CN105867125A - Optimization control method and apparatus of refining apparatus coupling unit - Google Patents

Abstract

The invention discloses an optimization control method and apparatus of a refining apparatus coupling unit. The method comprises the following steps: obtaining a closed-loop dynamic feedback decoupling compensator D(s) paired based on a relative gain matrix loop, decoupling a controlled object G(s), and converting the G(s) into a generalized object G(s)<^>; designing an inner-mold controller C(s) for the generalized object G(s)<^>; calculating a fuzzy input amount, a process output and process model output deviation e(t) and a deviation change rate e(t)<.>; determining a fuzzy rule, and obtaining a fuzzy relation between an inner-mold controller parameter and e(t) and e(t)<.>; correcting the value of a controller parameter lambda i in an online mode through performing analyzing, processing, table-querying and operation on the fuzzy rule according to the following formula: lambda i= lambda i0+{Ei,ECi}lambda i, wherein lambda i0 is an initial value set for the filter time constant lambda i according to an identification result of an object model parameter and operation experience, and {Ei,ECi} lambda i represents online correction on the lambda i after fuzzy logic reasoning calculation. The method can realize full-flow automatic control of a controlled system and further inhibit interference better.

Description

The optimal control method of refinery device coupling unit and device

Technical field

The present invention relates to control technical field, particularly relate to optimal control method and the dress of a kind of refinery device coupling unit Put.

Background technology

Currently, there are the oil refining of set more than 1000 and chemical production device in the whole nation, due to reaction complexity in chemical process, in oil refiningization The production scene of work, exists that automatic control rate is low, bad, the plant operator labor intensity that puts into the control loop effect automatically run The problem such as greatly, plant running is unstable.For these problems, numerous scientific research personnel never ipsilateral has carried out various research, Problem is attributed to a lot of sides such as " large dead time ", " non-linear ", " object is complicated ", " interference is frequently ", " PID controller is the most handy " Face, and propose numerous solutions, but still do not solve common problem.Single especially for its coupling Unit, owing to there is correlation between each loop, simple unity loop control can not reach preferably to control effect, engineering staff Generally Multivariable decoupling is controlled for separate single loop object.But often there is decoupler and set in conventional decoupling method Meter is complicated, can not full decoupled, control the problems such as effect is undesirable.

In production scene, device when carrying out the amount of putting forward, fall amount according to the market demand, raw material state, factory's schedule etc., as Really a lot of loop of device is in manually, needs the amount of operative employee's intensive adjusting apparatus each position: such as furnace outlet temperature, tower Charging and liquid level, tower backflow etc., operative employee's labor intensity is big, device control accuracy is low, fluctuation is big, is also unfavorable for saving energy and reduce the cost. Even if when device normal operating, owing to having that manual loop is more, part put into that automatic loop effect is bad, device and list It is not reasonable and perfect that element device controls loop, and the labor intensity of operative employee is the biggest, and device exists fluctuation, control accuracy relatively Low, energy consumption is higher.If implement full-range automatically, operative employee labor intensity, device smooth operation just can be greatly lowered high Effect, energy consumption are reduced.

Internal model control is as a kind of Novel Control in Dynamic matrix control, and because of it, to have design principle the easiest, and It is prone to the features such as adjustment, strong robustness, is widely applied in complex industrial process multivariable Control.But internal model control is a kind of Control method based on plant model, the order of accuarcy of model, the change of parameter, all can directly affect systematic function.And Be often time-varying in industry spot object parameters, the mismatch of model thus constantly change, and in the process of control In be constantly present random disturbances, if using fixing controller parameter, the precision of control is just difficult to be guaranteed.Fuzzy Control System is a kind of Based Intelligent Control reflecting wisdom of humanity thinking, it is easy to processes and has complexity, the controll plant of ambiguity or system, Without knowing the mathematical models of controlled device, it is only necessary to provide expertise and the experience of skilled operation personnel or on-the-spot behaviour Make data, so that its control mechanism and decision-making should be readily appreciated that and receive, be beneficial to application and promote.But the essence of fuzzy control Degree is limited by empirical rule and quantification gradation, it addition, for common fuzzy control, it is similar to proportion differential Control mode, easily produces the steady-state error of non-zero, so it belongs to droop control.

Summary of the invention

The purpose of the present invention is intended to solve one of above-mentioned technical problem the most to a certain extent.

To this end, the first of the present invention purpose is to propose the optimal control method of a kind of refinery device coupling unit, should Method is capable of the whole process of controlled system and automatically controls, and the most preferably suppresses interference.

Second object of the present invention is to propose the optimal control device of a kind of refinery device coupling unit.

For reaching above-mentioned purpose, first aspect present invention embodiment proposes the optimal control of a kind of refinery device coupling unit Method, comprises the following steps: obtain closed-loop dynamic feedback decoupling compensator D (s) based on the pairing of Relative increasing rate loop right Controlled device G (s) decouples, and described G (s) is changed into generalized objectFor described generalized objectDesign Its internal mode controller C (s)；Calculate Indistinct Input amount, the output of process and process model output bias e (t) and deviation variation rateDetermine fuzzy rule, obtain internal mode controller parameter and e (t) andBetween fuzzy relation；By to described fuzzy The analysis of rule, process, table look-up and computing carrys out on-line amending controller parameter λ_iValue, revises and carries out according to following formula:Wherein, λ_i0When coming wave filter for the identification result according to object model parameter and operating experience Between constant, λ_iSet an initial value,Represent after fuzzy logic inference calculates λ_iOn-line amending.

The optimal control method of refinery device coupling unit according to embodiments of the present invention, design is based on Relative increasing rate Closed-loop dynamic feedback decoupling compensator controlled device is decoupled, for decoupling after generalized object characteristics design in mould control Device processed, finally utilizes fuzzy control principle, controller parameters setting, and the method is capable of the whole process of controlled system and automatically controls System, the most preferably suppresses interference.

In one embodiment of the invention, the closed-loop dynamic feedback that described acquisition is matched based on Relative increasing rate loop Controlled device G (s) is decoupled by decoupling compensator D (s), and described G (s) is changed into generalized objectSpecifically include: institute State Relative increasing rate (RGA) in the several control loops intercoupled, select the controlled major loop T of each passage_i-k, i.e. by Kth input controls i-th output；I-th row kth column element D of described closed-loop dynamic feedback decoupling device D (s)_ikS () element sets It is set to 1, i.e. D_ik(s)=1；Described closed-loop dynamic feedback decoupling device D (s) removes D_ikS the element beyond () is designed as:When described control object contains non-minimum phase bit position, described closed-loop dynamic feedback is solved Coupling device D (s) carries out time lag compensation, and described closed-loop dynamic feedback decoupling device D (s) after compensation is represented by:Wherein, τ_ik=τ (G_ik)-τ_i,Ensure described generalized object Relative increasing rateEither element all not less than zero, i.e.Wherein,

In one embodiment of the invention, when described control object contains non-minimum phase bit position, described closed loop is moved State feedback decoupling device D (s) carries out time lag compensation and specifically includes: calculate τ_i:Determine generalized ensemble element H_ik (s):Wherein, G_ik-S () is G_ikS the minimum phase part of (), calculates D_ijTime lag τ of (s) (D_ij), and then determine decoupler element D_ij(s): τ (D_ij)=τ (G_ij)-τ(H_ik)。

For reaching above-mentioned purpose, second aspect present invention embodiment proposes the optimal control of a kind of refinery device coupling unit Device, it is characterised in that including: decoupling module, for obtaining closed-loop dynamic feedback based on the pairing of Relative increasing rate loop Decoupling compensator D (s) carries out decoupling to controlled device G (s) and changes into generalized objectDesign module, for described broad sense ObjectDesign its internal mode controller C (s)；Control module, calculates fuzzy reason input quantity the output of process defeated with process model Deviate e (t) and deviation variation rateDetermine fuzzy rule, obtain internal mode controller parameter and e (t) andBetween mould Paste relation, by the analysis of fuzzy rule, process, table look-up and computing carrys out on-line amending controller parameter λ_iValue, revises foundation Following formula is carried out:Wherein, λ_i0It is right to come for the identification result according to object model parameter and operating experience Filter time constant λ_iSet an initial value,Represent after fuzzy logic inference calculates λ_iRepair online Just.

The optimal control device of refinery device coupling unit according to embodiments of the present invention, decoupling module designs based on relatively Controlled device is decoupled by the closed-loop dynamic feedback decoupling compensator of gain matrix, and design module is for the broad sense pair after decoupling As characteristics design internal mode controller, last control module utilizes fuzzy control principle, controller parameters setting, and this device can be real The whole process of existing controlled system automatically controls, and the most preferably suppresses interference.

In one embodiment of the invention, described decoupling module specifically for: described Relative increasing rate is in phase mutual coupling The several control loops closed select the controlled major loop T of each passage_i-k, i.e. controlled i-th output by kth input；Described I-th row kth column element D of closed-loop dynamic feedback decoupling device D (s)_ikS () element is set to 1, i.e. D_ik(s)=1；Described closed loop is moved State feedback decoupling device D (s) removes D_ikS the element beyond () is designed as:Described in j ≠ k, control object contains During non-minimum phase bit position, described closed-loop dynamic feedback decoupling device D (s) being carried out time lag compensation, the described closed loop after compensation is moved State feedback decoupling device D (s) is represented by:Wherein, τ_ik=τ (G_ik)-τ_i,Ensure described generalized objectRelative increasing rateEither element all not less than zero, i.e.Wherein,

In one embodiment of the invention, described decoupling module is additionally operable to: calculate τ_i:Determine wide Justice system element H_ik(s):Wherein, G_ik-S () is G_ikS the minimum phase part of (), calculates D_ij Time lag τ (the D of (s)_ij), and then determine decoupler element D_ij(s): τ (D_ij)=τ (G_ij)-τ(H_ik)。

Aspect and advantage that the present invention adds will part be given in the following description, and part will become from the following description Obtain substantially, or recognized by the practice of the present invention.

Accompanying drawing explanation

The present invention above-mentioned and/or that add aspect and advantage will become from the following description of the accompanying drawings of embodiments Substantially with easy to understand, wherein,

Fig. 1 is the flow chart of the optimal control method of the refinery device coupling unit according to one embodiment of the invention；

Fig. 2 is the structural representation of the decoupling and controlling system of the closed-loop dynamic feedback decoupling device according to one embodiment of the invention Figure；

Fig. 3 is the structural representation of the fuzzy controller according to one embodiment of the invention；

Fig. 4 is the structural representation of the optimal control device of the refinery device coupling unit according to one embodiment of the invention Figure.

Detailed description of the invention

Embodiments of the invention are described below in detail, and the example of described embodiment is shown in the drawings, the most from start to finish Same or similar label represents same or similar element or has the element of same or like function.Below with reference to attached The embodiment that figure describes is exemplary, it is intended to is used for explaining the present invention, and is not considered as limiting the invention.

Below with reference to the accompanying drawings optimal control method and the device of the refinery device coupling unit of the embodiment of the present invention are described.

Fig. 1 is the flow chart of the optimal control method of the refinery device coupling unit according to one embodiment of the invention.

As it is shown in figure 1, the optimal control method of this refinery device coupling unit may include that

S1, obtains closed-loop dynamic feedback decoupling compensator D (s) based on the pairing of Relative increasing rate loop to controlled device G (s) carries out decoupling and changes into generalized object

Wherein, such as, for given n × n dimension controlled device G (s), the RGA of system is:

Carry out loop pair principle according to Relative increasing rate, select λ_ijLoop close to 1 is optimum pairing loop as far as possible, Avoid λ_ij=0 or λ_ijThe loop of ＜ 0 is matched.And then can compensate according to pair principle design closed-loop dynamic feedback decoupling Device.

It should be noted that in an embodiment of the present invention, Relative increasing rate is in the several control loops intercoupled The controlled major loop T of each passage of middle selection_i-k, i.e. controlled i-th output by kth input；Closed-loop dynamic feedback decoupling device D The i-th row kth column element D of (s)_ikS () element is set to 1, i.e. D_ik(s)=1；Closed-loop dynamic feedback decoupling device D (s) removes D_ik(s) Element in addition is designed as:When control object contains non-minimum phase bit position, closed loop is moved State feedback decoupling device D (s) carries out time lag compensation, and closed-loop dynamic feedback decoupling device D (s) after compensation is represented by:Wherein, τ_ik=τ (G_ik)-τ_i,Ensure generalized object Relative increasing rateEither element all not less than zero, i.e.Wherein,

It should be noted that in an embodiment of the present invention, when control object contains non-minimum phase bit position, closed loop is moved State feedback decoupling device D (s) carries out time lag compensation and specifically includes: calculate τ_i:Determine generalized ensemble element H_ik (s),Wherein, G_ik-S () is G_ikS the minimum phase part of (), calculates D_ijTime lag τ of (s) (D_ij), and then determine decoupler element D_ij(s): τ (D_ij)=τ (G_ij)-τ(H_ik)。

For example, it is assumed that controlled device is 3 × 3, and after loop is matched, optimum control loop is T_i-i, i =1,2,3, the decoupling and controlling system block chart of closed-loop dynamic feedback decoupling device D (s) is as shown in Figure 2.As shown in Figure 2, system becomes Measure and meet following relation:

U (s)=D (s) M (s) (2)

Wherein,

$U\left(s\right)=\left[\begin{array}{c}{u}_1\\ u_2\\ u_3\end{array}\right],M\left(s\right)=\left[\begin{array}{c}{m}_1\\ m_2\\ m_3\end{array}\right],D\left(s\right)=\left[\begin{array}{ccc}1& -D_{12}\left(s\right)& -D_{13}\left(s\right)\\ -D_{21}\left(s\right)& 1& -D_{23}\left(s\right)\\ -D_{31}\left(s\right)& -D_{32}\left(s\right)& 1\end{array}\right]$

Then

M (s)=D^-1(s)U(s) (3)

This compensator is applied in control target, the generalized object after being decoupled

$\hat{G}\left(s\right)=G\left(s\right)D^{-1}\left(s\right)---\left(4\right)$

By D_ijS () brings into, and decompose D (s), has:

$D\left(s\right)=\left[\begin{array}{ccc}\frac{1}{G_{11}\left(s\right)}& 0& 0\\ 0& \frac{1}{G_{22}\left(s\right)}& 0\\ 0& 0& \frac{1}{G_{33}\left(s\right)}\end{array}\right]\left[\begin{array}{ccc}{G}_{11}\left(s\right)& G_{12}\left(s\right)& G_{13}\left(s\right)\\ G_{21}\left(s\right)& G_{22}\left(s\right)& G_{23}\left(s\right)\\ G_{31}\left(s\right)& G_{32}\left(s\right)& G_{33}\left(s\right)\end{array}\right]---\left(5\right)$

D (s) is inverted, has

D^-1(s)=G^-1(s)G_Λ(s) (6)

Wherein,

$G_{\mathrm{\&Lambda;}}\left(s\right)={\left[\begin{array}{ccc}{G}_{11}^{-1}\left(s\right)& 0& 0\\ 0& G_{22}^{-1}\left(s\right)& 0\\ 0& 0& G_{33}^{-1}\left(s\right)\end{array}\right]}^{-1}=\left[\begin{array}{ccc}{G}_{11}\left(s\right)& 0& 0\\ 0& G_{22}\left(s\right)& 0\\ 0& 0& G_{33}\left(s\right)\end{array}\right]$

And then the object after can being decoupled is:

$\hat{G}\left(s\right)=G_{\mathrm{\&Lambda;}}\left(s\right)=\left[\begin{array}{ccc}{G}_{11}\left(s\right)& 0& 0\\ 0& G_{22}\left(s\right)& 0\\ 0& 0& G_{33}\left(s\right)\end{array}\right]---\left(7\right)$

If open-loop transfer function H (s) after using closed-loop dynamic Decoupling nonlinear feedback decoupling device that system carries out decoupling compensation, Then from the character of compensator:

$\begin{array}{c}H\left(s\right)=\frac{Y\left(s\right)}{U\left(s\right)}=G\left(s\right)D^{-1}\left(s\right)=diag\{H_{11},H_{22},H_{33}\}\\ =diag\{G_{11},G_{22},G_{33}\}\end{array}$

When controlled device comprises time lag part, in order to ensure D_ijThe realizability of (s), its time lag must meet:

$\mathrm{\&tau;}\left(D_{ij}\right)\&GreaterEqual;0,\&ForAll;i,j\&Element;n,i\&NotEqual;j$

ByUnderstand,

τ(D_ij)=τ (G_ij)-τ(G_ii)≤τ(G_ij)-τ_i (8)

Wherein,

From the decoupling-structure analysis of closed-loop dynamic feedback decoupling device,

D (s)=H^-1(s)G(s) (9)

τ(D_ij) be represented by

τ(D_ij)=τ (G_ij)-τ(H_ii) (10)

τ (D to be ensured_ij) >=0, as long as making the minimum time lag that H (s) comprises control object, i.e.

τ(H_ii)=τ_i (11)

Thus the diagonal element H after decoupling_iiCan be expressed as:

$H_{ii}\left(s\right)=e^{-{\mathrm{\&tau;}}_{i}s}{G}_{ii-}\left(s\right)---\left(12\right)$

Wherein G_ii-S () is G_iiThe minimum phase part of (s).

It is to say, when control object contains minimum phase part, in order to ensure stability and the realizability of system, Decoupler D (s) must be carried out time lag compensation, decoupler D (s) after compensation is represented by

$D\left(s\right)=\left[\begin{array}{ccc}{e}^{-{\mathrm{\&tau;}}_{11}}& 0& 0\\ 0& e^{-{\mathrm{\&tau;}}_{22}}& 0\\ 0& 0& e^{-{\mathrm{\&tau;}}_{nn}}\end{array}\right]\left[\begin{array}{ccc}1& -D_{12}\left(s\right)& -D_{13}\left(s\right)\\ -D_{21}\left(s\right)& 1& -D_{23}\left(s\right)\\ -D_{31}\left(s\right)& -D_{32}\left(s\right)& 1\end{array}\right]---\left(13\right)$

The uneoupled control of multi-variable system to be realized, the decoupling controller of system to be designed, and system to be analyzed The stability of system.Only on the premise of ensureing system stability, design decoupler, just calculate and complete the decoupling control having showed system System.The character of internal model control shows, wants to realize the internal model control of system, it is only necessary to ensure that: one is controlled device at 2 Being stable, the internal mode controller that two is designed is stable.It follows that be able to demonstrate that employing class feed forward decoupling control is made Generalized object after withIt is stable, it is ensured that system can realize uneoupled control.

The Relative increasing rate of system can characterize an input impact on other output channels of system.Normal conditions Under, it is intended that it is that each element is the diagonally dominant matrix more than zero.If a certain element in matrix is less than zero, system There is negative coupling channel.When we other loops are carried out open and close ring cutting change time, may result in negative coupling channel and instability occur Phenomenon, thus the stability of whole system is produced impact.Thus, it is ensured that the generalized ensemble after Xie OuStablize Property, we have only to orderRelative increasing rateEither element all not less than zero.It is to say, forRelative increasing rate

${\mathrm{\&Lambda;}}^N\left(\hat{G}\right)=\left\{{\widehat{\mathrm{\&lambda;}}}_{ij}\right|i,j=1,2,...,n\}=\hat{G}\left(0\right)\&CircleTimes;{\&lsqb;{\hat{G}}^{+}\left(0\right)\&rsqb;}^T---\left(14\right)$

Only needJust the stability contorting of feasible system.

S2, for generalized objectDesign its internal mode controller C (s).

Specifically,Wherein,ForMinimum phase part, F (s), it is ensured that control The attainable low pass filter of device processed, has a following form:

$F\left(s\right)={\&lsqb;F_{ij}\left(s\right)\&rsqb;}_{n\&times;n},F_{ij}\left(s\right)=\left\{\begin{array}{cc}{F}_{ik}\left(s\right),& j=k\\ 0,& j\&NotEqual;k\end{array}\right.---\left(15\right)$

Wherein,λ_iFor unique adjustable parameter of controller, n_iFor Relative order Secondary.

In order to ensure the realizability of controller, n_iNeed sufficiently large.When system decoupling is diagonal matrix, it is possible to by monotropic The internal mode controller of amount system design control system.From analysis above, the internal model of system after decouplingFor

$\hat{G}\left(s\right)=H\left(s\right)=diag\{H_{11},H_{22},...,H_{nn}\}---\left(16\right)$

CauseContaining non-minimum phase bit position, so needing to be broken down into minimum phase partWith non-minimum Phase place portionThus obtain internal mode controller and be:

$C\left(s\right)={\hat{G}}_{-}^{-1}\left(s\right)F\left(s\right)---\left(17\right)$

Wherein, F (s)=diag{F₁₁,F₂₂,…,F_nnFor ensureing the attainable low pass filter of controller,λ_iFor unique adjustable parameter of controller, n_iFor relative order, in order to ensure controller Realizability, n_iNeed sufficiently large.

And then, can obtain internal mode controller is:

C (s)=diag{C₁₁(s),C₂₂(s),C₃₃(s)} (18)

Wherein,

S3, calculates fuzzy reason input quantity the output of process and process model output bias e (t) and deviation variation rate

Set up fuzzy rule, export Y according to the output of process Y (s) with process model_mDeviation E (s) of (s)=Y (s)-Y_m(s) On-line tuning is carried out with deviation variation rate EC (s).

S4, determines fuzzy rule, obtain internal mode controller parameter and e (t) andBetween fuzzy relation.

S5, by the analysis of fuzzy rule, process, table look-up and computing carrys out constantly on-line amending controller parameter λ_i Value, revises and carries out according to following formula:Wherein, λ_i0For the identification result according to object model parameter and Operating experience is come filter time constant, λ_iSet a suitable initial value,Represent through fuzzy logic inference To λ after calculating_iOn-line amending.

For example, structure of fuzzy controller is as shown in Figure 3.The controlled device of decoupling is 3 separate loop sets Become, controller can be designed by the control method of single-variable system.For i-th loop, filter parameter λ_iBasis of adjusting The output of process y_iY is exported with process model_miDeviation E_i=y_i-y_miWith deviation variation rate EC_iCarry out on-line tuning.To fuzzy The input quantity of controller: deviation E_iWith deviation variation rate EC_i, carry out seven grades of linguistic quantifications, its linguistic variable according to fuzzy set theory For NB, NM, NS, Z, PS, PM, PB}, represent respectively negative big, and negative in, negative little, zero, the least, center, honest }.E_iAnd EC_iPerson in servitude Genus degree function selects the form that triangle, Z-type, S-type membership function combine, and domain is carried out according to actual controlled device Select, λ_iMembership function be taken as trigonometric function.

λ_iInitial value λ_i0Determination with domain is based on a large amount of emulation to system, initial value λ_i0Model according to system loses Join degree and robust performance analysis determines, it is ensured that λ_iValue is not negative.For large dead time or the system of big inertia time constant, λ_iInitial value λ_i0Higher value should be taken, to ensure that system has stronger robustness, can be more stable when bringing into operation；Work as system When time lag or inertia time constant are less, λ_i0Can take is smaller, to ensure that system has response speed faster.λ_iLower bound The Robust Stability Analysis based on system determines, upper bound robust performance based on the system analysis of λ determines.Apply in reality In, for different process objects, it is also possible to adjust λ by changing quantizing factor_iDomain scope.

Adjust λ_iFuzzy rule set up according to being: deviation E exported when the output of process and process model_iWith deviation Rate of change EC_iTime the biggest, need bigger λ_i, when deviation E_iRate of change EC with deviation_iTime the least, need less λ_i.Then Suitably adjusted by substantial amounts of emulation experiment, finally give adjustment λ_iFuzzy reasoning table, first row represents E_i, the first row Represent EC_i, as shown in table 1:

Table 1 fuzzy control rule

In the actual application of industry spot, can come filtering according to the identification result of object model parameter and operating experience Device time constant λ_iSet a suitable initial value λ_i0, then by the analysis of fuzzy rule, process, table look-up and computing comes not Disconnected ground on-line amending λ_iValue, revises and carries out according to following formula:

${\mathrm{\&lambda;}}_i={\mathrm{\&lambda;}}_{i0}+{\{E_i,{\mathrm{EC}}_i\}}_{{\mathrm{\&lambda;}}_i}$

In formula,Represent after fuzzy logic inference calculates λ_iOn-line amending.

The optimal control method of refinery device coupling unit according to embodiments of the present invention, design is based on Relative increasing rate Closed-loop dynamic feedback decoupling compensator controlled device is decoupled, for decoupling after generalized object characteristics design in mould control Device processed, finally utilizes fuzzy control principle, controller parameters setting, and the method is capable of the whole process of controlled system and automatically controls System, the most preferably suppresses interference.

Corresponding with the optimal control method of the refinery device coupling unit that above-described embodiment provides, the present invention's is a kind of real Execute example and also provide for the optimal control device of a kind of refinery device coupling unit, the refinery device coupling provided due to the embodiment of the present invention The optimal control method closing unit is corresponding with the optimal control device of the refinery device coupling unit that above-described embodiment provides, because of This is also applied for, at the embodiment of the optimal control method of aforementioned refinery device coupling unit, the refinery dress that the present embodiment provides Put the optimal control device of coupling unit, be not described in detail in the present embodiment.Fig. 4 is according to one embodiment of the invention The structural representation of the optimal control device of refinery device coupling unit.As shown in Figure 4, the optimization of this refinery device coupling unit Control device and may include that decoupling module 10, design module 20 and control module 30.

Wherein, decoupling module 10 compensates for obtaining closed-loop dynamic feedback decoupling based on the pairing of Relative increasing rate loop Device D (s) carries out decoupling to controlled device G (s) and changes into generalized object

Design module 20 is for for generalized objectDesign its internal mode controller C (s).

Control module 30 is used for calculating fuzzy reason input quantity the output of process and becomes with process model output bias e (t) and deviation RateDetermine fuzzy rule, obtain internal mode controller parameter and e (t) andBetween fuzzy relation, by fuzzy The analysis of rule, process, table look-up and computing carrys out on-line amending controller parameter λ_iValue, revises and carries out according to following formula:

${\mathrm{\&lambda;}}_i={\mathrm{\&lambda;}}_{i0}+{\{E_i,{\mathrm{EC}}_i\}}_{{\mathrm{\&lambda;}}_i}$

Wherein, λ_i0Come filter time constant λ for the identification result according to object model parameter and operating experience_iSet One initial value,Represent after fuzzy logic inference calculates λ_iOn-line amending.

In one embodiment of the invention, decoupling module 10 is several intercouple specifically for Relative increasing rate Control loop selects the controlled major loop T of each passage_i-k, i.e. controlled i-th output by kth input；Closed-loop dynamic is fed back I-th row kth column element D of decoupler D (s)_ikS () element is set to 1, i.e. D_ik(s)=1；Closed-loop dynamic feedback decoupling device D (s) Except D_ikS the element beyond () is designed as:When control object contains non-minimum phase bit position, Closed-loop dynamic feedback decoupling device D (s) is carried out time lag compensation, and closed-loop dynamic feedback decoupling device D (s) after compensation is represented by:Wherein, τ_ik=τ (G_ik)-τ_i,Ensure generalized object Relative increasing rateEither element all not less than zero, i.e.Wherein,

In one embodiment of the invention, decoupling module 10 is additionally operable to control object when containing non-minimum phase bit position, Closed-loop dynamic feedback decoupling device D (s) is carried out time lag compensation specifically include: calculate τ_i:Determine broad sense system System element H_ik(s),Wherein, G_ik-S () is G_ikS the minimum phase part of (), calculates D_ij(s) Time lag τ (D_ij), and then determine decoupler element D_ij(s): τ (D_ij)=τ (G_ij)-τ(H_ik)。

The optimal control device of refinery device coupling unit according to embodiments of the present invention, decoupling module designs based on relatively Controlled device is decoupled by the closed-loop dynamic feedback decoupling compensator of gain matrix, and design module is for the broad sense pair after decoupling As characteristics design internal mode controller, last control module utilizes fuzzy control principle, controller parameters setting, and this device can be real The whole process of existing controlled system automatically controls, and the most preferably suppresses interference.

In describing the invention, it is to be understood that term " first ", " second " are only used for describing purpose, and can not It is interpreted as instruction or hint relative importance or the implicit quantity indicating indicated technical characteristic.Thus, define " the One ", the feature of " second " can express or implicitly include at least one this feature.

In the description of this specification, reference term " embodiment ", " some embodiments ", " example ", " specifically show Example " or the description of " some examples " etc. means to combine this embodiment or example describes specific features, structure, material or spy Point is contained at least one embodiment or the example of the present invention.In this manual, to the schematic representation of above-mentioned term not Identical embodiment or example must be directed to.And, the specific features of description, structure, material or feature can be in office One or more embodiments or example combine in an appropriate manner.Additionally, in the case of the most conflicting, the skill of this area The feature of the different embodiments described in this specification or example and different embodiment or example can be tied by art personnel Close and combination.

Although above it has been shown and described that embodiments of the invention, it is to be understood that above-described embodiment is example Property, it is impossible to being interpreted as limitation of the present invention, those of ordinary skill in the art within the scope of the invention can be to above-mentioned Embodiment is changed, revises, replaces and modification.

Claims (6)

1. the optimal control method of a refinery device coupling unit, it is characterised in that comprise the following steps:

Obtain closed-loop dynamic feedback decoupling compensator D (s) based on the pairing of Relative increasing rate loop controlled device G (s) is entered Row decoupling, changes into generalized object by described G (s)

For described generalized objectDesign its internal mode controller C (s)；

Calculate Indistinct Input amount, the output of process and process model output bias e (t) and deviation variation rate

Determine fuzzy rule, obtain internal mode controller parameter and e (t) andBetween fuzzy relation；

By to the analysis of described fuzzy rule, process, table look-up and computing carrys out on-line amending controller parameter λ_iValue, revises foundation Following formula is carried out:

${\mathrm{\&lambda;}}_i={\mathrm{\&lambda;}}_{i0}+{\{E_i,{\mathrm{EC}}_i\}}_{{\mathrm{\&lambda;}}_i}$

Wherein, λ_i0Come filter time constant, λ for the identification result according to object model parameter and operating experience_iSet one Individual initial value,Represent after fuzzy logic inference calculates λ_iOn-line amending.

Optimal control method the most according to claim 1, it is characterised in that described acquisition is based on Relative increasing rate loop Controlled device G (s) is decoupled by closed-loop dynamic feedback decoupling compensator D (s) of pairing, and described G (s) is changed into broad sense pair AsSpecifically include:

Described Relative increasing rate selects the controlled major loop T of each passage in the several control loops intercoupled_i-k, i.e. by Kth input controls i-th output；

I-th row kth column element D of described closed-loop dynamic feedback decoupling device D (s)_ikS () element is set to 1, i.e. D_ik(s)=1；

Described closed-loop dynamic feedback decoupling device D (s) removes D_ikS the element beyond () is designed as:

$D_{ij}\left(s\right)=-\frac{G_{ij}\left(s\right)}{G_{ik}\left(s\right)},j\&NotEqual;k$

When described control object contains non-minimum phase bit position, described closed-loop dynamic feedback decoupling device D (s) is carried out time lag benefit Repaying, described closed-loop dynamic feedback decoupling device D (s) after compensation is represented by:

$D_{ij}\left(s\right)=\left\{\begin{array}{cc}{e}^{-{\mathrm{\&tau;}}_{ik}},& j=k\\ -\frac{G_{ij}\left(s\right)e^{-{\mathrm{\&tau;}}_{ik}}}{G_{ik}\left(s\right)},& j\&NotEqual;k\end{array}\right.$

Wherein, τ_ik=τ (G_ik)-τ_i,

Ensure described generalized objectRelative increasing rateEither element all not less than zero,Wherein,

Optimal control method the most according to claim 2, it is characterised in that described control object contains non-minimum phase portion Timesharing, carries out time lag compensation to described closed-loop dynamic feedback decoupling device D (s) and specifically includes:

Calculate τ_i:

${\mathrm{\&tau;}}_i=\underset{j}{\min }\mathrm{\&tau;}\left(G_{ij}\right)$

Determine generalized ensemble element H_ik(s):

$H_{ik}\left(s\right)=e^{-{\mathrm{\&tau;}}_{i}s}{G}_{ik-}\left(s\right)$

Wherein, G_ik-S () is G_ikThe minimum phase part of (s),

Calculate D_ijTime lag τ (the D of (s)_ij), and then determine decoupler element D_ij(s):

τ(D_ij)=τ (G_ij)-τ(H_ik)。

4. the optimal control device of a refinery device coupling unit, it is characterised in that including:

Decoupling module, for obtaining closed-loop dynamic feedback decoupling compensator D (s) based on the pairing of Relative increasing rate loop to quilt Control object G (s) carries out decoupling and changes into generalized object

Design module, for for described generalized objectDesign its internal mode controller C (s)；

Control module, is used for calculating fuzzy reason input quantity the output of process and process model output bias e (t) and deviation variation rateDetermine fuzzy rule, obtain internal mode controller parameter and e (t) andBetween fuzzy relation, by fuzzy rule Analysis, process, table look-up and computing carrys out on-line amending controller parameter λ_iValue, revises and carries out according to following formula:

${\mathrm{\&lambda;}}_i={\mathrm{\&lambda;}}_{i0}+{\{E_i,{\mathrm{EC}}_i\}}_{{\mathrm{\&lambda;}}_i}$

Wherein, λ_i0Come filter time constant λ for the identification result according to object model parameter and operating experience_iSet one Initial value,Represent after fuzzy logic inference calculates λ_iOn-line amending.

5. optimal control device as claimed in claim 4, it is characterised in that described decoupling module specifically for:

Described Relative increasing rate selects the controlled major loop T of each passage in the several control loops intercoupled_i-k, i.e. by Kth input controls i-th output；

I-th row kth column element D of described closed-loop dynamic feedback decoupling device D (s)_ikS () element is set to 1, i.e. D_ik(s)=1；

Described closed-loop dynamic feedback decoupling device D (s) removes D_ikS the element beyond () is designed as:

$D_{ij}\left(s\right)=-\frac{G_{ij}\left(s\right)}{G_{ik}\left(s\right)},j\&NotEqual;k$

When described control object contains non-minimum phase bit position, described closed-loop dynamic feedback decoupling device D (s) is carried out time lag benefit Repaying, described closed-loop dynamic feedback decoupling device D (s) after compensation is represented by:

$D_{ij}\left(s\right)=\left\{\begin{array}{cc}{e}^{-{\mathrm{\&tau;}}_{ik}},& j=k\\ -\frac{G_{ij}\left(s\right)e^{-{\mathrm{\&tau;}}_{ik}}}{G_{ik}\left(s\right)},& j\&NotEqual;k\end{array}\right.$

Wherein, τ_ik=τ (G_ik)-τ_i,

Ensure described generalized objectRelative increasing rateEither element all not less than zero,Wherein,

6. optimal control device as claimed in claim 5, it is characterised in that described decoupling module is additionally operable to:

Calculate τ_i:

${\mathrm{\&tau;}}_i=\underset{j}{\min }\mathrm{\&tau;}\left(G_{ij}\right)$

Determine generalized ensemble element H_ik(s):

$H_{ik}\left(s\right)=e^{-{\mathrm{\&tau;}}_{i}s}{G}_{ik-}\left(s\right)$

Wherein, G_ik-S () is G_ikThe minimum phase part of (s),

Calculate D_ijTime lag τ (the D of (s)_ij), and then determine decoupler element D_ij(s):

τ(D_ij)=τ (G_ij)-τ(H_ik)。