JPH04267401A - Controller - Google Patents

Abstract

PURPOSE:To provide a controller which can reduce overshoot without oscillation and with the rise better than the limit of the rise in which the scheduling effect for the set value can be obtained in the conventional control system. CONSTITUTION:This controller is provided with a set value generating part 2 designating a set value SP, a set value changing part 3 sheduling the set value from the set value SP generated in the set value generating part 2 as well as an observation value PV from the object of control, a PID parameter changing part 4 scheduling PID parameters Kp', Ti', and Td' from a pseudo-set value 'pseudo SP' generated in the set value generating part 2 and the observation value PV, and a PID operation part 5 calculating and outputting an amount of operation MV for the object of control from the pseudo-set value 'pseudo SP' outputted from the set value changing part 2, the PID parameters Kp', Ti' and Td' outputted from the PID parameter changing part 3 and the observation value PV.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention applies proportional + integral + differential calculations (these are collectively called The present invention relates to a control device having a function of suppressing overshoot that tends to occur in PID control that applies a manipulated variable obtained by performing PID calculation to a controlled object.

0002

2. Description of the Related Art When controlling systems such as plants and various operating devices, it is often required to suppress overshoot. Therefore, the following control method has been proposed.

First, as shown in FIG. 12, when a set value SP is given, in normal PID control, the observed value PV changes as shown by the broken line at the time of rise, and overshoot occurs for the controlled object. , PID calculation is performed using the auxiliary target value (SSP) as a set value instead of the actual SP. This SSP is set to a value lower than SP in order to prevent overshoot when the observed value PV increases rapidly. This is called setting value scheduling. Then, as shown by the solid line, as PV approaches SP, SSP approaches the same value as the actual set value SP, and eventually PV reaches SP without overshooting.

0004

[Problem to be Solved by the Invention] However, in the overshoot suppression method using the scheduling of set values as described above, as shown in FIG. In the case of (some vibrations occur), there is a problem that the vibrations remain. That is, in order to obtain the effect of scheduling the set value, it is necessary to suppress the rise to a level that does not cause vibration, and it is not possible to sufficiently cope with a control system that has a steep rise and causes vibration.

Therefore, an object of the present invention is to provide a control device that can eliminate vibration and reduce overshoot with a rise that is even better than the limit of rise that allows the scheduling effect of set values to be obtained using the conventional control method as described above. It is about providing.

[0006]

[Means for Solving the Problems] The present invention provides PID control for the deviation between an observed value from a controlled object and a set value that is a target value for control.
In the control device that applies the manipulated variable obtained by the calculation to the controlled object, by performing two operations: the scheduling of the setting value and the parameter of the PID calculation (PID parameter), the operation can be performed quickly and without vibration. Characterized by obtaining quantity. PID parameters are gain, integral time, and differential time.

The present invention is also characterized in that scheduling of PID parameters is performed by fuzzy reasoning.

Furthermore, the control device of the present invention includes a set value generating section that specifies the set value, and a set value that schedules the set value by fuzzy inference from the set value generated by the set value generating section and the observed value. a PID that schedules PID parameters by fuzzy inference from the setting value generated in the setting value generation unit and the observed value;
a parameter changing section, a setting value output from the setting value changing section, and a PID output from the PID parameter changing section.
It can be configured with a PID calculation section that calculates and outputs the manipulated variable for the controlled object from the ID parameter and the observed value.

[0009]

[Operation] In the present invention, PID parameter scheduling is performed in addition to setting value scheduling. In other words, from the given setting values and observed values, an appropriate setting value and PI
D parameters are determined, PID calculation is performed based on these set values, PID parameters, and observed values, and a manipulated variable is output. This makes the rise steep and normal PID
Even in a control system that generates vibrations, overshoot is reduced and vibration-free control is achieved.

Furthermore, by using fuzzy inference for scheduling PID parameters, parameters can be switched smoothly. Furthermore, scheduling of setting values does not have to be performed precisely.

[0011]

Embodiment FIG. 1 is a diagram showing the configuration of an embodiment of the apparatus of the present invention.

This control device 1 includes an SP generation section 2, an SP change section 3, a PID parameter change section 4, and a PID calculation section 5.
It consists of

The SP generating section 2 is composed of a man-machine interface including a keyboard that can be operated by the user and a display that displays information necessary for control, and is used to specify the set value SP by the user's operation.

[0014] The SP change unit 3 schedules the set value SP, and as shown in FIG. Based on the observed value (process amount) PV and set value SP, a temporary set value (pseudo
SP calculation unit 32 that calculates SP).

The operation of this SP changing section 3 is as shown in FIG. That is, it is determined whether the observed value PV is sufficiently far from the set value SP (step 11), and "Yes" is determined.
”, set the temporary set value equal to or larger than SP (step 12). If “No”, then set the observed value P
It is determined whether V is approaching the set value SP (step 13), and if "Yes", a temporary set value is set smaller than SP (step 14). If “No”, then
It is determined whether the observed value PV is close to the set value SP (step 15), and if "Yes", the temporary set value is set to the same value as SP (step 16). If "No", the provisional set value is set to be the same as the previous value (step 17).

The PID parameter changing unit 4 changes the PID parameters (gain Kp, integral time Ti, and differential time T
d), and as shown in Figure 3, it is a PI that stores PID parameter change data, PID parameter change rules, and PID parameter change formulas.
D parameter change memory 41 and an appropriate P from the observed value (process amount) PV and set value SP based on the stored contents.
PI to calculate ID parameters Kp', Ti', Td'
It is composed of a D parameter calculation section 42.

The operation of the PID parameter changing section 4 is shown in FIG.
It becomes as shown in . In other words, the observed value PV is the set value SP
Determine whether it is sufficiently far away from (step 21),
If “Yes”, PID
The parameters are set larger than the standard values (step 22). If “No”, then it is determined whether the observed value PV is likely to overshoot (step 23), and if “Yes”, the PID parameter is set larger than the above in order to suppress overshoot (step 24). . If "No", then it is determined whether the observed value PV is close to the set value SP (step 25), and if "Yes", the PID parameter is set smaller than the standard value so as not to vibrate (step 26). If "No", the PID parameter is set to the same value as the previous value (step 27).

In this embodiment, the SP changing section 3 and the PID
Scheduling of setting values and PID parameters in the parameter changing unit 4 is performed according to fuzzy reasoning.

First, steps 11 and 13 in FIGS. 4 and 5
, 15 and 21, 23, 25, “N
The fuzzy rules shown in Table 1 below are used to determine “o”.

[0020]

[Table 1] 1) If En = PB and dEn =
PB then dMV = {PID(1)}, p
seudo SP = SP(1) 2) If En
= PB and dEn = PS then d
MV = {PID(1)}, pseudo SP
= SP(1) 3) If En = PB and
dEn = ZO then dMV = {PID
(1)}, pseudo SP = SP(1)
4) If En = PB and dEn = NS
then dMV = {PID(2)}, pse
udo SP = SP(2) 5) If En =
PB and dEn = NB then dMV
= {PID(2)}, pseudo SP =
SP(2) 6) If En = ZO and d
En = PB then dMV = {PID(2
)}, pseudo SP = SP(2) 7)
If En = ZO and dEn = PS t
hen dMV = {PID(3)}, pseudo
o SP = SP(3) 8) If En = Z
O and dEn = ZO then dMV =
{PID(3)}, pseudo SP = SP
(3) 9) If En = ZO and dEn
= NS then dMV = {PID(3)}
, pseudo SP = SP(3) 10)I
f En = ZO and dEn = NB th
en dMV = {PID(2)}, pseudo
SP = SP(2) 11) If En = N
B and dEn = PB then dMV =
{PID(2)}, pseudo SP = SP
(2) 12) If En = NB and dE
n = PS then dMV = {PID(2)
}, pseudo SP = SP(2) 13)
If En = NB and dEn = ZO t
hen dMV = {PID(1)}, pseudo
o SP = SP(1) 14) If En =
NB and dEn = NS then dMV
= {PID(1)}, pseudo SP = S
P(1) 15) If En = NB and d
En = NB then dMV = {PID(1
)}, pseudo SP = SP (1) In the above table, En: Normalized deviation En = E
/(SP - PVINT) dEn: Normalized deviation change dEn
= dE / (SP - PVINT) E: Deviation dE: Change in deviation PB:
Positive Big PS
:Positive Small
ZO: 0 NS: Small on the negative side (Negative Sm
all) NB: Large on the negative side (Negati
ve Big) SP: Set value pseudo SP: Temporary set value PVINT: Initial value of control amount dMV: Change amount of manipulated variable PID (1): PID parameter PI with good rise
D(2): P that cancels out the rise of PID(1)
ID parameter PID (3): PID without overshoot and vibration
Parameter { }: Value obtained by performing PID calculation using the PID parameter SP(1): Original setting value or setting value farther from the initial value SP(1) = PVINT + (SP - PVI
NT)・αα: Appropriate value of 1.0 or more SP(2): Setting value closer to the initial value than the original setting value SP(2) = PVINT + (SP − PVI
NT)・ββ: Appropriate value in the range of 0.8 to 0.9 SP(3): Original setting value The membership function of the above normalized deviation En and normalized deviation change dEn is as follows: Figures 6(a) and (b)
It is determined as shown below.

That is, as shown in FIG. 6(a), the normalized deviation En is "NB" in the range of En≦-0.2, and "NB" in the range of -0.6≦En≦0.6. ”Z
"PB" in the range of En≧0.2, and each fitness degree is set to a value between 0 and 1, respectively.

Furthermore, regarding the normalized deviation change dEn, as shown in FIG. 6(b), dEn ≦
"NB" in the range of -0.2, -0.6≦dEn≦0
"NS" in the range of -0.2≦dEn≦0.2, "ZO" in the range of 0≦dEn≦0.6.
PS'' and ``PB'' in the range of En≧0.2, each fitness level is set to a value between 0 and 1, respectively.

By defining the fuzzy rules as described above, the judgment in step 11 of FIG.
) to 3) and 13) to 15), rules 4) to 6) and 10) to 12) are used for the judgment in step 13, and rules 7) to 9) are used for the judgment in step 15. .

Similarly, the rules 1) to 3) and 13) to 15) are used for the judgment in step 21 of FIG. Rules 7) to 9) are used for the determination in 25.

Referring again to FIG. 1, the SP changing section 3 and the PI
The D parameter changing unit 4 refers to the set value SP sent from the SP generating unit 2 and the detected observation value PV, and sets a temporary set value (pseudoSP) and appropriate PID parameters Kp', Ti', and Td'. is determined, and the PID calculation unit 5
send to

The PID calculation section 5 calculates the temporary setting value (pseudoSP) output from the SP change section 3 and the PID parameter (Kp') output from the PID parameter change section 4.
, Ti', Td') and the current observed value PV, the manipulated variable MV is calculated and output to the process to be controlled.

Next, FIG. 7 shows a process model to be controlled. This controls the water level of the tank in the "secondary delay + dead time" system, and includes a valve 51 connected to a water supply source (not shown) and an upper tank 5 that is supplied with water through this valve.
2, a lower tank 53 that is supplied with water from this tank, and a fuzzy controller 54 that detects the water level of the lower tank and controls the opening degree of the valve 51.

Note that the "second-order delay + dead time" system can approximate many objects to be controlled in existing processes.

The fuzzy controller 54 executes control using fuzzy inference as described above, and inputs the deviation E of the water level of the lower tank 53 and the change in the deviation of the water level of the lower tank 53 (each sampling time). (For example, 1
．． The output is the change in valve opening degree dMV.

The transfer function of this control system is as follows.

[0031]

[Equation 1] K is gain, K=4.0, T1 and T2 are time constants, T1 = 12.8 (seconds), T2 = 20.0 (seconds), L is dead time, L = 2 .0 (seconds).

However, the above equation is the standard setting value in the process model shown in FIG.
), and the tank model itself is nonlinear.

The specific method for determining numerical values varies depending on the process and is determined mainly based on experience.

FIGS. 8 to 11 show five different setting values SP.
FIG. 4 is a diagram showing a comparison of the rise (step response) when the control according to the present invention and the conventional PID control are given. In these figures, ■ is the response when conventional linear PI control is performed, ■ is the response when only scheduling of PID parameters is performed, and ■ is the response when scheduling of PID parameters and setting values is performed as shown below according to the present invention. .

In the case of FIG. 8, PID parameter scheduling Kp: 0.8 → 1.0 → 0.1Ti:
42 → 42 → 23 Td: 0
→ 0 → 0 Setting value scheduling SP: 2.0 → 1.9 → 2.0 In the case of Figure 9, PID parameter scheduling Kp: 0.8 → 1.1 → 0.1Ti:
42 → 42 → 23 Td: 0
→ 0 → 0 Setting value scheduling SP: 3.0 → 2.9 → 3.0 In the case of Figure 10, PID parameter scheduling Kp: 0.8 → 1.2 → 0.1Ti:
42 → 42 → 23 Td: 0
→ 0 → 0 Setting value scheduling SP: 4.0 → 3.8 → 4.0 In the case of Figure 11, PID parameter scheduling Kp: 0.8 → 1.2 → 0.1Ti:
42 → 42 → 23 Td: 0
→ 0 → 0 Scheduling of set values SP: 5.0 → 4.5 → 5.0 From these figures, we can see that the method of scheduling the PID parameters and set values in ■ allows control with a good start-up without overshoot and vibration. It can be seen that this has been achieved.

[0036]

As described above, according to the present invention, scheduling of PID parameters is performed in addition to scheduling of set values. Control with even better start-up can be achieved. Therefore, the settling time until the process amount becomes stable is also shortened.

Furthermore, by applying fuzzy inference to the scheduling of PID parameters, parameters can be changed or switched smoothly, and setting values do not need to be precisely scheduled, making it possible to easily change setting values. Effects can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

FIG. 2 is a diagram showing the configuration of an SP changing section in the embodiment of FIG. 1;

FIG. 3 is a diagram showing the configuration of a PID parameter changing section in the embodiment of FIG. 1;

FIG. 4 is a flowchart showing the operation of the SP changing unit in FIG. 2;

FIG. 5 is a flowchart showing the operation of the PID parameter changing section in FIG. 3;

FIG. 6 is a diagram showing membership functions of fuzzy rules used in the example.

FIG. 7 is an explanatory diagram of a process model to which the present invention is applied.

FIG. 8 is a diagram showing a step response when the set value is 2.

FIG. 9 is a diagram showing a step response when the set value is 3.

FIG. 10 is a diagram showing a step response when the set value is 4.

FIG. 11 is a diagram showing a step response when the set value is 5.

FIG. 12 is a diagram showing a response when setting value scheduling is added to normal PID control.

FIG. 13 is a diagram showing a response when setting value scheduling is added to a controlled object that has a better start-up.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1...Control device, 2...SP generation part, 3...SP change part, 4...
PID parameter change unit, 5...PID calculation unit, 31...S
P change memory, 32...SP calculation unit, 41...PID parameter change memory, 42...PID parameter calculation unit, 51
...Valve, 52...Upper tank, 53...Lower tank, 54
...fuzzy controller.

Claims (3)

[Claims] 1. A control device that provides the controlled object with a manipulated variable obtained by performing proportional + integral + differential calculations on the deviation between an observed value from the controlled object and a set value serving as a control target value,
A control device characterized in that a manipulated variable with a good rise and no vibration is obtained by performing two operations: scheduling of the set value and scheduling of the parameters of the proportional + integral + differential calculation. 2. The control device according to claim 1, wherein the scheduling of parameters for the proportional + integral + differential calculations is performed by fuzzy inference. 3. A control device that provides the controlled object with a manipulated variable obtained by performing proportional + integral + differential calculations on the deviation between an observed value from the controlled object and a set value serving as a target value for control,
a set value generation unit that specifies the set value; a set value change unit that schedules the set value by fuzzy inference from the set value generated in the set value generation unit and the observed value; a PID parameter changing unit that schedules parameters for proportional + integral + differential calculations by fuzzy inference from the set value and the observed value; a set value output from the set value changing unit; and a parameter output from the PID parameter changing unit. and a PID calculation unit that calculates and outputs the manipulated variable from the observed value.