CN101893232A - Improved method of limitation generalized predictive control for thermal power unit reheat steam temperature - Google Patents

Abstract

The invention relates to an improved method of limitation generalized predictive control for thermal power unit reheat steam temperature. In the invention, an interior model for improving limitation predictive control is obtained through a reheat steam temperature set value step experience; a reheat steam temperature control system records a spray desuperheating valve opening instruction sequence and a reheat steam temperature measured value sequence in real time; and an opening increment instruction of a current spray desuperheating valve is obtained through simple calculation. The control method solves the problems that the traditional limitation generalized predictive control has complicated optimization problem and large calculated amount and the regulation of the reheat steam temperature control system is lagged and has slow response, thereby causing rapid and stable temperature control of a reheater without dynamic deviation.

Description

Thermal power unit reheat steam temperature improves method of limitation generalized predictive control

Technical field

The present invention a kind ofly regulates the thermal power unit reheat steam temperature control system by improved constrained predictive control strategy, make reheat steam temperature fast, a kind of method of stable, agonic tracking setting value, belong to thermal technology's automation field.

Background technology

The thermal power unit reheat steam temperature system is one of most important control system in the thermal power plant, also is one of difficult point of thermal power plant's control.Controlled variable is a reheat steam temperature, the too high meeting of temperature makes the reheater booster, temperature is low excessively to reduce unit economy, can make steam turbine band water generates cavitation what is more, jeopardize unit safety operation, so it is extremely important to keep reheat steam temperature to be stabilized in rated value, because reheat steam temperature is to liking the large time delay thermal object, and increase along with unit capacity, these characteristics can be more remarkable, the traditional PID control algorithm often is difficult to satisfy promptly fast steady again requirement, so research and design advanced person's Switching Logic Control of Reheat Steam Temperature algorithm has the meaning of particular importance.

PREDICTIVE CONTROL is one of advanced control strategy of tool using value in process control industries, but conventional constrained predictive control strategy, the algorithm complexity, amount of calculation is difficult to engineering greatly and uses, so the research algorithm is simple, amount of calculation is easy to constrained predictive control that engineering uses for a short time and just seems and be even more important.

In order to improve the Switching Logic Control of Reheat Steam Temperature quality, the present invention is on the basis of the traditional constrained predictive control strategy of further investigation, rationally simplify constraints, it is simple to have proposed a kind of algorithm, amount of calculation is easy to the improvement constrained predictive control strategy realized for a short time, this control algolithm is introduced Reheated-steam Temperature Control System first, obtained effective Switching Logic Control of Reheat Steam Temperature.

Summary of the invention

Technical problem: the objective of the invention is to propose a kind of thermal power unit reheat steam temperature and improve method of limitation generalized predictive control, be used for Reheated-steam Temperature Control System, solve that traditional constrained predictive is difficult to that engineering is used and the problem of reheat steam temperature thermal object large time delay, non-linear and time variation.

Technical scheme: in order to overcome the problems referred to above, by adopting improved constrained predictive generalized predictive control, solve traditional constrained predictive control algolithm complexity, amount of calculation is big and be difficult to the shortcoming that engineering is used, be used for Reheated-steam Temperature Control System, make system response time fast, performance of dynamic tracking is good, system's nonoscillatory, dynamically zero deflection.

The technical scheme of improved limitation generalized PREDICTIVE CONTROL can adopt following steps to realize:

Step 1: according to the pure delay time of reheat steam temperature system controlled device, length of field P makes it greater than the pure delay time when prolonging prediction; Control time domain length M gets 1 or 2; The control weighted matrix R is got the diagonal matrix that diagonal entry is λ; Sampling time T at first satisfies Shannon's theorems, satisfies empirical formula T then ₉₅/ T=10～25, T ₉₅Rise to the adjusting time of reheat steam temperature setting value 95% for the reheat steam temperature degree;

Step 2: by the CARIMA model A (z of controlled device ^-1) Δ y (t)=B (z ^-1) the middle parametric polynomial of Δ u (t-1)+ξ (t) With

Do not have the constraint generalized forecast control method by tradition and calculate spray desuperheating valve optimum control increment Delta ua, wherein Δ=1-z ^-1, z ^-1Be backward shift operator, y (t) is a current t reheat steam temperature value constantly, and u (t-1) is a t-1 spray desuperheating valve controlled quentity controlled variable constantly, and ξ (t) is a noise, a _i, b _iBe respectively A (z ^-1), B (z ^-1) z in the multinomial ^-iCoefficient, na, nb are respectively multinomial A (z ^-1), B (z ^-1) order;

Step 3: spray desuperheating valve control increment constraint and control constraint by the current t moment, reach the t+1 constraint of reheat steam temperature constantly, calculate three constraintss of Δ u (t) respectively

Δ u _Min＜Δ u (t)＜Δ u _MaxFormula a

Δ u _Min1＜Δ u (t)＜Δ u _Max1Formula b

Δ u _Min2＜Δ u (t)＜Δ u _Max2Formula c

Δ u _Min, Δ u _MaxBe the restriction of reheater spray desuperheating valve control increment, Δ u _Min1, Δ u _Max1Be the constraint that goes out by spray desuperheating valve controlled quentity controlled variable limit calculation, Δ u _Min2, Δ u _Max2Be the constraint that goes out by reheat steam temperature temperature limit calculation, calculate Δ u _Min, Δ u _Min1, Δ u _Min2In maximum and be designated as Δ ub, calculate Δ u _Max, Δ u _Max1, Δ u _Max2In minimum of a value and be designated as Δ uc;

Step 4: if Δ ub-Δ uc＞0, then optimum reheater spray water control valve control increment is Δ ub, if Δ ua-Δ uc＞0 optimum reheater spray water control valve control increment is Δ uc, otherwise optimum reheater spray water control valve control increment is output as Δ ua.

Beneficial effect: utilize improved constrained predictive control method, amount of calculation is little, algorithm is simple, has solved traditional constrained predictive control method algorithm complexity, and amount of calculation is difficult to the shortcoming that engineering is used greatly, the method is applied to Reheated-steam Temperature Control System, make system responses rapid, system's nonoscillatory, no dynamic deviation, effectively overcome Reheated-steam Temperature Control System owing to the system responses that large time delay causes is slow, problems such as adjustment process vibration.

Description of drawings

Fig. 1 Reheated-steam Temperature Control System schematic diagram.

The specific embodiment

The present invention is a kind of at Reheated-steam Temperature Control System large time delay, non-linear and time-varying Characteristics, adopts improved limitation generalized PREDICTIVE CONTROL, makes the control system response rapidly, no dynamic deviation, the non-oscillating method of system.Specific implementation method is as follows:

Before algorithm is implemented, the Switching Logic Control of Reheat Steam Temperature object is done the test of setting value step, draw the transfer function model of Switching Logic Control of Reheat Steam Temperature object, draw the CARIMA model of controlled device by bilinear transformation, perhaps by the direct match CARIMA of test data model.

The technical scheme of improved limitation generalized PREDICTIVE CONTROL can adopt following steps to realize:

Step 1: according to the pure delay time of reheat steam temperature system controlled device, length of field P makes it greater than the pure delay time when prolonging prediction; Control time domain length M gets 1 or 2; The control weighted matrix R is got the diagonal matrix that diagonal entry is λ; Sampling time T at first satisfies Shannon's theorems, satisfies empirical formula T then ₉₅/ T=10～25, T ₉₅Rise to the adjusting time of reheat steam temperature setting value 95% for the reheat steam temperature degree;

Step 2: by the CARIMA model A (z of controlled device ^-1) Δ y (t)=B (z ^-1) the middle parametric polynomial of Δ u (t-1)+ξ (t) With Do not have the constraint generalized forecast control method by tradition and calculate spray desuperheating valve optimum control increment Delta ua, wherein Δ=1-z ^-1, z ^-1Be backward shift operator, y (t) is a current t reheat steam temperature value constantly, and u (t-1) is a t-1 spray desuperheating valve controlled quentity controlled variable constantly, and ξ (t) is a noise, a _i, b _iBe respectively A (z ^-1), B (z ^-1) z in the multinomial ^-iCoefficient, na, nb are respectively multinomial A (z ^-1), B (z ^-1) order, the specific embodiment is as follows:

CARIMA model A (z by controlled device ^-1) Δ y (t)=B (z ^-1) the middle parametric polynomial of Δ u (t-1)+ξ (t)

With And with following formula 1 and formula 2

1=R _j(z ^-1) A (z ^-1) Δ+z ^-jS _j(z ^-1) formula 1

R _j(z ^-1) B (z ^-1)=G _j(z ^-1) formula 2

Obtain multinomial

By above multinomial coefficient and obtain matrix G with following formula 3-formula 5 formulas, F and S,

$G={\left[\begin{array}{cccc}{g}_{\mathrm{1,0}}& & & \\ g_{\mathrm{2,1}}& g_{\mathrm{1,0}}& & 0\\ .& .& & \\ g_{P,P-1}& g_{P-1,P-2}& .& .g_{P-M+1,P-M}\end{array}\right]}_{P\×M}$ Formula 3

${F=\left[\begin{array}{cccc}{g}_{1,\mathrm{nb}}& g_{1,\mathrm{nb}-1}& ....& g_{\mathrm{1,1}}\\ g_{2,\mathrm{nb}+1}& g_{2,\mathrm{nb}}& ....& g_{\mathrm{2,2}}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ g_{P,\mathrm{nb}+P-1}& g_{P,\mathrm{nb}+P-2}& ....& g_{P,P}\end{array}\right]}_{P\×\mathrm{nb}}$ Formula 4

$S={\left[\begin{array}{cccc}{s}_{\mathrm{1,0}}& s_{\mathrm{1,1}}& ....& s_{1,\mathrm{na}}\\ s_{\mathrm{2,0}}& s_{\mathrm{2,1}}& ....& s_{2,\mathrm{na}}\\ .& .& & .\\ .& .& & .\\ .& .& & .\\ s_{P,0}& s_{P,1}& ....& s_{P,\mathrm{na}}\end{array}\right]}_{P\×\mathrm{na}}$ Formula 5

With above matrix G, F and S deposit in the controller module, are used for the control increment of calculating control system, wherein Δ=1-z ^-1, z ^-1Be backward shift operator, y (t) is t control system output constantly, and u (t-1) is a t-1 controlled quentity controlled variable constantly, and ξ (t) is a noise, a _i, b _iBe respectively A (z ^-1), B (z ^-1) z in the multinomial ^-iCoefficient, R _j(z ^-1), S _j(z ^-1) be respectively S in the formula 1 _j(z ^-1) the preceding product factor gets z ^-jThe time multinomial, G _j(z ^-1) be corresponding to R in the formula 2 _j(z ^-1) multinomial that draws, r _{J, i}s _{J, i}, g _{J, i}Be respectively R _j(z ^-1), S _j(z ^-1), G _j(z ^-1) z in the multinomial ^-iCoefficient, na, nb are respectively multinomial A (z ^-1), B (z ^-1) order; The online real time record of Reheated-steam Temperature Control System till the current time t spray desuperheating valve opening control increment signal [Δ u (t-nb) ..., Δ u (t-1)] ^TBe designated as Δ U ', t-1 moment valve opening amount u (t-1) and reheat steam temperature measurement data [y (t) ..., y (t-na)] ^TBe designated as Y, and receive the reheat steam temperature desired value [y that the reheat steam temperature master controller provides _r(t+1) ..., y _r(t+P)] ^TBe designated as Y _R, by the control increment of following formula 6 controlled systems:

Δ U=[G ^TG+R] ^-1G ^T[Y _R-SY-F Δ U '] formula 6

The control increment that first element of Δ U vector is current time is designated as Δ ua, the Δ u (t-nb) among the Δ U ' ..., Δ u (t-1) is respectively the valve opening control increment size corresponding to the moment in the bracket, the y among the Y (t) ..., y (t-na) and Y _RIn y _r(t+1) ..., y _r(t+P) be respectively corresponding to constantly reheat steam temperature measured value and reheat steam temperature desired value in the bracket;

Step 3: spray desuperheating valve control increment constraint and control constraint by the current t moment, reach the t+1 constraint of reheat steam temperature constantly, calculate three constraintss of Δ u (t) respectively

Δ u _Min＜Δ u (t)＜Δ u _MaxFormula a

Δ u _Min1＜Δ u (t)＜Δ u _Max1Formula b

Δ u _Min2＜Δ u (t)＜Δ u _Max2Formula c

Δ u _Min, Δ u _MaxBe the restriction of reheater spray desuperheating valve control increment, Δ u _Min1, Δ u _Max1Be the constraint that goes out by spray desuperheating valve controlled quentity controlled variable limit calculation, Δ u _Min2, Δ u _Max2Be the constraint that goes out by reheat steam temperature temperature limit calculation, calculate Δ u _Min, Δ u _Min1, Δ u _Min2In maximum and be designated as Δ ub, calculate Δ u _Max, Δ u _Max1, Δ u _Max2In minimum of a value and be designated as Δ uc, specifically implement as follows:

Calculate Δ u by following formula 7-formula 10 respectively _Min1, Δ u _Min2, Δ u _Max1, Δ u _Max2

Δ u _Min1=u _Min-u (t-1) formula 7

Δ u _Max1=u _Max-u (t-1) formula 8

$\mathrm{\Δ}{u}_{\min 2}=\frac{y_{\min }-S_1\·y\left(t\right)-F_1\·\mathrm{\Δu}(t-1)}{g_{\mathrm{1,0}}}$ Formula 9

$\mathrm{\Δ}{u}_{\max 2}=\frac{y_{\max }-S_1\·y\left(t\right)-F_1\·\mathrm{\Δu}(t-1)}{g_{\mathrm{1,0}}}$ Formula 10

Calculate Δ u _Min, Δ u _Min1, Δ u _Min2In maximum and be designated as Δ ub,

Calculate Δ u _Max, Δ u _Max1, Δ u _Max2In minimum of a value and be designated as Δ uc,

Δ u _Min, Δ u _MaxBe the restriction of Reheated-steam Temperature Control System spray desuperheating valve opening increment, u _Min, u _MaxBe respectively the position restriction of spray desuperheating valve, y _Min, y _MaxBe respectively the restriction of reheat steam temperature temperature, S ₁And F ₁Correspond respectively to R _j(z ^-1) and S _j(z ^-1) in j get 1 o'clock multinomial;

Step 4: if Δ ub-Δ uc＞0, then optimum reheater spray water control valve control increment is Δ ub, if Δ ua-Δ uc＞0 optimum reheater spray water control valve control increment is Δ uc, otherwise optimum reheater spray water control valve control increment is output as Δ ua.

Claims (1)

1. a thermal power unit reheat steam temperature improves method of limitation generalized predictive control, it is characterized in that this control method adopts the constrained predictive control of simplifying, and implementation step is as follows:

Step 1: according to the pure delay time of reheat steam temperature system controlled device, length of field P makes it greater than the pure delay time when prolonging prediction; Control time domain length M gets 1 or 2; The control weighted matrix R is got the diagonal matrix that diagonal entry is λ; Sampling time T at first satisfies Shannon's theorems, satisfies empirical formula T then ₉₅/ T=10～25, T ₉₅Rise to the adjusting time of reheat steam temperature setting value 95% for the reheat steam temperature degree;

Step 2: by the CARIMA model A (z of controlled device ^-1) Δ y (t)=B (z ^-1) the middle parametric polynomial of Δ u (t-1)+ξ (t)

With

Do not have the constraint generalized forecast control method by tradition and calculate spray desuperheating valve optimum control increment Delta ua, wherein Δ=1-z ^-1, z ^-1Be backward shift operator, y (t) is a current t reheat steam temperature value constantly, and u (t-1) is a t-1 spray desuperheating valve controlled quentity controlled variable constantly, and ξ (t) is a noise, a _i, b _iBe respectively A (z ^-1), B (z ^-1) z in the multinomial ^-iCoefficient, na, nb are respectively multinomial A (z ^-1), B (z ^-1) order;

Step 3: spray desuperheating valve control increment constraint and control constraint by the current t moment, reach the t+1 constraint of reheat steam temperature constantly, calculate three constraintss of Δ u (t) respectively

Δ u _Min＜Δ u (t)＜Δ u _MaxFormula a

Δ u _Min1＜Δ u (t)＜Δ u _Max1Formula b

Δ u _Min2＜Δ u (t)＜Δ u _Max2Formula c

Δ u _Min, Δ u _MaxBe the restriction of reheater spray desuperheating valve control increment, Δ u _Min1, Δ u _Max1Be the constraint that goes out by spray desuperheating valve controlled quentity controlled variable limit calculation, Δ u _Min2, Δ u _Max2Be the constraint that goes out by reheat steam temperature temperature limit calculation, calculate Δ u _Min, Δ u _Min1, Δ u _Min2In maximum and be designated as Δ ub, calculate Δ u _Max, Δ u _Max1, Δ u _Max2In minimum of a value and be designated as Δ uc;

Step 4: if Δ ub-Δ uc＞0, then optimum reheater spray water control valve control increment is Δ ub, if Δ ua-Δ uc＞0 optimum reheater spray water control valve control increment is Δ uc, otherwise optimum reheater spray water control valve control increment is output as Δ ua.

Images