JP3340926B2 - SAC controller - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD The present invention relates to a SAC control.
It is about LA.

[0002]

2. Description of the Related Art In the early 1980's, the basic form of an adaptive control system was determined, and the boundedness of the signal within the adaptive loop and the asymptotic stability of the output error were theoretically proved. However, it has been pointed out that the control algorithm is complicated, contains many parameters to be set in practical use, and furthermore, there is no clear guideline for adjusting these parameters.
Therefore, SAC (Simple Adaptive Control) has been studied as a solution to such a problem (Literature: Iwai, "Simple Adaptive Control (SAC)", Measurement and Control,
Vol. 35, No. 6, 1996).

[0003] SAC was established in 1982 by Sobel.
), A method proposed by Kaufman and Mabius, and subsequently studied by Bar-Kana et al. This method adaptively realizes an ideal control input (operating amount) such that the control amount matches the reference model output, and at the same time, secures the stability of the closed loop system in an adaptive output feedback form. CGT
(Command Generator tracker).

[0004] As a precondition that this SAC can be configured,
The control target is ASPR (Almost Strictly Positive Real
) It is necessary to satisfy the conditions. In conclusion,
If the control target process includes a dead time, the ASPR condition is not satisfied. Therefore, as a technique for satisfying the ASPR condition, parallel feedforward compensation (Parallel
Feedforward Compensator (hereinafter abbreviated as PFC) has been proposed. If this PFC is simply expressed, it is a method of adding an appropriate pseudo response to the measured value of the control amount in order to fake apparently that there is no response for the dead time of the control target.

FIG. 5 is a block diagram of a control system using such a conventional SAC controller. The SAC controller includes a reference model calculation unit 13, an operation amount calculation unit 14,
It comprises a PFC operation unit 15. First, the PFC operation unit 15
Calculates the output value yf (k) as follows. zf (k) = [b1 × zf (k−1) + Kf × {u (k−1) −u (k−2)}] / (b1 + dT) (1) yf (k) = ｛b2 × yf (k−1) + dT × zf (k)｝ / (b2 + dT) (2)

In equations (1) and (2), zf (k) is an internal variable, zf (k-1) is a value of the internal variable zf (k) one control cycle before, and yf (k-1) is an output. 1 of the value yf (k)
The values u (k−1) and u (k−2) before the control cycle are the values of the manipulated variable u (k) output from the manipulated variable calculation unit 1 before and after the control cycle, respectively. DT is a control cycle parameter, b1 and b2 are constants, and Kf is a PFC gain. Subsequently, the reference model calculation unit 13 sets the set value sp
From (k) to output values z _n-1 (k), z _n-2 (k) ... z ₁
(K) and z ₀ (k) are calculated as in the following equations.

[0007] In equation (3), z _n-1 (k-1), z _n-2 (k-1)
... z ₁ (k-1) and z ₀ (k-1) are output values z _n-1 (k) and z _n-2 (k) ... z ₁ (k) and z _{0, respectively.}
One control period previous value of _{(k), a m, a} 0, a 1 ··· a
_n-2 and a _n-1 are constants.

Next, the manipulated variable calculator 14 calculates the manipulated variable u (k) as follows. First, the addition unit 21 in the operation amount calculation unit 14 adds the control amount y (k) of the control target process 30 and the output value yf (k) of the PFC calculation unit 15, and further, the subtraction unit 22 includes the addition unit 21. The output value z ₀ (k) of the reference model calculation unit 13 is subtracted from the output of. Therefore,
The output value ey (k) of the subtraction unit 22 is represented by the following equation. ey (k) = y (k ) + yf (k) -z 0 (k) ··· (4)

The operation units 23 to 25 output the output value K _P1
(K), K _P2 (k) ... K _{Pn + 2} (k), K _J1 (k),
K _J2 (k)... K _{Jn + 2} (k) are calculated as follows.

[0010]

In equations (5) and (6), K _J1 (k−
_{1), K J2 (k-} 1) ··· K Jn + 2 (k-1) each output value K _J1 (k), 1 control _{K J2 (k) ··· K Jn} + 2 (k) Values before the cycle, γ _P1 , γ _P2 ... γ _Pn , γ _{Pn + 1} , γ
_{Pn + 2} , γ _J1 , γ _J2 ... γ _Jn , γ _{Jn + 1} , γ _{Jn + 2} are SAC
Gain. Finally, the adder 26 adds the outputs of the calculators 23 and 24, and the adder 27 further adds
6 and the output of the operation unit 25. As a result, the output of the adding unit 27 becomes the operation amount u (k) which is the output of the operation amount calculation unit 14.

U (k) = ｛K _P1 (k) + K _J1 (k)｝ × ey (k) + ｛K _P2 (k) + K _J2 (k)｝ × z ₀ (k) + ｛K _P3 (k ) + K _J3 (k)｝ × z ₁ (k): + ｛K _Pn (k) + K _Jn (k)｝ × z _n-2 (k) + ｛K _{Pn + 1} (k) + K _{Jn + 1} (k )｝ × z _n-1 (k) + ｛K _{Pn + 2} (k) + K _{Jn + 2} (k)｝ × sp (k) (7)

The actual control cycle for outputting the manipulated variable u (k) to the control target process 30 is determined from various hardware functions such as control of the user interface of the controller and checking for signal disconnection. 13, the parameters of the manipulated variable calculator 14 and the PFC calculator 15 are adjusted according to the process to be controlled. Therefore, due to the time required for processing such as communication with the outside,
SAC algorithm in one control cycle (PFC operation,
The ratio occupied by the reference model calculation and the operation amount calculation) is small, and an actual control cycle may be longer than a cycle desired for parameter setting adjusted in accordance with the control target process 30.

FIG. 6 and FIG. 7 show the set value followability of the conventional SAC controller. FIGS. 6 and 7 are simulation results in which the control amount y (k) is obtained when the set value sp (k) is changed from 0 to 30% at 0 second, and b
1 = 25 seconds. When the actual control cycle is 1 second, good control characteristics are obtained as shown in FIG.
If the actual control cycle (here, 5 seconds) is longer than the cycle desired for the AC algorithm, the control becomes unstable as shown in FIG. 7, and the operation as a controller cannot be guaranteed. This is because the PFC operation is given by equation (1),
As shown in (2), the operation amount u (k−
This is because 1) and u (k-2) are input, and if the control cycle is long and the manipulated variable change is roughly calculated, the operation does not operate as designed.

[0015]

As described above, the conventional S
In an AC controller, if the actual control period is longer than the desired period for the SAC algorithm, the PFC
There is a problem that control becomes unstable due to insufficient calculation accuracy, and if such a situation occurs, not only the control characteristics will deteriorate, but also the control target will fall into an extremely abnormal and dangerous state. There was a problem that it was possible. The present invention has been made in order to solve the above-described problem, and is capable of avoiding instability of control even when control must be performed at an inappropriate control cycle.
It is intended to provide a C controller.

[0016]

According to a first aspect of the present invention, there is provided a PFC operation unit for performing a parallel feedforward compensation operation on an operation amount based on a PFC gain and a control cycle parameter. A control amount, a set value, and a PFC are determined by a reference model calculation unit that performs a reference model calculation on the set value based on the control cycle parameter and a SAC algorithm based on the SAC gain and the control cycle parameter.
An operation amount calculation unit that calculates an operation amount from an output of the operation unit and an output of the reference model operation unit and outputs the operation amount to the PFC operation unit;
The value of each control cycle parameter of the C calculation unit, the reference model calculation unit, and the operation amount calculation unit is set shorter than the actual control period, and the operation amount calculated by the operation amount calculation unit is controlled for each actual control period. An internal control cycle change unit that outputs to the target process, and repeatedly executes parallel feedforward compensation calculation, reference model calculation and manipulated variable calculation for each internal control cycle indicated by the control cycle parameter, and controls only the manipulated variable output. It is designed to be executed periodically. Thus, the PFC calculation unit performs the parallel feedforward compensation calculation on the operation amount, the reference model calculation unit performs the reference model calculation on the set value, and the operation amount calculation unit
The control target process is controlled by calculating the operation amount from the control amount, the set value, and the outputs of the PFC operation unit and the reference model operation unit using an algorithm. At this time, the control must be performed in an inappropriate control cycle by repeatedly executing the parallel feedforward compensation calculation, the reference model calculation, and the manipulated variable calculation in each internal control cycle, and executing only the manipulated variable output in the actual control cycle. Even in the case where it is necessary, the control can be prevented from becoming unstable.

According to a second aspect of the present invention, the internal control cycle changing section includes an adjustment coefficient for determining the internal control cycle.
m (m> 1: an integer) and adjusting coefficient storage unit for storing, adjusting
From the coefficient m and the actual control cycle dT, the actual control cycle d
The correction value dTi of the internal control cycle shorter than T is given by dTi = dT
/ M, PFC calculation unit, reference model calculation unit
And the control cycle parameters of the manipulated variable calculation unit
a control cycle adjusting unit to change dTi, Thailand manipulated variable output
Variable n, which is initialized to 0 by
And the internal variable n after the addition becomes an adjustment coefficient m and
If they do not match, it is determined that it is not the timing of MV output.
And the internal variable n after the addition becomes equal to the adjustment coefficient m.
In the case, the internal cycle determination unit determines that it is the timing of the operation amount output, and an operation amount output unit that outputs the operation amount from the operation amount calculation unit to the control target process according to the determination result of the internal cycle determination unit. It is.

[0018]

FIG. 1 is a block diagram of a SAC controller showing a first embodiment of the present invention, and FIG. 2 is a flowchart for explaining the operation of this controller. Reference model calculation unit 3 in this controller,
Since the operations of the manipulated variable calculator 4 and the PFC calculator 5 are the same as those in the example of FIG. 5, a block diagram of a control system using them is shown in FIG.
Is the same as

In the present embodiment, the manipulated variable output unit 6, the adjustment coefficient storage unit 7, the control cycle adjustment unit 8, and the internal cycle determination unit 9
The characteristics of the control system are determined by setting an internal control cycle described later by an internal control cycle changing unit consisting of. The operation of the controller as the control system will be described first.

In the conventional SAC controller, the actual control cycle and the internal control cycle indicated by the control cycle parameter are the same, whereas in the present embodiment, the actual control cycle (operating amount output cycle) The internal control cycle (operating amount calculation cycle) may not match. Therefore, when the control cycle is simply described in the following description, it indicates the actual control cycle. First, the set value sp (k) is set by the operator of the controller, and is input to the reference model calculator 3 and the manipulated variable calculator 4 via the set value input unit 1 serving as an interface with the outside (FIG. 2). Step 103).

A control amount y (k) is input from the control target process 30 (actually, a sensor for detecting the control amount) to the control amount input unit 2 also serving as an interface (step 104). Next, the PFC calculation unit 5 calculates the output value yf (k) from the operation amount output from the operation amount calculation unit 4.
Is performed, and its transfer function Gf can be expressed as in the following equation.

Gf = Kf × Tf × s / (1 + Tf × s) ² (8) where Kf is a PFC gain and Tf is a time constant. FIG. 3 shows a response characteristic of the PFC operation unit 5 to the unit step input 1 / s. The response output yf (k) shows a local maximum value at time Tf, and this local maximum value yf (Tf) becomes Kf / e (e is 2.7182...). As described above, the PFC calculation unit 5 outputs yf (k), which is an appropriate pseudo response to be added to the control amount y (k).

The PFC operation unit 5 actually operates as follows in order to realize the PFC operation represented by the transfer function Gf of the continuous time system of the equation (8) in the discrete time system. At first,
The PFC operation unit 5 calculates an internal variable zf (k) as in the following equation. zf (k) = [b1 × zf (k−1) + Kf × {u (k−1) −u (k−2)}] / (b1 + dT) (9) Subsequently, the PFC operation unit 5 , The output value yf as
(K) is calculated. yf (k) = {b2 × yf (k−1) + dT × zf (k)} / (b2 + dT) (10)

In equations (9) and (10), zf (k−
1) is the value of the internal variable zf (k) one internal control cycle before, y
f (k-1) is a value one internal control cycle before the output value yf (k), and u (k-1) and u (k-2) are operation amounts u (k) output from the operation amount calculation unit 4. ) Are values before and after the internal control cycle, dT is a control cycle parameter, and b1 and b2 are constants corresponding to the time constant Tf. Thus, PFC
The calculation unit 5 performs a PFC calculation (step 105).

Next, the reference model embodies ideal characteristics of a control system in which the SAC controller and the control target process 30 are integrated, and the characteristics of the control system match the characteristics of the reference model. The controller parameters are adjusted. The reference model calculation unit 3 performs a reference model calculation as the reference model, and calculates the output of the set value sp (k) input from the set value input unit 1 using Expression (3) (Step 106). .

Where z _n-1 (k−
1), z _n-2 (k-1) ... z ₁ (k-1), z ₀ (k
-1) are the output values z _n-1 (k) and z _n-2 (k) here.
... Z ₁ (k) and z ₀ (k) are values one internal control cycle before.

Next, the operation inside the manipulated variable calculator 4 will be described with reference to FIG.
This will be described with reference to FIG. First, the adder 21 in the manipulated variable calculator 4 calculates the control amount y (k) of the control target process 30 and the PF
The output value yf (k) of the C operation unit 5 is added, and the subtraction unit 2
2 subtracts the output value z ₀ (k) of the reference model calculation unit 3 from the output of the addition unit 21. Therefore, the output value ey (k) of the subtraction unit 22 is expressed by the following equation. ey (k) = y (k ) + yf (k) -z 0 (k) ··· (11)

The operation units 23 to 25 output the output value K _P1
(K), K _P2 (k) ... K _{Pn + 2} (k), K _J1 (k),
K _J2 (k)... K _{Jn + 2} (k) are calculated as follows.

[0029]

In equations (12) and (13), K_J1(K
-1), K_J2(K-1) ... K_{Jn +2}(K-1) is it
Each output value K_J1(K), K_J2(K) ... K_{Jn + 2}(K)
The value before one internal control cycle of γ_P1, Γ_P2... γ_Pn, Γ
_{Pn + 1}, Γ_{Pn + 2}, Γ_J1, Γ_J2... γ_Jn, Γ_{Jn + 1}, Γ_{Jn + 2}
Is the SAC gain. The outputs of the arithmetic units 23 to 25
Of which, K_P1, K_J1Is calculated by the arithmetic unit 25, and K
_P2, K_P3... K_{Pn + 1}, K_J2, K_J3... K_{Jn + 1}About
The calculation unit 24 calculates_{Pn + 2}, K_{Jn + 2}About
The calculation unit 23 calculates.

Finally, the adder 26 adds the outputs of the calculator 23 and the calculator 24, and the adder 27 further adds
And the output of the operation unit 25 are added. As a result, the output of the addition unit 27 is the operation amount u which is the output of the operation amount calculation unit 4.
(K).

U (k) = ｛K _P1 (k) + K _J1 (k)｝ × ey (k) + ｛K _P2 (k) + K _J2 (k)｝ × z ₀ (k) + ｛K _P3 (k ) + K _J3 (k)｝ × z ₁ (k) _:: + ｛K _Pn (k) + K _Jn (k)｝ × z _n-2 (k) + ｛K _{Pn + 1} (k) + K _{Jn + 1} ( k)｝ × z _n-1 (k) + ｛K _{Pn + 2} (k) + K _{Jn + 2} (k)｝ × sp (k) (14)

As described above, the manipulated variable calculator 4 calculates the manipulated variable u (k) (step 107). Next, an operation amount output unit 6, an adjustment coefficient storage unit 7, a control cycle adjustment unit 8,
The operation of the internal cycle determination unit 9 will be described.

First, the control cycle adjusting section 8 and the internal cycle determining section 9 read out an adjustment coefficient m (m> 1 : an integer) for determining the internal control cycle from the adjustment coefficient storage section 7 (step 101). Then, the control cycle adjusting unit 8 calculates a correction value dTi of the control cycle parameter as in the following equation. dTi = dT / m (15)

Subsequently, the control cycle adjuster 8 outputs the calculated correction value dTi of the control cycle parameter to the reference model calculator 3, the manipulated variable calculator 4, and the PFC calculator 5. As a result, the control cycle parameter dT of these configurations is changed to the correction value dTi (step 102).

After the operations of steps 103 to 107 described above are performed, the internal cycle determination section 9 determines whether or not it is the timing of the manipulated variable output to be executed for each control cycle as follows. First, the internal cycle determination unit 9 calculates an internal variable n (k) as in the following equation (Step 108). n (k) = n (k-1) +1 (16)

In equation (16), n (k-1) is the value of the internal variable n (k) one internal control cycle before, and the initial value is 0. Subsequently, the internal cycle determination unit 9 determines whether or not the internal variable n (k) obtained by Expression (16) is equal to the adjustment coefficient m (Step 109).

If the internal variable n (k) is not equal to the adjustment coefficient m, the process returns to step 105 and the internal cycle determination unit 9 outputs a determination result signal to the operation amount output unit 6 to instruct the omission of the operation amount output. Is done. The operation amount output unit 6 serving as an interface with the outside receives the operation amount u (k) calculated by the operation amount calculation unit 4 when the determination result signal instructing the omission of the operation amount output is input. Is not output to

Therefore, unless the internal variable n (k) is equal to the adjustment coefficient m, the operations of steps 105 to 109 are repeated every internal control cycle, and the variable n (k) is incremented by 1 every internal control cycle. , The operation amount output is omitted.

Subsequently, in step 109, when the internal variable n (k) becomes equal to the adjustment coefficient m, the internal cycle determination section 9 initializes the internal variable n (k) to 0, and instructs execution of the manipulated variable output. (Step 1)
10). When the determination result signal instructing execution of the operation amount output is input, the operation amount output unit 6 outputs the operation amount u (k) calculated by the operation amount calculation unit 4 to the control target process 30 (Step 111). ).

The above operations of steps 101 to 111 are repeated every control cycle until the controller is stopped by an instruction from an operator or the like (step 112).
This is the operation of the SAC controller. In addition,
During one control cycle, the set value sp (k) and the value of the control amount y (k) input in steps 103 and 104 are maintained as they are.

In the conventional SAC controller shown in FIG. 5, the control cycle parameter dT is set to the same value as the actual control cycle (that is, the actual control cycle and the internal control cycle are the same). Therefore, if the actual control cycle is longer than the desired cycle for the reference model calculation unit, the operation amount calculation unit, and the PFC calculation unit, the control becomes unstable as shown in FIG.

On the other hand, in the SAC controller of the present embodiment, for example, when the adjustment coefficient m is set to 5, the initially set control cycle parameter dT equal to the actual control cycle is set.
Is changed to the correction value dTi = dT / 5, so that the PFC calculation (step 105) and the reference model calculation (step 10)
6), the operation amount calculation (step 107) is performed in the internal control cycle d.
It is repeated for each Ti.

Then, the internal variable n
Since the manipulated variable u (k) is output when (k) reaches 5, the manipulated variable output is executed in the actual control cycle. Therefore, the PFC calculation can be executed more finely than the actual control cycle only by setting the adjustment coefficient m in the adjustment coefficient storage unit 7, so that the calculation accuracy of the PFC does not become insufficient, and the control becomes inefficient. Stabilization can be avoided.

FIG. 4 shows the set value followability of the SAC controller of this embodiment. FIG. 4 is a simulation result in which the control amount y (k) is obtained under the same conditions as those in FIGS. 6 and 7, where the control cycle is 5 seconds and the adjustment coefficient m is 5. Thus, it can be seen that good control characteristics can be obtained even when control must be performed at an inappropriate control cycle.

Thus, the control amount y, which is a feature of the SAC,
(K) is matched with the reference model output ym (k). (In the present invention, since the PFC is used, to be precise, the control amount y
(K) and the result of adding the output yf (k) of the PFC operation unit match the reference model output ym (k). The feature of this SAC is that it provides a valuable solution in a demand for a steel process or the like in which a control amount is desired to be changed at a desired value and a desired speed.

It is to be noted that there is already an example in which the cycle for calculating the manipulated variable is equal to the cycle for outputting the manipulated variable to the control target process, and the cycle for updating the manipulated variable is longer than this control cycle. The operation amount update cycle is, for example, a period in which an operation device such as a heater takes in an operation amount output from the controller,
It is determined regardless of the specifications of the controller.

As a result, when such a heater is combined with a controller whose control cycle is already set for its own convenience, the above case may occur. On the other hand, the present invention executes the SAC algorithm in an internal control cycle obtained by subdividing an actual control cycle, as a problem to be solved in that the control cycle becomes coarse depending on parameter setting. Therefore, the above case is different from the above case because the control period <the control amount update period, and the present invention satisfies the internal control period <the control period <the control amount update period.

[0049]

According to the present invention, the internal control cycle changing section sets the value of the control cycle parameter to be shorter than the actual control cycle and calculates the value by the operation amount calculating section. By outputting the manipulated variable to the process to be controlled in each actual control cycle, the parallel feedforward compensation calculation, the reference model calculation, and the manipulated variable calculation are repeatedly executed for each internal control cycle shorter than the actual control cycle. Since only the quantity output is executed in the actual control cycle, the control can be prevented from becoming unstable even when the control must be performed in an inappropriate control cycle due to hardware or the like. As a result, it is possible to realize control in which the control amount, which is a feature of the SAC, matches the reference model output (accurately, the result of adding the control amount and the output of the PFC operation unit matches the reference model output). Can be.

Further, the internal control cycle changing section is composed of an adjustment coefficient storage section, a control cycle adjusting section, an internal cycle determining section, and an operation amount output section. Can be set shorter than the actual control cycle, and the internal control cycle changing unit that outputs the operation amount to the process to be controlled in each actual control cycle can be realized with a simple configuration.

[Brief description of the drawings]

FIG. 1 is a block diagram of a SAC controller according to a first embodiment of the present invention.

FIG. 2 is a flowchart for explaining the operation of the SAC controller in FIG. 1;

FIG. 3 is a response waveform diagram of a PFC calculation unit.

FIG. 4 is a diagram showing a set value tracking property of the SAC controller of FIG. 1;

FIG. 5 is a block diagram of a control system using a conventional SAC controller.

FIG. 6 is a diagram showing a set value tracking property of a conventional SAC controller.

FIG. 7 is a diagram showing a set value tracking property of a conventional SAC controller.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Set value input part, 2 ... Control amount input part, 3 ... Reference model calculation part, 4 ... Operation amount calculation part, 5 ... PFC calculation part, 6 ... Operation amount output part, 7 ... Adjustment coefficient storage part, 8 ... Control cycle adjuster, 9 ... internal cycle determiner.

Continuation of the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G05B 11/32 G05B 13/02-13/04

Claims (2)

(57) [Claims] A value obtained by adding a result of performing a parallel feedforward compensation operation to an operation amount output to a control target process and a control amount of a process is a value of a reference model in which a desired characteristic of a control system is mathematically expressed. to match the output, the SAC controller for calculating the operation amount, the a PFC arithmetic unit that performs parallel feedforward compensation calculation on the basis of the PFC gain and control cycle parameters to the operation amount, with respect to the input setting value a reference model calculation unit which performs reference model calculation on the basis of the control cycle parameters, SAC gain and SA based on the control cycle parameters
According to the C algorithm, the control amount, the set value, the
Whether the output of the PFC operation unit and the output of the reference model operation unit
And an operation amount calculating unit that calculates the al the manipulated variable output to the PFC operation unit, the PFC operation unit, the reference model calculation unit and the operation
An internal control cycle change in which the value of each control cycle parameter of the quantity calculation unit is set shorter than the actual control cycle, and the manipulated variable calculated by the manipulated variable calculation unit is output to the control target process for each actual control cycle. A parallel feedforward compensation calculation, a reference model calculation and a manipulated variable calculation are repeatedly executed for each internal control cycle represented by the control cycle parameter, and only the manipulated variable output is executed in an actual control cycle. SAC controller. 2. The SAC controller according to claim 1, wherein the internal control cycle changing unit determines the internal control cycle.
An adjustment coefficient storage unit for storing an adjustment coefficient m (m> 1: an integer) for performing the actual control based on the adjustment coefficient m and the actual control cycle dT.
The correction value dTi of the internal control cycle shorter than the control cycle dT is set to dT
i = dT / m, and the PFC calculation unit,
Each control cycle parameter of the reference model calculation unit and the manipulated variable calculation unit
A control cycle adjusting unit for changing the parameter to the correction value dTi ; and an internal variable n initialized to 0 at the timing of the operation amount output.
Is added to each of the internal control cycles, and the internal
If the variable n does not match the adjustment coefficient m,
It is determined that it is not the output timing, and the internal
If the variable n is equal to the adjustment coefficients m, the operation
An internal cycle determining section determines that the timing of the quantity output, is made of a manipulated variable output section for outputting an operation amount from the operation amount calculation unit in the controlled process in accordance with the determination result of the internal cycle determining unit A SAC controller, characterized in that: