CN1026444C - PID control system - Google Patents
PID control system Download PDFInfo
- Publication number
- CN1026444C CN1026444C CN 88101903 CN88101903A CN1026444C CN 1026444 C CN1026444 C CN 1026444C CN 88101903 CN88101903 CN 88101903 CN 88101903 A CN88101903 A CN 88101903A CN 1026444 C CN1026444 C CN 1026444C
- Authority
- CN
- China
- Prior art keywords
- time
- variable
- pid controller
- response
- control
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Abstract
A PID controller system comprises a PID controller for PID controlling a process standing for an object to be controlled, and an automatic adjuster being responsive to a variable relating to the manipulation of the PID controller and the response shape of the control variable from the process to obtain necessary performance indexes and preparing, on the basis of the performance indexes, optimum control parameters for tuning the actual control variable to the control command value, the optimum control parameters being fedback from the automatic adjuster to the PID controller.
Description
The present invention relates to the proportional plus integral plus derivative controller in the FEEDBACK CONTROL, relate in particular to the PID control system that the wherein adjustment of pid control parameter can automatically perform.
The invention still further relates to the system of energy automatic adjusting pid control parameter, this system can be used in the process of pure delay time to the large percentage of time constant, these processes for example can be: burning control in the boiler and temperature control, PH neutralization and flow speed control in chemistry and the general industry.
In the past, the adjustment of the pid control parameter in proportion integration differentiation (PID) controller is manually carried out by the operator, and in the operating process, the operator observes the various variations of various control variable.Caused a problem like this: i.e. adjustment work becomes very time-consuming, and adjusts the different influence that the result can be subjected to the operator personal considerations.
On the other hand, proposed various systems according to control theory, comprising adding that one sets test signal setting the dynamic perfromance of controlled object for the Be Controlled object, and controlled variable is adjusted to optimum value according to setting as a result.Yet, in these are proposed, can reckon with that control variable fluctuates because used the setting test signal, thereby make quality decline or particularly in highly nonlinear equipment, disadvantageous abnormality will take place.In addition, unless the every change of the dynamic perfromance of controlled object once, set and detect also change thereupon, the optimum value of controlled variable just can not reach, and can cause the processing procedure of trouble like this.
As described in " expert's self tuning controller " " measuring technique " the 66th～72 in November, 1986, disclosed a kind of heuristic method (expert method).Wherein the adjustment of controlled variable has considered that the shape of the response of various control variable carries out.According to this method, the real response shape is compared with the basic response shape of a plurality of preprepareds, thereby from a plurality of regulation rules, select a best rule to tend to according to real response shape or its transient process, obtain a basic response shape that is complementary, pid control parameter can be corrected like this.Unfavorable place in this method is, can reckon with that the number of regulation rule has increased, and caused the increase of memory capacity.
For example, relevant as what describe in " based on the PID self-correcting of expert method " " measuring technique " in November in 1986 the 52nd page to the 59th page with this kind method.
When comprising that controlled object (for example process) and one are used for the device starting of PID controller of controlling object, when the control variable step of the input signal that puts on process changes, usually set up with pure delay time response and first-order lag characteristic the time response of process variable, and pid control parameter is according to the process gain K in these characteristics, pure delay time L and time constant T and adjust to optimum value.The method of the best adjustment of this pid control parameter based on step response comprises: Ziegler-Ni Kuersi (Ziegler-Nichols) is method and Qi En-Huo En-come (Chein-Hrones-Reswick) method of thinking gram (CHR) (ZN).These methods have been simplified to calculate still following shortcoming.For example: when pid control parameter is adjusted according to the ZN method, for the controlled object with pure delay time and second-order lag characteristic, the variation of control response is shown in the L/T=1.0 of the L/T=0.14 of Figure 1A and Figure 1B when command value office step changes.Say that again when pid control parameter was adjusted according to the CHR method, the variation of control response was shown in the L/T=5 of the L/T=1 of the L/T=0.14 of Fig. 3 A and Fig. 3 B and Fig. 3 C when command value office step changes.Like this, in the ZN method, control response just departs from stable limit value at the L/T=1 place, and in the CHR method, control response is pulsed at the L/T=5 place.This shows that when for L/T higher value being arranged, any of above-mentioned two kinds of methods all can not be realized Optimal Control.
On the other hand, universal method as the adjustment of pid control parameter, at for example proceedings of " based on the method for designing of the control system of the part knowledge of controlled object " instrument and control engineering Shi Xuehui, August the 5th in 1979, volume was the 4th, 549 pages/555 pages described department pattern matching methods.
Below with the main points of brief description department pattern matching method.The device of realization department pattern matching method with the block diagram form as shown in Figure 2.Referring to this Fig. 2, PID controller of label 1 expression, label 2 expressions are corresponding to a process of a controlled device, and 7 expressions are the transport function G of process
p(s) the process setting device of the value of putting, 11 are expressed as the transport function G that determines the PID controller
pThe controlled variable determiner of the controlled variable optimum value (s).
According to the department pattern matching method, the controlled variable of PID controller 1 is definite like this, makes the closed loop transfer function, W(S that comprises command value SV and control variable PV) with the transport function G of the reference model of expression control variable PV ideal response
m(s, σ) consistent, wherein s is a Laplace operator, and σ is the time scale coefficient.The transport function G of the process 1 that obtains by process setting device 7
p(s) be expressed from the next:
The transport function G of PID controller 1
c(s) have following form:
Gc（S）＝Kp（1+ (l)/(Ti·s) +T
D·s） …（2.1）
＝ (Kp)/(Ti·s) （1+Ti·s+Ti·T
D·S
2）…（2.2）
K wherein
p, T
iAnd T
DBe controlled variable, they are known as proportional gain respectively, integral time and derivative time.
Therefore, closed loop transfer function, W(s) can be expressed from the next:
W（s）＝ (Gc（s）·Gp（s）)/(1+Gc（s）·Gp（s）) …（3）
The transport function Gm(s of reference model σ) can be provided by following formula:
Gm（s·σ）＝ 1/(1+σ·s+α
2(σ·s)
2+α
3(σ·s)
3+α
4(σ·s)
4)
…（4）
α in the formula
2, α
3, α
4Be the constant relevant with response wave shape.
By equation (3) and (4) are mapped the transport function G of PID controller
c(s) be simplified to:
Gc(s) = (Gm(s·σ))/(Gp(s){1-Gm(s·σ)})
= (g
0+g
1s+g
2s 2+g
3s 3+…)/(σ·s{1+α
2σ·s+α
3(σ·s)
2+α
4(σs)
3+…)
…（5）
, draw divided by molecule with the denominator in the equation (5)
Gc(s)= (g
0)/(σ·s) 〔(1+ (g
1)/(g
0) -α
2σ)s
+{ (g
2)/(g
0) -α
2(g
1)/(g
0) σ+(α
2 2-α
3)σ
2}s
2
+{ (g
3)/(g0) -α
2(g
2)/(g
0) σ+(α
2 2-α
3) (g
1)/(g
0) σ
2
+(2
α2α3-α
2 3-α
4)σ
3)s
3}+… 〕（6）
Consider that equation (1) equals equation (22), can draw following equation:
(Kp)/(Ti) = (g
0)/(σ) …（7）
Ti= (g
1)/(g
0) -α
2·σ …（8）
Ti·T
D= (g
2)/(g
0) -α2 (g
1)/(g
0) σ+(α
2 2-α
3)σ
2…（9）
0= (g
3)/(g
0) -α2 (g
2)/(g
0) σ+(α
2 2-α
3) (g
1)/(g0) σ
2
+(2
α2α3-α
2 3-α
4)σ
3…（10）
From equation (10), determine the positive minimum real root of σ, it by in the substitution equation (8) determining Ti, so positive minimum real root of determining and Ti by generation in equation (7) and (9) with definite K
pTo T
DAccording to this method, closed loop transfer function, W(s) and the reference model Gm(s of equation (4), the consistent bipyramid that continues until the S in the equation (4) σ).
Because the cubic algebra equation shown in equation (10) has solved the problem of definite σ, this method needs complicated calculating, and these calculate, and computing is inappropriate for microcomputer.
The present invention tries hard to solve the problem in the above-mentioned prior art, and provide a PID controller system, wherein the adjustment work of being finished by the operator can be reduced in large quantities to prevent that adjusting the result is influenced by each operator's individual character, the best adjustment of controlled variable needn't be adopted the setting signal that will disturb controlled object, and the variation of the dynamic perfromance of controlled object can be detected rapidly and needn't be relied on any manual operation, and total energy keeps best control characteristic.
Another object of the present invention provides a kind of PID control system, even this system also can obtain fabulous control characteristic for having the process of big pure delay time with the ratio of time constant, even and the optimum value that can determine pid control parameter according to suitable simple calculations with guarantee the computing unit of PID controller system by low capacity for example microcomputer also can carry out computing.
According to the present invention, above-mentioned first purpose can be finished by a PID controller system, this system comprises: observe the control variable response shape that occurs when command value or set point change or external disturbance the put on system control variable response shape recognition device with definite necessary performance index, by fuzzy reasoning qualitatively the estimated performance index calculating or to estimate the controlled variable correcting device of the modified value of pid control parameter, and the controllability determination device that the controlled variable correcting device is moved.
As for realizing second purpose, observation according to habitual department pattern matching method is to be difficult to find the solution the cubic algebra equation to obtain to determine necessary of time scale coefficient, have now found that the controlled object that sets by pure delay time and first-order lag characteristic can be similar to by a multistage hysteresis characteristic, by the transport function of the closed loop transfer function, of control variable and reference model is mapped, determined time scale coefficient just can be similar to by the linear equation of pure delay time for command value.
By by approximate result, the optimum value of pid control parameter can be used the gain corresponding to process, and the function that pure delay time and time constant are formed is represented, and utilized particular kind of relationship to determine.
Figure 1A and 1B have represented to change by the step of command value in habitual PID controller when the controlled object with pure delay time and second-order lag characteristic the figure of the example of the prior art of being controlled resulting control response, wherein pid control parameter is determined with habitual ZN method, Figure 1A is L/T=0.14, and Figure 1B is L/T=1's;
Fig. 2 is a block scheme of realizing controlling the prior art device according to the department pattern matching method;
Fig. 3 A, 3B, 3C show the figure that changes the prior art example of the control response that obtains when being controlled when the controlled object with pure delay time and second-order lag characteristic in habitual PID controller by the step of command value, wherein pid control parameter is to determine with habitual CHR method, Fig. 3 A illustrates L/T=0.14, Fig. 3 B illustrates L/T=1, and Fig. 3 C is L/T=5;
Fig. 4 is the block scheme according to first embodiment of the present invention explanation PID controller system;
Fig. 5 A and 5B are the synoptic diagram that the subordinate function of estimation overshoot is described respectively and estimates the subordinate function of damping ratio, and these subordinate functions are used in the embodiment shown in fig. 4;
Fig. 6 is the example of explanation regulation rule;
Fig. 7 is the synoptic diagram that is used for the subordinate function of controlled variable correction factor;
Fig. 8 is used to illustrate the synoptic diagram of determining controlled variable correction factor method;
Fig. 9 is the process flow diagram that functional steps is adjusted in explanation automatically;
Figure 10 is the synoptic diagram of explanation according to automatic adjustment result of the present invention;
Figure 11 is the block scheme that PID controller system according to a second embodiment of the present invention is described;
Figure 12 A and 12B illustrate when the command value step changes resulting control variable response shape respectively and the synoptic diagram of resulting control variable response shape when adding that outside step is disturbed;
Figure 13 A, 13B and 13C are the subordinate functions that the estimation overshoot is described respectively, the subordinate function of estimation damping ratio and the synoptic diagram of the subordinate function in estimation cycle;
Figure 14 is the synoptic diagram of an example of the regulation rule of the expression subordinate function that is used for respective type;
Figure 15 is the synoptic diagram of the subordinate function of expression controlled variable correction factor;
Figure 16 is the graphical representation that is used to illustrate the mode of determining the controlled variable correction factor;
Figure 17 is the process flow diagram of the step of another automatic regulatory function;
Figure 18 A adjusts the synoptic diagram of response automatically when being the big overshoot that occurs of explanation and vibration in initial control response shape;
Figure 18 B is the synoptic diagram of the automatic adjustment response the when overdamping that occurs in initial control response shape is described;
Figure 19 is the block diagram that PID controller system according to another embodiment of the present invention is described;
Figure 20 is the synoptic diagram how explanation interrelates L/T and σ/T and various reference model when the applying portion model matching method is determined pid control parameter for the controlled object with pure delay time and first-order lag characteristic;
Figure 21 A, 21B and 21C be illustrate respectively when pure delay time and second-order lag characteristic at L/T=0.4, be subjected to the synoptic diagram of the resulting control response of control that the step of command value changes under L/T=1 and the L/T=5 condition.
Below with reference to accompanying drawings, the present invention will be described by example.
Shown in Fig. 4 block scheme, comprise PID controller 1, controlled device 2 and automatic regulator 3 according to the PID controller system of the first embodiment of the present invention.
The PID that this PID controller 1 is carried out by comparison order value SV and the resulting control residual quantity of control variable PV e calculates, and produces the PID operation result as manipulated variable MV that is added on the controlled device 2.Automatic regulator 3 comprises control variable response shape recognizer 4, control determinant 5 and controlled variable corrector 6, and this controlled variable corrector 6 comprises controlled variable correction factor estimator 6a, regulation rule device 6b and controlled variable adjustment calculation device 6c.
Describe above-mentioned each parts below in detail.Control variable response shape recognizer 4 is observed or is checked command value SV or set point and control variable PV, and begins to observe the control variable response after control variable PV is adjusted to command value SV when control residual quantity e surpasses a predetermined value.When beginning to observe, recognizer 4 recovers the extreme value of (retrieve) PV and stops to observe when control variable PV adjusts to command value SV, and then recognizer 4 calculates overshoot and damping ratio according to the varying width of a plurality of extreme values that obtain and command value SV.
When not having the extreme value of control variable PV, set negative pseudo-value.If the overshoot of gained and damping ratio determine that at its permissible range inner control determinant 5 controlled variable is an optimum value respectively, and finish to adjust.If any in overshoot and the damping ratio be not in its permissible range, determinant 5 triggers controlled variable corrector 6.
According to fuzzy reasoning opinion method controlled variable correction factor estimator 6a is described below,, has determined subordinate function as Fig. 5 A and 5B in order to estimate overshoot and damping ratio amplitude qualitatively.In Fig. 5 A and 5B, E(i) (i=1～5) and D(i) (wherein i=1～3) be the constant of regulation subordinate function shape, PB, ZE and NB are the titles that is assigned to subordinate function according to qualitative estimation amplitude viewpoint.These names have following meaning:
PB: forward is big
ZE: zero
NB: negative sense is big
In these accompanying drawings, ordinate represents to represent the degree of membership G of quantitative grade, and the mark among Fig. 6 is to represent example by the regulation rule of the regulation rule device 6b of the PID control system of a plurality of ready control variable response shapes with subordinate function.
For example, referring to rule 1, the meaning of this rule is that " if E is PB and D is PB; then CKP is NB, and CTD is that PB CTD is PB " " if statement " is called as conditional statement " then statement " and is called as concluding sentence, and wherein E is the correction factor of overshoot, D is the correction factor of damping ratio, CKP is the correction factor of proportional gain, and CTI is the correction factor of integral time, and CTD is the correction factor of derivative time.Fig. 7 is the subordinate function synoptic diagram that explanation is used for qualitative definite controlled variable correction factor is transformed into quantitative values.In Fig. 7, C(i) (wherein i=125) is the constant of regulation subordinate function shape, PB, ZE and NB are the titles of giving subordinate function according to the viewpoint assignment of qualitative representation controlled variable correction factor amplitude, and they are corresponding to the title among Fig. 5 A and the 5B, and ordinate is represented degree of membership.
For example, below, with reference to the applicable cases of rule 1 and 2, the method for determining the controlled variable correction factor is described.Pay special attention to Fig. 8, it has illustrated the method for determining the correction factor CKP of proportional gain according to fuzzy reasoning method.Utilization each subordinate function shown in Fig. 5 A and 5B is determined the overshoot E from control variable response shape recognizer 4
oWith damping ratio D
o Degree qualitatively.In rule 1, E
oBe G
Ep, D
oBe G
Dp, E is G in rule 2
EpAnd D
oBe D
Dz, a common factor (minimum value) calculates and carries out in each rule to determine the adaptability of every rule, is G to rule 1
EpAnd 2 are G to rule
DeThen, be included in subordinate function in the conclusion statement of every rule by the adaptability institute weighting of every rule, the subordinate function that is weighted carries out and collects (maximal value) calculating and the center-of-gravity value of collection result calculated is determined output valve CKP as the proportional gain correction factor.Integral time correction factor CTI and derivative time correction factor CTD output valve can determine by similar methods.
Controlled variable adjustment calculation device 6c multiply by the existing value of pid control parameter correction factor that obtains and pid control parameter mutually and determines existing adjusted value.
Fig. 9 has illustrated the process flow diagram of the treatment step in automatic regulator 3.In step 10, at a predetermined period, SV and PV are transfused to, and each SV and PV be when being transfused to, and a status indication of the treatment state of automatic regulator 3 was determined in the 11st step.The status indication of " 0 " has been represented the state of supervision or supervision and control response, and the status indication of " 1 " has represented to observe or estimate the state of control response, and the status indication of " 2 " has been represented the state of calculation control parameter.For the status indication of " 0 ", will judge that in the 12nd step whether the control difference surpasses a predetermined value, if the control difference surpasses predetermined value, then status indication is by set in the 13rd goes on foot, and step enters into the control response observation state.If the control difference surpasses predetermined value, the then monitor state that retentive control is responded.If status indication is judged as " 1 " in the 11st step, the extreme value of PV is resumed (retrieve) in the 14th step.This process is carried out for each PV input value, and till judgement PV has been adjusted to SV in the 15th step (observe and finish).After observe finishing, performance index (overshoot and damping ratio) judge by the result with the extreme value of recovering in 14 steps in the 16th step, and status indication is put " 2 " and step enters into the controlled variable computing mode in the 17th goes on foot.The explanation correspondence of above-mentioned treatment scheme the operational process of control response shape recognizer 4.
If status indication is " 2 " in the 11st step, then in the 18th step by according to the performance index of in 16 steps, judging whether the conclusion within permissible range judge whether observed control response best.This processing is corresponding to the working condition of control determinant 5.Not under the situation of the best only in control response, then the correction factor of controlled variable will be determined in the 19th step, and the adjusted value of controlled variable will be determined in the 20th step, and the control that the controlled variable adjusted value that obtains like this is used in the PID controller 1 is calculated.Step 16 is corresponding to the working condition of controlled variable correction factor estimator 6a and regulation rule device 6b, step 20 is corresponding to the working condition of controlled variable adjustment calculation device 6c, when control response in the 18th step is judged as the best, and handle when in the 20th step, finishing, then status indication is reset to " 0 " in the 21st step, and step is got back to the control response monitor state.
The PID controller system of present embodiment moves the controlled object with second-order lag characteristic and pure delay time response, and the result of generation as shown in figure 10.Especially, the time response of resulting control variable is as shown in figure 10 when command value SV changes.When SV begins the step variation, the change of SV is being followed in the initial control response of PV, its overshoot E is 50%, and damping ratio D is 0.6, after this resulting performance index of observations (overshoot and damping ratio) of the second trial that changes corresponding to the secondary step of SV can drop in the predetermined permissible range of performance index, show to obtain best the adjustment, therefore, be adjusted at off-test for the third time.
As mentioned above, the automatic adjustment of controlled variable can reach by the regulation rule than peanut.
Although for illustrative purposes, all subordinate functions are that they are not limited thereto with the triangle pattern, and they can be realized with rectangular curve or index curve and not influence essence of the present invention.In addition, can set the number of subordinate function if desired.
Referring to accompanying drawing 11 to Figure 18 A and 18B, the PID controller system according to second embodiment of the present invention will be described below.Present embodiment is that with the difference that combines first embodiment that Fig. 4 to 10 does period ratio is to be used as a controlled variable except overshoot and damping ratio.Because the square frame form of the composition as Figure 11 is equal to first embodiment shown in Figure 4 in fact, only except the former period ratio is revised as an additional observation parameter, reference number shown in Figure 4 adds that any represents block diagram or element corresponding to Figure 11 of the block diagram of Fig. 4 or element.The actual implementation status of observing and controlling will be described below.
Control variable response shape recognizer 4 ' is always checked command value SV and control variable PV, and begins to observe the control variable response after control variable PV has adjusted to command value SV when the control difference e surpasses a predetermined value.Simultaneously when observing beginning, the extreme value of recognizer 4 ' recoverys control variable PV, and end observation when control variable is adjusted to command value SV.Then, Time Calculation overshoot, damping ratio and the period ratio of recognizer 4 ' take place according to a plurality of extreme values of gained and extreme value.Definite method of these controlled variable is with 12A and 12B explain in conjunction with the accompanying drawings.Figure 12 A has illustrated as command value SV to especially constantly from Y
0Step changes to Y
1The time resulting control variable PV the example of time response, shown at moment t
1.t
2And t
3The time extreme value X that occurs
1, X
2And X
3, and at moment t
4In time, adjust to finish.Overshoot E, damping ratio D and period T are following to be provided:
E＝（X
1-Y
1）/（Y
1-Y
0）
D＝（X
3-X
2）/（X
1-X
2）
T＝t
3-t
1
Therefore, known periods T
1Preceding value, period ratio R can followingly represent:
R＝T
1/T
Figure 12 B has specified when the external disturbance example of time response of resulting control variable PV that is added to command value when remaining on the input end of the controlled object on the Yo, has shown at t
1, t
2, t
3And t
4Extreme value X constantly
1, X
2, X
3And X
4, and at t
5Adjust constantly and finish.Overshoot E, damping ratio D and period T can be provided by following formula:
E＝（Y
0-Y
2）/（X
1-X
0）
D＝（X
3-X
4）/（X
3-X
2）
T＝t
4-t
2
If when command value changes, in the control variable response shape, do not have extreme value to occur or, when external disturbance adds for the second time extreme value does not appear, then overshoot is put with negative value, and the cycle is set to zero.If in the control variable response shape, only have an extreme value to occur changing, two extreme values perhaps in the control variable response shape, only occur and disturb with response external with the response instruction value; Then the time difference with reference to the time of adjustment is used as the cycle.
Control determinant 5 ' differentiating resulting overshoot and damping ratio respectively in the scope that it is allowed the time, controlled variable is optimum value and finishes to adjust.If any in overshoot and the damping ratio not in its permissible range, then determinant 5 ' triggering controlled variable corrector 6 '.
To controlled variable correction factor estimator 6a ' be described according to fuzzy reasoning method below.For amplitude and the period ratio of estimating overshoot and damping ratio qualitatively, stipulated A, 13B, the subordinate function shown in the 13C as Figure 13.At Figure 13 A in 13C, E(i) (wherein i=1～5), D(i) (wherein i=1～3) and R(i) (wherein i=1～3) be the constant of regulation subordinate function shape, and PB, PM, ZE and NB are the titles that is assigned to subordinate function according to the viewpoint of qualitative estimation amplitude.Under the situation of first embodiment, these titles are that following meaning is arranged:
PB: forward is big
PM: forward is medium
ZE: zero
NB: negative sense is big
In these figure, ordinate is represented degree of membership G, and it has been represented qualitatively and has measured.
Label among Figure 14 be for the used regulation rule device 66 of pid control parameter ' an example of regulation rule, pid control parameter is for adopting subordinate function ready various control variable response shape defineds.
For example, referring to rule 2, the meaning of this rule is: " if E is PB; D is that PM and R are PR; then CKP is NR, CTI is that NB and CTD are ZE " " if statement " is called as conditional statement, and " then statement " is called as the conclusion statement, wherein E is the correction factor of overshoot, D is the correction factor of damping ratio, and R is the correction factor of period ratio, and CKP is the correction factor of proportional gain, CTI is the correction factor of integral time, and CTD is the correction factor of derivative time.Figure 15 shows the subordinate function that is used for qualitative definite controlled variable correction factor is transmitted into quantitative values.In Figure 15, C(i) (wherein i=1～5) are to be used for the constant of shape of regulation subordinate function, PB, ZE and NB represent to specify (distribution) to give the title of subordinate function according to the viewpoint of qualitative representation controlled variable correction factor amplitude, they are corresponding to Figure 13 A, used title among 13B and the 13C, ordinate is represented degree of membership.
For example, consider to utilize the situation of rule 2 and 3, the method for determining the controlled variable correction factor will be described below.Especially referring to accompanying drawing 16, it has illustrated according to the fuzzy reasoning method method of the correction factor CKP of certainty ratio gain really.From control variable shape recognizer 4 ' and overshoot E
o, damping ratio D
oWith period ratio R
oQualitative measure respectively by Figure 13 A, the subordinate function shown in 13B and the 13C is determined.In rule 2, E
oBe G
Ep, D
oBe G
Dm, R
oBe G
Rp, and in rule 3, E
oBe G
Ep, D
oBe G
Dm, R
oBe G
RzA common factor (minimum value) calculates and carries out in rule separately, and to determine each regular adaptability, each rule is G for rule 2
Rp, then be G for rule 3
RzThen, be included in subordinate function in the conclusion statement of every rule by the adaptability institute weighting of every rule, the center-of-gravity value that the subordinate function after being weighted carries out and collection (maximal value) computing and collection result of calculation is confirmed as proportional gain correction factor K
PoOutput valve.Integral time correction factor T
1The output valve of correction factor CTD can be determined by similarity method with derivative time.
Controlled variable adjustment calculation device 6c ' multiply by the existing value of pid control parameter correction factor that obtains and PID control coefrficient mutually and determines present adjusted value.
Figure 17 be automatic regulator 3 ' in the process flow diagram of treatment step.
Figure 17 show automatic regulator 3 ' in the schematic flow sheet of treatment step.The 10th ' step in, be transfused at a predetermined period SV and PV, and in the moment that SV and PV are transfused at every turn, show automatic regulator 3 ' status indication decision in the 11st ' step of process status.Status indication " 0 " has been represented the state of supervision or supervision and control response, and the status indication of " 1 " represents to observe or estimate the state of control response, and the status indication of " 2 " is represented the state of calculation control parameter.For status indication is the situation of " 0 ", will judge in the 12nd ' step whether the control difference surpasses a predetermined value.If the control difference surpasses predetermined value, by set, program enters into the control response observation state to status indication in the 13rd ' step.If the control difference does not surpass predetermined value, the control response monitor state will be held.If status indication is judged as " 1 " in the 11st ' step, then the extreme value of PV is resumed in the 14th ' step.Such processing is all carried out for each PV input value, up to the 15th ' be judged as PV adjusted (observe and finish) till the SV in the step.Under situation about observe finishing, performance index (overshoot and damping ratio) will utilize the 14th ' go on foot the result of the extreme value of having recovered to determine in the 16th ' step, the 17th ' in, status indication is put " 2 ", and handling procedure enters into the controlled variable computing mode.More than the explanation handling procedure corresponding to control response shape recognizer 4 ' the course of work.
If the 11st ' step in status indication differentiated and be " 2 ", then the 18th ' in by utilize the 16th ' in the performance index judged whether decide with interior result whether observed control response is optimum value in its scope of allowing.This processing procedure corresponding to control determinant 5 ' the course of work.Only under control response is not best situation, the correction factor of controlled variable the 19th ' in be determined, and the adjusted value of controlled variable will be determined in the 20th ' step.The control that the controlled variable adjusted value that obtains like this will be used in the PID controller 1 is calculated.The 16th ' step, the 20th ' step was corresponding to the operation of controlled variable adjustment calculation device 6c ' corresponding to the course of work of controlled variable correction factor estimator 6a ' and regulation rule device 6b '.When the 18th ' step in control response be judged as the best, and the processing in the 20th ' step is when finishing, status indication is reset to " 0 " in 21 ' step, and treatment step is got back to the monitor state of control response.
The PID controller system of present embodiment is controlled for the controlled object with second-order lag and pure delay time response, and the result who is produced is shown in Figure 18 A and 18B.Especially, the time response of the control variable PV that obtains when command value SV changes is shown in Figure 18 A, the overshoot E that it has represented initial control response is 50% and damping ratio is an example of 0.6, and Figure 18 B has shown the example of initial control response under the overdamping situation.Be appreciated that in any one of these examples, best adjustment can reach by second trial.
As mentioned above, according to second embodiment of the present invention, not only the regulation rule of available peanut can reach the automatic adjustment of controlled variable, and can reach quick response regulation by the estimation control variable according to period ratio.Control variable response shape recognizer also can adopt the 4th performance index, control variable corresponding to the rise time of above-mentioned control variable, the for the first time preceding value of extreme value time of occurrence and the ratio between the existing value, the perhaps preceding value of the adjustment time of control variable and the ratio between the existing value, the subordinate function of controlled variable corrector 6 ' available triangular form, but the form of subordinate function is not limited thereto, and it can be rectangular curve or index curve and do not influence essence of the present invention.In addition, can set the number of subordinate function as required.
Referring to accompanying drawing 19 and its figure, PID controller system according to another embodiment of the present invention will be described below.As shown in Figure 19, the PID controller system of present embodiment comprises 1, one 2, one automatic regulator of process " 3 " as a controlled device of a PID controller "; it comprises process setting device 7, proportional gain counter 8, integral time counter 9 and derivative time counter 10.
Embodiment shown in Figure 19 is particularly suitable at the equipment initial period controlled variable of PID controller 1 being adjusted on optimum value.Process setting device 7 comes the dynamic perfromance of assignment procedure 2 by pure delay time/first-order lag characteristic, and according to the gain K of pure delay time and first-order lag characteristic deterministic process, pure delay time L and time constant T.Proportional gain counter 8 is determined proportional gain K
p, as handling the gain K, the pure delay time L that obtain and the function of time constant T from process setting device 7, integral time, counter 9 was determined T integral time it
I, and it as from the pure delay time L of process setting device 7 and the function of time constant T, derivative time, counter 10 was determined T derivative time
D, and it as from the pure delay time L of process setting device 7 and a function of time constant.
The pure delay set by process setting device 7 time/first-order lag characteristic G
p(S) as follows
Gp(s)= (K·e
- LS)/(1+T·s) …（11）
The pure delay time transfer function of equation (11) can be launched and is simplified to by Maclaurin expansion:
By dynamic characteristic of the course applying portion model matching method, can obtain following result to (10) from equation (7) to equation (12):
(Kp)/(Ti) = (g
0)/(σ) …（13）
Ti=L+T-α
2σ…（14）
Ti·T
D= (L
2)/2 +LT-α
2(L+T)σ+(α
2 2-α
3)σ
2
…（15）
0= (L
3)/6 + (L
2)/2 T-α
2( (L
2)/2 +LT)σ
+(α
2 2-α
3)(L+T)σ
2
+(2
α2α3-α
2 3-α
4)σ
3…（16）
Unless in equation (16), have the positive real root of σ, otherwise the department pattern matching method can not deal with to the represented instruction control response of equation (4).
Then, time ratio L/T is utilized various reference models to time constant and time scale coefficient ratio σ/T to the relation between the time coefficient and checks to obtain result as shown in figure 20 as parameter.The transport function of used reference model is as follows in the inspection:
(1) Kitamori model
Gm（s·σ）＝ 1/(1+σs+0.5(σs)
2+0.15(σs)
3+0.03(σs)
4)
…（17）
(2) Butter wooth model
Gm（s·σ）＝ 1/(1+σs+0.503(σs)
2+0.1479(σs)
3+0.02188(σs)
4)
…（18）
(3) Biromial model
Gm（s·σ）＝ 1/(1+σs+0.375(σs)
2+0.0625(σs)
3+0.003906(σs)
4)
…（19）
(4) ITAE least model
Gm（s·σ）＝ 1/(1+σs+0.4664(σs)
2+0.1067(σs)
3+0.01882(σs)
4)
…（20）
Variable σ is corresponding to the rise time of control response, and it will increase and increase with L.Like this,, be appreciated that the Kitamori model is used as reference model, can in the wide region of L/T from 0 to 10, obtain the rational value of σ/T by research Figure 20.This σ/T can be approximate by following formula:
(σ)/(T) ＝1.37· (L)/(T) …（21）
From equation (21) as can be seen, the time scale coefficient is reduced to:
σ＝1.37·L …（22）
Then, replace equation (22) equation (13) in (15), proportional gain K
p, integral time T
iWith T derivative time
DCan be defined as following form:
Kp＝ (0.215L+T)/(1.37K·L) …（23）
Ti＝0.315L+T …（24）
T
D= (0.315L·T+0.003L
2)/(0.315L+T) …（25）
Therefore, proportional gain counter 8 is according to K, and L and T calculation equation (23) are to determine proportional gain K
pOptimum value, counter 9 calculation equation integral time (24) are to determine T integral time
iOptimum value, counter 10 calculation equation derivative time (25) are to determine T derivative time
DOptimum value, guarantee the proportional gain K of PID controller 1 with this
p, integral time T
IWith T derivative time
DBe set respectively and revise.
In equation (25), because L is littler than T, so L
2Can ignore, derivative time T
DCan calculate according to following formula:
T
D＝ (0.315L·T)/(0.315L+T) …（26）
According to embodiment shown in Figure 19, have as Figure 1A, 1B, 3A, 3B, the control response of the process of pure delay time shown in the 3C and second-order lag characteristic is examined by the command value that changes unit step, to obtain the L/T=0.4 shown in Figure 21 A, the result of L/T=5 shown in L/T=1 shown in Figure 21 B and Figure 21 C even proved when L/T is big, also can obtain the better controlled characteristic.
Although the explanation of the foregoing description is carried out with hardware, purpose is to make explanation clear, understands that significantly they are under the help of computing machine, realize with software with similar methods.
Claims (9)
1, a kind of PID controller system comprises:
One receives the residual quantity between set point and the controlled process control variable and a performance variable is offered the PID controller of this process, and described process is subjected to the work of described performance variable in order to produce described control variable; With
Automatic regulating apparatus is characterized in that, this adjusting gear comprises:
Be used to receive the control variable of set point and described process, and determine the control variable response shape recognition device of a plurality of performance index according to the response shape of the control variable that when set point changes or applies disturbance, obtains;
A controlled variable correcting device, it has a plurality of estimating devices that utilize a plurality of performance index, this estimating device illustrates the amplitude of a plurality of performance index, determine the amplitude of controlled variable correction factor and by using the described performance index of a plurality of expressions and, calculating the controlled variable correction factor by the value of using a plurality of qualitative estimations through fuzzy reasoning by the regulation rule of mutual relationship between the determined parameter correction factor of the value of a plurality of qualitative estimations;
Described automatic regulating apparatus is according to relevant with the manipulation of a described PID controller variable with from the control variable obtained performance index of described process, and according to described performance index for described control variable being adjusted to described set point preparation optimization control parameter, described optimization control parameter feeds back to described PID controller from described automatic regulating apparatus.
2, PID controller system as claimed in claim 1 is characterized in that: described a plurality of performance index are overshoot and damping ratio.
3, PID controller system as claimed in claim 1 is characterized in that: described a plurality of performance index are overshoot, the preceding value of damping ratio and indication cycle and the period ratio of the ratio between existing value.
4, PID controller system as claimed in claim 1 is characterized in that: described a plurality of performance index are overshoot, the rise time ratio of ratio between damping ratio and the preceding value of expression rise time and existing value.
5, PID controller system as claimed in claim 1 is characterized in that: described a plurality of performance index are adjustment time ratios of ratio between overshoot, damping ratio and the preceding value of expression adjustment time and existing value.
6, a kind of PID controller system comprises:
Residual quantity between control variable that receives set point and controlled process also offers the PID controller of this process to a performance variable, and described process is subjected to the work of described performance variable in order to produce described control variable, it is characterized in that this controller also comprises:
The automatic regulating apparatus that comprises the process estimating device, this process estimating device are used for controlling described process and are used to be created in the gain that this process is used, pure delay time and time constant characteristic;
In response to gain, pure delay time and the time constant characteristic of described process estimating device generation, with the proportional gain calculation element of preset proportion gain;
In response to the pure delay time and the time constant characteristic of described process estimating device generation, to set calculation element integral time of integral time; With
In response to the pure delay time and the time constant characteristic of described process estimating device generation, to set calculation element derivative time of derivative time.
7, PID controller system as claimed in claim 6 is characterized in that: described integral time, calculation element was determined integral time to pure delay time response and the addition of time constant characteristic; Described proportional gain calculation element is that determined proportional gain divided by the product term of described gain and pure delay time response the integral time that obtains to add up; And described derivative time, calculation element was divided by determining derivative time the integral time that obtains by adding up the product term of described gain and pure delay time response.
8, PID controller system as claimed in claim 6 is characterized in that the output K of described proportional gain calculation element
p, described integral time calculation element output Ti and described derivative time calculation element output T
DDetermine by following formula respectively:
Kp＝ (0.315L+T)/(1.37K·L)
Ti＝0.315L+T
T
D＝ (0.315L·T+0.003L
2)/(0.315L+T)
Wherein K is a gain characteristic, and L is the pure delay time response, and T is the time constant characteristic.
9, PID controller system as claimed in claim 6, it is characterized in that, described automatic regulating apparatus is according to relevant with the manipulation of a described PID controller variable with from the control variable obtained performance index of described process, and according to described performance index for described control variable being adjusted to described set point preparation optimization control parameter, described optimization control parameter feeds back to described PID controller from described automatic regulating apparatus.
Applications Claiming Priority (8)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP80976/87 | 1987-04-03 | ||
| JP62-80976 | 1987-04-03 | ||
| JP62080976A JPS63247801A (en) | 1987-04-03 | 1987-04-03 | Pid controller |
| JP62-201733 | 1987-08-14 | ||
| JP201733/87 | 1987-08-14 | ||
| JP62237201A JP2585294B2 (en) | 1987-09-24 | 1987-09-24 | PID controller |
| JP62-237201 | 1987-09-24 | ||
| JP237201/87 | 1987-09-24 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1036278A CN1036278A (en) | 1989-10-11 |
| CN1026444C true CN1026444C (en) | 1994-11-02 |
Family
ID=26421943
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 88101903 Expired - Fee Related CN1026444C (en) | 1987-04-03 | 1988-04-02 | PID control system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1026444C (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110529874A (en) * | 2019-09-03 | 2019-12-03 | 杭州和利时自动化有限公司 | A kind of boiler oil investment amount control method, system and associated component |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008079441A (en) * | 2006-09-22 | 2008-04-03 | Matsushita Electric Ind Co Ltd | Motor controller and controller comprising the same |
| CN1945468B (en) * | 2006-09-30 | 2010-12-08 | 中国科学院电工研究所 | Self searching optimum control of scanning type PI parameter |
| JP5075701B2 (en) * | 2008-03-25 | 2012-11-21 | アズビル株式会社 | Control device and power estimation method |
| CN101533281B (en) * | 2008-05-28 | 2012-01-18 | 大连连城数控机器股份有限公司 | Multi-wire cutting machine mortar temperature intelligent PID control method |
| DE102010001203B4 (en) * | 2010-01-26 | 2011-12-08 | Ford Global Technologies, Llc | Regulation arrangement and procedure |
| KR101849509B1 (en) | 2012-12-07 | 2018-05-28 | 오므론 가부시키가이샤 | Adjustment device, control method, and medium of control program |
| DE102016011865A1 (en) * | 2016-10-01 | 2018-04-05 | Khd Humboldt Wedag Gmbh | Regulating device with adjustability of the control behavior |
-
1988
- 1988-04-02 CN CN 88101903 patent/CN1026444C/en not_active Expired - Fee Related
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110529874A (en) * | 2019-09-03 | 2019-12-03 | 杭州和利时自动化有限公司 | A kind of boiler oil investment amount control method, system and associated component |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1036278A (en) | 1989-10-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1044522C (en) | Process control device | |
| CN1839356A (en) | PID parameter adjustment device | |
| CN1026444C (en) | PID control system | |
| CN1080409A (en) | The two-phase method that is used for real-time process control | |
| CN1490691A (en) | Arrangement of integrated module predicting control and optimizing block and observation display | |
| CN1894637A (en) | Use of core process models in model predictive controller | |
| CN1066618A (en) | The statistical Process Control that is used for air separation process | |
| CN1192222A (en) | Method for controlling chemical synthesis processes | |
| CN1975611A (en) | Constraint and limit feasibility handling in a process control system optimizer | |
| CN1484118A (en) | Model-free self-adaptation of process controller | |
| CN1300651C (en) | Constraint and limit feasibility process in process control system optimizer procedure | |
| CN1082203C (en) | Process controlling method and apparatus | |
| CN101078913A (en) | Posterior control and multi-sensor infusion method and central conditioner energy-saving control device | |
| CN1460908A (en) | Control equipment | |
| CN1418172A (en) | Advanced control strategies for chlorine dioxide generating processes | |
| CN1451106A (en) | Controlling device and controlling method | |
| CN115591493A (en) | Reaction kettle temperature control method | |
| CN1099060C (en) | Universal multi-variable quantity model pre-estimating coordinating control method | |
| CN1447931A (en) | Controlling device and controlling method | |
| CN100347542C (en) | Compound-fertilizer nutrient-content soft-measuring modeling method beased on hybrid multiple models | |
| CN1687921A (en) | Rare-earth cascade extraction separation component content soft measuring method | |
| CN1758161A (en) | Optimum control method based on non-linear restraint predic control | |
| CN111650829A (en) | Embedded PID module parameter adjusting method, system and device based on artificial intelligence | |
| JP2585294B2 (en) | PID controller | |
| CN1724810A (en) | Width position corresponding relation deciding method and manufacturing device using it |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C06 | Publication | ||
| PB01 | Publication | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |