CN103389746A - Prediction function control optimized control method for furnace pressure of waste plastic oil refining cracking furnace - Google Patents

Abstract

The invention discloses a prediction function control optimized control method for furnace pressure of a waste plastic oil refining cracking furnace. The method comprises the steps that a model of a furnace pressure object is established based on step response data of the furnace pressure object of the cracking furnace; the basic object characters are excavated; then a parameter of a corresponding PI (Proportional-Integral) controller is set according to the characteristics of prediction function control; and PI control is implemented on the furnace pressure object of the cracking furnace finally. According to the method, the performance of the prediction function control is assigned to the PI control, so that the performance of the traditional control method is improved effectively, and an application of an advanced control method is facilitated.

Description

The waste plastic oil-refining pyrolysis furnace hearth pressure control method that Predictive function control is optimized

Technical field

The invention belongs to technical field of automation, relate to a kind of waste plastic oil-refining pyrolysis furnace fire box temperature proportional integral (PI) control method of optimizing based on Predictive function control.

Background technology

In the control of actual industrial object, due to the restriction of the aspects such as hardware, cost, enforcement difficulty, although some advanced control methods have obtained application to a certain degree, the PID that is still that accounts at present main flow controls.The oil refining pyrolysis furnace is the important device in petrochemical production process, and wherein furnace pressure has very important impact to cracking process, and furnace pressure is excessive may bring danger, and the too low meeting of furnace pressure causes the lysis efficiency step-down.Because the advanced control method application is limited, the common adoption rate integration of control (PI) of pyrolysis furnace furnace pressure is controlled at present.Predictive function control is a kind of as advanced control method, compare PI control and have better control performance in the control of pyrolysis furnace furnace pressure, if the performance of Predictive function control can be assigned to PI controls, that will further advance the application of advanced control method, also can obtain better Actual Control Effect of Strong simultaneously.

Summary of the invention

The objective of the invention is the application weak point for existing advanced control method, provide a kind of pyrolysis furnace furnace pressure PI control method of optimizing based on Predictive function control, to obtain better working control performance.The method, by in conjunction with Predictive function control and PI, controlling, has obtained a kind of control method of PI with the Predictive function control performance.The method also guarantees form simply and meets the needs of actual industrial process when inheriting the Predictive function control premium properties.

At first the inventive method sets up the model of furnace pressure object based on the step response data of pyrolysis furnace furnace pressure object, excavate basic plant characteristic; Then go the parameter of adjusting corresponding PI controller according to the characteristic of Predictive function control; Finally pyrolysis furnace furnace pressure object being implemented PI controls.

Technical scheme of the present invention is to set up, predict the means such as mechanism, optimization by data acquisition, model, has established a kind of PI control method of optimizing based on Predictive function control, utilizes the method can effectively improve precision and the stability of control.

The step of the inventive method comprises:

Step (1). set up the model of controlled device by the real-time step response data of pyrolysis furnace furnace pressure object, concrete grammar is:

A. the pi controller of process is rested on manual operation state, the operation driver plate makes its output have individual step to change, and is recorded the output valve of real process by recorder, with real process output valve y _p(k) response curve converts Dimensionless Form y to _p ^*(k), specifically:

y _p ^*(k)＝y _p(k)/y _p(∞)

Wherein, y _p(∞) be the output of the proportional plus integral plus derivative controller real process output y while having step to change _p(k) steady-state value.

B. choose and meet y _p ^*(k ₁)=0.39 and y _p ^*(k ₂Two calculation level k of)=0.63 ₁And k ₂, according to the model parameter K of following formula computation process object _m, T and τ:

K _m＝y _p(∞)/q

T＝2(k ₁-k ₂)

τ＝2k ₁-k ₂

The transport function of the process object that obtains finally is:

$G\left(s\right)=\frac{K_m}{\mathrm{Ts}+1}{e}^{-\mathrm{\τs}}$

Wherein, q is the step amplitude of variation of the proportional plus integral plus derivative controller output of process, and G (s) is the transport function of process object, and s is the Laplace transform operator, K _mFor the gain coefficient of model, T is the time constant of model, and τ is parameter retardation time of model.

Step (2). the PI controller of design process object, concrete grammar is:

A. to the transport function that obtains at sampling time T _sUnder add a zero-order holder discretize, obtain discrete model and be

y _m(k)＝a _my _m(k-1)+K _m(1-a _m)u(k-1-L)

y _m(k) be k process object model prediction output constantly,

U (k-1-L) is the control inputs of k-1-L process object constantly, and L is the time lag of discrete transfer function model, L=τ/T _s.

B. the computation process object removes the pure hysteresis step prediction of the P under Predictive function control later output, and form is as follows:

y _mav(k)＝a _my _mav(k-1)+K _m(1-a _m)u(k-1)

y _mav(k+P)＝a _m ^Py _mav(k)+K _m(1-a _m ^P)u(k)

Wherein, P is prediction step, y _mav(k+P) constantly remove the P step prediction output of process object under Predictive function control of pure hysteresis, y for k _mav(k) constantly remove the process model output of pure hysteresis for k.

C. the actual output of revising current time obtains comprising the new process border output valve of future anticipation information, and form is as follows:

y _Pav(k)＝y _P(k)+y _mav(k)-y _mav(k-L)

Wherein, y _Pav(k) constantly comprise the new the output of process value of future anticipation information, y for proofreading and correct the k that obtains _P(k) be k real output value constantly.

D. choose the reference locus y of predictive functional control algorithm _r(k+P) and objective function J, form is as follows:

y _r(k+P)＝β ^Py _p(k)+(1-β ^P)c(k)

J＝min(y _r(k+P)-y _mav(k+P)-e(k)) ²

e(k)＝y _pav(k)-y _mav(k)

Wherein, β is reference locus softening coefficient, and c (k) is k setting value constantly, and e (k) is the error amount of k time correction.

E. solve parameter in the PI controller according to the objective function in steps d, controlled quentity controlled variable u (k) carried out conversion here:

u(k)＝u(k-1)+K _p(e ₁(k)-e ₁(k-1))+K _ie ₁(k)

e ₁(k)＝βy _p(k-1)+(1-β)c(k-1)-y _p(k)

Further abbreviation is:

u(k)＝u(k-1)+w(k) ^ΤE(k)

w(k)＝[w ₁(k),w ₂(k)] ^Τ

w ₁(k)＝K _p+K _i,w ₂(k)＝-K _p

E(k)＝[e ₁(k),e ₁(k-1)] ^Τ

Wherein, K _p, K _iBe respectively ratio, the integral parameter of PI controller, e ₁(k) be the constantly error between reference locus value and real output value of k, Τ is the transpose of a matrix symbol.

In conjunction with above-mentioned formula, can in the hope of:

$w\left(k\right)=\frac{(y_r(k+P)-{a_m}^P{y}_{\mathrm{mav}}\left(k\right)-K_m(1-{a_m}^P)u(k-1))E}{K_m(1-{a_m}^P)E^{T}E}$

Further can obtain:

K _p＝-w ₂(k)

K _i＝w ₁(k)-K _P

F. obtain the parameter K of PI controller _p, K _iForm controlled quentity controlled variable u (k) later and act on controlled device, u (k)=u (k-1)+K _p(e ₁(k)-e ₁(k-1))+K _ie ₁(k).

G. at next constantly, according to b, to the step in f, continue to solve the new parameter K of PI controller _p, K _i, circulation successively.

A kind of pyrolysis furnace of waste plastic oil-refining based on Predictive function control optimization furnace pressure PI control method that the present invention proposes has been assigned to PI control with the performance of Predictive function control, effectively improve the performance of traditional control method, also promoted the application of advanced control method simultaneously.

Embodiment

Take the process control of waste plastic oil-refining pyrolysis furnace furnace pressure as example:

The pyrolysis furnace furnace pressure is the important parameter in the pyrolysis furnace cracking process, and regulating measure adopts the aperture of stack damper.

Step (1). set up the model of controlled device by the real-time step response data of pyrolysis furnace furnace pressure object, concrete grammar is:

A. the pi controller of pyrolysis furnace furnace pressure process is rested on manual operation state, the operation driver plate makes its output have individual step to change, and is recorded the real output value of furnace pressure process by recorder, with real output value y _p(k) response curve converts Dimensionless Form y to _p ^*(k), specifically:

y _p ^*(k)＝y _p(k)/y _p(∞)

Wherein, y _p(∞) be the output of the pi controller furnace pressure the output of process y while having step to change _p(k) steady-state value.

B. choose and meet y _p ^*(k ₁)=0.39 and y _p ^*(k ₂Two calculation level k of)=0.63 ₁And k ₂, according to the model parameter K of following formula calculating furnace pressure process _m, T and τ:

K _m＝y _p(∞)/q

T＝2(k ₁-k ₂)

τ＝2k ₁-k ₂

The transport function of the process model that obtains finally is:

$G\left(s\right)=\frac{K_m}{\mathrm{Ts}+1}{e}^{-\mathrm{\τs}}$

Wherein, q is the step amplitude of variation of the proportional plus integral plus derivative controller output of furnace pressure process, and G (s) is the transport function of furnace pressure process, and s is the Laplace transform operator, K _mFor the gain coefficient of furnace pressure process model, T is the time constant of furnace pressure process model, and τ is parameter retardation time of furnace pressure process model.

Step (2). the PI controller of design furnace pressure process, concrete grammar is:

A. to the transport function of the furnace pressure process model that obtains at sampling time T _sUnder add a zero-order holder discretize, obtain discrete model and be

y _m(k)＝a _my _m(k-1)+K _m(1-a _m)u(k-1-L)

y _m(k) be k furnace pressure process model prediction output constantly,

U (k-1-L) is the control inputs of k-1-L furnace pressure process model constantly, and L is the time lag of discrete transfer function model, L=τ/T _s.

B. calculate the furnace pressure process model and remove the pure hysteresis step prediction of the P under Predictive function control later output, form is as follows:

y _mav(k)＝a _my _mav(k-1)+K _m(1-a _m)u(k-1)

y _mav(k+P)＝a _m ^Py _mav(k)+K _m(1-a _m ^P)u(k)

Wherein, P is prediction step, y _mav(k+P) constantly remove the P step prediction output of furnace pressure process model under Predictive function control of pure hysteresis, y for k _mav(k) constantly remove the furnace pressure process model output of pure hysteresis for k.

C. the furnace pressure process real output value of revising current time obtains comprising the new process border output valve of future anticipation information, and form is as follows:

y _Pav(k)＝y _P(k)+y _mav(k)-y _mav(k-L)

Wherein, y _Pav(k) constantly comprise the new the output of process value of the furnace pressure process of future anticipation information, y for proofreading and correct the k that obtains _P(k) be k furnace pressure process real output value constantly.

D. choose the reference locus y of predictive functional control algorithm _r(k+P) and objective function J, form is as follows:

y _r(k+P)＝β ^Py _p(k)+(1-β ^P)c(k)

J＝min(y _r(k+P)-y _mav(k+P)-e(k)) ²

e(k)＝y _pav(k)-y _mav(k)

Wherein, β is reference locus softening coefficient, and c (k) is the setting value of k furnace pressure process constantly, and e (k) is the error amount of k time correction.

E. solve parameter in the PI controller according to the objective function in steps d, the stack damper aperture controlled quentity controlled variable u (k) of furnace pressure process carried out conversion here:

u(k)＝u(k-1)+K _p(e ₁(k)-e ₁(k-1))+K _ie ₁(k)

e ₁(k)＝βy _p(k-1)+(1-β)c(k-1)-y _p(k)

Further abbreviation is:

u(k)＝u(k-1)+w(k) ^ΤE(k)

w(k)＝[w ₁(k),w ₂(k)] ^Τ

w ₁(k)＝K _p+K _i,w ₂(k)＝-K _p

E(k)＝[e ₁(k),e ₁(k-1)] ^Τ

Wherein, K _p, K _iBe respectively ratio, the integral parameter of furnace pressure process PI controller, e ₁(k) be the reference locus value of k moment furnace pressure process and the error between real output value, Τ is the transpose of a matrix symbol.

In conjunction with above-mentioned formula, can in the hope of:

$w\left(k\right)=\frac{(y_r(k+P)-{a_m}^P{y}_{\mathrm{mav}}\left(k\right)-K_m(1-{a_m}^P)u(k-1))E}{K_m(1-{a_m}^P)E^{T}E}$

Further can obtain:

K _p＝-w ₂(k)

K _i＝w ₁(k)-K _P

F. obtain the parameter K of PI controller _p, K _iForm controlled quentity controlled variable u (k) later and act on the stack damper aperture valve of pyrolysis furnace burner hearth, u (k)=u (k-1)+K _p(e ₁(k)-e ₁(k-1))+K _ie ₁(k).

G. at next constantly, according to b, to the step in f, continue to solve the new parameter K of pyrolysis furnace furnace pressure process PI controller _p, K _i, circulation successively.

Claims (1)

1. the waste plastic oil-refining pyrolysis furnace hearth pressure control method optimized of Predictive function control is characterized in that the concrete steps of the method are:

Step (1). set up the model of controlled device by the real-time step response data of pyrolysis furnace furnace pressure object, concrete grammar is:

I. the pi controller of process is rested on manual operation state, and the operation driver plate makes its output have individual step to change, and is recorded the output valve of real process by recorder, with real process output valve y _p(k) response curve converts Dimensionless Form y to _p ^*(k), specifically:

y _p ^*(k)＝y _p(k)/y _p(∞)

Wherein, y _p(∞) be the output of the proportional plus integral plus derivative controller real process output y while having step to change _p(k) steady-state value;

II. choose and meet y _p ^*(k ₁)=0.39 and y _p ^*(k ₂Two calculation level k of)=0.63 ₁And k ₂, according to the model parameter K of following formula computation process object _m, T and τ:

K _m＝y _p(∞)/q

T＝2(k ₁-k ₂)

τ＝2k ₁-k ₂

The transport function of the process object that obtains finally is:

$G\left(s\right)=\frac{K_m}{\mathrm{Ts}+1}{e}^{-\mathrm{\τs}}$

Wherein, q is the step amplitude of variation of the proportional plus integral plus derivative controller output of process, and G (s) is the transport function of process object, and s is the Laplace transform operator, K _mFor the gain coefficient of model, T is the time constant of model, and τ is parameter retardation time of model;

Step (2). the PI controller of design process object, concrete grammar is:

A. to the transport function that obtains at sampling time T _sUnder add a zero-order holder discretize, obtain discrete model and be

y _m(k)＝a _my _m(k-1)+K _m(1-a _m)u(k-1-L)

y _m(k) be k process object model prediction output constantly,

U (k-1-L) is the control inputs of k-1-L process object constantly, and L is the time lag of discrete transfer function model, L=τ/T _s

B. the computation process object removes the pure hysteresis step prediction of the P under Predictive function control later output, and form is as follows:

y _mav(k)＝a _my _mav(k-1)+K _m(1-a _m)u(k-1)

y _mav(k+P)＝a _m ^Py _mav(k)+K _m(1-a _m ^P)u(k)

Wherein, P is prediction step, y _mav(k+P) constantly remove the P step prediction output of process object under Predictive function control of pure hysteresis, y for k _mav(k) constantly remove the process model output of pure hysteresis for k;

C. the actual output of revising current time obtains comprising the new process border output valve of future anticipation information, and form is as follows:

y _Pav(k)＝y _P(k)+y _mav(k)-y _mav(k-L)

Wherein, y _Pav(k) constantly comprise the new the output of process value of future anticipation information, y for proofreading and correct the k that obtains _P(k) be k real output value constantly;

D. choose the reference locus y of predictive functional control algorithm _r(k+P) and objective function J, form is as follows:

y _r(k+P)＝β ^Py _p(k)+(1-β ^P)c(k)

J＝min(y _r(k+P)-y _mav(k+P)-e(k)) ²

e(k)＝y _pav(k)-y _mav(k)

Wherein, β is reference locus softening coefficient, and c (k) is k setting value constantly, and e (k) is the error amount of k time correction;

E. solve parameter in the PI controller according to the objective function in steps d, controlled quentity controlled variable u (k) carried out conversion here:

u(k)＝u(k-1)+K _p(e ₁(k)-e ₁(k-1))+K _ie ₁(k)

e ₁(k)＝βy _p(k-1)+(1-β)c(k-1)-y _p(k)

Further abbreviation is:

u(k)＝u(k-1)+w(k) ^ΤE(k)

w(k)＝[w ₁(k),w ₂(k)] ^Τ

w ₁(k)＝K _p+K _i,w ₂(k)＝-K _p

E(k)＝[e ₁(k),e ₁(k-1)] ^Τ

Wherein, K _p, K _iBe respectively ratio, the integral parameter of PI controller, e ₁(k) be the constantly error between reference locus value and real output value of k, Τ is the transpose of a matrix symbol;

In conjunction with above-mentioned formula, can in the hope of:

$w\left(k\right)=\frac{(y_r(k+P)-{a_m}^P{y}_{\mathrm{mav}}\left(k\right)-K_m(1-{a_m}^P)u(k-1))E}{K_m(1-{a_m}^P)E^{T}E}$

Further can obtain:

K _p＝-w ₂(k)

K _i＝w ₁(k)-K _P

F. obtain the parameter K of PI controller _p, K _iForm controlled quentity controlled variable u (k) later and act on controlled device, u (k)=u (k-1)+K _p(e ₁(k)-e ₁(k-1))+K _ie ₁(k);

G. at next constantly, according to b, to the step in f, continue to solve the new parameter K of PI controller _p, K _i, circulation successively.