Patent Application
20200272178
The present invention relates to a valve control apparatus and a vacuum valve.
In order to automatically control a valve disc opening and quickly converge a chamber pressure to a target pressure, a valve apparatus described in Patent Literature 1 (JP-A-2018-106718) is configured in such a manner as to bring the pressure close to the target pressure with a coarse adjustment in open-loop control, and further drive the pressure to the target pressure with a fine adjustment by switching to closed-loop control. Moreover, as closed-loop control in the valve apparatus, such valve opening control that performs feedback control in such a manner as to converge a pressure error signal between a target pressure Ps and a detected chamber pressure Pr to zero as described in Patent Literature 2 (JP-A-2018-112933) is known.
If closed-loop control described in Patent Literature 2 is applied in closed-loop control after the switching in Patent Literature 1, when open-loop control is performed until coming close to the target pressure, and is then switched to closed-loop control, a difference Δ=|Pr−Ps| between the pressure Pr measured and the target pressure Ps is not zero. Accordingly, the pressure error signal of closed-loop control changes discontinuously and stepwise before and after the switching. As a result, a large overshoot may occur in closed-loop control to increase the pressure regulation time.
A valve control apparatus that performs first valve opening control and second valve opening control in succession on a vacuum valve between a vacuum chamber and a vacuum pump, and controls a chamber pressure of the vacuum chamber to a target pressure, the valve control apparatus comprises: a generator configured to generate a first set pressure signal; and a valve opening controller configured to perform the first valve opening control on the basis of the target pressure and, after the first valve opening control, perform the second valve opening control performing feedback control on the basis of a difference between the first set pressure signal and the chamber pressure. The generator generates the first set pressure signal on the basis of a first set pressure locus converging from the chamber pressure at the time of switching from the first valve opening control to the second valve opening control to the target pressure with a predetermined time constant.
The time constant is calculated on the basis of the chamber pressure at the time of the switching and a time derivative value of the chamber pressure, and the target pressure.
The generator corrects the time constant of the first set pressure locus every control period on the basis of a change in the first set pressure signal over time and a change in the chamber pressure over time, and generates a second set pressure signal on the basis of a second set pressure locus converging to the target pressure with the corrected time constant, and
the valve opening controller performs third valve opening control performing feedback control, instead of the second valve opening control, on the basis of a difference between the second set pressure signal and the chamber pressure.
The corrected time constant is set to be less than the time constant before the correction upon the magnitude of the change in the first set pressure signal over time being less than the magnitude of the change in the chamber pressure over time, be greater than the time constant before the correction upon the magnitude of the change in the first set pressure signal over time being greater than the magnitude of the change in the chamber pressure over time, and be equal to the time constant before the correction upon the magnitude of the change in the first set pressure signal over time being equal to the magnitude of the change in the chamber pressure over time.
The time constant is calculated by an equation “|(Ps−Pr0)|/|(dPr/dt|0|)|” where the target pressure is Ps, the chamber pressure is Pr, and the chamber pressure at the time of the switching is Pr0.
The generator uses a time constant obtained by correcting the time constant on the basis of a control error in the first valve opening control, instead of the time constant, and generates the first set pressure signal on the basis of a set pressure locus converging to the target pressure with the corrected time constant.
A vacuum valve comprises: a disc; a disc driver configured to drive the disc to open and close the disc; and the valve control apparatus according to claim 1 configured to control the opening and closing drive by the disc driver.
According to the present invention, it is possible to quickly converge a chamber pressure to a target pressure after switching to feedback control.
Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
The vacuum valve 1 includes a valve plate 11, and a valve motor 12 that drives the valve plate 11 to open and close it. The vacuum valve 1 is an automatic pressure regulator valve that adjusts a pressure Pr of the vacuum chamber 3 measured by the vacuum gauge 31 to a target pressure Ps inputted into the valve control apparatus 2, on the basis of the target pressure Ps and the pressure Pr of the vacuum chamber 3. The target pressure Ps is inputted from, for example, a controller (not illustrated) of the vacuum apparatus 100 into the valve control apparatus 2.
The valve control apparatus 2 includes a set pressure generator 20, a valve opening controller 21, and a motor controller 25. The valve opening controller 21 includes a target valve opening estimator 22, a feedforward controller 23, and a feedback controller 24. The valve control apparatus 2 executes first valve opening control to rapidly make a coarse adjustment to the valve opening of the valve plate 11 in the first stage of valve opening control, and second valve opening control to make a fine adjustment to the valve opening with high accuracy in the second stage of the valve opening control after the first valve opening control.
In the embodiment, open-loop control based on the target pressure Ps is performed as the first valve opening control by the feedforward controller 23. Closed-loop control based on a pressure deviation ε being a difference between the pressure Pr and a set pressure Pset1 generated by the set pressure generator 20 is performed as the second valve opening control by the feedback controller 24. A valve opening setting output Δθ1 from the feedforward controller 23 and a valve opening setting output Δθ2 from the feedback controller 24 are added to be inputted as a valve opening setting output θset into the motor controller 25. The motor controller 25 controls and drives the valve motor 12 on the basis of the valve opening setting output θset.
A correlation between a valve opening θ and a pressure P is stored in a storage 221 of the target valve opening estimator 22. A computer 220 of the target valve opening estimator 22 computes and estimates an estimated target valve opening value θse on the basis of the target pressure Ps and the correlation. A method for computing and estimating the estimated target valve opening value θse is described below. The feedforward controller 23 outputs the valve opening setting output Δθ1 on the basis of the estimated target valve opening value θse computed and estimated by the target valve opening estimator 22. In this manner, in open-loop control, the estimated target valve opening value θse at which the target pressure Ps is attained is obtained separately, and the setting of the valve opening is outputted through a path decided appropriately to attain the estimated target valve opening value θse in the end. When the valve opening θr reaches the estimated target valve opening value θse in the end in open-loop control, the valve opening control switches from open-loop control to closed-loop control while the valve opening setting output Δθ1 outputted from the feedforward controller 23 is fixed at the estimated target valve opening value θse.
The set pressure generator 20 generates a set pressure used to calculate the pressure deviation ε on the basis of the target pressure Ps and the pressure Pr inputted from the vacuum gauge 31. In open-loop control in the first stage of the valve opening control, the pressure Pr is outputted as the set pressure from the set pressure generator 20. In closed-loop control, the set pressure generator 20 computes and estimates a set pressure locus on the basis of a pressure Pr0 at the timing when open-loop control is switched to closed-loop control, and the target pressure Ps, and outputs the set pressure Pset1 based on the set pressure locus. A method for calculating the set pressure Pset1 is described in detail below.
The feedback controller 24 outputs the valve opening setting output Δθ2 based on the pressure deviation ε=Pr−(the set pressure). The feedback controller 24 normally includes proportional and integral gains (what is called PI gains). As described above, in open-loop control, the pressure Pr is outputted as the set pressure from the set pressure generator 20. Accordingly, the pressure deviation ε=0 is inputted into the feedback controller 24. Δθ2=0 is outputted from the feedback controller 24. On the other hand, when the valve opening setting output Δθ1 outputted from the feedforward controller 23 is fixed at the estimated target valve opening value θse and open-loop control is switched to closed-loop control, the set pressure generator 20 outputs the set pressure Pset1. As a result, the pressure deviation ε=Pr−Pset1 is inputted into the feedback controller 24, and the feedback controller 24 outputs the valve opening setting output Δθ2 on the basis of the pressure deviation ε. The motor controller 25 controls the valve opening on the basis of the valve opening setting output θset=Δθ1 (fixed)+Δθ2.
As described above, the computer 220 of the target valve opening estimator 22 computes and estimates the estimated target valve opening value θse on the basis of the correlation between the valve opening stored in the storage 221 and the pressure of the vacuum chamber 3, and the inputted target pressure Ps. A case where a plant gain Gp defined by the following equation (1) is used as the correlation between the valve opening θ and the pressure P is described here as an example. As can be seen from equation (1), the plant gain Gp represents the pressure response characteristics to the change of the valve opening θ. |(ΔP/Δθ)| indicates the absolute value of (ΔP/Δθ).
Gp=|(ΔP/Δθ)|/P (1)
Assume that the valve opening is the input and the pressure of the vacuum chamber 3 is the output. In this case, the quantity ((ΔP/Δθ)/P) obtained by normalizing the change of the pressure for the change of the valve opening, that is, (ΔP/Δθ) by the pressure P represents a static gain characteristic of the pressure response of lean gas in the vacuum chamber 3 being a plant. The absolute value |(ΔP/Δθ)|/P of (ΔP/Δθ)/P of (ΔP/Δθ)/P is referred to as the plant gain Gp here. The plant gain Gp expressed by equation (1) can also be considered to be the quantity obtained by differentiating a logarithmic value (LogeP(θ)) of the pressure P(θ) with respect to the valve opening θ.
The plant gain Gp is a characteristic of a system including the vacuum valve 1, the vacuum chamber 3, and the vacuum pump 4. Even if the vacuum valve 1 is the same, the plant gain Gp varies slightly if the vacuum chamber 3 and the vacuum pump 4 are different.
An effective pumping speed Se of the vacuum chamber 3 is generally used to control the valve opening of the vacuum valve 1. When the vacuum valve 1 is used, mounted on the vacuum chamber 3, an initial calibration operation related to the effective pumping speed Se, that is, an initial calibration operation related to valve control is normally performed. The gain of a controller is generally calibrated in accordance with the volume of the vacuum chamber 3, the sensitivity of the valve plate 11, and the like with a typical gas of a process condition applied or under average gas conditions (a gas type and a gas flow rate). For example, an average molecular weight of mixed gas is obtained, and a gas type that can be relatively easily handled is used for the average condition in many cases.
An initial value (data table) of the plant gain Gp is stored in the storage 221. However, an initial calibration is also required for the plant gain Gp as in the known case of control using the effective pumping speed Se. The initial calibration allows obtaining a plant gain Gpi at each valve opening θi in predetermined specific gas type and flow rate value conditions. The storage 221 is caused to store the obtained data table (θi, Gpi) as the calibrated plant gain Gp.
Next, an example of the method for calculating the estimated target valve opening value θse is described. Here, the current pressure regulation state (θ, P) is expressed as (θ0, P0), and the next pressure regulation state to shift to is expressed as (θ1, P1). As described above, the data table (θ, Gp) representing the relationship between the valve opening θ and the plant gain Gp is stored in the storage 221. In other words, the plant gain Gp can be expressed as a function Gp(θ) of the valve opening θ. When the definition equation of equation (1) is used, the relationship between the valve opening change Δθ and the pressure change ΔP is expressed as in equation (2). The pressure P decreases with increasing valve opening θ. Accordingly, the minus sign is assigned to the right side of equation (2).
ΔP=−P×Gp(θ)×Δθ (2)
The relationship between the current pressure regulation state (θ0, P0) and the next pressure regulation state (θ1, P1) can be expressed by the following equations (3) and (4).
θ1=θ0+∫dθ (the integration range is from θ0 to θ1) (3)
P1=P0+∫dP (the integration range is from P0 to P1) (4)
In reality, the integrals of equations (3) and (4) are expressed as an addition of the increment Δθ of the valve opening and the increment ΔP of the pressure in each quasistatic state from the pressure regulation state (θ0, P0) to the pressure regulation state (θ1, P1). Accordingly, θ1 and P1 can be expressed as in the following equations (5) and (6). Δθ_n(θ_n) in equation (5) represents an increment Δθ_n of the valve opening at a valve opening θ_n. For example, Δθ_1(θ_1) is an increment of the valve opening when the valve opening changes from a valve opening θ0 to a valve opening θ_1. ΔP_n(p_n) represents an increment of the pressure at a pressure P_n likewise.
θ1=θ0+ΣΔθ_n(θ_n) (5)
P1=P0+ΣΔP_n(P_n) (6)
In terms of Δθ_n(θ_n) of equation (5), it is simply required to give the increment Δθ_n(θ_n) in accordance with each valve opening θ_n as in equation (7). In terms of ΔP_n(P_n) of equation (6), it is simply required to give the increment ΔP_n(P_n) at the pressure P_n corresponding to the increment Δθ_n(θ_n) at the valve opening θ_n, using the correlation between the valve opening θ and the pressure P. If the plant gain Gp is used as the correlation, the increment ΔP_n=ΔP_n(P_n) at the pressure P_n is expressed as in the following equation (8), applying the above-mentioned equation (2). In equation (8), Δθ_n=Δθ_n(θ_n), and Gp_n=Gp(θ_n).
Δθ_n=Δθ_n(θ_n) (7)
ΔP_n=ΔP_n(P_n)=−P_n×Gp_n×Δθ_n (8)
When equations (7) and (8) are substituted into equations (5) and (6) to compute a cumulative sum, it is possible to reach approximately the valve opening value θ1 and the pressure value P1 in the pressure regulation state (θ1, P1). Moreover, when equations (5) and (6) are modified and expressed as described below, it is possible to sequentially perform computations.
θ_n+1=θ_n+Δθ_n (9)
P_n+1=P_n+ΔP_n (10)
If the target pressure Ps is inputted as the next pressure regulation condition into the valve control apparatus 2 in the current pressure regulation state (θ0, P0), at least whether the next target pressure is higher or lower than the current target pressure at this point in time is determined. Therefore, if the next target pressure Ps is higher (Ps=P1>P0), as long as there is no extreme change in flow rate, the disc driving direction is a direction in which the valve opening is reduced as compared to the current valve opening. Conversely, if the next target pressure Ps is lower (Ps=P1<P0), the disc driving direction is a direction in which the valve opening is increased as compared to the current valve opening. In other words, if the next pressure regulation condition (the target pressure Ps) is inputted into the valve control apparatus 2, it is possible to determine the direction in which the increment of the valve opening is monotonously increased or reduced from the current pressure regulation state (θ0, P0).
If the computation processing capability of the target valve opening estimator 22 is high, the value of Δθ_n in equation (7) is set at a minute quantity as much as possible to increase the computational accuracy. For example, Δθ_n is set at 0.1% or below for 100% full valve opening. A cumulative sum of the valve opening θ is computed from equations (7) and (9) in monotone increments of +0.1% or −0.1% from the pressure regulation state (θ0, P0). A cumulative sum of the pressure P is sequentially computed from equations (8) and (10). If the value “P0+ΣΔP_n(P_n)” on the right side of equation (6) being the computed cumulative pressure value reaches (or exceeds) the next pressure value (the target pressure value) P1, the computation ends. The cumulative valve opening computed when the value has reached the pressure value P1 corresponds to the estimated target valve opening value θse. The example of the method for calculating the estimated target valve opening value θse is described above. However, the plant gain is also the quantity obtained by differentiating the logarithmic value of the pressure with respect to the valve opening as described above. Therefore, although the details are not described, it also is possible to obtain the estimated target valve opening value θse, applying a logarithmic computation or an exponential computation.
In step S12, it is determined whether or not the valve opening θr has reached the estimated target valve opening value θse. If not, execution proceeds to step S13. If so, execution proceeds to step S14. In step S13, it is determined whether or not a predetermined time Δt being the control period (the sampling period) has passed. If it is determined the time has passed, execution returns to step S11.
On the other hand, if it is determined in step S12 that the valve opening θr has reached the estimated target valve opening value θse to proceed to step S14, open-loop control ends in a state where the valve opening is fixed at the estimated target valve opening value θse. The valve opening is then adjusted in closed-loop control in step S14. In other words, the pressure deviation ε=Pr−Pset1 between the set pressure Pset1 and the pressure Pr, which have been outputted from the set pressure generator 20, is inputted into the feedback controller 24. The valve opening is controlled on the basis of the valve opening setting output Δθ2 outputted from the feedback controller 24.
In step S15, it is determined whether or not the predetermined time Δt being the control period (the sampling period) has passed. If it is determined that the time has passed, execution proceeds to step S16. In step S16, it is determined whether or not the target pressure Ps has been changed. If the target pressure Ps has been changed, execution returns from step S15 to step S11. The adjustment of the valve opening in open-loop control starts again. On the other hand, if the target pressure Ps has not been changed, execution returns to step S14 to continue closed-loop control.
The respect that in closed-loop control in the embodiment, the set pressure generator 20 generates the set pressure Pset1 to perform feedback control on the basis of the pressure deviation ε=Pr−Pset1 between the set pressure Pset1 and the pressure Pr is different from before. In conventional closed-loop control, the valve opening is controlled on the basis of a difference between the target pressure Ps and the measured pressure Pr as described in, for example, FIG. 3 in JP-A-2018-12263. This corresponds to that the set pressure generator 20 of
In the embodiment, the set pressure Pset1 outputted from the set pressure generator 20 is set at a set pressure described below. Accordingly, the occurrence of, for example, an overshoot illustrated in
If the vacuum apparatus 100 is, for example, an apparatus that performs semiconductor processes, many pressure regulation events are performed in which various process conditions including different gas types, flow rates, and target pressures Ps of gas introduced into the vacuum chamber 3 are switched and changed every predetermined time. In each pressure regulation event, the flow rate is converged by the flow controller 32 to a predetermined flow rate value immediately after the start, and concurrently, the valve opening of the vacuum valve 1 is adjusted and the effective pumping speed of the exhaust system is controlled. Accordingly, the pressure Pr of the vacuum chamber 3 converges to the target pressure Ps. An equation of the exhaust indicated by the following equation (11) holds for the pressure Pr of the vacuum chamber 3. In equation (11), V denotes the volume of the vacuum chamber 3, Se denotes the effective pumping speed of the exhaust system including conductance of the vacuum valve 1, and Qin denotes the flow rate of gas introduced into the vacuum chamber 3.
V×(dPr/dt)+Se×Pr=Qin (11)
In a general process, the timing when the flow controller 32 completes the control and convergence of the flow rate is earlier than the timing when the vacuum valve 1 completes the control and convergence of the pressure. Hence, it can be considered that at the timing when closed-loop control starts, the flow rate Qin of gas introduced has already converged to a predetermined flow rate Qin0. Moreover, before the start of closed-loop control, the valve opening θr is fixed at the final valve opening value (the estimated target valve opening value θse) of open-loop control. Accordingly, it can be considered that the effective pumping speed Se has also reached a fixed value Se0 at the timing when closed-loop control starts.
In open-loop control, the valve opening θr is controlled to the final valve opening value (the estimated target valve opening value θse) that brings the pressure Pr close to the target pressure Ps. The above Se0 is the effective pumping speed at the valve opening value θse. The correlation between the valve opening θ and the effective pumping speed Se is stored in advance in the storage 221 (refer to
On the other hand, equation (11) of the exhaust at the timing when closed-loop control starts is expressed as in equation (12). In equation (12), “dPr/dt|0” is the value of a time derivative “dPr/dt” of the pressure at the timing when closed-loop control starts.
V×(dPr/dt|0)+Se0×Pr0=Qin0 (12)
When Qin0=Ps×Se0 is applied to equation (12), the following equation (13) is obtained.
V/Se0=(Ps−Pr0)/(dPr/dt|0) (13)
The left side=V/Se0 of equation (13) is a time constant τ0 at the timing when closed-loop control starts. In other words, the time constant τ0 can be calculated by the following equation (14), using the pressure Pr (=Pr0) at the timing when closed-loop control starts and the time derivative value (dPr/dt|0) of the pressure. The time derivative value (dPr/dt|0) of the pressure can be calculated by ΔPr/Δt, using the pressure difference (ΔPr) during the control period Δt.
τ0=(Ps−Pr0)/(dPr/dt|0) (14)
The time derivative value (dPr/dt) of the pressure receives a large influence of noise. Accordingly, it is better to perform the smoothing process (filtering such as a moving average) in accordance with the level of noise mixed in a detection signal. As described above, the use of equation (14) allows calculating the time constant τ0 irrespective of the gas type and the flow rate that are unknown on the valve control apparatus 2 side.
Here, a case is considered in which in
Equation (15) is a pressure response that converges to the target pressure Ps only with open-loop control without closed-loop control if the estimated target valve opening value θse is calculated without an error and agrees with the ideal target pressure Ps. In other words, equation (15) can be interpreted as an optimal value as the set pressure Pset1 that does not cause an unnecessary pressure deviation in closed-loop control. In the embodiment, line L15 of
Pset1_i=A×Pset_i−1+Ps×(1−A) for i≥1 (16)
where Δ=exp(−Δt/τ0)
Line L20 of
In closed-loop control, the valve opening setting output Δθ1 of the feedforward controller 23 is fixed at the estimated target valve opening value θse. Accordingly, the valve opening setting output θse (=Δθ1+Δθ2) inputted into the motor controller 25 is also θse. The valve opening θr is maintained constant (θse) as in line L21 of
As described above, in the embodiment, the setting is not changed discontinuously (stepwise) to the target pressure Ps at the timing when open-loop control is switched to closed-loop control as in the comparative example illustrated in
In the above-mentioned description, the example of the case where the target pressure Ps is higher than the pressure at the start of the pressure regulation (hereinafter referred to as the up case) has been described. However, naturally, the same shall apply to a case where the target pressure Ps is lower than the pressure at the start of the pressure regulation (hereinafter referred to as the down case), and a similar effect can be obtained.
In the examples illustrated with line L15 of
However, there is an error in the estimated target valve opening value θse in reality. If the error is such a large error as cannot be ignored, when the valve opening θr is put in the state of being maintained in the state of the estimated target valve opening value θse, it does not bring a pressure response that converges to the target pressure Ps like line L15 of
In
Hence, in the second embodiment, it is configured in such a manner as to give a set pressure that can converge appropriately and quickly to the target pressure Ps even if the final estimated target valve opening value θse of open-loop control includes an error.
The set pressure generator 20A further generates a second set pressure Pset2 based on the above-mentioned set pressure Pset1, in addition to the set pressure Pset1, on the basis of the inputted target pressure Ps and the pressure Pr. The set pressure generator 20A outputs the pressure Pr as the set pressure in open-loop control in the first stage of the valve opening control and outputs the set pressure Pset2 in closed-loop control.
As described above, at the timing when closed-loop control starts, it is not clear whether the final value of the pressure response with the final valve opening value (the estimated target valve opening value θse) of open-loop control indicated by lines L30 and L31 of
Pset2_i=A_i×Pset2_i−1+Ps×(1−A_i) for i≥1 (17)
where A_i=exp(−Δt·α_i/τ0))
In equation (17), in a case of a α_i=1, A_i is the same as A of equation (16), and the set pressure Pset2 agrees with the set pressure Pset1. τ0/α_i corresponds to the time constant related to the set pressure Pset2, if α_i>1, represents the time constant of a set pressure locus that changes more quickly than the case of the time constant τ0 and, if α_i<1, represents the time constant of a set pressure locus that changes more slowly than the case of the time constant τ0. A recurrence relation of the set pressure Pset2 indicated by equation (17) is derived on the basis of the set pressure locus of the pressure response indicated by the following equation (18).
Pr_open=Pr0 exp(−t·α_i/τ0)+Ps[1−exp(−t·α_i/τ0)] (18)
α_i is a coefficient value determined every control period Δt, and is set as in the following equations (19a) to (19c), comparing a rate of change (ΔPset1_i/Δt) at a control timing i of the set pressure Pset1 as the reference and a rate of change (ΔPr_i/Δt) of the pressure Pr at the control timing i. α_0 in a case of i=0 is α_0=1, and Δα is a positive constant.
α_i=a_i−1−Δα for |(ΔPset1_i/Δt)|<|(ΔPr_i/Δt)| (19a)
α_i=α_i−1 for |(ΔPset1_i/Δt)|=|(ΔPr_i/Δt)| (19b)
α_i=α_i−1+Δα for |(ΔPset1_i/Δt)|>|(ΔPr_i/Δt)| (19c)
Equation (19a) is a case where the rate of change of the actual pressure Pr is greater than the rate of change of the set pressure Pset1, and corresponds to, for example, lines L40 and L30 of
Equation (19c) is a case where the rate of change of the actual pressure Pr is less than the rate of change of the set pressure Pset1, and corresponds to, for example, lines L41 and L31 of
The following first and second modifications can be applied to the above-mentioned first and second embodiments. cl First Modification
If the pressure Pr0 at the timing when closed-loop control starts is increased as compared to the target pressure Ps in the case of the up case, or is reduced as compared to the target pressure Ps in the case of the down case, due to an error in the estimated target valve opening value θse or a disturbance, the time constant τ0 calculated by equation (14) has a minus value. In this case, the set pressure Pset1 results in being set in a direction deviating from the target pressure Ps, which results in increasing the pressure regulation time.
Hence, in the first modification, instead of equation (14), the following equation (20) is used to calculate the time constant τ0.
τ0=|Ps−Pr0)|/|(dPr/dt|0)| (20)
The time constant τ0 is calculated by equation (20). Accordingly, even if an irregularity occurs in the timing when closed-loop control starts, the time constant always has a plus value, and the appropriate set pressure Pset1 is outputted and the reliability of pressure regulation improves.
The second modification is a countermeasure in a case where there is a tendency in the target valve opening estimator 22 to calculate the estimated target valve opening value θse at which a virtual response in open-loop control becomes like line L30 (or line L31) of
Pset1_i=A2×Pset1_i−1+Ps×(1−A2) for i≥1 (21)
where A2=exp(−Δt·β/τ0) and β≠1.
A recurrence relation of the set pressure Pset1 indicated by equation (21) is also derived on the basis of the set pressure locus of a pressure response indicated by the following equation (22) as in the case of the recurrence relation of the set pressure Pset2 indicated by equation (17).
Pr_open=Pr0 exp(−t·β/τ0)+Ps[1−exp(−t·β/τ0)] (22)
If it is known beforehand that the tendency that the virtual response exceeds the target pressure Ps as in line L30 of
The operations and effects of the above-mentioned embodiments and modifications are summarized as follows:
(1) As described in the first embodiment, the valve control apparatus 2 of
It is configured in this manner that the set pressure Pset1 in the second valve opening control is generated on the basis of the set pressure locus that converges from the pressure Pr0 of the vacuum chamber 3 at the time of switching to the second valve opening control to the target pressure Ps with the predetermined time constant τ0. Accordingly, an overshoot of the pressure response after switching to closed-loop control indicated by line L12 of
In the above-mentioned embodiments, the case where the first valve opening control is open-loop control (feedforward control) has been described as an example. However, the first valve opening control is not limited to open-loop control. For example, also if feedback control of coarse adjustments is performed in the first valve opening control and feedback control of fine adjustments is performed in the second valve opening control, such a set pressure Pset1 described above can be applied.
(2) It is preferable that the above-mentioned predetermined time constant τ0 be calculated on the basis of, for example, the pressure Pr (=Pr0) of the vacuum chamber 3 at the time of switching from the first valve opening control to the second valve opening control and the time derivative value (dPr/dt|0) of the pressure Pr, and the target pressure Ps. Consequently, it is possible to make the set pressure locus in the second valve opening control (that is, the set pressure locus indicated by equation (15)) substantially the same as the pressure locus by open-loop control alone.
(3) Furthermore, as in the second embodiment, the time constant τ0 of a first set pressure locus related to the set pressure Pset1 (that is, the set pressure locus indicated by equation (15)) is corrected to, for example, (τ0/α_i) every control period Δt, applying equations (19a) to (19c), in other words, on the basis of a change in the set pressure Pset1 over time and a change in the pressure Pr of the vacuum chamber 3 over time, and the second set pressure Pset2 is generated on the basis of a second set pressure locus that converges to the target pressure Ps with the corrected time constant (τ0/α_i) (that is, the set pressure locus indicated by equation (18)). It is preferable that instead of the second valve opening control using the set pressure Pset1, third valve opening control that performs feedback control on the basis of a difference ε between the set pressure Pset2 and the pressure Pr of the vacuum chamber 3 be performed.
When an error occurs in the final valve opening value (the estimated target valve opening value θse) in open-loop control (the first valve opening control), even if the valve opening is controlled on the basis of the set pressure Pset1 as in lines L40 and L41 of
(4) It is preferable that the above-mentioned corrected time constant (τ0/α_i) be set to be less than the time constant τ0 if the magnitude of the change in the set pressure Pset1 over time is less than the magnitude of the change in the pressure Pr of the vacuum chamber 3 over time as in equation (19a), be set to be greater than the time constant τ0 if the magnitude of the change in the set pressure Pset1 over time is greater than the magnitude of the change in the pressure Pr of the vacuum chamber 3 over time, and be set to be equal to the time constant τ0 if the magnitude of the change in the set pressure Pset1 over time is equal to the magnitude of the change in the pressure Pr of the vacuum chamber 3 over time.
(5) Furthermore, it is preferable that the time constant τ0 be calculated by an equation “τ0=|(Ps−Pr0)|/|(dPr/dt|0)|”. As a result, even if an irregularity occurs in the timing when closed-loop control starts, the time constant τ0 always has a plus value, the appropriate set pressure Pset1 is outputted, and the reliability of pressure regulation improves.
(6) Moreover, it is preferable that if a control error in the first valve opening control (open-loop control) is known beforehand, the time constant (τ0/β) obtained by correcting the time constant τ0 on the basis of the control error be used instead of the time constant τ0 to generate the first set pressure Pset1 on the basis of the set pressure locus that converges to the pressure Pr of the vacuum chamber 3 with the corrected time constant (τ0/β) (the set pressure locus indicated by equation (22)). As a result, the influence of the control error in the first valve opening control can be reduced.
Various embodiments and modifications have been described above. However, the present invention is not limited to these contents. Other aspects that are conceivable within the scope of the technical idea of the present invention are also included within the scope of the present invention. For example, the vacuum valve 1 where the valve plate 11 is driven and swung to change the valve opening has been described as an example. However, the present invention can also be applied similarly to an automatic pressure regulating vacuum valve of another configuration.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2019-031724 | Feb 2019 | JP | national |
