Information processing methods, computer programs, and information processing devices.

Abstract

This invention provides an information processing method, a computer program, and an information processing apparatus that enable control of substrate processing using a substrate processing model that models the behavior of a substrate processing apparatus. [Solution] The information processing method includes a substrate processing model which includes a higher-level model and a lower-level model, where the higher-level model and the lower-level model are models that receive a target trajectory and output an estimated trajectory. The information processing device acquires the higher-level target trajectory, which is the target output of the higher-level model, generates a lower-level target trajectory, which is the target output of the lower-level model, provides the lower-level target trajectory to the lower-level model to acquire the lower-level estimated trajectory output by the lower-level model, performs a simulation based on the lower-level estimated trajectory to calculate the higher-level estimated trajectory related to the output value of the higher-level model, updates the lower-level target trajectory of the lower-level model and recalculates the higher-level estimated trajectory if the higher-level estimated trajectory does not satisfy predetermined conditions, and outputs the lower-level target trajectory when the higher-level estimated trajectory satisfies predetermined conditions as the target trajectory of the lower-level model.

Description

Technical Field

[0005] ,

[0001] The present disclosure relates to an information processing method, a computer program, and an information processing apparatus.

Background Art

[0002] In Patent Document 1, an information processing system has been proposed that executes a simulation of a process state during execution in a semiconductor manufacturing apparatus using a simulation model of the semiconductor manufacturing apparatus. This information processing system acquires physical sensor data measured in a semiconductor manufacturing apparatus that is executing a process according to process parameters, executes a simulation by a simulation model according to the process parameters to calculate virtual sensor data and virtual process result data, and uses the physical sensor data, virtual sensor data, and virtual process result data to visualize the process state of the semiconductor manufacturing apparatus during the execution of the process and display it on a display unit. <​​​​​​​​​​​​​​​​​​​​​​​​​​​​An information processing method according to one embodiment is an information processing method in which an information processing device generates a target trajectory for substrate processing performed by a substrate processing device using a substrate processing model relating to substrate processing performed by the substrate processing device, wherein the substrate processing model includes a higher-level model and a lower-level model connected to the higher-level model, and the higher-level model and the lower-level model are models that receive a target trajectory which is the target of the output and output an estimated trajectory estimated for the target trajectory, wherein the information processing device acquires a higher-level target trajectory which is the target of the output of the higher-level model, generates a lower-level target trajectory which is the target of the output of the lower-level model based on the acquired higher-level target trajectory, provides the generated lower-level target trajectory to the lower-level model to acquire a lower-level estimated trajectory output by the lower-level model, performs a simulation based on the acquired lower-level estimated trajectory to calculate a higher-level estimated trajectory relating to the output value of the higher-level model, updates the lower-level target trajectory of the lower-level model and recalculates the higher-level estimated trajectory if the calculated higher-level estimated trajectory does not satisfy predetermined conditions, and outputs the lower-level target trajectory when the calculated higher-level estimated trajectory satisfies predetermined conditions as the target trajectory of the lower-level model. [Effects of the Invention]

[0006] According to this disclosure, it is expected that control of substrate processing can be realized using a substrate processing model that models the behavior of the substrate processing apparatus. [Brief explanation of the drawing]

[0007] [Figure 1] This is a schematic diagram illustrating the overview of the information processing system according to this embodiment. [Figure 2] This is a schematic diagram illustrating the overview of control of a substrate processing apparatus using a substrate processing model. [Figure 3] This is a schematic diagram showing one example configuration of a substrate processing model. [Figure 4] This is a block diagram showing one example configuration of the information processing device according to this embodiment. [Figure 5] This flowchart shows an example of the procedure for the optimization process related to the overall function Ω performed by the information processing device in this embodiment. [Figure 6] This flowchart shows an example of the procedure for the optimization process for the functional unit X, which is performed within the optimization process for the overall function Ω. [Figure 7] This flowchart shows an example of the procedure for the optimization process related to the control system Z, which is performed within the optimization process related to the functional unit X. [Figure 8] This is a schematic diagram showing the correspondence between each layer of the substrate processing model and the target period for the optimization process. [Modes for carrying out the invention]

[0008] Specific examples of information processing systems according to the embodiments of this disclosure will be described below with reference to the drawings. However, this disclosure is not limited to these examples and is intended to include all modifications within the meaning and scope of the claims as indicated by the claims.

[0009] <System Overview> Figure 1 is a schematic diagram illustrating the outline of the information processing system according to this embodiment. The information processing system according to this embodiment is configured to include an information processing device 1 and a substrate processing device 3, etc. The substrate processing device 3 is a device that performs various substrate processing on semiconductor substrates (wafers), such as CVD (Chemical Vapor Deposition), sputtering, or etching. The information processing device 1 is a device that monitors and controls the operation of the substrate processing device 3. The information processing device 1 causes the substrate processing device 3 to perform the desired substrate processing based on setting information (so-called recipes) related to substrate processing that have been prepared in advance by the user. The information processing device 1 and the substrate processing device 3 may be an integrated device or separate devices. If they are separate devices, the information processing device 1 and the substrate processing device 3 can exchange information, for example, by wired or wireless communication.

[0010] In this embodiment, the information processing device 1 generates target trajectories for various hardware units such as controllers provided by the board processing device 3 based on a given recipe, and provides the board processing device 3 with multiple target trajectories generated for each of the multiple controllers, etc. In this embodiment, the target trajectory is information that collects target values ​​of the output values ​​of the controllers, etc. at multiple points in time, and can be treated as data, for example, in which the target values ​​of the controllers, etc. are arranged in chronological order. The board processing device 3 provides the target trajectories given by the information processing device 1 to the corresponding controllers, etc., and each controller, etc. operates to realize a trajectory with the same output value as the given target trajectory.

[0011] The substrate processing device 3 is equipped with various sensors and provides the results of sensor measurements of various values ​​related to substrate processing to the information processing device 1. In this embodiment, the substrate processing device 3 measures the output value of a controller or the like that has been given a target trajectory using sensors and provides the information processing device 1 with a trajectory of the output value obtained by collecting the measured values ​​of the output value of the controller or the like at multiple points in time. The information processing device 1 can calculate control errors, etc., by comparing the target trajectory provided to the substrate processing device 3 with the trajectory of the output value provided from the substrate processing device 3 as sensor measurements, and can use the calculated errors, etc., to generate the next target trajectory.

[0012] In the information processing system according to this embodiment, a substrate processing model 5 that reproduces the behavior of the substrate processing apparatus 3 when performing substrate processing is generated in advance, and the generated substrate processing model 5 is stored in advance by the information processing apparatus 1. When controlling the substrate processing apparatus 3, the information processing apparatus 1 performs optimization processing using the substrate processing model 5 to generate an optimal target trajectory that achieves the substrate processing goals set in the recipe, etc. By providing the substrate processing apparatus 3 with the optimal target trajectory generated by the information processing apparatus 1 using the substrate processing model 5, it is expected that the substrate processing apparatus 3 will perform the optimal substrate processing and achieve the goals set in the recipe, etc.

[0013] The substrate processing model 5 can be called a digital twin of the substrate processing device 3, and various models can be employed, such as a physical model expressed by a predetermined calculation formula or a machine learning model such as a neural network. In this embodiment, the substrate processing model 5 is pre-generated by the information processing device 1 or other server device, etc., based on data obtained by performing substrate processing in the substrate processing device 3 (data such as setting value data input to the substrate processing device 3 and sensor measurement values). Since the method for generating these models is an existing technology, a detailed explanation is omitted in this embodiment.

[0014] Figure 2 is a schematic diagram illustrating the control of the substrate processing apparatus 3 using the substrate processing model 5. In the illustrated example, the substrate processing apparatus 3 is assumed to be equipped with hardware units such as a device for performing "frequency optimization calculations," a "frequency controller," a "variable capacitor controller," an "exhaust valve," and an "intake valve." The substrate processing model 5 in this example includes a frequency optimization calculation model 5a, a frequency controller model 5b, a variable capacitor controller model 5c, an exhaust valve model 5d, and an intake valve model 5e, which correspond to these hardware units provided by the substrate processing apparatus 3. Furthermore, the substrate processing model 5 includes a frequency optimization model 5f that optimizes the frequency using the frequency optimization calculation model 5a and the frequency controller model 5b, a gas pressure control model 5g that controls the gas pressure using the exhaust valve model 5d and the intake valve model 5e, a high-frequency power optimization model 5h that optimizes the high-frequency power using the frequency optimization model 5f and the variable capacitor controller model 5c, and an overall optimization model 5i that optimizes the entire substrate processing using the gas pressure control model 5g and the high-frequency power optimization model 5h.

[0015] The information processing device 1 can exchange information between the user and the substrate processing model 5 via an appropriate interface. For example, the information processing device 1 receives input of target values ​​for substrate processing from the user via the interface, calculates the optimal target trajectory that can achieve the received target values ​​through optimization processing using the substrate processing model 5, and provides the calculated target trajectory to the substrate processing device 3 to perform the substrate processing. The information processing device 1 also provides the user with information such as sensor measurement values ​​obtained as a result of the substrate processing via an appropriate interface.

[0016] Figure 3 is a schematic diagram showing one example configuration of the substrate processing model 5. The configuration of the substrate processing model 5 shown in Figure 3 is a generalization and simplification of the configuration of the substrate processing model 5 shown in Figure 2. The substrate processing model 5 is composed of multiple models connected hierarchically. In the illustrated example, the substrate processing model 5 is composed of three models corresponding to the control system equipment of the substrate processing device 3 (hereinafter referred to as control systems Y, Z, and W), a model corresponding to the higher-level functions of control systems Z and W (hereinafter referred to as functional unit X), and a model that is a higher-level function of functional unit X and control system Y and corresponds to the overall function of the substrate processing device 3 (hereinafter referred to as overall function Ω).

[0017] The multiple models included in the board processing model 5 are connected in a hierarchical structure in the form of a tree. Specifically, in the board processing model 5, the overall function Ω is located at the top level, the functional unit X and control system Y are connected as the level below the overall function Ω, and the control system Z and control system W are connected as the level below the functional unit X. In the tree structure, the overall function Ω corresponds to the root node, the control systems Y, Z, and W correspond to the leaf nodes, and the functional unit X corresponds to the other internal nodes. For example, the leaf node models can be models that reproduce the behavior of the physical hardware units of the board processing device 3, and the models of the other nodes can be models that calculate abstract or logical values ​​based on the results of the physical behavior by the leaf node models. Note that the illustrated tree structure is just one example and is not limited to this, and the board processing model 5 may have any hierarchical structure.

[0018] When comparing the configuration of the substrate processing model 5 shown in FIG. 3 with the configuration of the substrate processing model 5 shown in FIG. 2, for example, the overall function Ω can be regarded as corresponding to the high-frequency power optimization model 5h. In this case, the functional unit X can be regarded as corresponding to the frequency optimization model 5f, the control system Y corresponds to the varicon controller model 5c, the control system Z corresponds to the frequency optimization calculation model 5a, and the control system W can be regarded as corresponding to the frequency controller model 5b. Also, for example, the overall function Ω can be regarded as corresponding to the overall optimization model 5i. In this case, the functional unit X can be regarded as corresponding to the high-frequency power optimization model 5h, the control system Y corresponds to the gas pressure control model 5g, the control system Z corresponds to the frequency optimization model 5f, and the control system W can be regarded as corresponding to the varicon controller model 5c. Note that the correspondence relationships of these models are only examples and are not limited thereto.

[0019] In any case, the substrate processing model 5 according to the present embodiment is a model in which a plurality of models such as these overall functions Ω, functional units X, and control systems Y, Z, W are hierarchically connected. Each model included in the substrate processing model 5 outputs an estimated trajectory that estimates the behavior of the substrate processing apparatus 3 corresponding thereto when a target trajectory is input. The information processing apparatus 1 generates a target trajectory that can realize optimal substrate processing by performing an optimization process using this substrate processing model 5.

[0020] At this time, the information processing device 1 according to this embodiment generates an initial value for the target trajectory of a model in a lower-level layer (sub-layer) based on the target trajectory given from the higher-level layer (upper-layer) for a certain layer (target layer) of the substrate processing model 5. The information processing device 1 gives the generated initial value for the target trajectory to the lower-level model and obtains an estimated trajectory from the lower-level model. The information processing device 1 calculates the estimated trajectory of the target-level model based on the estimated trajectory of the lower-level model, calculates the value of the objective function set for the target-level model based on the calculated estimated trajectory, and determines whether the target trajectory given to the lower-level model is optimal based on the calculated value. If it is determined that it is not optimal, the information processing device 1 updates the target trajectory to the lower-level model as appropriate and gives it to the lower-level model, and repeats the same process. If it is determined that it is optimal, the information processing device 1 sets the target trajectory of the lower-level model at this time as the optimal target trajectory and gives the estimated trajectory of the target-level model at this time to the higher-level model.

[0021] The information processing device 1 according to this embodiment performs the optimization process for the target hierarchy described above recursively, from the highest level to the lowest level of the substrate processing model 5. This allows the information processing device 1 to generate initial values ​​for the target trajectory sequentially from the highest level to the lowest level, and then generate the optimal target trajectory sequentially while performing the optimization process from the lowest level to the highest level. The information processing device 1 can then provide the substrate processing device 3 with the target trajectory for each generated level, thereby causing the substrate processing device 3 to perform the desired substrate processing. It is not necessary for the information processing device 1 to provide the substrate processing device 3 with the target trajectory for all generated levels; for example, it may select and provide the target trajectory for the model corresponding to the physical hardware unit provided by the substrate processing device 3.

[0022] <Device configuration> Figure 4 is a block diagram showing an example configuration of the information processing device 1 according to this embodiment. The information processing device 1 according to this embodiment can be realized by installing a predetermined application program on a general-purpose information processing device such as a personal computer or a server computer. The information processing device 1 according to this embodiment is configured to include a processing unit 11, a storage unit 12, a communication unit 13, a display unit 14, and an operation unit 15, etc. The information processing device 1 may also be a computer such as a quantum annealer, a CMOS (Complementary Metal Oxide Semiconductor) annealer, a gate-type quantum computer, or a reservoir computer.

[0023] The processing unit 11 is composed of a CPU (Central Processing Unit), an MPU (Micro-Processing Unit), a GPU (Graphics Processing Unit), or a quantum processor, as well as ROM (Read Only Memory) and RAM (Random Access Memory). The processing unit 11 reads and executes a program 12a stored in the memory unit 12, performing various processes such as generating a target trajectory to be given to the substrate processing device 3, and acquiring sensor measurement values ​​from the substrate processing device 3.

[0024] The storage unit 12 is configured using a large-capacity storage device such as a hard disk or an SSD (Solid State Drive). The storage unit 12 stores various programs executed by the processing unit 11, and various data necessary for the processing of the processing unit 11. In this embodiment, the storage unit 12 stores the program 12a executed by the processing unit 11. The storage unit 12 is also provided with a model information storage unit 12b that stores information related to models such as the substrate processing model 5 used for substrate processing.

[0025] In this embodiment, the program (computer program, program product) 12a is provided in a form recorded on a recording medium 99 such as a memory card or optical disc, and the information processing device 1 reads the program 12a from the recording medium 99 and stores it in the storage unit 12. However, the program 12a may also be written to the storage unit 12 during the manufacturing stage of the information processing device 1, for example. Alternatively, the program 12a may be distributed by a remote server device or the like and acquired by the information processing device 1 via communication. For example, the program 12a may be read from the recording medium 99 by a writing device and written to the storage unit 12 of the information processing device 1. The program 12a may be provided by distribution via a network, or it may be provided in a form recorded on the recording medium 99.

[0026] The model information storage unit 12b stores information about various models, including the pre-generated board processing model 5. The model information may include, for example, information indicating the model's configuration (such as mathematical formulas) and information such as predetermined internal parameter values. In this embodiment, the model information storage unit 12b stores information about the configuration and parameters of each of the multiple models included in the board processing model 5, as well as information indicating how these multiple models are connected. Each model is generated, for example, by pre-collecting input / output data to the corresponding hardware unit of the board processing device 3, and determining the parameters of a pre-defined model, such as a physical calculation formula or a neural network, to reproduce the input / output correspondence based on the collected data. The generation of these models may be performed by the information processing device 1 or by other devices.

[0027] The communication unit 13 is connected to the substrate processing device 3 via a wired or wireless network N and exchanges data with the substrate processing device 3. In this embodiment, the communication unit 13 transmits data such as target trajectories provided by the processing unit 11 to the substrate processing device 3. The communication unit 13 also receives data such as sensor measurements transmitted from the substrate processing device 3 and provides it to the processing unit 11.

[0028] The display unit 14 is configured using a liquid crystal display or the like, and displays various images and characters based on the processing of the processing unit 11. The display unit 14 displays various information such as the progress of substrate processing by the substrate processing device 3, whether or not there are any abnormalities, the target trajectory transmitted to the substrate processing device 3, or sensor measurement values ​​acquired from the substrate processing device 3. The operation unit 15 receives user operations and notifies the processing unit 11 of the received operations. For example, the operation unit 15 receives user operations via an input device such as a mechanical button or a touch panel provided on the surface of the display unit 14. Alternatively, the operation unit 15 may be an input device such as a mouse and a keyboard, and these input devices may be configured to be detachable from the information processing device 1.

[0029] The memory unit 12 may be an external storage device connected to the information processing device 1. The information processing device 1 may be a multicomputer comprising multiple computers, or it may be a virtual machine virtually constructed by software. Furthermore, the information processing device 1 is not limited to the above configuration, and may not include, for example, a display unit 14 and an operation unit 15.

[0030] Furthermore, in the information processing device 1 according to this embodiment, the processing unit 11 reads and executes the program 12a stored in the storage unit 12, thereby realizing the target trajectory generation unit 11a, the control processing unit 11b, the display processing unit 11c, etc., as software-based functional units in the processing unit 11. In this figure, the functional units of the processing unit 11 that perform processing related to the control of substrate processing by the substrate processing device 3 are shown, and functional units related to other processing are omitted from the illustration.

[0031] The target trajectory generation unit 11a generates a target trajectory to be provided to the substrate processing device 3 based on the setting information (recipe) related to substrate processing input by the user and the substrate processing model 5 stored in the model information storage unit 12b. In this embodiment, for example, if the substrate processing model 5 is a tree configuration as shown in Figure 3, the target trajectory generation unit 11a determines the initial value of the target trajectory of the overall function Ω, which corresponds to the root node of the substrate processing model 5, based on the input recipe. Next, the target trajectory generation unit 11a determines the initial value of the target trajectory of the lower-level models connected to the overall function Ω, namely the functional unit X and the control system Y, based on the initial value of the target trajectory of the overall function Ω. Similarly, the target trajectory generation unit 11a determines the initial value of the target trajectory of the lower-level models connected to the functional unit X, namely the control systems Z and W, based on the initial value of the target trajectory of the functional unit X.

[0032] The control systems Z and W, which correspond to the leaf nodes of the substrate processing model 5, calculate and output an estimated trajectory of the output value by simulating their operation based on the initial values ​​of the given target trajectory. The target trajectory generation unit 11a simulates the operation of the functional unit X based on the estimated trajectory output by the control systems Z and W and the initial values ​​of the target trajectory given to the functional unit X. Based on the results of the simulation for the functional unit X, the target trajectory generation unit 11a calculates the value of the objective function defined for the functional unit X and determines whether the target trajectory given to the control systems Z and W is optimal based on the calculated value. If it is not optimal, the target trajectory generation unit 11a updates the target trajectory to the control systems Z and W from the initial value as appropriate and repeats the same process to determine the optimal target trajectory. If optimal target trajectories are obtained for control systems Z and W, the target trajectory generation unit 11a calculates the estimated trajectory for functional unit X at that time, and, based on the estimated trajectory of control system Y obtained in the same way as the estimated trajectory of functional unit X, optimizes the target trajectories of functional unit X and control system Y in the same way using the objective function of the overall function Ω.

[0033] In this way, the target trajectory generation unit 11a determines the initial values ​​of the target trajectory sequentially from the root node down to the leaf nodes in the tree structure of multiple models included in the substrate processing model 5, and repeatedly calculates the estimated trajectory and optimizes the target trajectory sequentially from the root node up to the higher levels, thereby generating the optimal target trajectory to be provided to the substrate processing device 3 through final optimization at the root node.

[0034] The target trajectory generated by the target trajectory generation unit 11a is provided to the substrate processing apparatus 3 as data in which multiple target values ​​for substrate processing over a predetermined period are arranged in time series. The target trajectory generation unit 11a repeatedly generates target trajectories by repeatedly performing the optimization process described above at predetermined intervals. In this embodiment, when the target trajectory generation unit 11a generates a target trajectory for the next lower level (second trajectory) based on the target trajectory for a predetermined period (first predetermined period) given for a certain level (first trajectory), it generates a target trajectory for the second trajectory for each second predetermined period, which is obtained by dividing the first predetermined period into multiple parts. The target trajectory generation unit 11a obtains corresponding estimated trajectories for each of the multiple target trajectories for each second predetermined period, and after the estimated trajectory for the first predetermined period is obtained from the second trajectory, it calculates the estimated trajectory for the first trajectory. In this way, the target trajectory generation unit 11a divides the period of the target trajectory sequentially from the upper trajectory to the lower trajectory, shortening the target period for optimization of the target trajectory in each trajectory. As described above, the target trajectory generation unit 11a can shorten the target period from a higher level to a lower level, but it may also adopt a target period of the same length throughout the hierarchy.

[0035] The control processing unit 11b transmits the target trajectory generated by the target trajectory generation unit 11a through optimization processing to the substrate processing unit 3, thereby controlling the substrate processing unit 3 to perform substrate processing according to this target trajectory. The optimization processing by the target trajectory generation unit 11a provides the optimal target trajectory for each model at each level of the substrate processing model 5. From the obtained target trajectories at each level, the control processing unit 11b acquires, for example, the target trajectory at the level corresponding to the hardware unit of the substrate processing unit 3, and transmits this target trajectory corresponding to the hardware unit to the substrate processing unit 3.

[0036] The control processing unit 11b also acquires information such as sensor measurements output by the substrate processing device 3 as a result of performing substrate processing, for example, when a target trajectory is provided. The control processing unit 11b may store the acquired sensor measurements in the storage unit 12 as log information for substrate processing, or it may provide the information to the user by providing it to the display processing unit 11c for display. Alternatively, the control processing unit 11b may provide the sensor measurements acquired from the substrate processing device 3 to the target trajectory generation unit 11a, and the target trajectory generation unit 11a may optimize the next target trajectory based on the provided sensor measurements.

[0037] As described above, the target trajectory generation unit 11a calculates a target trajectory at predetermined intervals. When sensor measurements obtained from the substrate processing device 3 are used to generate the next target trajectory, for example, at a predetermined cycle for generating the highest-level target trajectory of the substrate processing model 5, the control processing unit 11b provides the target trajectory generated by the target trajectory generation unit 11a to the substrate processing device 3, and the control processing unit 11b repeatedly optimizes the target trajectory generation unit 11a based on the sensor measurements obtained from the substrate processing device 3 to generate the target trajectory for the next predetermined cycle. When sensor measurements obtained from the substrate processing device 3 are not used to generate the next target trajectory, for example, the target trajectory generation unit 11a can generate target trajectories for the entire substrate processing period in advance, and the control processing unit 11b can provide the pre-generated target trajectories to the substrate processing device 3.

[0038] The control processing unit 11b also calculates the difference (error) between the target trajectory given to the substrate processing device 3 and the actual trajectory based on sensor measurements obtained from the substrate processing device 3, and determines whether the calculated error exceeds a predetermined threshold. If the error exceeds the threshold, the control processing unit 11b determines that there is an abnormality in the substrate processing being performed by the substrate processing device 3, and may take actions such as giving a command to stop the substrate processing to the substrate processing device 3 or outputting a warning message to the user.

[0039] The display processing unit 11c performs processing to display various characters and images on the display unit 14. In this embodiment, the display processing unit 11c graphs, for example, the target trajectory given to the substrate processing device 3 and the actual trajectory based on sensor measurements obtained from the substrate processing device 3, and displays them on the display unit 14. Alternatively, the display processing unit 11c may schematically illustrate the hierarchical structure of the substrate processing model 5 and display the operating status of hardware units of the substrate processing device 3 in correspondence with the models of each hierarchical level. The display processing unit 11c may also display various other information on the display unit 14.

[0040] <Target trajectory generation process> The following describes a detailed example of the target trajectory generation process performed by the information processing device 1. In this example, the substrate processing device 3 controlled by the information processing device 1 is assumed to be represented by the configuration of the substrate processing model 5 shown in Figure 3. In this explanation, the transmission period of the target trajectory from the information processing device 1 to the substrate processing device 3 is called a "run," and the period obtained by dividing the time equivalent to one run into multiple parts is called a "step."

[0041] Figure 5 is a flowchart showing an example of the optimization process for the overall function Ω performed by the information processing device 1 in this embodiment. Note that the steps in the flowchart in Figure 5 represent the process for one run, and the information processing device 1 repeats this process. Figure 6 is a flowchart showing an example of the optimization process for the functional unit X performed within the optimization process for the overall function Ω. Figure 7 is a flowchart showing an example of the optimization process for the control system Z performed within the optimization process for the functional unit X. Note that the optimization processes for control systems Y and W are the same as the optimization process for control system Z, so the flowcharts are omitted. Figure 8 is a schematic diagram showing the correspondence between each layer of the substrate processing model 5 and the target period of the optimization process.

[0042] As shown in Figure 5, the target trajectory generation unit 11a of the processing unit 11 of the information processing device 1 according to this embodiment acquires a target trajectory related to the optimization process of the overall function Ω of the substrate processing model 5 (step S1). The target trajectory of the overall function Ω may be set by the user, for example, or generated based on a substrate processing recipe created by the user, or may be determined by other methods. In step S1, the target trajectory generation unit 11a acquires a target trajectory for one run.

[0043] The target trajectory generation unit 11a generates the target trajectories (initial values) of the functional unit X and control system Y connected to the overall function Ω based on the target trajectory acquired in step S1 (step S2).

[0044] Various methods can be used to generate the initial value of the target trajectory in step S2. For example, the target trajectory generation unit 11a may use a predetermined trajectory as the initial value of the target trajectory. Alternatively, the target trajectory generation unit 11a may generate a random trajectory and use this as the initial value of the target trajectory. In this case, the target trajectory generation unit 11a may generate the trajectory within predetermined conditions, such as generating a random trajectory that follows a normal distribution with a mean of 0 and a standard deviation of 2.5 at each time step.

[0045] For example, the target trajectory generation unit 11a may store the optimal target trajectory obtained from the previous optimization process and use the stored target trajectory as the initial value for the current target trajectory. In this case, there may be cases where it is inappropriate to use the previous optimal target trajectory as the initial value for the current trajectory, for example, if the settings for the substrate processing recipe are different between the previous and current processes. Therefore, the target trajectory generation unit 11a can store information such as the recipe settings, sensor measurements obtained from the substrate processing apparatus 3 during substrate processing, or feature quantities obtained from these, in association with the optimal target trajectory obtained from the optimization process. The target trajectory generation unit 11a can search among these previously stored information for something similar to the current recipe settings, sensor measurements, or feature quantities, and use the target trajectory associated with the recipe settings, sensor measurements, or feature quantities obtained through the search as the initial value for the current target trajectory.

[0046] Furthermore, if the system is configured to store past recipe settings, sensor measurements, or feature quantities, as well as past optimal target trajectories, as described above, the information processing device 1 can calculate, for example, the error or similarity between past recipe settings, sensor measurements, or feature quantities and the current recipe settings, sensor measurements, or feature quantities. If the information processing device 1 satisfies predetermined conditions, such as an error of ±0.1% or less or a similarity of 99% or more, it may adopt the stored target trajectory as the optimal target trajectory for the current trajectory without performing the processing from step S2 onward in the flowchart, i.e., the processing in steps S3 to S9 described later, and proceed to step S10.

[0047] The method by which the target trajectory generation unit 11a generates the initial value of the target trajectory in step S2 is not limited to the above. For example, the target trajectory generation unit 11a may pre-generate a learning model that has learned the correspondence between recipe settings, sensor measurements or feature quantities, etc., and the target trajectory using machine learning, and use this learning model to generate the initial value of the target trajectory. Alternatively, the target trajectory generation unit 11a may generate the initial value of the target trajectory by appropriately combining the above-mentioned methods.

[0048] The target trajectory generation unit 11a adjusts the target trajectories of the functional unit X and control system Y generated in step S2 to conform to predetermined constraints (step S3). The constraints are, for example, upper and lower limits of the target trajectory and are predetermined based on the device characteristics of the substrate processing apparatus 3. If, for example, a portion of the target trajectory exceeds the upper limit, the target trajectory generation unit 11a can conform to the constraints by replacing this portion with a value less than the upper limit. Note that the constraints are not limited to upper and lower limits, but may include various other constraints such as an upper limit on the amount of displacement, and conforming to the constraints may be done using various methods other than value replacement.

[0049] Next, the target trajectory generation unit 11a performs optimization processing of the target trajectory in functional unit X based on the target trajectory of functional unit X that was adapted to the constraints in step S3 (step S4). The target trajectory generation unit 11a also performs estimation processing of the target trajectory in control system Y based on the target trajectory of control system Y that was adapted to the constraints in step S3 (step S5). The target trajectory generation unit 11a performs estimation processing at the hierarchy corresponding to leaf nodes in the tree structure of substrate processing model 5, and optimization processing at the other hierarchical levels. The target trajectory generation unit 11a may perform the optimization processing in functional unit X in step S4 and the estimation processing in control system Y in step S5 in any order, or it may perform the two processes simultaneously. Details of the optimization processing in functional unit X are shown in Figure 6, and details of the estimation processing in control system Y are the same as the estimation processing in control system Z shown in Figure 7, so they are not shown.

[0050] In addition, there are cases where the processing for functional unit X in step S4 can be omitted. For example, if it is known that functional unit X can follow the target trajectory without problems within a certain range of target trajectories, it is determined whether the target trajectory of functional unit X generated in step S2 falls within that range, and if it does, this target trajectory can be used as the optimal target trajectory and the processing in step S4 can be omitted. The target trajectory generation unit 11a stores predetermined conditions for determining whether such omission is possible, and if it is determined that the target trajectory generated in step S2 satisfies the predetermined conditions, the processing in step S4 is omitted and the processing to the next step is performed.

[0051] The target trajectory generation unit 11a can obtain the optimal target trajectory of the functional unit X (estimated trajectory of the functional unit X) for the input target trajectory through the optimization process of the functional unit X in step S4. Similarly, the target trajectory generation unit 11a can obtain the estimated trajectory of the control system Y (optimal target trajectory of the control system Y) for the input target trajectory through the estimation process of the control system Y in step S5. Based on the optimal target trajectory of the functional unit X obtained as a result of step S4 and the estimated trajectory of the control system Y obtained as a result of step S5, the target trajectory generation unit 11a performs a simulation of substrate processing using a model of the overall function Ω (step S6). At this time, the target trajectory generation unit 11a may also use sensor measurement values ​​obtained from the substrate processing device 3 in the simulation.

[0052] The target trajectory generation unit 11a can obtain an estimated trajectory for the target trajectory input to the overall function Ω as a result of the simulation in step S6. Based on the results of this simulation, the target trajectory generation unit 11a calculates a predetermined objective function for the overall function Ω (step S7). For example, based on whether the value of the objective function calculated in step S7 exceeds a threshold, the target trajectory generation unit 11a determines whether the target trajectories of the functional unit X and the control system Y at that time are optimal (step S8).

[0053] If the target trajectory is determined to be not optimal (S8: NO), the target trajectory generation unit 11a updates the target trajectories of the functional unit X and control system Y at that point in time, for example, so that the value of the objective function becomes smaller (or larger) (step S9), and returns to step S3. The target trajectory generation unit 11a repeats the processes of steps S3 to S9 until the target trajectory becomes optimal.

[0054] The target trajectory generation unit 11a can update the target trajectory using methods such as gradient methods including the steepest descent method or projection gradient method, Bayesian optimization such as CMA-ES (Covariance Matrix Adaptation Evolution Strategy), genetic algorithms, or neural networks. Furthermore, the hyperparameters required for calculations using these methods may be optimized using a separate system from this system, such as through meta-learning.

[0055] If the target trajectory generation unit 11a determines that the target trajectory is optimal (S8: YES), it outputs the target trajectory of the functional unit X and the control system Y at that point in time, and the estimated trajectory of the overall function Ω calculated based on this target trajectory, as the optimal target trajectory for one run (step S10), and terminates the process.

[0056] Figure 6 shows the details of the optimization process performed in functional unit X at step S4 of the flowchart shown in Figure 5. In the optimization process for functional unit X, the target trajectory generation unit 11a obtains the target trajectory of functional unit X given by the optimization process of the overall function Ω shown in Figure 5 (step S21). Next, the target trajectory generation unit 11a generates the target trajectories (initial values) of control systems Z and W connected to the functional unit X based on the target trajectory obtained in step S21 (step S22). The target trajectory generation unit 11a can generate the initial values ​​of the target trajectories in step S22 in the same way as described in step S2.

[0057] In this embodiment, the target trajectory provided from the overall function Ω to the functional unit X is for one run. However, the target trajectory generation unit 11a divides this one run period into multiple steps and generates a target trajectory for each step. Figure 8 shows the target trajectory periods handled by each layer: the overall function Ω, the functional unit X, the control system Z, the control system W, and the control system Y. The overall function Ω is provided with a target trajectory for a period of one run. In this embodiment, the overall function Ω is not divided into periods, and the functional unit X and the control system Y, which are lower levels of the overall function Ω, are provided with a target trajectory for a period of one run. In this embodiment, the functional unit X divides the one run period into three steps, and the control system Z and the control system W, which are lower levels of the functional unit X, are provided with a target trajectory for a period of one step. However, the period division shown in Figure 8 is just an example and is not limited to this. Whether or not to divide the target trajectory and how many parts to divide it into may be appropriately determined depending on the configuration of the substrate processing device 3, etc. In this embodiment, the target trajectory generation unit 11a generates target trajectories for control systems Z and W for 3 steps based on the target trajectory of functional unit X for 1 run in step S22.

[0058] The target trajectory generation unit 11a adapts the target trajectories of control system Z and control system W for one step, generated in step S22, to predetermined constraints for control system Z and control system W, respectively (step S23). Next, the target trajectory generation unit 11a performs estimation processing of the target trajectory in control system Z based on the target trajectory of control system Z for one step adapted to the constraints in step S23 (step S24). The target trajectory generation unit 11a also performs estimation processing of the target trajectory in control system W based on the target trajectory of control system W for one step adapted to the constraints in step S23 (step S25). The target trajectory generation unit 11a may perform the estimation processing for control system Z in step S24 and the estimation processing for control system W in step S25 in any order, or it may perform the two processes simultaneously. Details of the estimation processing for control system Z are shown in Figure 7, and details of the estimation processing for control system W are the same as those for control system Z shown in Figure 7, so they are not shown.

[0059] The target trajectory generation unit 11a can obtain the estimated trajectory of the control system Z (the optimal target trajectory of the control system Z) for the input target trajectory through the estimation process in the control system Z in step S24. Similarly, the target trajectory generation unit 11a can obtain the estimated trajectory of the control system W (the optimal target trajectory of the control system W) for the input target trajectory through the estimation process in the control system W in step S25. Based on the estimated trajectory of the control system Z obtained as a result of step S24 and the estimated trajectory of the control system Y obtained as a result of step S25, the target trajectory generation unit 11a performs a simulation of substrate processing using a model of the functional unit X (step S26).

[0060] The target trajectory generation unit 11a can obtain an estimated trajectory for the target trajectory input to the functional unit X as a result of the simulation in step S26. Based on the results of this simulation, the target trajectory generation unit 11a calculates a predetermined objective function for the functional unit X (step S27). For example, based on whether the value of the objective function calculated in step S27 exceeds a threshold, the target trajectory generation unit 11a determines whether the target trajectories of the control system Z and the control system W at that time are optimal (step S28).

[0061] If the target trajectory is determined to be not optimal (S28: NO), the target trajectory generation unit 11a updates the target trajectories of control system Z and control system W at that point in time, for example, so that the value of the objective function becomes smaller (or larger) (step S29), and returns to step S23. The target trajectory generation unit 11a repeats the process from steps S23 to S29 until the target trajectory becomes optimal.

[0062] If the target trajectory generation unit 11a determines that the target trajectory is optimal (S28: YES), it determines whether the optimization process for one run has been completed (step S30). If the optimization process for one run has not been completed (S30: NO), the target trajectory generation unit 11a moves the target of the process to the next step (step S31) and returns to step S23. The target trajectory generation unit 11a repeats the process from steps S23 to S31 until the optimization process for one run, i.e., the optimization process for three steps, is completed.

[0063] When the optimization process for one run is completed (S30:YES), the target trajectory generation unit 11a outputs the target trajectories of control system Z and control system W for that run at that point, and the estimated trajectory of functional unit X calculated based on these target trajectories, as the optimal target trajectory for that run (step S32), and then terminates the optimization process for functional unit X and returns to the optimization process for the overall function Ω.

[0064] Figure 7 shows the details of the estimation process in the control system Z, which takes place in step S24 of the flowchart shown in Figure 6. In the estimation process in the control system Z, the target trajectory generation unit 11a obtains the target trajectory of the control system Z, which is given by the optimization process in the functional unit X shown in Figure 6 (step S41). Next, the target trajectory generation unit 11a adapts the target trajectory of the control system Z obtained in step S41 to the constraints predetermined for the control system Z (step S42).

[0065] Next, the target trajectory generation unit 11a performs a simulation of substrate processing using a model of the control system Z, based on the target trajectory of the control system Z that was adapted to the constraints in step S42 (step S43). As a result of the simulation in step S43, the target trajectory generation unit 11a can obtain an estimated trajectory for the target trajectory input to the control system Z. The target trajectory generation unit 11a outputs the estimated trajectory of the control system Z obtained as a result of the simulation (step S44), terminates the estimation process in the control system Z, and returns the process to the optimization process in the functional unit X.

[0066] As described above, the information processing device 1 according to this embodiment can generate the optimal target trajectory at each level included in the substrate processing model 5 by recursively repeating the optimization process from the overall function Ω, which is the highest level of the substrate processing model 5, down to the lower levels. In this embodiment, the substrate processing model 5 has a tree-like hierarchical structure as shown in Figure 3. When generating a target trajectory by optimization processing using the substrate processing model 5 with such a hierarchical structure, the processing shown in the flowchart of Figure 5 is performed at the level corresponding to the root node of the tree structure, the processing shown in the flowchart of Figure 7 is performed at the level corresponding to the leaf node, and the processing shown in the flowchart of Figure 6 is performed at the other intermediate levels.

[0067] The control processing unit 11b of the information processing device 1 according to this embodiment provides the substrate processing device 3 with the target trajectory generated by the target trajectory generation unit 11a through optimization processing using the substrate processing model 5, thereby causing the substrate processing device 3 to perform the desired substrate processing. As described above, the target trajectory generation unit 11a can generate target trajectories corresponding to each layer of the model included in the substrate processing model 5, and the control processing unit 11b may provide all of these target trajectories to the substrate processing device 3, or may appropriately select and provide the necessary target trajectories to the substrate processing device 3.

[0068] For example, the information processing device 1 may provide the target trajectory of the functional unit X to the substrate processing device 3. In this case, the substrate processing device 3 generates target trajectories for control system Z and control system W from the given target trajectory of functional unit X, and operates control system Z and control system W by inputting the generated target trajectories to control system Z and control system W, respectively. Alternatively, the information processing device 1 may provide the target trajectories of control system Z and control system W to the substrate processing device 3. In this case, the substrate processing device 3 operates control system Z and control system W by inputting the target trajectories provided by the information processing device 1 to control system Z and control system W, respectively. Furthermore, in this embodiment, the information processing device 1 provides the substrate processing device 3 with target trajectories for each 1-run period, but this is not limited to this, and for example, target trajectories for each 1-step period may be provided.

[0069] In this embodiment, the information processing device 1 generates the optimal target trajectory by repeatedly updating the initial value of the target trajectory through optimization processing, but it is not limited to this. For example, the information processing device 1 may generate the optimal target trajectory by using the inverse model of one or more models included in the substrate processing model 5. The inverse model is a model that accepts data such as the value of the objective function, the target trajectory of other related functional units, or operating status as input, and outputs a target trajectory that achieves the value of the objective function. The inverse model is generated in advance based on information obtained from, for example, the substrate processing device 3, and stored in the storage unit 12 of the information processing device 1. The information processing device 1 may also perform optimization processing using the target trajectory output by the inverse model as the initial value of the iterative calculation described above.

[0070] Furthermore, when the information processing device 1 calculates the objective function at each level or performs simulations using the model, it may take into account constraints appropriately input by the user or another system. For example, the information processing device 1 may interrupt the iterative calculation when it is determined that the input constraints are not met, update the target trajectory, and re-execute the optimization process.

[0071] Furthermore, the information processing device 1 may update the target trajectory using a function that includes adjustable parameters, such as a neural network, which accepts inputs such as the previous target trajectory, the value of the objective function obtained from this target trajectory, and data indicating the target trajectory or operating status of other related functional units, and outputs a new target trajectory. The information processing device 1 may configure the adjustable parameters of this function by pre-learning them to optimize the rate at which the value of the objective function decreases through optimization processing. Similarly, objective functions other than the overall objective function, constraint fitting functions, or simulation models may include adjustable parameters, and these adjustable parameters may be configured by pre-learning them to optimize the rate at which the objective function decreases through optimization.

[0072] Furthermore, the information processing device 1 may, for example, calculate time-series data of control variables or power consumption in the processing of the control system shown in the flowchart of Figure 7, and use the calculation results to perform simulation or optimization calculations. For example, it may perform calculations such as adding to the overall optimization objective function the result obtained by applying a monotonically increasing function to the power consumed by each control system.

[0073] Furthermore, for control systems Y, Z, and W corresponding to leaf nodes, if the models of these control systems use only the values ​​specified in the target trajectory, and if the difference between the target trajectory and the estimated trajectory of the control result can be ignored as a result of fitting the constraints, then the information processing device 1 can omit processes such as simulation in the optimization process shown in the flowcharts of Figures 5 to 7. That is, for example, for control system Z, the information processing device 1 can omit the process related to control system Z shown in Figure 7 by assuming that the target trajectory (initial value) and the estimated trajectory are the same.

[0074] Furthermore, the information processing device 1 can adopt conditions for terminating repetitions in the optimization processes of each level, such as when the value of the objective function falls below a threshold, when the value of the objective function falls below a threshold that depends on the number of repetitions, or when the number of repetitions exceeds a threshold. In addition, when the information processing device 1 performs multiple optimization processes in parallel, it may terminate the repetitions of other optimization processes when the repetition of one optimization process is completed, or extend the repetitions of other optimization processes until the repetition of one optimization process is completed.

[0075] Furthermore, the information processing device 1 may calculate a predicted error for the target trajectory calculated as optimal in the optimization process at each level. For example, the information processing device 1 may use ±1 standard deviation of the calculation result as the error. The information processing device 1 can propagate the calculated error from lower levels to higher levels and perform processing according to the error, such as repeating the optimization process if the final error exceeds an acceptable value.

[0076] Furthermore, the information processing device 1 may transmit and store information such as the results of optimization processing or the results of simulations using the board processing model 5 to a server device, and the server device may use the stored information to perform processing such as updating the board processing model 5. The server device may also acquire and store information about multiple board processing devices 3 from the corresponding information processing device 1, and use the stored information to generate an average board processing model 5 for the multiple board processing devices 3 and transmit it to each information processing device 1. The information processing device 1 may acquire the average board processing model 5 from the server device and, based on the acquired average board processing model 5, generate a board processing model 5 suitable for the board processing device 3 under its control.

[0077] Furthermore, in the optimization process at each level, the information processing device 1 may limit the search for the optimal target trajectory to the vicinity of the given initial value of the target trajectory. In this case, for example, a server device may optimize the initial value of the target trajectory with respect to the standard settings of the board processing and deliver the optimal initial value of the target trajectory to the information processing device 1. If the settings in each board processing device 3 deviate from the standard, the information processing device 1 may use perturbation theory related to the deviation and generate the optimal target trajectory by solving a lower-order optimization problem. For example, if the target trajectory of the heater output when the temperature of the device is 25°C is X1, the information processing device 1 may set the target trajectory of the heater output when the temperature of the device is 27°C as X = X1 + δX, linearize the optimization problem by ignoring terms of second order or higher related to δX and the temperature difference δT = 2°C, and calculate δX.

[0078] <Summary> The information processing system according to this embodiment, configured as described above, has an information processing device 1 that has a substrate processing model 5 relating to substrate processing by a substrate processing device 3, and uses this substrate processing model 5 to generate a target trajectory for substrate processing by the substrate processing device 3. The substrate processing model 5 has a hierarchical structure in which one or more lower-level models are connected to a higher-level model. Each model included in the substrate processing model 5 receives a target trajectory as input and outputs an estimated trajectory estimated for this target trajectory.

[0079] For example, if the higher-level model is a functional unit X and the lower-level models are control systems Z and W, the information processing device 1 obtains the target trajectory of functional unit X, generates the target trajectories (initial values) of control systems Z and W based on the obtained target trajectory, and provides the generated target trajectories to control systems Z and W respectively. The information processing device 1 obtains the estimated trajectories (lower estimated trajectories) output by control systems Z and W, performs a simulation using the model of functional unit X based on the obtained estimated trajectories to calculate the estimated trajectory (higher estimated trajectory) of functional unit X, calculates the value of the objective function of functional unit X based on the calculated estimated trajectory, and determines whether the calculated value satisfies predetermined conditions. If the calculated value does not satisfy the predetermined conditions, the information processing device 1 updates the target trajectories of control systems Z and W and recalculates the estimated trajectory of functional unit X. If the calculated value satisfies the predetermined conditions, the information processing device 1 outputs the target trajectory at that time as the optimal target trajectory. As a result, the information processing system according to this embodiment can control the operation of the substrate processing apparatus 3 using the generated target trajectory, and it is expected that control of substrate processing using the substrate processing model 5 can be realized.

[0080] Furthermore, in the information processing system according to this embodiment, the information processing device 1 acquires the target trajectory for a first predetermined time (1 run) of the higher-level model (functional unit X), generates the target trajectory for the lower-level models (control system Z and control system W) for a second predetermined time (1 step) obtained by dividing 1 run into multiple parts, and performs the processing for each step in multiple steps. As a result, the information processing system according to this embodiment can have different granularities for the optimization processing for multiple layers of models included in the board processing model 5, and is expected to improve the efficiency of the optimization processing calculations compared to the case where the optimization processing for all layers is performed in the same time unit.

[0081] Furthermore, in the information processing system according to this embodiment, the board processing model 5 has a tree structure in which multiple lower-level models are connected to a higher-level model. The model corresponding to the leaf node of the tree structure is a model that reproduces the behavior of the hardware unit of the board processing device 3. The information processing device 1 controls the board processing by the board processing device 3 by providing the board processing device 3 with the target trajectory generated for the model corresponding to the root node, the target trajectory generated for the model corresponding to the leaf node, and / or the target trajectory generated for the model corresponding to nodes other than the root node and leaf node. Thus, the information processing system according to this embodiment is expected to realize control of board processing using the tree-structured board processing model 5.

[0082] However, the substrate processing model 5 is not limited to a tree structure. The substrate processing model 5 may include a structure in which, for example, one lower model is connected to two higher models. For example, if the temperature control function of the substrate processing apparatus 3 includes a heater and a cooler, the model of the temperature control function may be configured such that the heater target trajectory is given by the first higher model and the cooler target trajectory is given by the second higher model. In this case, for example, the first higher model may be configured to be given the cooler target trajectory by the second higher model and generate the heater target trajectory based on this cooler target trajectory.

[0083] Furthermore, in the information processing system according to this embodiment, the information processing device 1 adapts the acquired target trajectory of the higher hierarchy to predetermined constraints for the higher hierarchy model, and generates a target trajectory for the lower hierarchy model based on the target trajectory adapted to the constraints. As a result, the information processing system according to this embodiment is expected to generate a target trajectory that takes into account, for example, hardware constraints of the substrate processing device 3.

[0084] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of this disclosure is indicated by the claims, not in the sense described above, and all modifications within the sense and scope equivalent to the claims are intended.

[0085] The matters described in each embodiment can be combined with each other. Furthermore, the independent and dependent claims described in the claims can be combined with each other in any combination, regardless of the form of reference. In addition, the claims use a form in which claims referencing two or more other claims (multi-claim form), but are not limited to this. A form in which multi-claims referencing at least one multi-claim (multi-multi-claim) may also be used. [Explanation of Symbols]

[0086] 1. Information processing equipment (computer) 3. Substrate Processing Equipment 5. PCB Processing Models 5a Frequency Optimization Calculation Model 5b Frequency Controller Model 5c Variable Capacitor Controller Model 5d Exhaust Valve Model 5e Intake Valve Model 5f frequency optimization model 5g gas pressure control model 5h High-Frequency Power Optimization Model 5i Overall Optimization Model 11 Processing Section 11a Target trajectory generation section 11b Control Processing Unit 11c Display Processing Unit 12 Storage section 12a Program (Computer Program) 12b Model Information Storage Unit 13 Communications Department 14 Display section 15 Control section 99 Recording media

Claims

1. An information processing method comprising an information processing device that generates a target trajectory for substrate processing performed by a substrate processing device using a substrate processing model relating to substrate processing performed by the substrate processing device, The aforementioned substrate processing model includes a higher-level model and a lower-level model connected to the higher-level model. The aforementioned higher-level model and lower-level model are models that receive a target trajectory, which is the target of the output, and output an estimated trajectory estimated for that target trajectory. The aforementioned information processing device The higher-level target trajectory, which is the output target of the aforementioned higher-level model, is obtained. Based on the acquired higher-level target trajectory, a lower-level target trajectory, which is the target output of the lower-level model, is generated. The generated lower target trajectory is given to the lower model, and the lower estimated trajectory output by the lower model is obtained. Based on the acquired lower-level estimated trajectory, a simulation is performed to calculate the higher-level estimated trajectory related to the output value of the higher-level model. If the calculated upper-level estimated trajectory does not satisfy the predetermined conditions, the lower-level target trajectory of the lower-level model is updated and the upper-level estimated trajectory is recalculated. The lower target trajectory, when the calculated upper-level estimated trajectory satisfies predetermined conditions, is output as the target trajectory of the lower-level model. Information processing methods.

2. The aforementioned information processing device The higher target trajectory is obtained during the first predetermined time, Multiple lower target trajectories are generated in the second predetermined time, which is obtained by dividing the first predetermined time. The calculation of the aforementioned higher-level estimated trajectory is performed in multiple steps. The information processing method according to claim 1.

3. The aforementioned information processing device Based on the calculated higher-level estimated trajectory, the value of the objective function defined for the higher-level model is calculated. Determine whether the calculated value satisfies the predetermined conditions. The information processing method according to claim 1 or claim 2.

4. The substrate processing model has a tree structure in which multiple lower-level models are connected to the higher-level model. The information processing method according to claim 1 or claim 2.

5. The model corresponding to the leaf node in the tree structure of the substrate processing model is a model that reproduces the behavior of the hardware unit of the substrate processing apparatus. The information processing method according to claim 4.

6. The aforementioned information processing device Based on the target trajectory generated for the model corresponding to the root node of the tree structure of the substrate processing model, the target trajectory generated for the model corresponding to the leaf node, or the target trajectory generated for the model corresponding to nodes other than the root node and the leaf node, the substrate processing is controlled. The information processing method according to claim 4.

7. The aforementioned information processing device The acquired higher-level target trajectory is modified to a trajectory that conforms to predetermined constraints for the higher-level model. The corrected trajectory is used as the higher-level target trajectory to generate the lower-level target trajectory. The information processing method according to claim 1 or claim 2.

8. The aforementioned information processing device Determine whether the generated lower target trajectory satisfies predetermined conditions. If the above predetermined conditions are met, the generated lower target trajectory shall be the lower estimated trajectory. The information processing method according to claim 1 or claim 2.

9. The aforementioned information processing device The target trajectory of the lower-level model that was output is stored, When generating the aforementioned sub-target trajectory, the target trajectory of the stored sub-model is used as the sub-target trajectory. The information processing method according to claim 1 or claim 2.

10. The aforementioned substrate processing model has a hierarchical structure of three or more layers in which multiple models are connected. The aforementioned information processing device For each pair of models that have a connection relationship between the higher-level model and the lower-level model included in the substrate processing model, the process for generating the target trajectory is performed. The information processing method according to claim 1 or claim 2.

11. A computer program that causes a computer to execute a process to generate a target trajectory for substrate processing by a substrate processing apparatus, using a substrate processing model relating to substrate processing performed by the substrate processing apparatus, The aforementioned substrate processing model includes a higher-level model and a lower-level model connected to the higher-level model. The aforementioned higher-level model and lower-level model are models that receive a target trajectory, which is the target of the output, and output an estimated trajectory estimated for that target trajectory. To the aforementioned computer, The higher-level target trajectory, which is the output target of the aforementioned higher-level model, is obtained. Based on the acquired higher-level target trajectory, a lower-level target trajectory, which is the target output of the lower-level model, is generated. The generated lower target trajectory is given to the lower model, and the lower estimated trajectory output by the lower model is obtained. Based on the acquired lower-level estimated trajectory, a simulation is performed to calculate the higher-level estimated trajectory related to the output value of the higher-level model. If the calculated upper-level estimated trajectory does not satisfy the predetermined conditions, the lower-level target trajectory of the lower-level model is updated and the upper-level estimated trajectory is recalculated. If the calculated upper-level estimated trajectory satisfies predetermined conditions, the lower-level target trajectory is output as the target trajectory of the lower-level model. A computer program that executes a process.

12. An information processing apparatus comprising a substrate processing model relating to substrate processing performed by a substrate processing apparatus, and a processing unit that generates a target trajectory for substrate processing by the substrate processing apparatus using the substrate processing model, The aforementioned substrate processing model includes a higher-level model and a lower-level model connected to the higher-level model. The aforementioned higher-level model and lower-level model are models that receive a target trajectory, which is the target of the output, and output an estimated trajectory estimated for that target trajectory. The aforementioned processing unit, The higher-level target trajectory, which is the output target of the aforementioned higher-level model, is obtained. Based on the acquired higher-level target trajectory, a lower-level target trajectory, which is the target output of the lower-level model, is generated. The generated lower target trajectory is given to the lower model, and the lower estimated trajectory output by the lower model is obtained. Based on the acquired lower-level estimated trajectory, a simulation is performed to calculate the higher-level estimated trajectory related to the output value of the higher-level model. If the calculated upper-level estimated trajectory does not satisfy the predetermined conditions, the lower-level target trajectory of the lower-level model is updated and the upper-level estimated trajectory is recalculated. The lower target trajectory, when the calculated upper-level estimated trajectory satisfies predetermined conditions, is output as the target trajectory of the lower-level model. Information processing device.

Images