JP2022092690A - Controller, adjustment method, lithography apparatus, and manufacturing method of article - Google Patents

Abstract

To adjust proper control characteristics in a short time, in a controller using a neural network.SOLUTION: A controller for generating a control signal for controlling a control object comprises: a first compensator for generating a first signal on the basis of a control deviation of the control object; a corrector for correcting the control deviation with one control part among a plurality of control parts for generating a correction signal by correcting the control deviation according to a calculation formula capable of adjusting factors; a second compensator for generating a second signal by a neural network on the basis of the correction signal; and a calculator for generating the control signal on the basis of the first signal and the second signal.SELECTED DRAWING: Figure 3

Description

The present invention relates to a control device, an adjustment method, a lithography device, and a method for manufacturing an article.

In a photolithography process when manufacturing a device such as a semiconductor device or a flat panel display (FPD), an exposure device that transfers a mask pattern to a substrate is used. The exposure apparatus is required to perform highly accurate position control and synchronization control of the mask stage that holds the mask and the substrate stage that holds the substrate, for example, in order to align the mask and the substrate.

The requirements for accuracy required for position control and synchronous control of the stage and the like as described above are becoming stricter as the device becomes higher in definition, and the required accuracy may not be reached only by the conventional feedback control. Therefore, in addition to the conventional controller, an effort is being made to configure a neural network controller in parallel (Patent Document 1). Further, a method has been devised in which the neural network controller is switched according to the state of the controlled object and compensation is performed according to the controlled object (Patent Document 2).

Special Table No. 7-503563 Gazette Japanese Unexamined Patent Publication No. 7-277286

However, although the accuracy can be expected to be improved by configuring a plurality of neural network controllers, the control calculation time increases. Further, although the parameters of the neural network controller are adjusted by machine learning, a lot of time is required to learn the parameters of a plurality of neural networks. Further, when the state change of the controlled object or the change of the disturbance environment occurs, the parameters of the predetermined neural network become non-optimal, so that it takes a lot of time to readjust the parameters.

Therefore, an object of the present invention is to provide a control device using a neural network, which is advantageous for adjusting appropriate control characteristics in a short time.

In order to achieve the above object, the control device as one aspect of the present invention is a control device that generates a control signal for controlling the control target, and a first signal is generated based on the control deviation of the control target. A correction that corrects the control deviation using one of the first compensator to be generated and a plurality of adjustment units that generate a correction signal by correcting the control deviation according to an arithmetic expression that can adjust the coefficient. It includes a device, a second compensator that generates a second signal by a neural network based on the correction signal, and a calculator that generates the control signal based on the first signal and the second signal. It is characterized by that.

According to the present invention, in a control device using a neural network, it is possible to provide an advantageous control device for adjusting appropriate control characteristics in a short time.

It is a figure which shows the configuration example of the system in 1st Embodiment. It is a figure which shows the configuration example of the system in 1st Embodiment. It is a figure which shows the structural example of the controller in the system of 1st Embodiment. It is a figure which shows the structural example of the controller in Example 6. It is a figure which shows the structural example of the controller in Example 7. FIG. It is a figure which shows the structural example of the controller in Example 8. FIG. It is a figure which shows the configuration example of the system in 1st Embodiment. It is a flowchart which shows the operation example when the system of 1st Embodiment is applied to a production apparatus. It is a figure which shows the example of the measurement result of the disturbance suppression characteristic. It is a figure which shows the structural example of the stage control apparatus in 2nd Embodiment. It is a figure which shows the structural example of the control board in the system of 2nd Embodiment. It is a flowchart which shows the adjustment of a corrector. It is a figure which illustrates the position control deviation. It is a figure which shows the example of the result of the frequency analysis. It is a figure which shows the structural example of an exposure apparatus.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In addition, in each figure, the same reference number is given to the same member, and duplicate description is omitted.

<First Embodiment>
FIG. 1 shows the configuration of the system SS in this embodiment. The system SS is applied, for example, to a manufacturing apparatus for manufacturing an article. The manufacturing apparatus includes, for example, a processing apparatus for processing an article or a member constituting a part of the article. The processing device may be, for example, a lithography device that transfers a pattern to a material or member, a film forming device that forms a film on the material or member, an apparatus that etches the material or member, or a heating device that heats the material or member. Can be.

The system SS includes, for example, a sequence unit 101, a control device 100, and a control target 103. The control device 100 includes a controller 102. The control device 100 or the controller 102 generates a control signal MV for controlling the control target 103. When the system SS is applied to a production system, the sequence unit 101 is provided with a production sequence. The production sequence defines the procedure for production. The sequence unit 101 generates a target value R for controlling the control target 103 based on the production sequence, and provides the target value R to the control device 100 or the controller 102.

The control device 100 or the controller 102 feedback-controls the control target 103. Specifically, in the control device 100, the control amount CV of the control target 103 is set based on the control deviation which is the difference between the target value R provided by the sequence unit 101 and the control amount CV provided by the control target 103. The control target 103 is controlled so as to follow the target value R. The controlled object 103 can have a sensor for detecting the controlled variable CV, and the controlled variable CV detected by the sensor can be provided to the controller 102. The target value R, the control signal MV, and the control amount CV can be time-series data whose values change with the passage of time.

As illustrated in FIG. 2, the learning unit 201 may be incorporated in the system SS. The learning unit 201 may be configured as a part of the control device 100 or may be configured as an external device of the control device 100. When the learning unit 201 is configured as an external device of the control device 100, the learning unit 201 may be separated from the control device 100 after the learning is completed. The learning unit 201 is configured to send a pre-prepared learning sequence to the sequence unit 101. The sequence unit 101 generates a target value R according to a learning sequence and provides it to the controller 102.

The controller 102 generates a control signal MV based on a control deviation which is a difference between the target value R generated and provided by the sequence unit 101 according to the learning sequence and the controlled variable CV provided by the controlled object 103. Here, the controller 102 has a neural network, and the control signal MV is generated using the neural network. The control signal MV generated by the controller 102 is provided to the control target 103, and the control target 103 operates according to the control signal MV. The control amount CV as a result of this operation is provided to the controller 102. The controller 102 provides the learning unit 201 with an operation history showing an operation history of the controller 102 based on the target value R. The learning unit 201 determines the parameter value of the neural network of the controller 102 based on the operation history, and sets the parameter value in the neural network. The parameter value is determined by machine learning such as reinforcement learning, for example.

FIG. 3 is a diagram showing one of the configuration examples of the controller 102. The controller 102 includes a first compensator 301 that generates the first signal S1 based on the control deviation E, and a corrector 303 that generates a correction signal CS by calculating the control deviation E according to an arithmetic expression whose coefficient can be adjusted. And include. Further, the controller 102 is an operation to generate a control signal MV based on the second compensator 302 that generates the second signal S2 by the neural network based on the correction signal CS, and the first signal S1 and the second signal S2. Includes vessel 306.

The corrector 303 has a plurality of correctors including a first adjustment unit 303a and a second adjustment unit 303b, and an adjustment unit (connected adjustment unit) to be used can be selected according to a control state. .. The control signal MV is the sum of the first signal S1 and the second signal S2, and the arithmetic unit 306 may be composed of an adder. Further, the control signal MV is a signal obtained by correcting the first signal S1 based on the second signal S2. The controller 102 includes a subtractor 305 that generates a control deviation E, which is a difference between the target value R and the control amount CV. The controlled variable CV is acquired by being measured by a sensor (not shown) included in the controlled object 103. Further, the control amount and the target which are the results of controlling the control target 103 based on the control signal MV are compared with the difference between the control amount and the target value R which are the result of controlling the control target 103 based on the first signal S1. The difference from the value R is smaller.

The controller 102 further includes an operation history recording unit 304. The learning unit 201 in FIG. 2 is configured to perform learning for determining the parameter value of the neural network of the second compensator 302 in FIG. For learning by the learning unit 201, the operation history recording unit 304 records the operation history required for learning by the learning unit 201, and provides the recorded operation history to the learning unit 201. The operation history is, for example, a correction signal CS which is input data to the second compensator 302 and a second signal S2 which is output data of the second compensator 302, but the control deviation and the output of the second compensator 302. It may be the second signal S2 which is data, or it may be other data. The first adjustment unit 303a and the second adjustment unit 303b can perform learning with arbitrary parameters as initial values.

In the following Examples 1 to 5, a configuration example of the corrector 303 will be described. In Examples 1 to 5, an example of an arithmetic expression used by the corrector 303 to generate a correction signal CS based on the control deviation E is shown. The arithmetic expression can be, for example, a monomial expression or a polynomial.

(Example 1)
In the first embodiment, the first adjusting unit 303a and the second adjusting unit 303b have the control characteristics represented by the following equation (1). Here, the input (E) to the corrector 303 is x, the output (CS) of the corrector 303 is y, and an arbitrary coefficient (constant) is Kp.

(Example 2)
In the second embodiment, the first adjusting unit 303a and the second adjusting unit 303b have the control characteristics represented by the following equation (2). Here, the input (E) to the corrector 303 is x, the output (CS) of the corrector 303 is y, the time is t, and an arbitrary coefficient (constant) is Ki. The integration may be performed a plurality of times. The integral may be a definite integral in a certain time interval or an indefinite integral.

(Example 3)
In the third embodiment, the first adjusting unit 303a and the second adjusting unit 303b have the control characteristics represented by the following equation (3). Here, the input (E) to the corrector 303 is x, the output (CS) of the corrector 303 is y, the time is t, and an arbitrary coefficient (constant) Kd is used. The differentiation may be performed a plurality of times.

(Example 4)
In the fourth embodiment, the first adjusting unit 303a and the second adjusting unit 303b have the control characteristics represented by the following equation (4). Here, the input (E) to the corrector 303 is x, the output (CS) of the corrector 303 is y, and the arbitrary coefficients (constants) are Kp, Ki, and Kd. The integration and differentiation may be performed a plurality of times.

(Example 5)
In the fifth embodiment, the first adjusting unit 303a and the second adjusting unit 303b have control characteristics represented by the arithmetic expression of the following equation (5). Here, the input (E) to the corrector 303 is x, the output (CS) of the corrector 303 is y, the integral order of the multiple integral is n, the derivative order is m, and any coefficient (constant) is Kp, n multiple integral. Let Ki_n be an arbitrary coefficient (constant) at the time of, and Kd_m be an arbitrary constant at the time of the m-th order derivative.

In the first to fifth embodiments, the arithmetic expression used by the corrector 303 to generate the correction signal CS has at least one of a term proportional to the control deviation E, a term for integrating, and a term for differentiating. It can be understood as an example that includes.

The coefficients (constants) Kp, Ki, Kd, Ki_n, and Kd_m of the arithmetic expressions given in Examples 1 to 5 are examples of adjustable parameters of the corrector 303. The first adjusting unit 303a and the second adjusting unit 303b determine the optimum parameters using any one of the first to fifth embodiments according to the change in the control state assumed in advance. The control state is, for example, switching of synchronous control, switching of controllers, switching of operation patterns, changes in environment and disturbance such as temperature, noise, and floor vibration. Optimal control characteristics can be obtained by selecting the first adjustment unit 303a and the second adjustment unit 303b according to the control state. Since the adjustment time due to the configuration of a plurality of correctors is shorter than the adjustment time due to the configuration of a plurality of neural networks, it is advantageous from the viewpoint of time reduction.

Further, when the state of the controlled object 103 or the disturbance environment changes during the operation of the system SS, the value (parameter value) of the arithmetic expression (coefficient) exemplified as Examples 1 to 5 is adjusted. Can respond to changes. The time required for adjusting the value of the arithmetic expression (coefficient) of the corrector 303 is shorter than the time required for retraining the neural network. Therefore, the control accuracy can be maintained without reducing the productivity of the system SS. That is, by introducing the corrector 303, it is possible to improve the tolerance of the controlled object 103 to changes in the state and the disturbance environment.

(Example 6)
Examples 6 to 8 describe the relationship between the change in the control state and the switching of the adjusting unit used in the corrector 303.

FIG. 4 is a diagram showing a configuration example of the controller 102 in the sixth embodiment. In the sixth embodiment, as shown in FIG. 4, it is possible to switch whether to control the plurality of control targets including the control target 103a and the control target 103b individually or in synchronization with each other. The control state in the sixth embodiment is a state determined by whether a plurality of control targets are individually controlled or synchronously controlled, and an appropriate adjustment unit is switched depending on whether or not the plurality of control targets are synchronously controlled. Run. Further, in the sixth embodiment, the corrector 303 is switched according to the state of the synchronous control switching unit 402, and it is possible to select whether to use the first adjusting unit 303a or the second adjusting unit 303b. .. Further, in the sixth embodiment, the axis in which the control target 103a is controlled is set as the master axis, the axis in which the control target 103b is controlled is set as the slave axis, and the slave axis follows the master axis. explain.

The controller 102 acquires the control quantities CVa and CVb of the control targets 103a and 103b measured by the sensors (not shown) included in the control target 103, and sets the difference from the target values Ra and Rb, respectively, as the control deviation. Calculated as Ea and Eb.

The control deviation Ea is input to the controller 301a. The input to the controller 301b and the corrector 303 provided in front of the neural network 302 configured in parallel with the controller 301b is switched depending on whether or not the control target 103a and the control target 103b are synchronously controlled. be able to. For the input to the corrector 303, select the control deviation Eb switched by the synchronization control switching unit 402 that switches the synchronization control between the control target 103a and the control target 103b, or the synchronization deviation Ec that is the difference between the control deviation Eb and the control deviation Ea. can do. For the output of the corrector 303, it is possible to select whether to use the first adjustment unit 303a or the second adjustment unit 303b according to the state of the synchronization control switching unit 402.

As a specific example of the sixth embodiment, for example, when applied to an exposure apparatus, different adjustment units may be selected depending on whether the plate stage and the mask stage are synchronized with each other and when other operations are performed. At this time, the parameters of the first adjusting unit 303a and the second adjusting unit 303b are optimized when the plate stage and the mask stage are synchronized and when the other operations are performed. The outputs of the first adjustment unit 303a and the second adjustment unit 303b selected according to the state of the synchronization control switching unit 402 are input to the neural network 302 (second compensator). The output of the compensator 301a is a control signal MVa. The output of the compensator 301b and the output of the neural network 302 are added to obtain a control signal MVb. The controller 102 outputs the control signals MVa and MVb to the controlled objects 103a and 103b, respectively.

The first adjusting unit 303a and the second adjusting unit 303b determine the optimum parameters using any one of Examples 1 to 5 according to the state of the synchronization control switching unit 402. Optimal control characteristics can be obtained by selecting the first adjustment unit 303a and the second adjustment unit 303b according to the state of the synchronization control switching unit 402.

The increase in the adjustment time due to the configuration in which the corrector 303 selects the optimum adjustment unit from the plurality of adjustment units is shorter than the increase in the adjustment time due to the configuration of the plurality of neural networks. Further, when the state of the controlled object 103 or the disturbance environment changes during the operation using any of the first to fifth embodiments, it is possible to cope with the change by adjusting the parameters of the first to fifth embodiments. can. The time required for adjustment of the first adjustment unit 303a and the second adjustment unit 303b is shorter than the time required for re-learning the neural network. In the sixth embodiment, even when a plurality of compensations are performed according to the state of the controlled object and the disturbance environment, the increase in the calculation time and the learning time can be suppressed, and the state change of the controlled object and the disturbance environment occur. However, the appropriate control characteristics can be adjusted in a short time.

(Example 7)
FIG. 5 is a diagram showing a configuration example of the controller 102 in the seventh embodiment. In the seventh embodiment, the compensator 301 is switched according to the state and operation of the controlled object 103, and it is possible to select whether to use the compensator 301a or the compensator 301b. The control state in the seventh embodiment is a state determined by which of the plurality of compensators is used, and the adjustment unit is switched depending on whether the compensator 301a or the compensator 301b is used. do. Further, in the seventh embodiment, the compensator 303 is switched depending on the state of the compensator 301, and it is possible to select whether to use the first adjusting unit 303a or the second adjusting unit 303b.

As a specific example of the seventh embodiment, for example, when applied to an exposure apparatus, the compensator 301a is used during the exposure operation of the plate stage, and the compensator 301b is used during the plate transfer operation. May be good. That is, the adjustment unit used for correcting the control deviation E may be selected from the plurality of adjustment units based on whether or not the gain of the control target 103 has been switched. At this time, the first adjusting unit 303a and the second adjusting unit 303b determine the optimum parameters for the compensator 301a and the compensator 301b using any one of the first to fifth embodiments. Optimal control characteristics can be obtained by selecting the first adjustment unit 303a and the second adjustment unit 303b according to the state of the compensator 301.

The increase in the adjustment time due to the configuration in which the corrector 303 selects the optimum adjustment unit from the plurality of adjustment units is shorter than the increase in the adjustment time due to the configuration of the plurality of neural networks. Further, when the state of the controlled object 103 or the disturbance environment changes during the operation using any of the first to fifth embodiments, it is possible to cope with the change by adjusting the parameters of the first to fifth embodiments. can. The time required for adjustment of the first adjustment unit 303a and the second adjustment unit 303b is shorter than the time required for re-learning the neural network. In the seventh embodiment, even when a plurality of compensations are performed according to the state of the controlled object and the disturbance environment, the increase in the calculation time and the learning time can be suppressed, and the state change of the controlled object and the disturbance environment occur. However, the appropriate control characteristics can be adjusted in a short time.

(Example 8)
FIG. 6 is a diagram showing a configuration example of the controller 102 in the eighth embodiment. The control state in the eighth embodiment is a state determined by whether or not the operation pattern 403 to be controlled is changed. A specific example of the operation pattern 403 will be described later. In the eighth embodiment, the adjustment unit is switched according to the state of the operation pattern 403, and it is possible to select whether to use the first adjustment unit 303a or the second adjustment unit 303b.

As a specific example of the seventh embodiment, for example, when it is applied to a stage device used for an exposure device or the like, it may be applied by switching between an acceleration section when the stage is driven and an operation pattern other than that. At this time, the first adjusting unit 303a and the second adjusting unit 303b determine the optimum parameters using any of the first to fifth embodiments according to the state of the operation pattern 403 of the controlled object 103. Optimal control characteristics can be obtained by selecting the first adjustment unit 303a and the second adjustment unit 303b according to the state of the operation pattern 403.

The increase in the adjustment time due to the configuration in which the corrector 303 selects the optimum adjustment unit from the plurality of adjustment units is shorter than the increase in the adjustment time due to the configuration of the plurality of neural networks. Further, when the state of the controlled object 103 or the disturbance environment changes during the operation using any of the first to fifth embodiments, it is possible to cope with the change by adjusting the parameters of the first to fifth embodiments. can. The time required for adjustment of the first adjustment unit 303a and the second adjustment unit 303b is shorter than the time required for re-learning the neural network. In the eighth embodiment, even when a plurality of compensations are performed according to the state of the controlled object and the disturbance environment, the increase in the calculation time and the learning time can be suppressed, and the state change of the controlled object and the disturbance environment occur. However, the appropriate control characteristics can be adjusted in a short time.

As illustrated in FIG. 7, the control device 100 may include a setting unit 202 for selecting whether to use the first adjustment unit 303a or the second adjustment unit. Further, the setting unit 202 may have a role of setting the parameter value of the corrector 303.

The setting unit 202 may execute an adjustment process for switching the adjustment unit or adjusting the parameter value, and may determine and set the switching of the adjustment unit or the parameter value by this adjustment process, or based on a command from the user. You may switch the adjustment unit or set the parameter value. In the former, the setting unit 202 may send a confirmation sequence for confirming the operation of the controller 102 to the sequence unit 101, and cause the sequence unit 101 to generate the target value R based on this confirmation sequence. Then, the setting unit 202 acquires an operation history (for example, a control deviation) from the controller 102 that operates based on the target value R, and based on the operation history, whether or not there is a need to switch the corrector 303 and parameters. The value can be determined. The setting unit 202 having such a function can be understood as an adjusting unit for switching the corrector 303 and adjusting the parameter value.

The setting unit 202 acquires an operation history (for example, control deviation) from the controller 102 at the time of production in which the sequence unit 101 generates a target value R based on the production sequence, and the parameter of the corrector 303 is based on the operation history. You may decide whether to perform value adjustments. Alternatively, a determination unit for determining whether to adjust the parameter value of the corrector 303 by the setting unit 202 at the time of production in which the sequence unit 101 generates the target value R based on the production sequence is provided separately from the setting unit 202. May be done.

Next, an example in which production is performed by the system in this embodiment will be described. FIG. 8 is an operation example of the system SS when the system SS of the present embodiment is applied to the production apparatus. In step S501, the sequence unit 101 generates a target value R based on a given production sequence and provides the target value R to the control device 100 or the controller 102. The control device 100 or the controller 102 controls the control target 103 based on the target value R.

In the process S502, the setting unit 202 acquires the operation history (for example, control deviation) of the controller 102 in the process S501.

In the process S503, the setting unit 202 determines whether to perform adjustment of the corrector 303 such as switching of the adjustment unit and adjustment (or readjustment) of the parameter value based on the operation history acquired in the process S502. sell. For example, the setting unit 202 can determine that the adjustment (or readjustment) of the switching parameter value of the adjustment unit is executed when the operation history satisfies a predetermined condition. The predetermined condition is a condition for stopping production. For example, when the control deviation acquired as the operation history exceeds the specified value, it is determined that the correction device 303 needs to be adjusted. Then, when the adjustment of the corrector 303 is executed by the setting unit 202, the process proceeds to step S504, and when not, the process proceeds to step S505.

In step S504, the setting unit 202 executes the adjustment of the corrector 303. This adjustment is made in a state where the parameter value of the second compensator 302 is maintained in the previous state, and by this adjustment, for example, the parameter value (coefficient) of the corrector 303 is reset.

In the process S505, the sequence unit 101 determines whether or not to end the production according to the production sequence, returns to the process S501 if it does not end, and ends the production if it ends. According to the above processing, even when the production should be stopped, the parameter value of the corrector 303 can be promptly adjusted and the production can be restarted while minimizing the interruption of the production.

In step S504, the setting unit 202 sends the confirmation sequence to the sequence unit 101, causes the sequence unit 101 to execute the confirmation sequence, and can acquire the operation history (for example, control deviation) in the confirmation sequence from the controller 102. Then, the setting unit 202 performs frequency analysis of the operation history, determines a frequency to be improved based on the result, and sets the parameter value of the corrector 303 so that the control deviation at that frequency is within the specified value. Can be decided. A more specific example of step S504 will be described in the second embodiment.

FIG. 9 is a diagram illustrating the measurement results of the disturbance suppression characteristic. The result of measuring the frequency response when the control deviation when a sine wave is input as the control signal MV in FIG. 2 is used as an output is called a disturbance suppression characteristic. In FIG. 9, the horizontal axis represents the frequency and the vertical axis represents the gain of the disturbance suppression characteristic. Since the disturbance suppression characteristic represents the frequency response of the control deviation E when the disturbance is added to the control signal MV, a large gain indicates that the effect of suppressing the disturbance is low. On the other hand, a small gain indicates that the effect of suppressing disturbance is high. In FIG. 9, the broken line shows the disturbance suppression characteristic before adjustment, and the solid line shows the disturbance suppression characteristic after adjustment.

When the frequency shown by the alternate long and short dash line in FIG. 9 is set as the frequency at which the disturbance suppression characteristic should be improved and the step S504 is executed, for example, the disturbance suppression characteristic as shown by the solid line can be obtained. It can be seen that the gain of the disturbance suppression characteristic becomes smaller at the frequency to be improved, and the disturbance suppression characteristic is improved. In Examples 1 to 8, when the parameters of the corrector 303 provided in the front stage of the neural network are adjusted, the parameters may be adjusted using the disturbance suppression characteristic shown in FIG. 9 as an index.

<Second Embodiment>
In this embodiment, an example of applying the control system SS described in the first embodiment to the stage control device 800 will be described. Matters not mentioned as this embodiment follow the first embodiment. FIG. 10 is a diagram showing a hardware configuration when the control system SS shown in FIG. 1 is applied to the stage control device 800.

The stage control device 800 is configured to control the stage 804 while holding the object on the stage 804 in order to control the position of an object such as a substrate. The stage control device 800 includes a control board 801, a current driver 802, a motor 803, a stage 804, and a sensor 805. The control board 801 corresponds to the control device 100 or the controller 102 in the system SS of the first embodiment. The current driver 802, the motor 803, the stage 804, and the sensor 805 correspond to the controlled object 103 in the system SS of the first embodiment. However, the current driver 802 may be incorporated in the control board 801. Although not shown in FIG. 10, the stage control device 800 may include a sequence unit 101, a learning unit 201, and a setting unit 202.

A position target value as a target value can be supplied to the control board 801 from the sequence unit 101. The control board 801 can generate a current command as a control signal and supply it to the current driver 802 based on the position target value supplied from the sequence unit 101 and the position information supplied from the sensor 805. Further, the control board 801 can supply the operation history to the sequence unit 101.

The current driver 802 can supply the motor 803 with a current according to the current command. The motor 803 may be an actuator that converts the current supplied from the current driver 802 into thrust and drives the stage 804 with the thrust. The stage 804 may hold an object such as a plate or mask. The sensor 805 can detect the position of the stage 804 and supply the position information obtained thereby to the control board 801.

FIG. 11 shows a configuration example of the control board 801 as a block diagram. The control board 801 is corrected by correcting the control deviation E according to the first compensator 301 that generates the first signal S1 based on the position control deviation E of the stage 804 as the control target and the arithmetic expression that can adjust the coefficient. It may include a corrector 303 that generates a signal CS. Further, the control board 801 generates a current command as a control signal based on the second compensator 302 that generates the second signal S2 by the neural network based on the correction signal CS, and the first signal S1 and the second signal S2. It may include the arithmetic unit 306 to be used. Further, the control board 801 may include a subtractor 305 that generates a control deviation E which is a difference between the position target value PR and the position information.

The stage control device 100 of the second embodiment may also include the learning unit 201 as in the first embodiment described with reference to FIG. 7. The learning unit 201 may be configured to perform learning for determining the parameter value of the neural network of the second compensator 302. For learning by the learning unit 201, the operation history recording unit 304 may record the operation history required for learning by the learning unit 201 and provide the recorded operation history to the learning unit 201. The operation history may be, for example, a correction signal CS which is input data for the second compensator 302 and a second signal S2 which is output data of the second compensator 302, but other data may be used.

The stage control device 100 of the second embodiment can include a setting unit 202. The setting unit 202 may execute an adjustment process for adjusting the parameter value of the corrector 303, determine and set the parameter value of the corrector 303 by this adjustment process, or correct based on a command from the user. The parameter value of the device 303 may be set.

With reference to FIG. 8, the operation of the stage device 800 when the stage control device 800 of the second embodiment is applied to the production device will be exemplified. In step S501, the sequence unit 101 can generate a position target value PR based on a given production sequence and provide it to the stage control device 800. The stage control device 800 controls the position of the stage 804 based on the position target value PR.

In step S502, the setting unit 202 acquires the operation history (for example, control deviation) of the control board 801 in step S501.

In the process S503, the setting unit 202 determines whether to perform adjustment of the corrector 303 such as switching of the adjustment unit and adjustment (or readjustment) of the parameter value based on the operation history acquired in the process S502. sell. For example, the setting unit 202 can determine that the adjustment of the corrector 303 is executed when the operation history satisfies a predetermined condition. The predetermined condition is a condition for stopping production. For example, when the maximum value of the position control deviation during constant velocity driving of the stage 804 exceeds a predetermined predetermined value, the corrector 303 needs to be adjusted. Is judged. Then, when the adjustment of the corrector 303 is executed by the setting unit 202, the process proceeds to step S504, and when not, the process proceeds to step S505.

In step 504, the setting unit 202 may execute the adjustment of the corrector 303. In the process S505, the sequence unit 101 determines whether or not to end the production according to the production sequence, returns to the process S501 if it does not end, and ends the production if it ends.

FIG. 12 shows a specific example of the processing in the adjustment of the parameter value (or the readjustment of the parameter value) in the adjustment of the corrector 303 in the step S504. In step S601, the setting unit 202 may send a confirmation sequence for confirming the operation of the stage control device 800 to the sequence unit 101, and cause the sequence unit 101 to generate a position target value PR based on this confirmation sequence. In step S602, the setting unit 202 can acquire the position control deviation E as the operation history from the controller 102 that operates based on the position target value PR.

Here, with reference to FIG. 13, the change in the position control deviation E before and after the parameter adjustment will be described. FIG. 13 is a diagram illustrating the position control deviation before and after the parameter adjustment. In FIG. 13, the horizontal axis represents time and the vertical axis represents position control deviation E. Here, the curve shown by the dotted line is the position control deviation E before adjusting the parameter value of the corrector 303, and indicates that the position control accuracy is deteriorated. By adjusting the parameter value, the fluctuation of the position control deviation E can be reduced.

In step S603, the setting unit 202 can perform frequency analysis of the position control deviation E acquired in step S602. Here, the results of frequency analysis before and after parameter adjustment will be described with reference to FIG. FIG. 14 is a diagram illustrating the results of frequency analysis before and after parameter adjustment. In FIG. 14, the horizontal axis is frequency and the vertical axis is power spectrum. The dotted line indicates the frequency showing the maximum spectrum before adjustment. In step S604, the setting unit 202 can determine, for example, the frequency showing the maximum spectrum in the power spectrum as the frequency to be improved.

Steps S605 to S610 are specific examples of the adjustment process for adjusting the parameter value of the corrector 303. Here, an example is described by adopting the steepest descent method as a method for adjusting the parameter value, but other methods may be used. In step S605, the setting unit 202 initializes n to 1. For example, when the arithmetic expression of the corrector 303 is composed of three terms of a first-order integral term, a proportional term, and a first-order differential term, there are three parameters for which the parameter values should be adjusted: Ki, Kp, and Kd. The parameter value pn in the nth adjustment is shown by the following equation (6).

In step S606, the setting unit 202 can set an arbitrary initial value for the parameter value p1 in the first adjustment of the parameter value pn. In the nth adjustment, the parameter value pn represented by the equation (8) described later can be set.

The objective function J (pn) for adjusting the parameter value pn can be, for example, the gain of the disturbance suppression characteristic at the frequency determined in step S604. In step S607, the setting unit 202 can measure the gradient vector grade J (pn) of the objective function J (pn). The gradient vector grade J (pn) can be given by the following equation (7). The gradient vector grade J (pn) can be measured by changing each element Ki-n, Kpn, Kd-n constituting the parameter value pn by a minute amount.

In step S608, the setting unit 202 can determine whether or not the value of each element of the gradient vector grade J (pn) is equal to or less than the specified value as the convergence test of the steepest descent method. If the value of each element of the gradient vector grade J (pn) is equal to or less than the specified value, the setting unit 202 may end the adjustment of the parameter value of the corrector 303. On the other hand, if the value of each element of the gradient vector grade J (pn) exceeds the specified value, the setting unit 202 can calculate the parameter value pn + 1 in step S609. Here, the parameter value pn + 1 can be calculated according to the following equation (8) using, for example, any constant α larger than 0. In step S610, the setting unit 202 adds 1 to the value of n and returns to step S606.

In step S611, the setting unit 202 may send a confirmation sequence for confirming the operation of the stage control device 800 to the sequence unit 101, and cause the sequence unit 101 to generate a position target value PR based on this confirmation sequence. In step S612, the setting unit 202 can acquire the position control deviation E as the operation history from the controller 102 that operates based on the position target value PR.

In step S613, the setting unit 202 determines whether or not the position control deviation E acquired in step S612 is equal to or less than the specified value, and if the position control deviation E exceeds the specified value, returns to step S601 and repeats the adjustment. If the position control deviation E is equal to or less than the specified value, the adjustment can be completed.

According to the present embodiment, when the state of the controlled object including the stage 804 or the disturbance changes, the change can be dealt with by adjusting the parameter value of the corrector 303. For example, in the example of FIG. 13, the position control deviation shown by the dotted line is reduced to the position control deviation shown by the solid line, and the control accuracy is improved.

In the example of the equation (6), the number of parameters of the corrector 303 is only 3, which is much smaller than the number of parameters of a general neural network. For example, when a deep neural network is used, if the number of dimensions of the input layer is 5, the number of dimensions of the hidden layer is 32, and the number of dimensions of the output layer is 8, the number of parameters is 1545. The adjustment can be completed in a shorter time by adjusting the parameter values of the corrector 303 than by determining the values of these 1545 parameters by re-learning. Therefore, the control accuracy can be maintained without reducing the productivity of the stage control device 800.

<Third Embodiment>
In this embodiment, an example of applying the control system SS described in the first embodiment to the exposure apparatus EXP will be described. Matters not mentioned as this embodiment follow the first embodiment. FIG. 15 schematically shows a configuration example of the exposure apparatus EXP of the present embodiment. The exposure apparatus EXP may be configured as a scanning exposure apparatus.

The exposure apparatus EXP may include, for example, an illumination light source 1000, an illumination optical system 1001, a mask stage 1003, a projection optical system 1004, and a plate stage 1006. The illumination light source 1000 may include, but is not limited to, a mercury lamp, an excimer laser light source or an EUV light source. The exposure light 1010 from the illumination light source 1000 is formed by the illumination optical system 1001 into the shape of the irradiation region of the projection optical system 1004 with uniform illuminance. In one example, the exposure light 1010 can be formed into a rectangle long in the X direction, which is an axis perpendicular to the plane by the Y-axis and the Z-axis. Depending on the type of projection optical system 1004, the exposure light 1010 may be formed into an arc shape. The molded exposure light 1010 is applied to the pattern of the mask (original plate) 1002, and the exposure light 1010 that has passed through the pattern of the mask 1002 is applied to the surface of the plate 1005 (substrate) via the projection optical system 1004 of the pattern of the mask 1002. Form an image.

The mask 1002 is held by the mask stage 1003 by vacuum suction or the like. The plate 1005 is held by the chuck 1007 of the plate stage 1006 by vacuum suction or the like. The positions of the mask stage 1003 and the plate stage 1006 can be controlled by a multi-axis position control device including a position sensor 1030 such as a laser interferometer or a laser scale, a drive system 1031 such as a linear motor, and a controller 1032. The position measurement value output from the position sensor 1030 can be provided to the controller 1032. The controller 1032 drives the mask stage 1003 and the plate stage 1006 by generating a control signal based on the position control deviation which is the difference between the position target value and the position measurement value and providing the control signal to the drive system 1031. .. By scanning and exposing the plate 1005 while synchronously driving the mask stage 1003 and the plate stage 1006 in the Y direction, the pattern of the mask 1002 is transferred to the plate 1005 (the upper photosensitive material).

A case where the second embodiment is applied to the control of the plate stage 1006 will be described. In FIG. 11, the control board 801 corresponds to the controller 1032, the current driver 802 and the motor 803 correspond to the drive system 1031, the stage 804 corresponds to the plate stage 1006, and the sensor 805 corresponds to the position sensor 1030. By applying a controller having a neural network to the control of the plate stage 1006, the position control deviation of the plate stage 1006 can be reduced. This makes it possible to improve the superposition accuracy and the like. Neural network parameter values can be determined by a predetermined learning sequence. However, when the state of the controlled object changes or the disturbance environment changes from the time of learning, the control accuracy of the plate stage 1006 deteriorates. Even in such a case, by adjusting the parameter value of the corrector, the adjustment can be completed in a shorter time than re-learning the neural network. As a result, the control accuracy can be maintained without reducing the productivity of the exposure apparatus.

A case where the second embodiment is applied to the control of the mask stage 1003 will be described. In FIG. 11, the control board 801 corresponds to the controller 1032, the current driver 802 and the motor 803 correspond to the drive system 1031, the stage 804 corresponds to the mask stage 1003, and the sensor 805 corresponds to the position sensor 1030.

Even when the second embodiment is applied to the control of the mask stage 1003, the position control deviation of the mask stage 1003 can be reduced. This makes it possible to improve the superposition accuracy and the like. Neural network parameter values can be determined by a predetermined learning sequence. However, when the state of the controlled object changes or the disturbance environment changes from the time of learning, the control accuracy of the mask stage 1003 deteriorates. Even in such a case, by adjusting the parameter value of the corrector, the adjustment can be completed in a shorter time than re-learning the neural network. As a result, the control accuracy can be maintained without reducing the productivity of the exposure apparatus.

The second embodiment can be applied not only to the control of the stage in the exposure apparatus but also to the control of the stage in other lithography apparatus such as an imprint apparatus and an electron beam lithography apparatus. The first embodiment or the second embodiment can also be applied to, for example, control of a movable portion in a transport mechanism for transporting an article, for example, a hand for holding the article.

<Embodiment of manufacturing method of article>
The method for manufacturing an article according to an embodiment of the present invention is suitable for manufacturing, for example, a flat panel display (FPD). The method for manufacturing an article of the present embodiment includes a step of forming a latent image pattern on a photosensitive agent applied on a substrate (a step of exposing a substrate) using the above-mentioned exposure apparatus, and a step of forming a latent image pattern in such a step. It includes a step of developing the processed substrate. Further, such a manufacturing method includes other well-known steps (oxidation, film formation, vapor deposition, doping, flattening, etching, resist peeling, dicing, bonding, packaging, etc.). The method for manufacturing an article of the present embodiment is advantageous in at least one of the performance, quality, productivity, and production cost of the article as compared with the conventional method.

Although the preferred embodiments of the present invention have been described above, it goes without saying that the present invention is not limited to these embodiments, and various modifications and changes can be made within the scope of the gist thereof.

100 Control device 103 Control target 301 1st compensator 302 2nd compensator 303 Compensator 303a 1st adjustment unit 303b 2nd adjustment unit 306 Computer S1 1st signal S2 2nd signal

Claims (23)

A control device that generates a control signal to control a controlled object.
A first compensator that generates a first signal based on the control deviation of the controlled object,
A corrector that corrects the control deviation by using one of the plurality of adjustment units that generate a correction signal by correcting the control deviation according to an arithmetic expression that can adjust the coefficient.
A second compensator that generates a second signal by a neural network based on the correction signal,
An arithmetic unit that generates the control signal based on the first signal and the second signal,
A control device characterized by comprising. The control device according to claim 1, wherein in the corrector, an adjustment unit used for correcting the control deviation is selected from the plurality of adjustment units based on the control state of the control target. The feature is that the adjustment unit used for correcting the control deviation is selected from the plurality of adjustment units based on whether or not the control target and a control target different from the control target are controlled in synchronization. The control device according to claim 1 or 2. The control device according to claim 1 or 2, wherein an adjusting unit used for correcting the control deviation is selected from the plurality of adjusting units based on whether or not the gain of the controlled object is switched. .. The control device according to claim 1 or 2, wherein an adjusting unit used for correcting the control deviation is selected from the plurality of adjusting units based on the state of the first compensator. The control device according to claim 1 or 2, wherein an adjusting unit used for correcting the control deviation is selected from the plurality of adjusting units based on whether or not the operation pattern of the controlled object has changed. .. The control device according to any one of claims 1 to 6, wherein the arithmetic expression includes a term proportional to the control deviation. The control device according to any one of claims 1 to 7, wherein the arithmetic expression includes a term for integrating the control deviation. The control device according to any one of claims 1 to 8, wherein the arithmetic expression includes a term for differentiating the control deviation. The control according to any one of claims 1 to 6, wherein the arithmetic expression includes at least one of a term proportional to the control deviation, a term for integrating, and a term for differentiating. Device. Further provided with a setting unit for adjusting the corrector
The control device according to any one of claims 1 to 10, wherein the setting unit selects an adjustment unit used for correcting the control deviation from the plurality of adjustment units. The control device according to claim 11, wherein the setting unit sets the arithmetic expression. The control device according to claim 11 or 12, wherein the setting unit resets the coefficient of the calculation formula when the control deviation satisfies a predetermined condition. The control device according to claim 13, wherein the predetermined condition includes the control deviation exceeding a specified value. The control device according to any one of claims 1 to 14, further comprising a learning unit that determines a parameter value of the neural network by machine learning. The control signal is a signal obtained by correcting the first signal based on the second signal.
It is the result of controlling the control target based on the control signal as compared with the difference between the control amount which is the result of controlling the control target based on the first signal and the target value for controlling the control target. The control device according to any one of claims 1 to 15, wherein the difference between the controlled amount and the target value for controlling the controlled object is smaller. A control device that generates a control signal to control a controlled object.
A first compensator that generates a first signal based on the control deviation of the controlled object,
A corrector that corrects the control deviation using one of the plurality of adjustment units that generate a correction signal by correcting the control deviation according to an arithmetic expression.
A second compensator that generates a second signal by a neural network based on the correction signal,
An arithmetic unit that generates the control signal based on the first signal and the second signal,
A control device characterized by comprising. The control device according to claim 17, wherein in the corrector, an adjustment unit used for correcting the control deviation is selected from the plurality of adjustment units based on the control state of the control target. The control device according to claim 17 or 18, wherein the arithmetic expression includes at least one of a term proportional to the control deviation, a term for performing integration, and a term for performing differentiation. A stage control device that controls a stage that holds an object in order to control the position of the object.
A stage control device comprising the control device according to any one of claims 1 to 19. A lithography device that transfers the pattern of the original plate to the substrate.
The lithography apparatus according to any one of claims 1 to 19, which is configured to control the position of the substrate or the original plate. A transfer step of transferring the pattern of the original plate to the substrate using the lithography apparatus according to claim 21.
Including a processing step of processing the substrate that has undergone the transfer step,
A method for manufacturing an article, which comprises obtaining an article from the substrate that has undergone the processing step. The control is performed using one of a first compensator that generates a first signal based on the control deviation of the controlled object and a plurality of adjusting units that generate a correction signal by correcting the control deviation. A corrector that corrects the deviation, a second compensator that generates a second signal by a neural network based on the correction signal, and an operation that generates the control signal based on the first signal and the second signal. It is an adjustment method that adjusts the control device equipped with the device.
The corrector includes an adjustment step of selecting an adjustment unit to be used for correcting the control deviation from the plurality of adjustment units based on the control state of the control target.
The adjustment method characterized by that.

Images