TW202223562A - Control device, adjustment method, lithography device, and method for manufacturing article - Google Patents

Abstract

The present invention is a control device that generates a control signal for controlling a control object, said control device comprising: a first compensator that generates a first signal on the basis of the control deviation of the control object; a corrector that corrects the control deviation using one adjustment unit among a plurality of adjustment units that generate a correction signal by correcting the control deviation according to a calculation formula with which a coefficient can be adjusted; a second compensator that generates a second signal using a neural network, on the basis of the correction signal; and a calculation unit that generates the control signal on the basis of the first signal and the second signal.

Description

Control device, adjustment method, lithography device, and manufacturing method of articles

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 in the manufacture of devices such as semiconductor devices and flat panel displays (FPDs), an exposure device that transcribes a mask pattern to a substrate is used. In an exposure apparatus, for example, in order to align the mask and the substrate, position control and synchronization control of the mask stage holding the mask and the substrate stage holding the substrate need to be performed with high precision.

The requirements for the accuracy required for the above-described position control and synchronization control of the stage and the like have become stricter with the progress of high-definition devices, and the required accuracy may not be achieved with conventional feedback control alone. Therefore, in addition to the conventional controllers, efforts have been made to construct neural network controllers in parallel (Patent Document 1). In addition, a method of switching the neural network controller according to the state of the control object and performing compensation according to the control object has been proposed (Patent Document 2). [Prior Art Literature] [Patent Literature]

[Patent Document 1] Japanese Patent Publication No. Hei 7-503563 [Patent Document 2] Japanese Patent Application Laid-Open No. 7-277286

[The problem to be solved by the invention]

However, if a plurality of neural network controllers are formed, an improvement in accuracy can be expected, but the control calculation time increases. In addition, in the neural network controller, the parameters are adjusted by machine learning, and it takes a lot of time to learn the parameters of a plurality of neural networks. Furthermore, when a change in the state of the control object or a change in the external disturbance environment occurs, the parameters of the neural network determined in advance become non-optimal, and a large amount of time is required to readjust the parameters.

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

In order to achieve the above object, a control device as an aspect of the present invention is a control device that generates a control signal for controlling a control object, and the control device includes: a first compensator based on a control deviation of the control object to generate a first signal; a calibrator for correcting the control deviation by using one adjustment part among a plurality of adjustment parts for generating a correction signal by correcting the control deviation according to an arithmetic formula of an adjustable coefficient; a second compensator , which generates a second signal through a neural network based on the calibration signal; and a calculator, which generates the control signal based on the first signal and the second signal.

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

Hereinafter, preferred embodiments of the present invention will be described in detail based on the drawings. In addition, in each figure, the same reference numerals are attached to the same members, and overlapping descriptions are omitted.

<First Embodiment> In FIG. 1, the structure of the system SS in this embodiment is shown. The system SS can be applied, for example, to a manufacturing apparatus for manufacturing articles. The manufacturing apparatus includes, for example, a processing apparatus that processes an article or a member constituting a part of the article. The processing apparatus can be, for example, any of the following: a lithography apparatus that transfers a pattern to a material or member; a film forming apparatus that forms a film on a material or member; an apparatus that etches a material or member; A heating device for heating components.

The system SS includes, for example, a sequence unit 101 , a control device 100 , and a control object 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 object 103 . When the system SS is applied to a production system, a production sequence is provided to the sequence unit 101 . The production sequence defines the sequence used for production. The sequence part 101 generates a target value R for controlling the control object 103 based on the production sequence, and supplies the target value R to the control device 100 or the controller 102 .

The control device 100 or the controller 102 performs feedback control on the control object 103 . Specifically, the control device 100 controls the controlled object 103 so that the control amount CV of the controlled object 103 follows the target value R based on the control deviation, which is the target value R supplied from the sequence unit 101 and the controlled object 103 Provides the difference of the control quantity CV. The control object 103 may have a sensor for detecting the control amount CV, and the control amount CV detected by the sensor may be provided to the controller 102 . The target value R, the control signal MV, and the control amount CV may be time-series data whose values vary over time.

As exemplified in FIG. 2 , a learning unit 201 may also 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 disconnected from the control device 100 after the learning is completed. The learning unit 201 is configured to send a learning sequence prepared in advance to the sequence unit 101 . The sequence unit 101 generates the target value R according to the learning sequence and supplies it to the controller 102 .

The controller 102 generates the control signal MV based on the control deviation which is the difference between the target value R generated and supplied from the sequence unit 101 according to the learning sequence and the control amount CV supplied from the controlled object 103 . Here, the controller 102 has a neural network, and uses the neural network to generate the control signal MV. The control signal MV generated by the controller 102 is provided to the control object 103, and the control object 103 acts according to the control signal MV. The control amount CV as a result of this action is supplied to the controller 102 . The controller 102 supplies the operation history of the operation history of the controller 102 based on the target value R to the learning unit 201 . The learning unit 201 determines the parameter value of the neural network of the controller 102 based on the motion history, and sets the parameter value in the neural network. This parameter value is determined by machine learning such as reinforcement learning, for example.

FIG. 3 is a diagram illustrating an example of the configuration of the controller 102 . The controller 102 includes a first compensator 301 that generates a first signal S1 based on the control deviation E, and a corrector 303 that calculates the control deviation E according to an arithmetic formula of an adjustable coefficient to generate a correction signal CS. In addition, the controller 102 includes a second compensator 302 for generating the second signal S2 through the neural network based on the correction signal CS, and an calculator 306 for generating the control signal MV based on the first signal S1 and the second signal S2.

The corrector 303 includes a plurality of correctors including a first adjusting unit 303a and a second adjusting unit 303b, and the adjusting unit (connected adjusting unit) to be used according to the control state can be selected. The control signal MV is the sum of the first signal S1 and the second signal S2, and the calculator 306 can be configured as an adder. In addition, 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 the difference between the target value R and the control variable CV. The control amount CV is acquired by measuring through a sensor or the like, which is not shown, provided in the control object 103 . In addition, compared with the difference between the control amount and the target value R as a result of controlling the control object 103 based on the first signal S1, the difference between the control amount and the target value R as a result of controlling the control object 103 based on the control signal MV smaller.

The controller 102 further includes an action history recording unit 304 . The learning unit 201 in FIG. 2 is configured to perform learning for determining parameter values of the neural network of the second compensator 302 in FIG. 3 . In order to pass the learning of the learning part 201 , the motion history recording part 304 records the motion histories necessary for the learning of the learning part 201 , and supplies the recorded motion histories to the learning part 201 . The action history is, for example, the correction signal CS for the input data of the second compensator 302 and the second signal S2 for the output data of the second compensator 302 , or the control deviation and the output data of the second compensator 302 . The second signal S2 may also be other data. The first adjustment unit 303a and the second adjustment unit 303b can perform learning using arbitrary parameters as initial values.

In the following Embodiments 1 to 5, a configuration example of the corrector 303 will be described. In Embodiments 1 to 5, examples of the calculation formula used by the corrector 303 to generate the correction signal CS based on the control deviation E are shown. The arithmetic expression may be, for example, a monomial or a polynomial.

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

(Example 2) In Example 2, the first adjustment part 303a and the second adjustment part 303b have control characteristics represented by the following formula (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 K _i . In addition, integration can be performed multiple times. The integral can be a definite integral over a time interval or an indefinite integral.

(Example 3) In Example 3, the 1st adjustment part 303a and the 2nd adjustment part 303b have the control characteristic represented by the following formula (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) is K _d . In addition, differentiation can be performed a complex number of times.

(Example 4) In Example 4, the first adjustment part 303a and the second adjustment part 303b have control characteristics represented by the following formula (4). Here, the input (E) to the corrector 303 is x, the output (CS) of the corrector 303 is y, and arbitrary coefficients (constants) are K _p , K _i , and K _d . In addition, integration and differentiation can be performed multiple times.

(Example 5) In Example 5, the 1st adjustment part 303a and the 2nd adjustment part 303b have the control characteristic represented by the following formula (5). Here, let the input (E) to the corrector 303 be x, the output (CS) of the corrector 303 be y, the integration order of the multi-integration is n, the differentiation order is m, and an arbitrary coefficient ( A constant) is K _p , an arbitrary coefficient (constant) at the time of n multiple integration is K _{i — n} , and an arbitrary constant at the time of m-order differentiation is K _{d — m} .

Embodiments 1 to 5 can be understood as examples in which the arithmetic expression used by the corrector 303 to generate the correction signal CS includes at least one of a term proportional to the control deviation E, an integral term, and a differential term.

The coefficients (constants) K _p , K _i , K _d , K _{i — n} , and K _{d — m} of the arithmetic expressions described in Embodiments 1 to 5 are examples of parameters that can be adjusted by the corrector 303 . The first adjustment unit 303a and the second adjustment unit 303b determine in advance the optimum parameters using any one of the first to fifth embodiments in accordance with a change in the control state assumed in advance. The control state is, for example, switching of synchronous control, switching of controllers, switching of operation modes, changes in the environment such as temperature, noise, and floor vibration, and changes in external disturbances. By selecting the first adjustment unit 303a and the second adjustment unit 303b in accordance with the control state, optimum control characteristics can be obtained. Since the adjustment time by forming a plurality of correctors is shorter than the adjustment time by forming a plurality of neural networks, it is advantageous from the viewpoint of shortening the time.

In addition, when the state of the control object 103 and the external disturbance environment change during the operation of the system SS, the values (parameter values) of the (coefficients) of the arithmetic expressions exemplified in Embodiments 1 to 5 can be adjusted so as to respond to the change. Variety. The time required for adjustment of the value of the calculation formula (coefficient of) of the corrector 303 is shorter than the time required for relearning of 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, the tolerance to the state change of the control object 103 and the change of the external disturbance environment can be improved.

(Example 6) In Examples 6 to 8, the relationship between the change of the control state and the switching of the adjustment unit used in the corrector 303 will be described.

FIG. 4 is a diagram illustrating 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 between: individually controlling a plurality of control objects including the control object 103 a and the control object 103 b , or synchronizing the control. The control state in the sixth embodiment is a state that is determined by individually controlling or synchronizing a plurality of control objects, and switching of an appropriate adjustment unit is performed according to whether the synchronization control is performed on a plurality of control objects. In addition, in the sixth embodiment, the corrector 303 is switched according to the state of the synchronous control switching unit 402, and the first adjustment unit 303a or the second adjustment unit 303b can be selectively used. In the sixth embodiment, a description will be given of a synchronous control called a master-slave method in which the axis that controls the control object 103a is the master axis and the axis that controls the control object 103b is the slave axis. .

The controller 102 acquires the individual control variables CVa and CVb of the control objects 103a and 103b measured by the sensor (not shown) included in the control object 103, and calculates the difference from the individual target values Ra and Rb as the control deviation Ea, respectively. , Eb.

The control deviation Ea is input to the controller 301a. The input to the controller 301b and the input to the corrector 303 provided in the preceding stage of the neural network 302 formed in parallel with the controller 301b can be switched depending on whether the control object 103a and the control object 103b are controlled synchronously . As the input to the corrector 303, the control deviation Eb or the synchronization of the difference between the control deviation Eb and the control deviation Ea can be selected by the synchronous control switching unit 402 which switches the synchronous control of the control object 103a and the control object 103b. Deviation Ec. The output of the corrector 303 can be selected to use the first adjustment part 303a or the second adjustment part 303b according to the state of the synchronization control switching part 402 .

In the specific example of the sixth embodiment, when applied to an exposure apparatus, for example, different adjustment parts can be selected when synchronizing the platen stage and the mask stage and when performing other operations. At this time, the parameters of the first adjustment unit 303a and the second adjustment unit 303b are optimized respectively when the platen stage and the mask stage are synchronized, and when 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 the control signal MVa. The output of the compensator 301b and the output of the neural network 302 are added to serve as the control signal MVb. The controller 102 outputs the control signals MVa and MVb to the control objects 103a and 103b, respectively.

The first adjustment unit 303a and the second adjustment unit 303b predetermine optimal parameters using any one of the first to fifth embodiments in accordance with the state of the synchronization control switching unit 402 . By selecting the first adjustment unit 303a and the second adjustment unit 303b according to the state of the synchronous control switching unit 402, optimum control characteristics can be obtained.

The corrector 303 selects an optimum adjustment unit from the plurality of adjustment units, and the increase in adjustment time due to the configuration of the adjustment unit is shorter than the increase in adjustment time due to the configuration of the plurality of neural networks. In addition, when the state of the control object 103 and the external disturbance environment change during operation using any of the first to fifth embodiments, the parameters of the first to fifth embodiments can be adjusted to cope with the change. The time required for the adjustment by the first adjustment unit 303a and the second adjustment unit 303b is shorter than the time required for the relearning of the neural network. In Embodiment 6, even when a plurality of compensations are performed in accordance with the state of the control object and the external disturbance environment, the increase in the calculation time and the learning time can be suppressed, even if the state of the control object and the external disturbance environment change. , the appropriate control characteristics can still 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 control object 103, and the compensator 301a or the compensator 301b can be selected. The control state in Embodiment 7 is a state determined according to which compensator is used among the plurality of compensators, and the switching of the adjustment part is performed according to whether the compensator 301a or the compensator 301b is used. In addition, in the seventh embodiment, the corrector 303 is switched according to the state of the compensator 301, and the first adjustment part 303a or the second adjustment part 303b can be selected to be used.

The specific example of Embodiment 7, for example, when applied to an exposure apparatus, can also be applied to the switching of the gain such that the compensator 301a is used in the exposure operation of the platen and the compensator 301b is used in the plate conveyance operation. . That is, based on whether or not the gain of the control object 103 is switched, the adjustment unit used for the correction of the control deviation E may be selected from a plurality of adjustment units. At this time, the first adjustment unit 303a and the second adjustment unit 303b predetermine the optimal parameters using any of the first to fifth embodiments for the compensator 301a and the compensator 301b. By selecting the first adjustment part 303a and the second adjustment part 303b according to the state of the compensator 301, optimum control characteristics can be obtained.

The corrector 303 selects an optimum adjustment unit from the plurality of adjustment units, and the increase in adjustment time due to the configuration of the adjustment unit is shorter than the increase in adjustment time due to the configuration of the plurality of neural networks. In addition, when the state of the control object 103 and the external disturbance environment change during operation using any of the first to fifth embodiments, the parameters of the first to fifth embodiments can be adjusted to cope with the change. The time required for the adjustment by the first adjustment unit 303a and the second adjustment unit 303b is shorter than the time required for the relearning of the neural network. In Embodiment 7, even when a plurality of compensations are performed in accordance with the state of the control object and the external disturbance environment, the increase in the calculation time and the learning time can be suppressed, even if the state of the control object changes and the external disturbance environment changes. , the appropriate control characteristics can still 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 that depends on whether the action mode 403 of the control object changes. A specific example of the operation mode 403 will be described later. In the eighth embodiment, the switching of the adjustment part is performed according to the state of the operation mode 403, and the first adjustment part 303a or the second adjustment part 303b can be selected to be used.

The specific example of Embodiment 7 can be applied, for example, by switching between an acceleration section during driving of the stage and an operation mode other than that when applied to a stage apparatus used in an exposure apparatus or the like. At this time, the first adjustment unit 303a and the second adjustment unit 303b determine in advance the optimum parameters using any one of the first to fifth embodiments in accordance with the state of the operation mode 403 of the control object 103 . By selecting the first adjustment unit 303a and the second adjustment unit 303b in accordance with the state of the operation mode 403, optimum control characteristics can be obtained.

The corrector 303 selects an optimum adjustment unit from the plurality of adjustment units, and the increase in adjustment time due to the configuration of the adjustment unit is shorter than the increase in adjustment time due to the configuration of the plurality of neural networks. In addition, when the state of the control object 103 and the external disturbance environment change during operation using any of the first to fifth embodiments, the parameters of the first to fifth embodiments can be adjusted to cope with the change. The time required for the adjustment by the first adjustment unit 303a and the second adjustment unit 303b is shorter than the time required for the relearning of the neural network. In the eighth embodiment, even when a plurality of compensations are performed in accordance with the state of the control object and the external disturbance environment, the increase in the calculation time and the learning time can be suppressed, even if the state of the control object changes and the external disturbance environment changes. , the appropriate control characteristics can still be adjusted in a short time.

As shown in FIG. 7 as an example, the control device 100 may include a setting unit 202 that selects to use the first adjustment unit 303a or to use the second adjustment unit. In addition, the setting unit 202 may also have a role of setting parameter values of the corrector 303 .

The setting unit 202 executes adjustment processing for switching the adjustment unit and adjusting the parameter value. Through the adjustment processing, the switching of the adjustment unit and the parameter value can be determined and set, and the switching of the adjustment unit and the parameter value can also be set based on an instruction from the user. value. In the former aspect, 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 the confirmation sequence. Then, the setting unit 202 may acquire an operation history (eg, control deviation) from the controller 102 that operates based on the target value R, and determine the necessity of switching the corrector 303 and parameter values based on the operation history. The setting unit 202 having such a function can be understood as an adjustment unit that switches the corrector 303 and adjusts parameter values.

The setting unit 202 may acquire an operation history (for example, control deviation) from the controller 102 when the sequence unit 101 generates production of the target value R based on the production sequence, and determine whether or not to adjust the parameter value of the corrector 303 based on the operation history. . Alternatively, a determination unit may be provided separately from the setting unit 202 for determining whether to perform adjustment of the parameter values by the corrector 303 of the setting unit 202 when the sequence unit 101 generates production of the target value R based on the production sequence.

Next, an example of production 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 a production apparatus.

In program S501 , the sequence unit 101 generates the target value R based on the supplied production sequence, and supplies it to the control device 100 or the controller 102 . The control device 100 or the controller 102 controls the control object 103 based on the target value R.

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

In routine S503, the setting unit 202 can determine whether to perform adjustment of the corrector 303 such as switching of the adjustment unit and adjustment (or readjustment) of parameter values, based on the operation history acquired in the routine S502. For example, the setting unit 202 may determine that the adjustment (or readjustment) of the switching parameter value of the adjustment unit is executed when the operation history meets a predetermined condition. The predetermined condition is a condition in which production should be stopped. For example, when the control deviation acquired as the operation history exceeds a predetermined value, it can be determined that the adjustment of the corrector 303 is necessary. Then, when the adjustment through the corrector 303 of the setting unit 202 is performed, the process proceeds to a process S504, and when not, the process proceeds to a process S505.

In routine S504, the setting unit 202 performs adjustment of the corrector 303. This adjustment is performed while the parameter value of the second compensator 302 is maintained in the previous state, and through this adjustment, for example, the parameter value (coefficient) of the corrector 303 is reset.

In program S505, the sequence unit 101 determines whether or not the production in accordance with the production sequence is terminated, and if not terminated, the process returns to program S501, and if terminated, the production is terminated. According to the above process, even when the production should be stopped, the parameter value of the corrector 303 can be adjusted quickly, and the production can be restarted with the production interruption kept to a minimum.

In program S504 , the setting unit 202 may send the confirmation sequence to the sequence unit 101 , so that the sequence unit 101 executes the confirmation sequence, and obtains the operation history (eg, control deviation) of the confirmation sequence from the controller 102 . Then, the setting unit 202 can perform frequency analysis of the operation history, determine the frequency to be improved based on the result, and determine the parameter value of the corrector 303 so that the control deviation of the frequency is within a predetermined value. A more specific example of the procedure S504 will be described in the second embodiment.

In FIG. 9, it is a figure which shows the measurement result of the external 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 called external disturbance suppression characteristic. In FIG. 9 , the horizontal axis represents the frequency, and the vertical axis represents the gain of the external disturbance suppression characteristic. The external disturbance suppression characteristic shows the frequency response of the control deviation E when the external disturbance is added to the control signal MV, so a large gain means a low effect of suppressing the external disturbance. On the other hand, a small gain means that the effect of suppressing external disturbance is high. In FIG. 9 , the broken line represents the external disturbance suppression characteristic before adjustment, and the solid line represents the external disturbance suppression characteristic after adjustment.

When the routine S504 is executed with the frequency indicated by the dashed-dotted line in FIG. 9 as the frequency of the external disturbance suppression characteristic to be improved, the external disturbance suppression characteristic indicated by the solid line, for example, can be obtained. In terms of the frequency to be improved, it can be seen that the gain of the external disturbance suppression characteristic is reduced, and the external disturbance suppression characteristic is improved. In Embodiments 1 to 8, when the parameter adjustment of the corrector 303 provided in the preceding stage of the neural network is performed, the parameter adjustment can be performed using the external disturbance suppression characteristic shown in FIG. 9 as an index.

<Second Embodiment> In this embodiment, an example in which the control system SS described in the first embodiment is applied to the stage control device 800 will be described. Matters not mentioned in this embodiment conform to 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 apparatus 800 .

The stage control apparatus 800 is configured to control the position of the object such as the substrate by controlling the stage 804 in a state in which the object is held on the stage 804 . 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 control object 103 in the system SS of the first embodiment. Among them, the current driver 802 may be incorporated into the control substrate 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 .

The control board 801 is supplied with a position target value as a target value from the sequence unit 101 . The control substrate 801 can generate a current command as a control signal based on the position target value supplied from the sequencer 101 and the position information supplied from the sensor 805 , and supply the current command to the current driver 802 . In addition, the control board 801 can supply the action history to the sequence unit 101 .

The current driver 802 can supply the current according to the current command to the motor 803 . 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 can hold objects such as a flat plate or a mask, for example. The sensor 805 can detect the position of the stage 804 and supply the position information obtained thereby to the control substrate 801 .

In FIG. 11 , a configuration example of the control board 801 is shown by a block line diagram. The control board 801 may include a first compensator 301 for generating a first signal S1 based on a position control deviation E of the stage 804 to be controlled, and a calibrator for generating a calibration signal CS by correcting the control deviation E according to an arithmetic formula of an adjustable coefficient 303. In addition, the control substrate 801 may include a second compensator 302 for generating a second signal S2 through a neural network based on the calibration signal CS, and an calculator 306 for generating a current command as a control signal based on the first signal S1 and the second signal S2. In addition, the control substrate 801 may include a subtractor 305 that generates a control deviation E that is the difference between the position target value PR and the position information.

The stage control apparatus 100 of the second embodiment may 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 values of the neural network of the second compensator 302 . In order to pass the learning of the learning part 201 , the motion history recording part 304 may record the motion history required for the learning passed through the learning part 201 , and provide the recorded motion history to the learning part 201 . The operation history may be, for example, the correction signal CS for the input data of the second compensator 302 and the second signal S2 for the output data of the second compensator 302, or other data.

The stage control apparatus 100 of the second embodiment may include a setting unit 202 . The setting unit 202 executes an adjustment process for adjusting the parameter value of the corrector 303 , and can determine and set the parameter value of the corrector 303 through the adjustment process, or set the parameter value of the corrector 303 based on an instruction from the user.

The operation of the stage apparatus 800 in the case where the stage control apparatus 800 of the second embodiment is applied to a production apparatus will be exemplarily described with reference to FIG. 8 . In procedure S501 , the sequencer 101 may generate the position target value PR based on the supplied production sequence, and supply it to the stage control device 800 . Stage control device 800 controls the position of stage 804 based on this position target value PR.

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

In routine S503, the setting unit 202 can determine whether to perform adjustment of the corrector 303 such as switching of the adjustment unit and adjustment (or readjustment) of parameter values, based on the operation history acquired in the routine S502. For example, the setting unit 202 can determine that the adjustment of the corrector 303 is to be executed when the operation history meets a predetermined condition. The predetermined condition is a condition to stop production. For example, when the maximum value of the position control deviation during constant speed driving of the stage 804 exceeds a predetermined value, it is determined that the adjustment of the corrector 303 is necessary. Then, when the adjustment through the corrector 303 of the setting unit 202 is performed, the process proceeds to a process S504, and when not, the process proceeds to a process S505.

In routine S504, the setting unit 202 performs adjustment of the corrector 303. In program S505, the sequence unit 101 determines whether or not the production in accordance with the production sequence is terminated, and if not terminated, the process returns to program S501, and if terminated, the production is terminated.

FIG. 12 shows a specific example of processing in the adjustment of the parameter value (or the readjustment of the parameter value) in the adjustment of the corrector 303 in the routine S504. In procedure 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 may cause the sequence unit 101 to generate the position target value PR based on the confirmation sequence. In routine S602, the setting unit 202 acquires the position control deviation E as an operation history from the controller 102 operating based on the position target value PR.

Here, the change of the position control deviation E before and after the parameter adjustment will be described with reference to FIG. 13 . FIG. 13 is a diagram illustrating the position control deviation before and after parameter adjustment. In FIG. 13 , the horizontal axis represents time, and the vertical axis represents the position control deviation E. As shown in FIG. Here, the curve indicated by the dotted line is the position control deviation E before the parameter value of the corrector 303 is adjusted, 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 routine S603, the setting unit 202 can perform frequency analysis of the position control deviation E acquired in routine S602. Here, the results of frequency analysis before and after parameter adjustment will be described with reference to FIG. 14 . FIG. 14 is a diagram illustrating the results of frequency analysis before and after parameter adjustment. In FIG. 14 , the horizontal axis is the frequency, and the vertical axis is the power spectrum. The dotted line indicates the frequency at which the maximum spectrum is displayed before adjustment. In step S604, the setting unit 202 may determine, for example, the frequency at which the maximum spectrum is displayed in the power spectrum as the frequency to be improved.

Routines S605 to S610 are specific examples of the adjustment processing of the parameter value of the adjustment corrector 303 . Here, an example of using the steepest descent method as the adjustment method of the parameter value will be described, but other methods may be used. In routine S605, the setting unit 202 initializes n to 1. For example, when the calculation formula of the corrector 303 is composed of three terms of a first-order integral term, a proportional term, and a first-order differential term, the parameters whose parameter values should be adjusted are three of K _i , K _p , and K _d . The parameter value pn of the _n -th adjustment is represented by the following numeral 6.

In program S606, the setting unit 202 may set an arbitrary initial value for the parameter value p1 in the _first adjustment of the parameter value _pn . In the n-th adjustment, the parameter value _pn represented by the equation (8) to be described later can be set.

The objective function J( _pn ) for adjusting the parameter value _pn can be determined as, for example, the gain of the disturbance suppression characteristic outside the frequency determined in the procedure S604. In step S607, the setting unit 202 may measure the gradient vector grad J( _pn ) of the objective function J( _pn ). The gradient vector grad J( _pn ) can be given by the following equation (7). The gradient vector grad J( _pn ) can be measured by changing each element K _in , K _pn , and K _dn constituting the parameter value _pn by a small amount.

In step S608 , the setting unit 202 may determine whether or not the value of each element of the gradient vector grad J( _pn ) is equal to or less than a predetermined value as a convergence determination of the steepest descent method. When the value of each element of the gradient vector grad J( _pn ) is equal to or less than the predetermined value, the setting unit 202 may end the adjustment of the parameter value of the corrector 303 . On the other hand, when the value of each element of the gradient vector grad J( _pn ) exceeds the predetermined value, the setting unit 202 may calculate the parameter value _{pn + 1} in the procedure S609. Here, the parameter value _{pn + 1} can be calculated according to the following formula (8) using, for example, an arbitrary constant α larger than 0. In routine S610, the setting unit 202 adds 1 to the value of n, and returns to routine S606.

In program 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 the sequence unit 101 may generate the position target value PR based on the confirmation sequence. In routine S612, the setting unit 202 acquires the position control deviation E as an operation history from the controller 102 operating based on the position target value PR.

In routine S613, the setting unit 202 determines whether the position control deviation E acquired in the routine S612 is equal to or less than a predetermined value. If the position control deviation E exceeds the predetermined value, the process returns to the routine S601 to perform adjustment again. When the position control deviation E is equal to or less than the predetermined value, the End the adjustment.

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

In the example of formula (6), the number of parameters of the corrector 303 is only three, which is far less than the number of parameters of a general neural network. For example, when using a deep neural network, when the dimension of the input layer is 5, the dimension of the hidden layer is 32, and the dimension of the output layer is 8, the number of parameters is 1545. Compared with determining the values of these 1545 parameters through relearning, adjusting the parameter values of the corrector 303 can complete the adjustment in a shorter time. Therefore, the control accuracy can be maintained without reducing the productivity of the stage control apparatus 800 .

<Third Embodiment> In the present embodiment, an example in which the control system SS described in the first embodiment is applied to the exposure apparatus EXP will be described. Matters not mentioned in this embodiment conform to the first embodiment. In FIG. 15, the structural example of the exposure apparatus EXP of this embodiment is shown typically. 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 flat plate stage 1006 . The illumination light source 1000 may include a mercury lamp, an excimer laser light source or an EUV light source, but is not limited thereto. The exposure light 1010 from the illumination light source 1000 passes through the illumination optical system 1001 and is shaped into the shape of the irradiation area of the projection optical system 1004 with uniform illuminance. In one example, the exposure light 1010 may be shaped into a rectangle that is long in the X direction, which is an axis perpendicular to a plane based on the Y axis and the Z axis. Depending on the type of the projection optical system 1004, the exposure light 1010 may be shaped into an arc shape. The shaped exposure light 1010 is irradiated on the pattern of the mask (original) 1002, and the exposure light 1010 passing through the pattern of the mask 1002 forms an image of the pattern of the mask 1002 on the surface of the flat plate 1005 (substrate) via the projection optical system 1004.

The mask 1002 is held by vacuum suction or the like through the mask stage 1003 . The flat plate 1005 is held by a jig 1007 passing through the flat plate stage 1006 by vacuum suction or the like. The positions of the mask stage 1003 and the flat stage 1006 can be controlled by a position sensor 1030 including a laser interferometer, a laser scale, etc., a driving system 1031 such as a linear motor, and a multi-axis position control device of the controller 1032 . The position measurements output from the position sensor 1030 may be provided to the controller 1032 . The controller 1032 generates a control signal based on the position control deviation, which is the difference between the position target value and the position measurement value, and supplies the control signal to the drive system 1031 to drive the mask stage 1003 and the plate stage 1006 . The pattern of the mask 1002 is transferred to the flat plate 1005 (photosensitive material thereon) by scanning and exposing the flat plate 1005 while driving the mask stage 1003 and the flat plate stage 1006 in synchronization with the Y direction.

A case where the second embodiment is applied to the control of the plate stage 1006 will be described. The control board 801 in FIG. 11 corresponds to the controller 1032 , the current driver 802 and the motor 803 corresponds to the driving system 1031 , the stage 804 corresponds to the flat stage 1006 , and the sensor 805 corresponds to the position sensor 1030 . Applying a controller with a neural network to the control of the tablet stage 1006 can reduce the position control deviation of the tablet stage 1006 . Thereby, the superposition accuracy and the like can be improved. The parameter values of the neural network can be determined through a predetermined learning sequence. However, the control accuracy of the tablet stage 1006 is degraded when the state of the control object during learning changes or the external disturbance environment changes. In such a case, by adjusting the parameter values of the corrector, the adjustment can be completed in a shorter time than the re-learning of the neural network. As a result, the control accuracy can be maintained without lowering 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. The control board 801 in FIG. 11 corresponds to the controller 1032 , the current driver 802 and the motor 803 corresponds to the driving system 1031 , the stage 804 corresponds to the mask stage 1003 , and the sensor 805 corresponds to the position sensor 1030 .

When the second embodiment is applied to the control of the mask stage 1003, the position control deviation of the mask stage 1003 can also be reduced. Thereby, the superposition accuracy and the like can be improved. The parameter values of the neural network can be determined through a predetermined learning sequence. However, the control accuracy of the mask stage 1003 decreases when the state of the control object during learning changes or the external disturbance environment changes. In such a case, by adjusting the parameter values of the corrector, the adjustment can be completed in a shorter time than the re-learning of the neural network. As a result, the control accuracy can be maintained without lowering the productivity of the exposure apparatus.

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

<Embodiment of the manufacturing method of the article> The manufacturing method of the article which concerns on embodiment of this invention is suitable for manufacturing, for example, a flat panel display (FPD). The manufacturing method of the article of the present embodiment includes: a process of forming a latent image pattern on a photosensitive agent applied on a substrate using the above-mentioned exposure apparatus (a process of exposing the substrate); The process of developing the substrate. Furthermore, the manufacturing method includes other well-known procedures (oxidation, film formation, vapor deposition, doping, planarization, etching, resist stripping, dicing, bonding, packaging, etc.). The manufacturing method of the article of the present embodiment is advantageous in at least one of the performance, quality, productivity, and production cost of the article compared to the conventional method.

Although 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.

The present invention is not limited to the above-described embodiments, and various changes and modifications can be made without departing from the spirit and scope of the present invention. Therefore, the scope of the patent application has been prepared to disclose the scope of this invention.

In this case, the priority is claimed on the basis of Japanese Patent Application No. 2020-205546 filed on December 11, 2020, and the entire contents of its description are used here.

100: Controls 101: Sequence Department 102: Controller 103: Control Object 103a: Control objects 103b: Control Objects 201: Learning Department 202: Setting Department 301: 1st compensator 301a: Controller 301b: Controller 302: Neural Networks 303: Corrector 303a: Adjustment Section 1 303b: Adjustment Section 2 304: Action History Department 305: Subtractor 306: Calculator 402: Synchronous control switching section 403: Action Mode 800: Stage Controls 801: Control board 802: Current Driver 802a: Current Driver 802b: Current Driver 803: Motor 803a: Motor 803b: Motor 804: Stage 804a: Stage 804b: Stage 805: Sensor 805a: Sensors 805b: Sensor 1000: Lighting source 1001: Illumination Optical Systems 1002: Mask 1003: Masking Table 1004: Projection Optical Systems 1005: Tablet 1006: Flat Table 1007: Fixtures 1030: Position Sensor 1031: Drive System 1032: Controller CS: Calibration signal CV: control volume CVa: control amount CVb: control amount E: Control deviation Ea: control deviation Eb: Control deviation EXP: Exposure Device MV: control signal MVa: control signal MVb: control signal PR: Position target value R: target value Ra: target value Rb: target value S1: Signal 1 S2: 2nd signal SS: Control System

[ Fig. 1] Fig. 1 is a diagram illustrating a configuration example of the system in the first embodiment. [ Fig. 2] Fig. 2 is a diagram illustrating a configuration example of the system in the first embodiment. [ Fig. 3] Fig. 3 is a diagram showing a configuration example of a controller in the system of the first embodiment. [ Fig. 4] Fig. 4 is a diagram illustrating a configuration example of a controller in the sixth embodiment. [ Fig. 5] Fig. 5 is a diagram illustrating a configuration example of a controller in the seventh embodiment. [ Fig. 6] Fig. 6 is a diagram illustrating a configuration example of a controller in the eighth embodiment. [ Fig. 7] Fig. 7 is a diagram illustrating a configuration example of the system in the first embodiment. 8 is a flowchart showing an example of operation when the system of the first embodiment is applied to a production apparatus. [ Fig. 9] Fig. 9 is a graph showing an example of measurement results of external disturbance suppression characteristics. [ Fig. 10] Fig. 10 is a diagram showing a configuration example of the stage control device in the second embodiment. [ Fig. 11] Fig. 11 is a diagram showing a configuration example of a control board in the system of the second embodiment. [ FIG. 12 ] A flowchart illustrating adjustment of the corrector. [ Fig. 13 ] A diagram illustrating the position control deviation. [ Fig. 14 ] A diagram illustrating an example of the result of frequency analysis. [ Fig. 15] Fig. 15 is a diagram illustrating a configuration example of an exposure apparatus.

102: Controller

103: Control Object

301: 1st compensator

302: Neural Networks

303: Corrector

303a: Adjustment Section 1

303b: Adjustment Section 2

304: Action History Department

305: Subtractor

306: Calculator

CS: Calibration signal

CV: control volume

E: Control deviation

MV: control signal

R: target value

S1: Signal 1

S2: 2nd signal

Claims (23)

A control device that generates a control signal for controlling a control object, The aforementioned control device includes: a first compensator, which generates a first signal based on the control deviation of the control object; a calibrator, which corrects the control deviation by using one adjustment part among the plurality of adjustment parts for generating the correction signal by correcting the control deviation according to the calculation formula of the adjustable coefficient; a second compensator, which generates a second signal through a neural network based on the aforementioned correction signal; and The calculator generates the control signal based on the first signal and the second signal. The control device according to claim 1, wherein, in the corrector, an adjustment unit for correcting the control deviation is selected from the plurality of adjustment units based on the control state of the control object. The control device according to claim 1, wherein an adjustment unit for correcting the control deviation is selected from the plurality of adjustment units based on whether or not the control object is controlled in synchronization with the control object that is another control object. The control device according to claim 1, wherein an adjustment unit for correcting the control deviation is selected from the plurality of adjustment units based on whether or not the gain of the control object is switched. The control device according to claim 1, wherein an adjustment unit for correcting the control deviation is selected from the plurality of adjustment units based on the state of the first compensator. The control device according to claim 1, wherein an adjustment unit for correcting the control deviation is selected from the plurality of adjustment units based on whether or not the operation mode of the control object is changed. The control apparatus according to claim 1, wherein the arithmetic expression includes a term proportional to the control deviation. The control device according to claim 1, wherein the calculation formula includes a term for integrating the control deviation. The control device according to claim 1, wherein the calculation formula includes a term for differentiating the control deviation. The control device according to claim 1, wherein the arithmetic expression includes at least one of a term proportional to the control deviation, an integral term, and a derivative term. The control device according to claim 1, further comprising a setting unit for adjusting the corrector, The setting unit selects an adjustment unit for correcting the control deviation from the plurality of adjustment units. The control device according to claim 11, wherein the setting unit sets the calculation formula. The control device according to claim 11, wherein the setting unit re-sets 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 that the control deviation exceeds a predetermined value. The control device according to claim 1, further comprising a learning unit that determines parameter values of the neural network through machine learning. The control device of claim 1, wherein, The control signal is a signal obtained by correcting the first signal based on the second signal, Compared with the difference between the control amount as a result of controlling the control object based on the first signal and the target value for controlling the control object, the control amount as a result of controlling the control object based on the control signal is the difference between the control amount and the target value for controlling the control object based on the control signal. The difference in the target value of the control object is small. A control device that generates a control signal for controlling a control object, The aforementioned control device includes: a first compensator, which generates a first signal based on the control deviation of the control object; a calibrator, which corrects the control deviation by using one adjustment part among the plurality of adjustment parts for generating the correction signal by correcting the control deviation according to an algorithm; a second compensator, which generates a second signal through a neural network based on the aforementioned correction signal; and The calculator generates the control signal based on the first signal and the second signal. The control device according to claim 17, wherein, in the corrector, an adjustment unit for correcting the control deviation is selected from the plurality of adjustment units based on the control state of the control object. The control device of claim 17, wherein the calculation formula includes at least one of a term proportional to the control deviation, an integral term, and a derivative term. A stage control device that controls a stage holding the object in order to control the position of the object, The aforementioned stage control device includes the control device according to claim 1. A lithography device, which transfers the pattern of the original plate to a substrate, The aforementioned lithography apparatus includes the control device according to any one of claims 1 to 19 configured to control the position of the aforementioned substrate or the aforementioned master plate. A method of manufacturing an article, comprising: A transfer procedure for transferring the pattern of the master plate to a substrate using the lithography apparatus as claimed in claim 21; and a processing procedure for processing the aforementioned substrate subjected to the aforementioned transfer procedure; Articles are obtained from the aforementioned substrates subjected to the aforementioned processing procedures. An adjustment method, which adjusts a control device, wherein the control device includes: a first compensator that generates a first signal based on a control deviation of a control object; a corrector that corrects the control deviation to generate a plurality of correction signals One of the adjustment parts is used to correct the control deviation; a second compensator generates a second signal through a neural network based on the calibration signal; and an algorithm is based on the first signal and the second signal to generate a control signal; The adjustment method includes, in the aspect of the corrector, an adjustment program for selecting an adjustment unit for correcting the control deviation from the plurality of adjustment units based on the control state of the control object.

Images