Substrate processing device, information processing method, and computer program

Abstract

Provided are a substrate processing device, an information processing method, and a computer program that achieve substrate processing taking into account the characteristics of a unit having a learning function.　The substrate processing device includes a command value generating unit that generates a command value to be input to a target unit related to substrate processing, wherein the target unit has a function of adjusting a response characteristic thereof on the basis of the input command value and its own output value corresponding to the command value, and the command value generating unit: stores, in a storage unit, a first model of the target unit generated on the basis of data of the command value generated assuming that the response characteristic of the target unit is a predetermined characteristic and the output value output by the target unit in response to the command value; stores, in the storage unit, a second model of the target unit generated on the basis of data of the command value generated on the assumption that the response characteristic of the target unit is the characteristic of the first model and the output value output by the target unit in response to the command value; and generates the command value for the target unit on the basis of the first model and the second model.

Description

Substrate Processing Apparatus, Information Processing Method, and Computer Program

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

[0002] In Patent Document 1, there is proposed an exposure apparatus that controls the position of a stage on which a substrate is mounted using a position control apparatus including a detection unit that detects the position of a control target, a subtraction unit that subtracts the output of the detection unit from a target value, and a filter to which the deviation between the output of the detection unit and the target value is input, and that performs feedforward of a control input to the control target, and a device manufacturing method of exposing a pattern on a substrate and developing the exposed substrate.

[0003] Japanese Unexamined Patent Application Publication No. 2009 - 205641

[0004] The present disclosure provides a substrate processing apparatus, an information processing method, and a computer program that can be expected to realize substrate processing in consideration of the characteristics of a unit having a learning function such as ILC (Iterative Learning Control).

[0005] A substrate processing apparatus according to an embodiment is a substrate processing apparatus including a command value generation unit that generates a command value to be input to a target unit related to substrate processing, the target unit having a function of adjusting a response characteristic based on an input command value and its own output value corresponding to the command value, the command value generation unit storing, in a storage unit, a first model of the target unit generated based on data of a command value generated assuming that the response characteristic of the target unit is a predetermined characteristic and an output value output by the target unit in response to the command value, storing, in the storage unit, a second model of the target unit generated based on data of a command value generated assuming that the response characteristic of the target unit is the characteristic of the first model and an output value output by the target unit in response to the command value, and generating a command value to the target unit based on the first model and the second model.

[0006] According to the present disclosure, it can be expected to realize substrate processing in consideration of the characteristics of a unit having a learning function such as ILC.

[0007] This is a schematic diagram illustrating the outline of the substrate processing apparatus according to this embodiment. This is a schematic diagram illustrating an example of control of a lower-level control unit by a higher-level control unit. This is a block diagram illustrating an example configuration of the substrate processing apparatus according to this embodiment. This is a flowchart illustrating an example of the procedure for model generation processing by the higher-level control unit of the substrate processing apparatus according to this embodiment. This is a schematic diagram illustrating the method for generating the first model. This is a schematic diagram illustrating the method for generating the second model. This is a schematic diagram illustrating the method for generating command values ​​by the higher-level control unit using the first and second models. This is a schematic diagram showing an example configuration of a substrate processing apparatus according to a modified example. This is a schematic diagram illustrating the correspondence between command values ​​and output values ​​in each layer of the substrate processing apparatus according to a modified example.

[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 changes within the meaning and scope of the claims as indicated by the claims.

[0009] <System Overview> Figure 1 is a schematic diagram illustrating the overview of the substrate processing apparatus according to this embodiment. The substrate processing apparatus 1 according to this embodiment is an apparatus that performs various substrate processing on semiconductor substrates (wafers), such as CVD (Chemical Vapor Deposition), sputtering, or etching. The substrate processing apparatus 1 is composed of a combination of multiple units. In this embodiment, a unit is a constituent unit, component, or a set of devices realized by dividing hardware resources that are responsible for multiple functions necessary for the substrate processing apparatus 1, such as pressure control, power control, temperature control, and gas flow rate control, into a manner that allows for independent replacement or substitution for each function. A unit may also be called by names such as module, component, or assembly.

[0010] The substrate processing apparatus 1 according to this embodiment comprises a plurality of lower-level control units 5, such as a pressure control unit 5a and a power control unit 5b, and a higher-level control unit 3 that controls these plurality of lower-level control units 5. The substrate processing apparatus 1 also has at least one substrate processing chamber for processing one or more substrates, and the pressure and power supplied within the substrate processing chamber during substrate processing are controlled by these lower-level control units 5. Each of the plurality of lower-level control units 5 is connected to the higher-level control unit 3 in a communication manner and operates according to command values ​​(pressure command values, power command values, etc.) given from the higher-level control unit 3. In this embodiment, the substrate processing apparatus 1 has a two-tiered configuration of a higher-level control unit 3 and lower-level control units 5, but is not limited to this, and may have a three-tiered configuration of, for example, a higher-level control unit, an intermediate control unit and lower-level control units, or a configuration of four or more tiers.

[0011] One of the lower-level control units 5 provided in the substrate processing apparatus 1 is the pressure control unit 5a, which controls the pressure inside the chamber provided in the substrate processing apparatus 1. The power control unit 5b is a unit that controls the supply of power necessary for the operation of each unit provided in the substrate processing apparatus 1. Note that these lower-level control units 5 shown in Figure 1 are just examples and are not limited to these; various other units not shown may be provided in the substrate processing apparatus 1.

[0012] The higher-level control unit 3 is a unit that controls the operation of multiple lower-level control units 5 provided by the substrate processing apparatus 1 based on information (recipes) such as procedures or conditions for substrate processing that have been prepared in advance by the user. The higher-level control unit 3 determines the target value of the output of each lower-level control unit 5 in order to realize the substrate processing according to the given recipe, and provides the determined target value as a command value to the lower-level control unit. In this embodiment, the higher-level control unit 3 provides the "trajectory" of the target value of the output of the lower-level control unit 5 as a command value. The "trajectory" is information that shows the time-series change of values ​​such as input values ​​to the unit or output values ​​of the unit, and can be treated as data in which multiple values ​​are arranged in chronological order, for example. Each value included in the trajectory data may be a difference value from the value at the previous point in time.

[0013] Figure 2 is a schematic diagram illustrating an example of control of the lower control unit 5 by the upper control unit 3. The upper control unit 3 of the substrate processing apparatus 1 according to this embodiment generates command values ​​for controlling the operation of the lower control unit 5 by performing optimization processing based on, for example, a predetermined evaluation function and a model that reproduces the operation of the lower control unit 5 that has been generated in advance. The lower control unit 5 controls pressure or power, etc., using the command values ​​given by the upper control unit 3 as target values, and outputs the result of measuring the control result pressure or power, etc., with a sensor, etc., as an output value.

[0014] The lower part of Figure 2 shows an example of the correspondence between the command value for the lower control unit 5 and the output value output by the lower control unit 5, with the command value shown as a solid line and the output value as a dashed line. In this example, the command value requests that the output value of the lower control unit 5 be raised sharply and then lowered sharply, but the output value of the lower control unit 5 is unable to follow this, and the command value and the output value are different.

[0015] Each lower-level control unit 5 in the substrate processing apparatus 1 according to this embodiment has a function to adjust its response characteristics so that the output value approaches (follows) the given command value, based on the command value given by the upper-level control unit 3 and its own output value corresponding to this command value. One example of the function of this lower-level control unit 5 is a learning function called iterative learning control (ILC). For example, the lower-level control unit 5 can periodically acquire the command value and the output value and adjust its response characteristics by updating internal control parameters, etc., so that the difference between the command value and the output value becomes smaller.

[0016] Furthermore, the higher-level control unit 3 of the substrate processing apparatus 1 according to this embodiment generates command values ​​that take into account the changes in response characteristics due to the learning function described above and provides them to the lower-level control unit 5. To this end, the higher-level control unit 3 first assumes that the response characteristics of the lower-level control unit 5 are ideal (outputting a value equal to the command value) and generates command values ​​by performing optimization processing based on recipes, etc. The higher-level control unit 3 provides the generated command values ​​to the lower-level control unit 5 to obtain output values ​​and generates a first model showing the response characteristics of the lower-level control unit 5 based on the input and output data to the lower-level control unit 5.

[0017] Next, the higher-level control unit 3 performs optimization processing assuming that the response characteristics of the lower-level control unit 5 are not ideal, but rather as shown by the first model it generates, and generates command values. The higher-level control unit 3 provides the generated command values ​​to the lower-level control unit 5 to obtain output values, and based on the input and output data to the lower-level control unit 5, it generates a second model that shows the response characteristics of the lower-level control unit 5. This second model is a model that shows the change in response characteristics due to the learning function of the lower-level control unit 5.

[0018] Thereafter, the higher-level control unit 3 generates command values ​​for the lower-level control unit 5 using the generated first and second models. For example, the higher-level control unit 3 generates command values ​​by performing optimization processing based on the first model, and then converts the generated command values ​​using the inverse model of the second model to obtain the final command values ​​for the lower-level control unit 5. As a result, in the substrate processing apparatus 1 according to this embodiment, it is expected that the command values ​​given by the higher-level control unit 3 to the lower-level control unit 5 will match the output values ​​output by the lower-level control unit 5 in response to these command values.

[0019] <Device Configuration> Figure 3 is a block diagram showing an example configuration of the substrate processing apparatus 1 according to this embodiment. Although the substrate processing apparatus 1 is equipped with multiple lower-level control units 5, Figure 2 shows a simplified configuration in which only one lower-level control unit 5 is illustrated. The substrate processing apparatus 1 according to this embodiment is equipped with a higher-level control unit 3 and multiple lower-level control units 5, and the higher-level control unit 3 and each lower-level control unit 5 are connected by communication lines or the like to exchange data with each other. The substrate processing apparatus 1 may be equipped with multiple higher-level control units 3.

[0020] The upper-level control unit 3 is configured with a processing unit 31, a storage unit 32, a communication unit 33, and the like. The processing unit 31 is configured using a computing device such as a CPU (Central Processing Unit), MPU (Micro-Processing Unit), GPU (Graphics Processing Unit), or quantum processor, as well as ROM (Read Only Memory) and RAM (Random Access Memory). The processing unit 31 reads and executes a program 32a stored in the storage unit 32, performing various processes such as generating command values ​​to control the operation of the lower-level control unit 5, and generating a model necessary for generating these command values.

[0021] The storage unit 32 is configured using a storage device such as a hard disk or an SSD (Solid State Drive). The storage unit 32 stores various programs executed by the processing unit 31, and various data necessary for the processing of the processing unit 31. In this embodiment, the storage unit 32 stores the program 32a executed by the processing unit 31. The storage unit 32 is also provided with a model DB (database) 32b that stores information related to the model of the lower-level control unit 5 provided in the substrate processing device 1.

[0022] In this embodiment, the program (computer program, program product) 32a is provided in a form recorded on a recording medium 99 such as a memory card or optical disc, and the substrate processing device 1 reads the program 32a from the recording medium 99 and stores it in the storage unit 32. However, the program 32a may also be written to the storage unit 32 during the manufacturing stage of the substrate processing device 1 or the higher-level control unit 3, for example. Alternatively, the program 32a may be obtained by the substrate processing device 1 via communication from a remote server device or the like that distributes it. For example, the program 32a may be read from the recording medium 99 by a writing device and written to the storage unit 32 of the higher-level control unit 3. The program 32a may be provided in a form distributed via a network, or it may be provided in a form recorded on the recording medium 99.

[0023] Model DB32b is a database that stores information about the first and second models generated for each lower-level control unit 5. The first and second models are generated based on input and output data to the lower-level control unit 5 and represent the response characteristics between the input and output. The first and second models can be of various configurations, such as a physical model represented by a predetermined mathematical formula or a machine learning model such as a neural network. The information about the model stored in Model DB32b may include information about the structure of the model (such as the mathematical formula or how multiple neurons are connected) and information about the model's parameters (such as coefficients in the mathematical formula or weight parameters of a neural network).

[0024] The communication unit 33 is connected to the lower control unit 5, for example, via a communication line or a control signal line, and transmits and receives various data with the lower control unit 5. In this embodiment, it is assumed that the upper control unit 3 and the lower control unit 5 communicate on a one-to-one basis, but it is not limited to this, and the upper control unit 3 and multiple lower control units 5 may be connected in a network configuration such as a bus type or a ring type. The communication unit 33 transmits command values ​​generated by the processing unit 31 to the lower control unit 5, and also receives data such as output values ​​(for example, observed values ​​by sensors) transmitted by the lower control unit 5 as a response to the control, and provides it to the processing unit 31.

[0025] Furthermore, in the higher-level control unit 3 of the substrate processing apparatus 1 according to this embodiment, the processing unit 31 reads and executes the program 32a stored in the memory unit 32, thereby realizing the model generation unit 31a and the command value generation unit 31b, etc., as software-based functional units in the processing unit 31. In this figure, the functional units of the processing unit 31 are shown as functional units related to the processing for controlling the lower-level control unit 5, and functional units related to other processing are omitted from the illustration.

[0026] The model generation unit 31a performs the process of generating a first model and a second model of the subordinate control unit 5 to be controlled. The model generation unit 31a provides the command value obtained by performing optimization processing assuming the response characteristics are ideal to the target subordinate control unit 5 and obtains the output value of the subordinate control unit 5 corresponding to this command value. Based on the correspondence between the command value given to the subordinate control unit 5 and the output value obtained from the subordinate control unit 5, the model generation unit 31a generates a first model that reproduces this correspondence. At this time, the model generation unit 31a can generate the first model from the input and output to the subordinate control unit 5 using a general-purpose system identification algorithm. For example, by using the tfest function of MATLAB®, the continuous-time transfer function can be estimated, and the estimated transfer function can be used as the first model. However, any method may be used by the model generation unit 31a to generate the first model of the subordinate control unit 5 from the input and output data.

[0027] Next, the model generation unit 31a performs an optimization process based on the response characteristics of the generated first model and provides the command value obtained from this optimization process to the target lower-level control unit 5, and obtains the output value of the lower-level control unit 5 corresponding to this command value. Based on the correspondence between the command value given to the lower-level control unit 5 and the output value obtained from the lower-level control unit 5, the model generation unit 31a generates a second model that reproduces this correspondence. The model generation unit 31a can generate the second model in the same way as it generates the first model.

[0028] The model generation unit 31a stores information about the generated first and second models in the model database 32b, along with information such as the identification information of the corresponding lower-level control unit 5 and the date and time the models were generated. The model generation unit 31a may also generate the first and second models at an appropriate frequency, such as once a day or once every 100 wafers processed, and update the information in the model database 32b. In this case, the update frequency of the models may differ between the first and second models.

[0029] The command value generation unit 31b uses the first and second models generated by the model generation unit 31a to generate command values ​​for the lower-level control units 5. The command value generation unit 31b generates command values ​​for each of the multiple lower-level control units 5 based on the substrate processing recipe set by the user or the like. For example, the command value generation unit 31b can generate command values ​​(the trajectory of the target output value) by performing optimization processing on a lower-level control unit 5 whose response characteristics are represented in the first model, based on a predetermined evaluation function. Note that the method of generating command values ​​by performing optimization processing is an existing technology, so a detailed explanation is omitted.

[0030] Furthermore, the command value generation unit 31b generates a final command value by further converting the command value generated using the first model using the second model. In this embodiment, the command value generation unit 31b generates an inverse model of the second model, converts the value using the inverse model, and provides the command value converted by the inverse model to the lower control unit 5.

[0031] The second model is a model that outputs the output value of the lower control unit 5 in response to the input of a command value, and the inverse model of the second model is a model that outputs the command value input to the lower control unit 5 in response to the output value of the lower control unit 5. The command value generation unit 31b may generate the inverse model by, for example, calculating the reciprocal of the coefficients of the calculation formula of the second model, or by, for example, performing an inverse Fourier transform on the calculation formula of the second model, or by any other method. Alternatively, the inverse model may be generated by any other method, for example, the model generation unit 31a, instead of the command value generation unit 31b generating the inverse model of the second model, directly generating the inverse model of the second model from the input and output data of the lower control unit 5 (without generating the second model).

[0032] Alternatively, the command value generation unit 31b may directly generate command values ​​for the lower-level control unit 5 by performing optimization processing using both the first and second models, without generating an inverse model of the second model. In this case, the command value generation unit 31b combines the two models by taking the output of the first model as input to the second model and generates command values ​​by performing optimization processing in the same manner. It is not necessary to perform a conversion using the inverse model of the second model on the generated command values.

[0033] The lower-level control unit 5 of the substrate processing apparatus 1 is configured to include a processing unit 51, a storage unit 52, a communication unit 53, and the like. The processing unit 51 is configured to use an arithmetic processing unit such as a CPU or MPU, ROM, RAM, and the like. The processing unit 51 reads and executes a program 52a stored in the storage unit 52, thereby performing various processes such as control processing according to command values ​​from the higher-level control unit 3, and learning processing through repeated learning control.

[0034] The storage unit 52 is configured using a non-volatile storage device such as flash memory or EEPROM (Electrically Erasable Programmable Read-Only Memory). The storage unit 52 stores various programs executed by the processing unit 51, and various data necessary for the processing of the processing unit 51. In this embodiment, the storage unit 52 stores the program 52a executed by the processing unit 51. The storage unit 52 also stores its own control characteristic information 52b, which is updated by repeated learning control.

[0035] In this embodiment, the program (computer program, program product) 52a is provided in a form recorded on a recording medium 98 such as a memory card or optical disc, and the substrate processing device 1 reads the program 52a from the recording medium 98 and stores it in the storage unit 52. However, the program 52a may also be written to the storage unit 52 during the manufacturing stage of the substrate processing device 1 or the lower control unit 5, for example. Alternatively, the program 52a may be distributed by a remote server device or the like and acquired by the substrate processing device 1 via communication. For example, the program 52a may be read from the recording medium 98 by a writing device and written to the storage unit 52 of the lower control unit 5. The program 52a may be provided by distribution via a network, or it may be provided in a form recorded on the recording medium 98.

[0036] The control characteristic information 52b is information such as various setting values ​​or parameters necessary for each lower-level control unit 5 to control, for example, pressure or power. The lower-level control unit 5 operates based on the setting values ​​or parameters stored in the control characteristic value information 52b. In this embodiment, the lower-level control unit 5 repeatedly updates the control characteristic information 52b by performing repeated learning control or the like in order to output an output value that follows the command value given by the upper-level control unit 3.

[0037] The communication unit 53 is connected to the higher-level control unit 3 via, for example, a communication line or a control signal line, and transmits and receives various data with the higher-level control unit 3. The communication unit 53 receives command value data transmitted by the higher-level control unit 3 and provides it to the processing unit 51, and also transmits data such as observed values ​​from sensors provided by the processing unit 51 to the higher-level control unit 3.

[0038] Furthermore, in the lower-level control unit 5 of the substrate processing apparatus 1 according to this embodiment, the processing unit 51 reads and executes the program 52a stored in the memory unit 52, thereby realizing the control processing unit 51a and the learning processing unit 51b, etc., as software-based functional units in the processing unit 51. In this figure, the functional units of the processing unit 51 are shown as functional units related to control processing and learning processing related to substrate processing, and functional units related to other processing are omitted from the illustration.

[0039] The control processing unit 51a performs control processing related to substrate processing according to the command values ​​provided by the higher-level control unit 3. For example, if the lower-level control unit 5 is a pressure control unit 5a, the control processing unit 51a controls the pressure inside the chamber of the substrate processing apparatus 1 by controlling the operation of a pump or the like using the pressure command value provided as a command value as the target value. The control processing unit 51a may also measure the results of the substrate processing it has controlled using a sensor or the like and transmit the obtained measurement data to the higher-level control unit 3. For example, in the case of a pressure control unit 5a, the control processing unit 51a may measure the pressure inside the chamber, which is the output, and transmit the measurement data to the higher-level control unit 3.

[0040] The learning processing unit 51b updates its own response characteristics by updating the control characteristic information 52b stored in the memory unit 32. Based on the iterative learning control method, the learning processing unit 51b updates its own response characteristics in order to make its output value follow the command value given by the higher-level control unit 3. Since the iterative learning control method is an existing technology, a detailed explanation is omitted.

[0041] <Command Value Generation Process> Figure 4 is a flowchart showing an example of the procedure for model generation processing by the higher-level control unit 3 of the substrate processing apparatus 1 according to this embodiment. Figure 5 is a schematic diagram illustrating the method for generating the first model. The model generation unit 31a of the processing unit 31 of the higher-level control unit 3 of the substrate processing apparatus 1 according to this embodiment sets an ideal model that shows ideal characteristics as a model that shows the response characteristics of the lower-level control unit 5 to be controlled (step S1). The ideal characteristics are, for example, characteristics in which the lower-level control unit 5 outputs an output value that matches the input command value. The model generation unit 31a generates a command value that becomes the target value of the output value of the lower-level control unit 5 by performing optimization processing based on the ideal model set in step S1 and a predetermined evaluation function (step S2). The model generation unit 31a provides the command value generated in step S2 to the lower-level control unit 5 and obtains the output value that the lower-level control unit 5 outputs in response (step S3). Based on the command value and output value obtained in step S3, the model generation unit 31a generates a first model that shows the response characteristics of the lower-level control unit 5 that accepts a command value as input and outputs an output value (step S4).

[0042] Figure 6 is a schematic diagram illustrating the method for generating the second model. The model generation unit 31a sets the first model generated in step S4 as a model showing the response characteristics of the subordinate control unit 5 to be controlled (step S5). The model generation unit 31a generates a command value that will be the target value of the output value of the subordinate control unit 5 by performing optimization processing based on the first model set in step S5 and a predetermined evaluation function (step S6). The model generation unit 31a provides the command value generated in step S6 to the subordinate control unit 5 and obtains the output value output by the subordinate control unit 5 accordingly (step S7). Based on the command value and output value obtained in step S7, the model generation unit 31a generates a second model showing the response characteristics of the subordinate control unit 5 that receives the command value as input and outputs an output value (step S8). The model generation unit 31a stores information about the generated first model and second model in the model DB 32b (step S9) and terminates the process.

[0043] FIG. 7 is a schematic diagram for explaining a method of generating a command value by the upper control unit 3 using the first model and the second model. The command value generation unit 31b of the substrate processing apparatus 1 according to the present embodiment sets the first model generated in the above-described model generation process as a model indicating the response characteristics of the lower control unit 5 to be controlled, and performs an optimization process based on this first model and a predetermined evaluation function, thereby generating an initial value of the command value. The initial value of the command value generated at this stage is the same as the command value generated in step S6 of the above-described model generation process.

[0044] The command value generation unit 31b inputs the initial value of the command value generated by the optimization process to the inverse model of the second model. The second model is a model that outputs an output value of the lower control unit 5 in response to an input of a command value, and the inverse model of the second model is a model that outputs a command value input to the lower control unit 5 in response to an input of an output value of the lower control unit 5. The command value generation unit 31b acquires the command value output by the inverse model of the second model in response to the input of the initial value of the command value, and gives this command value to the lower control unit 5.

[0045] The lower control unit 5 given a command value from the upper control unit 3 controls the pressure, power, etc. related to substrate processing using this command value as a target value. When the measured value by a sensor such as pressure or power as a result of this control is used as the output value of the lower control unit 5, in the substrate processing apparatus 1 according to the present embodiment, it can be expected that the command value to the lower control unit 5 and the output value of the lower control unit 5 will match.

[0046] <Modification Example> FIG. 8 is a schematic diagram showing a configuration example of the substrate processing apparatus 1 according to the modification example. The substrate processing apparatus 1 according to the second modification example has a three-layer hierarchical structure of an upper control unit 3, a middle control unit 4, and a lower control unit 5. The middle control unit 4 and the lower control unit 5 according to the modification example shown in FIG. 8 correspond to the upper control unit 3 and the lower control unit 5 shown in FIG. 7, and the middle control unit 4 performs the model generation process and the command value generation process shown in FIGS. 4 to 7 with the lower control unit 5 as a control target.

[0047] The substrate processing apparatus 1 according to the modified example has a configuration in which a host control unit 3 is provided above a middle control unit 4 that controls a lower control unit 5. The host control unit 3 similarly performs model generation processing and command value generation processing with the middle control unit 4 as the control target. The host control unit 3 generates a first model and a second model of the middle control unit 4 in advance and stores them in a model DB 32b, generates a command value for the middle control unit 4 using the first model and the second model, and gives the generated command value to the middle control unit 4. The command value from the host control unit 3 becomes a target value for the evaluation function of the middle control unit 4, and the middle control unit 4 performs optimization processing so that the evaluation function satisfies this target value, and generates a command value for the lower control unit 5.

[0048] FIG. 9 is a schematic diagram for explaining the correspondence between command values and output values in each layer in the substrate processing apparatus 1 according to the modified example. In the substrate processing apparatus 1 according to the present embodiment, as described above, the middle control unit 4 generates a command value using the first model and the second model of the lower control unit 5, so that it can be expected that the command value to the lower control unit 5 and the output value from the lower control unit 5 will match. Similarly, in the substrate processing apparatus 1 according to the modified example, the host control unit 3 generates a command value using the first model and the second model of the middle control unit 4, so that it can be expected that the command value to the middle control unit 4 and the output value of the first evaluation function when the output value of the lower control unit 5 is input to the evaluation function (the first evaluation function in this figure) of the middle control unit 4 will match.

[0049] Furthermore, in the substrate processing apparatus 1 according to the modified example, when there is a unit above the host control unit 3 that controls the host control unit 3, this unit generates a command value using the first model and the second model of the host control unit 3 and inputs it to the host control unit 3, so that it can be expected that the command value to the host control unit 3 and the output value of the second evaluation function when the output value of the first evaluation function of the middle control unit 4 is input to the evaluation function (the second evaluation function in this figure) of the host control unit 3 will match.

[0050] The substrate processing apparatus 1 may have multiple substrate processing chambers, known as multi-chambers. These multiple substrate processing chambers may perform the same type of processing, or they may perform different processing. Multiple intermediate control units 4 may be assigned to each of the multiple substrate processing chambers, and each intermediate control unit 4 may control a lower control unit 5 that controls the processing in each substrate processing chamber.

[0051] Even when the substrate processing apparatus 1 is equipped with multiple layers of units, in the substrate processing apparatus 1 according to this embodiment, it is expected that the input / output command values ​​and output values ​​will match in the units of each layer. Therefore, the user can design each unit considering only the behavior of the unit in the layer below it.

[0052] <Summary> The substrate processing apparatus 1 according to this embodiment, having the above configuration, includes a command value generation unit (upper control unit 3) that generates command values ​​to be input to a target unit (lower control unit 5) involved in substrate processing. The lower control unit 5 has functions such as iterative learning control that adjusts its response characteristics based on the input command value and its own output value corresponding to the command value. The upper control unit 3 stores a first model of the lower control unit 5 in the model DB 32b of the storage unit 32, which is generated based on data of a command value generated assuming that the response characteristics of the lower control unit 5 are predetermined characteristics (ideal characteristics) and the output value output by the lower control unit 5 in response to this command value. The upper control unit 3 also stores a second model of the lower control unit 5 in the model DB 32b, which is generated based on data of a command value generated assuming that the response characteristics of the lower control unit 5 are the characteristics of the first model and the output value output by the lower control unit 5 in response to this command value. The upper control unit 3 generates command values ​​to the lower control unit 5 based on the first model and the second model stored in the model DB 32b. As a result, the substrate processing apparatus 1 according to this embodiment is expected to realize substrate processing that takes into account the characteristics of a unit having a learning function such as repeated learning control.

[0053] Furthermore, in the substrate processing apparatus 1 according to this embodiment, the higher-level control unit 3 converts the command values ​​generated by optimization processing using the first model based on the inverse model of the second model, and uses the value obtained from the conversion as the command value to the lower-level control unit 5. As a result, the substrate processing apparatus 1 according to this embodiment is expected to reduce the influence of the learning ability of the lower-level control unit 5 by using the generated second model.

[0054] Furthermore, in the substrate processing apparatus 1 according to this embodiment, the higher-level control unit 3 may generate command values ​​for the lower-level control unit 5 through optimization processing using the generated first and second models. As a result, the substrate processing apparatus 1 according to this embodiment is expected to reduce the influence of the learning ability of the lower-level control unit 5 by using the generated second model.

[0055] Furthermore, in the substrate processing apparatus 1 according to this embodiment, the upper control unit 3 generates a command value trajectory containing multiple command values ​​for multiple time points and provides it to the lower control unit 5. As a result, the substrate processing apparatus 1 according to this embodiment can provide command values ​​for multiple time points from the upper control unit 3 to the lower control unit 5 all at once, which is expected to facilitate the operation of the upper control unit 3 and the lower control unit 5 at different processing speeds.

[0056] Furthermore, in the substrate processing apparatus 1 according to this embodiment, the lower-level control unit 5 adjusts its response characteristics so that the difference between the command value input from the upper-level control unit 3 and the output value corresponding to this command value becomes smaller. As a result, in the substrate processing apparatus 1 according to this embodiment, the lower-level control unit 5 can change its own response characteristics through repeated learning control and is expected to output an output value that has a smaller difference from the given command value.

[0057] Furthermore, in the substrate processing apparatus 1 according to this embodiment, the intermediate control unit 4 generates command values ​​for the lower control unit 5 such that the difference between the target value given by the upper control unit 3 and the evaluation value calculated based on the output value of the lower control unit 5 becomes small. As a result, even if the substrate processing apparatus 1 according to this embodiment has a configuration in which units are provided in multiple layers of three or more layers, it can be expected that the units in each layer will generate command values ​​for the unit in the layer below, similar to a two-layer configuration of an upper control unit 3 and a lower control unit 5.

[0058] 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 meaning and scope equivalent to the claims are intended.

[0059] 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.

[0060] 1. Circuit board processing unit (computer) 3. Higher-level control unit (higher-level unit) 4. Intermediate-level control unit (higher-level unit, target unit) 5. Lower-level control unit (target unit) 5a. Pressure control unit 5b. Power control unit 31. Processing unit 31a. Model generation unit 31b. Command value generation unit 32. Storage unit 32a. Program 32b. Model DB 33. Communication unit 51. Processing unit 51a. Control processing unit 51b. Learning processing unit 52. Storage unit 52a. Program (computer program) 52b. Control characteristic information 53. Communication unit 98, 99. Recording medium

Claims

1. A substrate processing apparatus comprising a command value generation unit for generating command values ​​to be input to a target unit for substrate processing, wherein the target unit has a function to adjust its response characteristics based on the input command value and its own output value corresponding to the command value, the command value generation unit stores in a storage unit a first model of the target unit generated based on data of a command value generated assuming that the response characteristics of the target unit are predetermined characteristics and an output value output by the target unit according to the command value, the command value generation unit stores in the storage unit a second model of the target unit generated based on data of a command value generated assuming that the response characteristics of the target unit are the characteristics of the first model and an output value output by the target unit according to the command value, and generates command values ​​to the target unit based on the first model and the second model.

2. The substrate processing apparatus according to claim 1, wherein the command value generation unit converts the command value generated by the optimization process using the first model based on the inverse model of the second model, and the value obtained by the conversion is used as the command value for the target unit.

3. The substrate processing apparatus according to claim 1, wherein the command value generation unit generates the command value by optimization processing using the first model and the second model.

4. The substrate processing apparatus according to claim 1, wherein the command value generation unit generates a command value trajectory that includes multiple command values ​​for multiple time points.

5. The substrate processing apparatus according to claim 1, wherein the target unit adjusts the response characteristics so that the difference between the input command value and the output value corresponding to the command value becomes small.

6. The substrate processing apparatus according to claim 5, wherein the command value generation unit generates the command value such that the difference between the target value given by the higher-level unit and the evaluation value calculated based on the output value of the target unit is reduced.

7. The substrate processing apparatus according to claim 1, wherein the command value generation unit provides the generated command value to the target unit, and the target unit performs operations related to substrate processing to output an output value corresponding to the given command value.

8. An information processing method for a substrate processing apparatus to generate a command value to input to a target unit having a function to adjust response characteristics based on a command value and an output value corresponding to the command value, wherein the substrate processing apparatus stores in a storage unit a first model of the target unit generated based on data of a command value generated assuming that the response characteristics of the target unit are predetermined characteristics and an output value output by the target unit in accordance with the command value, stores in the storage unit a second model of the target unit generated based on data of a command value generated assuming that the response characteristics of the target unit are the characteristics of the first model and an output value output by the target unit in accordance with the command value, and generates a command value to the target unit based on the first model and the second model.

9. A computer program that causes a computer to execute a process to generate a command value to be input to a target unit having a function to adjust response characteristics based on a command value and an output value corresponding to the command value, wherein the computer stores in a storage unit a first model of the target unit generated based on data of a command value generated assuming that the response characteristics of the target unit are predetermined characteristics and an output value output by the target unit in accordance with the command value, and stores in the storage unit a second model of the target unit generated based on data of a command value generated assuming that the response characteristics of the target unit are the characteristics of the first model and an output value output by the target unit in accordance with the command value, and causes the computer to execute a process to generate a command value to the target unit based on the first model and the second model.

Images