 2 is a diagram showing the structure of the processing liquid supply mechanism included in the coating apparatus shown in Figure 1. Figure 3 is a block diagram showing the structure of the control unit included in the coating apparatus shown in Figure 1. Figure 4 is a flowchart showing the film thickness estimation method of the embodiment. Figure 5 is a diagram showing the structure of the basic model used in the learning step shown in Figure 4. Figure 6 is a diagram showing the structure of the sub-model used in the estimation step shown in Figure 4.

Claims

1. A film thickness estimation device for estimating the film thickness of a coated film applied to a substrate, the film thickness estimation device comprising: The learning department enables the basic model to learn; The learning unit includes a film thickness estimation unit that uses a sub-model based on the learned base model obtained by the learning unit to estimate film thickness data according to process parameters related to the coating process. The base model has an autoencoder and a second encoder unit. The autoencoder has: a first encoder unit that takes the film thickness data as input and outputs a first latent variable compressed into a low-dimensional form; and a decoder unit that takes the first latent variable output from the first encoder unit as input and outputs reconstructed film thickness data. The second encoder unit takes the process parameters as input and outputs a second latent variable with the same dimension as the first latent variable. The sub-model has: a second encoder unit; and a decoder unit connected to the output side of the second encoder unit. The learning unit performs a first learning process and a second learning process. The first learning process enables the autoencoder to learn in a manner that minimizes the error between the film thickness data input to the autoencoder and the film thickness data output based on the input. The second learning process enables the second encoder to learn in a manner that minimizes the error between the first latent variable obtained by inputting the film thickness data into the first encoder and the second latent variable obtained by inputting the process parameters corresponding to the film thickness data input into the first encoder into the second encoder.

2. The film thickness estimation device as described in claim 1, wherein, The process parameters include control parameters for controlling the coating apparatus, which applies a coating solution to the substrate.

3. The film thickness estimation device as described in claim 2, wherein, The process parameters also include device information inherent to the coating apparatus.

4. A film thickness estimation method for estimating the film thickness of a coating film applied on a substrate, the film thickness estimation method comprising: a) Learning steps to enable the basic model to learn; and b) an estimation step, using a sub-model based on the learned base model obtained from the learning step, to estimate film thickness data according to process parameters related to the coating process, wherein the base model has an autoencoder and a second encoder unit, the autoencoder having: a first encoder unit that takes the film thickness data as input and outputs a first latent variable compressed into a low dimension; and a decoder unit that outputs the reconstructed film thickness data as the first latent variable output from the first encoder unit, the second encoder unit taking the process parameters as input and outputting a second latent variable with the same dimension as the first latent variable, the sub-model having: the second encoder unit; The decoder unit is connected to the output side of the second encoder unit. The learning step includes: a1) a first learning step, in which the auto encoder learns in a manner that minimizes the error between the film thickness data input to the auto encoder and the film thickness data output based on the input; and a2) a step, in which the second encoder unit learns in a manner that minimizes the error between the first latent variable obtained by inputting the film thickness data to the first encoder unit and the second latent variable obtained by inputting the process parameters corresponding to the film thickness data input to the first encoder unit to the second encoder unit.

5. A computer program for causing a computer to perform the film thickness estimation method as described in claim 4.

Images