CN104950945A

CN104950945A - Self-adaptive temperature optimization control method under all working conditions of cement calcination decomposing furnace

Info

Publication number: CN104950945A
Application number: CN201510394016.XA
Authority: CN
Inventors: 王靖; 魏灿; 艾军
Original assignee: Tianjin Cement Industry Design and Research Institute Co Ltd
Current assignee: Tianjin Cement Industry Design and Research Institute Co Ltd
Priority date: 2015-07-07
Filing date: 2015-07-07
Publication date: 2015-09-30

Abstract

The invention discloses a self-adaptive temperature optimization control method under all working conditions of a cement calcination decomposing furnace. The method comprises steps as follows: a GPC (generalized predictive controller) is established on the basis of a predictive control algorithm of an input/output parameterized model, wherein input variables comprise the kiln tail coal feeding quantity, the kiln feeding quantity, the tertiary air temperature and the smoke chamber temperature, and the output variable comprises the decomposing furnace outlet temperature; control model parameters are set, and the GPC is initialized; input/output variable signals are acquired in real time and fed back to the GPC; the GPC performs self-correction on the control model parameters according to the actual value and the predicted value of the output variable, works out input variable controlled quantities and outputs control commands to an executing device; the executing device controls the kiln tail coal feeding quantity and performs self-adaptive adjustment on the supply quantity of the kiln feeding quantity according to the decomposing furnace outlet temperature; recycle and feedback control is performed until the technological process is ended. The mathematical model better conforming to actual production working conditions is established for control, and the control effect is more accurate.

Description

All-condition self-adaptive temperature optimization control method for cement burning decomposing furnace

Technical Field

The invention relates to a decomposing furnace outlet temperature control method, in particular to a full-working-condition self-adaptive temperature optimization control method for a cement burning decomposing furnace.

Background

In the production process of a cement precalciner system, the outlet temperature of a decomposing furnace is a very important technological parameter, and the parameter can represent the combustion in the cement kiln and the decomposition condition of materials. Therefore, the stabilization of the outlet temperature of the decomposing furnace has very important influence on the stabilization of the whole cement production line and the improvement of the product quality. However, the outlet temperature of the decomposing furnace is a complex control object with pure lag, large inertia, nonlinearity and mutual coupling of multiple variables, the influence factors are many, the coupling among the factors is strong, and the control is difficult to be carried out by a simple conventional control method.

The most important factor influencing the outlet temperature of the decomposing furnace is the coal feeding amount at the tail of the kiln. Generally, increasing the amount of coal fed increases the furnace exit temperature, and conversely decreases the furnace exit temperature. In addition, the kiln feed rate, tertiary air temperature and flue temperature all affect the decomposing furnace exit temperature. When the feeding amount of the kiln is increased, the heat required by the decomposition of the materials is correspondingly increased, so that the temperature of the outlet of the decomposing furnace is reduced; conversely, when the kiln feeding amount is reduced, the heat required for decomposing the material is correspondingly reduced, and the temperature at the outlet of the decomposing furnace is increased. The tertiary air refers to hot air introduced into the decomposing furnace from the grate cooler, and the temperature of the outlet of the decomposing furnace is increased due to the increase of the temperature of the tertiary air, and the temperature of the outlet of the decomposing furnace is decreased due to the increase of the temperature of the tertiary air. And the flue gas can enter the decomposing furnace due to the untight sealing of the flue gate valve, so that the temperature of the decomposing furnace is influenced. Therefore, in order to control the decomposing furnace outlet temperature more accurately, the influences of the kiln tail coal feeding amount, the kiln feeding amount, the tertiary air temperature, and the smoke chamber temperature must be considered. For the control of the outlet temperature of the cement decomposing furnace, a plurality of methods which are adopted at present are fuzzy control, simple PID control or single variable predictive control, and the methods do not take several important influence factors of the kiln tail coal feeding amount, the kiln feeding amount, the tertiary air temperature and the smoke chamber temperature into consideration.

In addition, due to the reasons that the coal pre-homogenization effect of most cement plants is poor or the coal taking mode is not fixed and the like, the coal calorific value can be changed frequently, and the parameters of the control model are seriously influenced, so that the control model is considered to be adaptive and more accord with the complex working condition of cement production.

Disclosure of Invention

The invention provides a full-working-condition self-adaptive temperature optimization control method for a cement burning decomposing furnace, which aims to solve the technical problems in the prior art.

The technical scheme adopted by the invention for solving the technical problems in the prior art is as follows: a full-working-condition self-adaptive temperature optimization control method for a cement burning decomposing furnace is characterized by comprising the following steps:

step a: establishing a GPC controller on an upper computer based on a predictive control algorithm of an input-output parameterized model; wherein the input variables comprise kiln tail coal feeding amount, kiln feeding amount, tertiary air temperature and smoke chamber temperature, and the output variables comprise: the outlet temperature of the decomposing furnace;

step b: setting control model parameters, initializing and setting the GPC controller and then running;

step c: acquiring the input variable and the output variable signals in real time, preprocessing the data and feeding back the preprocessed data to the GPC controller;

step d: the GPC controller self-corrects the control model parameters according to the actual value and the predicted value of the outlet temperature of the decomposing furnace, calculates the control quantity of the input variable by adopting an optimization algorithm, and outputs a corresponding control instruction to the execution device;

step e: the execution device executes the instruction of the GPC controller, controls the coal feeding amount at the tail of the kiln and adaptively adjusts the feeding amount of the kiln according to the temperature condition of the outlet of the decomposing furnace;

step f: and (c) returning to the step (c) until the technological process is finished.

Further, in the step a, the specific steps of establishing the GPC controller are as follows:

step a-1: under a steady-state working condition, respectively establishing a step response mathematical model of the outlet temperature of the decomposing furnace to the coal feeding quantity of the kiln tail, the outlet temperature of the decomposing furnace to the kiln feeding quantity, the outlet temperature of the decomposing furnace to the tertiary air temperature and the outlet temperature of the decomposing furnace to the smoke chamber temperature, and further discretizing;

let T_sObtaining a discrete equation corresponding to each step response model for the sampling time of the system:

in the formula (1), K is the proportionality coefficient of the system, T is the time constant, tau is the time lag, T_sIs the sampling time of the system; y (k) is a subject output variable; r (k) is a control input variable;

step a-2: a control model of a decomposing furnace outlet temperature control system is established by adopting a CARIMA model, and the expression is as follows:

A(z^-1)Y(k)＝B(z^-1)R(k-1)+ζ(k)/Δ (2)

A (z^{- 1}) = 1 + a_{1} z^{- 1} + a_{2} z^{- 2} + ... + a_{n_{a}} z^{- n_{a}}

B (z^{- 1}) = b_{0} + b_{1} z^{- 1} + b_{2} z^{- 2} + ... + b_{n_{b}} z^{- n_{b}}

wherein, y (k) is a control object output vector, r (k) is a control object input vector, and ζ (k) is a zero-mean noise sequence which is uncorrelated with each other; a (z)^-1)、B(z^-1) Respectively representing the operators z^-1A polynomial of (a); 1-z^-1Is a difference operator; then A (z)^-1)、B(z^-1) R (k-1) corresponds to an input variable: the corresponding matrix of the kiln tail coal feeding amount, the kiln feeding amount, the tertiary air temperature and the smoke chamber temperature is as follows:

B (z^{- 1}) = {[\begin{matrix} B_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * B_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * B_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * B_{V_{3}} (z^{- 1}) \end{matrix}]}^{T}

R (k - 1) = [\begin{matrix} u (k - 1) \\ v_{1} (k - 1) \\ v_{2} (k - 1) \\ v_{3} (k - 1) \end{matrix}]

wherein,

from the above, A (z)^-1) Of order n_a＝4，B(z^-1) Of order of

In equation (3), i is the number corresponding to the different input variables: 1 corresponds to the input variable of the coal feeding quantity at the tail of the kiln; i-v₁Inputting variables corresponding to the feeding amount of the kiln; i-v₂Inputting variables corresponding to the temperature of the tertiary air; i-v₃Inputting variables corresponding to the temperature of the smoke chamber; u (k-1) is an input variable value of the kiln tail coal feeding quantity at the k-1 moment; v. of₁(k-1) inputting a variable value for the kiln feeding amount at the time of k-1; v. of₂(k-1) inputting a variable value of the temperature of the tertiary air at the k-1 moment; v. of₃(k-1) inputting a variable value for the smoke chamber temperature at the moment of k-1; a. the_i、B_iIs an operator z corresponding to different input variables^-1Polynomial expression: tau is_iIs the time lag for different input variables: k_iIs the scaling factor for different input variables: t is_iIs a time constant corresponding to different input variables：T_sIs the sampling time of the system;

step a-3: definition E (z)^-1) And F (z)^-1) In the form of formula (4), A (z) is obtained from said step a-2^-1) Calculate outFurther, an intermediate variable E (z) is obtained^-1) And F (z)^-1) Wherein E (z)^-1) And F (z)^-1) The expression of (a) is as follows:

E(z^-1)＝[E₁(z^-1),···E_P(z^-1)]^T (4)

F(z^-1)＝[F₁(z^-1),···F_P(z^-1)]^T

E_j(z^-1)＝e₀+e₁z^-1+e₂z^-1+···+e_j-1z^-j+1

F_j(z^-1)＝f_j,0+f_j,1z^-1+f_j,2z^-2+f_j,3z^-3+f_j,4z^-4

\begin{matrix} f_{j + 1, i} = f_{j, i + 1 -} {\tilde{a}}_{i + 1} e_{j} & i = 1, 2, 3 \\ f_{j + 1, 4} = {\tilde{a}}_{5} e_{j} = - {\tilde{a}}_{5} f_{j, 0} & i = 4 \end{matrix}

E_j+1(z^-1)＝E_j(z^-1)+e_jz^-j；

when the initial value j is 1, E₁(z^-1)＝1，e₀＝1，

Wherein e is_i,f_i,0,f_i,1,f_i,2,f_i,3,f_i,4Is a vector of dimension 1 x 1, P is the prediction time domain,has the following forms:

step a-4: determining B (z) from said step a-2 and said step a-3^-1)、E(z^-1) And F (z)^-1) Obtaining a system step response matrix G (z)^-1) And a historical response matrix H (z)^-1)，G(z^-1) And H (z)^-1) The expression is as follows:

\begin{matrix} G_{j} (z^{- 1}) = g_{0} + g_{1} z^{- 1} + g_{2} z^{- 2} + ... + g_{n_{b} + j - 1} z^{- n_{b} - j + 1} & j = 1, 2, ..., P \end{matrix}

g_j＝e_jb₀+h_j,0

h_j+1,i-1＝e_jb₀+h_j,i 1≤i≤n_b

h_{j + 1, n_{b} - 1} = e_{j} b_{n_{b}}

when the initial value j is 1:

G₁(z^-1)＝g₀＝e₀b₀，

wherein g is_i,h_j,iIs a 1 x 4 dimensional vector; p is a prediction time domain, M is a control time domain;

step a-5: determining an intermediate variable F (z) from said step a-3 and said step a-4^-1) And H (z)^-1) The past response f (k) of the system is obtained in the form:

f(k)＝[f₁(k) f₂(k)…f_P(k)]^T

wherein f is_jIs a vector of 1 × 1 dimension, P is a prediction time domain, and M is a control time domain.

Further, in the step d, an optimum control increment Δ r (k) ═ d [ Y ] is calculated from the steps a to 4 and a to 5_s-f(k)-c*e(k)]Wherein d isFirst row of (2), matrixIs a matrix formed by the elements of the first row and the first column of each element in the matrix G in the step a-4, Y_s＝[y_s(k+1)^T,…,y_s(k+N)^T]^TIs the following control target curve:

y_s(k)＝y(k)

y_s(k+j)＝ay_s(k+j-1)+(1-a)W,j＝1,…,P

wherein W is a control target value, λ is a control weighting coefficient, a is a softening coefficient, c is an error correction parameter, and e (k) is a prediction error;

the prediction error e (k) is expressed as: e (k) ═ y (k) — y_p(k) Wherein y (k) is an actual value of the current decomposing furnace outlet temperature, y_p(k) Is the predicted value of the outlet temperature of the decomposing furnace at the last moment;

thereby, the optimum control amount r (k +1) ═ r (k) + Δ r (k) is calculated.

Further, in the step d, a specific method of self-correcting the control model parameters is as follows:

(1) the predicted value y of the outlet temperature of the decomposing furnace can be obtained by substituting the optimal control increment r (k) obtained in the step d into the formula (1)_p(K +1), the model proportionality coefficient K' can be obtained by comparing the decomposition furnace outlet temperature y (K +1) with the following equation:

in the calculation processing at the moment of K +1, adaptively adjusting the proportionality coefficient of the control model to K';

(2) if the temperature of the outlet of the decomposing furnace is always lower than the set target in the set period, automatically and quantitatively reducing the feeding amount of the kiln; and if the temperature of the outlet of the decomposing furnace is always higher than the set target in the set period, automatically and quantitatively increasing the feeding amount of the kiln.

Further, in the step c, the data preprocessing method includes: rejecting invalid data by setting an upper limit and a lower limit; for random noise, removing by high-pass filtering, low-pass filtering, data smoothing and Kalman filtering; and eliminating the significant errors by adopting a residual error analysis method, a correction value analysis method, a generalized likelihood ratio method, a Bayes method, an increment method and an principal component analysis method.

The invention has the advantages and positive effects that: the invention considers the influence of a plurality of variables such as kiln tail coal feeding quantity, kiln feeding quantity, tertiary air temperature, smoke chamber temperature and the like on the outlet temperature of the decomposing furnace, establishes a mathematical model which is more in line with the actual production working condition for control, and has more accurate control effect. The method can make corresponding adjustment according to the change of each disturbance variable, and can carry out self-adaptive control aiming at the whole working condition of the cement burning system; the fluctuation of the temperature of the decomposing furnace under the normal working condition is reduced to be within +/-5 ℃, and even can be kept within +/-2 ℃ for a long time, and the fluctuation of the whole working condition is within +/-10 ℃. Compared with manual control, the fluctuation range is reduced by 75%. The algorithm can be used for long-term stable and safe operation, and the fluctuation of the production process is reduced to the maximum extent. The system operation rate can reach more than 99 percent, and the labor intensity of operators is greatly reduced.

Drawings

FIG. 1: and (3) a closed-loop control system block diagram of the outlet temperature of the decomposing furnace.

In the figure: w (t) is a control target value in the control target curve; e (t) is the error value between the actual value and the target value; y is₁(t) is the decomposing furnace outlet temperature component influenced by the kiln tail coal feeding amount;is the temperature component at the outlet of the decomposing furnace influenced by the feeding amount of the kiln;the temperature component of the outlet of the decomposing furnace influenced by the temperature of the tertiary air;is the component of the outlet temperature of the decomposing furnace influenced by the temperature of the smoke chamber; u (t) is the coal feeding amount of the kiln tail; v. of₁(t) kiln feed rate; v. of₂(t) is the tertiary air temperature; v. of₃(t) is the smoke chamber temperature.

Detailed Description

In order to further understand the contents, features and effects of the present invention, the following embodiments are illustrated and described in detail with reference to the accompanying drawings:

referring to fig. 1, a method for controlling the optimal temperature of a cement burning decomposing furnace under all conditions in a self-adaptive manner is characterized by comprising the following steps:

The kiln tail coal feeding amount can be used as a main adjusting variable to perform key control, the kiln feeding amount is used as a secondary adjusting variable, and the tertiary air temperature and the smoke chamber temperature are used as disturbance variables to be not adjusted.

Further, in the step a, the specific steps of establishing the GPC controller may be as follows:

firstly, respectively obtaining a control model of kiln tail coal feeding quantity, kiln feeding quantity, tertiary air temperature and smoke chamber temperature to the outlet temperature of the decomposing furnace through step response test. Assuming that the outlet temperature of the decomposing furnace is an output variable y (k), and the kiln tail coal feeding quantity, the kiln feeding quantity, the tertiary air temperature and the smoke chamber temperature are respectively operation variables u (k), v₁(k)、v₂(k)、v₃(k) And describing the corresponding relation between the outlet temperature of the decomposing furnace and each operation variable by adopting a first-order inertia time-lag model. The transfer function takes the form

Where K is the scaling factor of the system, T is the time constant, and τ is the time lag.

step a-2: a CARIMA model can be adopted to establish a control model of a decomposing furnace outlet temperature control system, and the expression is as follows: a (z)^-1)Y(k)＝B(z^-1)R(k-1)+ζ(k)/Δ (2)

A (z^{- 1}) = 1 + a_{1} z^{- 1} + a_{2} z^{- 2} + ... + a_{n_{a}} z^{- n_{a}}

B (z^{- 1}) = b_{0} + b_{1} z^{- 1} + b_{2} z^{- 2} + ... + b_{n_{b}} b^{- n_{b}}

Wherein Y (k) is the control object output vector, R (k) is the control object input vector, and ζ (k) is the zero mean noise orderThe noise sequence is reflected in the prediction error after the actual system action, so that the noise sequence can not be considered in the calculation before the prediction error; a (z)^-1)、B(z^-1) Respectively representing the operators z^-1A polynomial of (a); 1-z^-1Is a difference operator; then A (z)^-1)、B(z^-1) R (k-1) corresponds to an input variable: the corresponding matrix of the kiln tail coal feeding amount, the kiln feeding amount, the tertiary air temperature and the smoke chamber temperature is as follows:

B (z^{- 1}) = {[\begin{matrix} B_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * B_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * B_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * B_{V_{3}} (z^{- 1}) \end{matrix}]}^{T}

R (k - 1) = [\begin{matrix} u (k - 1) \\ v_{1} (k - 1) \\ v_{2} (k - 1) \\ v_{3} (k - 1) \end{matrix}]

wherein,

from the above, A (z)^-1) Of order n_a＝4，B(z^-1) Of order of

In equation (3), i is the number corresponding to the different input variables: 1 corresponds to the input variable of the coal feeding quantity at the tail of the kiln; i-v₁Inputting variables corresponding to the feeding amount of the kiln; i-v₂Inputting variables corresponding to the temperature of the tertiary air; i-v₃Inputting variables corresponding to the temperature of the smoke chamber; u (k-1) is an input variable value of the kiln tail coal feeding quantity at the k-1 moment; v. of₁(k-1) inputting a variable value for the kiln feeding amount at the time of k-1; v. of₂(k-1) inputting a variable value of the temperature of the tertiary air at the k-1 moment; v. of₃(k-1) inputting a variable value for the smoke chamber temperature at the moment of k-1; a. the_i、B_iIs an operator z corresponding to different input variables^-1Polynomial expression: tau is_iIs the time lag for different input variables: k_iIs the scaling factor for different input variables: t is_iIs the time constant for the different input variables: t is_sIs the sampling time of the system; such as: a. the₁、B₁Operator z for corresponding kiln tail coal feeding quantity input^-1A polynomial;operator z input for corresponding kiln feed volume^-1A polynomial;operator z for input of corresponding tertiary air temperature^-1A polynomial;operator z for corresponding smoke chamber temperature^-1A polynomial; tau is₁Inputting time lag for the coal feeding amount of the corresponding kiln tail;the time lag is input corresponding to the feeding amount of the kiln;the time lag corresponds to the input of the tertiary air temperature;the time lag is input corresponding to the temperature of the smoke chamber; k₁Is the proportional coefficient corresponding to the input of the coal feeding amount at the tail of the kiln;is the proportional coefficient corresponding to the input of the kiln feeding amount;is a proportional coefficient corresponding to the temperature input of the tertiary air;is the proportionality coefficient corresponding to the smoke chamber temperature input; t is₁Is the time constant corresponding to the input of the coal feeding amount at the tail of the kiln;is the time constant corresponding to the input of the feeding amount of the kiln;is a time constant corresponding to the temperature input of the tertiary air;is the time constant corresponding to the smoke chamber temperature input;

step a-3: can define E (z)^-1) And F (z)^-1) In the form of formula (4), A (z) is obtained from said step a-2^-1) Calculate outFurther, an intermediate variable E (z) is obtained^-1) And F (z)^-1) Wherein E (z)^-1) And F (z)^-1) The expression of (a) is as follows:

E(z^-1)＝[E₁(z^-1),···E_P(z^-1)]^T (4)

F(z^-1)＝{F₁(z^-1),···F_P(z^-1)]^T

E_j(z^-1)＝e₀+e₁z^-1+e₂z^-2+···+e_j-1z^-j+1

F_j(z^-1)＝f_j，0+f_j,1z^-1+f_j,2z^-2+f_j,3z^-3+f_j,4z^-4

\begin{matrix} f_{j + 1, i} = f_{j, i + 1 -} {\tilde{a}}_{i + 1} e_{j} & i = 1, 2, 3 \\ f_{j + 1, 4} = {\tilde{a}}_{5} e_{j} = - {\tilde{a}}_{5} f_{j, 0} & i = 4 \end{matrix}

E_j+1(z^-1)＝E_j(z^-1)+e_jz^-j；

when the initial value j is 1, E₁(z^-1)＝1，e₀＝1，

step a-4: b (z) which can be determined from the steps a-2 and a-3^-1)、E(z^-1) And F (z)^-1) Obtaining a system step response matrix G (z)^-1) And a historical response matrix H (z)^-1)，G(z^-1) And H (z)^-1) The expression is as follows:

\begin{matrix} G_{j} (z^{- 1}) = g_{0} + g_{1} z^{- 1} + g_{2} z^{- 2} + ... + g_{n_{b} + j - 1} z^{- n_{b} - j + 1} & j = 1, 2..., P \end{matrix}

g_j＝e_jb₀+h_j,0

when the initial value j is 1:

G₁(z^-1)＝g₀＝e₀b₀，

step a-5: the intermediate variable F (z) can be determined from the steps a-3 and a-4^-1) And H (z)^-1) The past response f (k) of the system is obtained in the form:

f(k)＝[f₁(k) f₂(k) … f_P(k)]^T

Further, in the step d, an optimum control increment Δ r (k) ═ d [ Y ] may be calculated from the steps a to 4 and a to 5_s-f(k)-c*e(k)]Wherein d isFirst row of (2), matrixIs a matrix formed by the elements of the first row and the first column of each element in the matrix G in the step a-4, Y_s＝[y_s(k+1)^T,…,y_s(k+N)^T]^TIs the following control target curve:

y_s(k)＝y(k)

y_s(k+j)＝ay_s(k+j-1)+(1-a)W,j＝1,…,P

wherein, W is a control target value, lambda is a control weighting coefficient, a is a softening coefficient, c is an error correction parameter, and the four parameters are all given manually; e (k) is a prediction error, which is generated by applying the second term ζ (k)/Δ on the right side of the equation (2) to the actual system, and represents an error between the actual value and the predicted value;

Further, in the step d, a specific method for self-correcting the control model parameters may be as follows:

(2) if the temperature of the outlet of the decomposing furnace is always lower than the set target in the set period, the feeding amount of the kiln can be automatically and quantitatively reduced; if the temperature at the outlet of the decomposing furnace is always higher than the set target in the set period, the feeding amount of the kiln can be automatically and quantitatively increased.

Further, in the step c, the method for preprocessing data may include: rejecting invalid data by setting an upper limit and a lower limit; for random noise, removing by high-pass filtering, low-pass filtering, data smoothing and Kalman filtering; and eliminating the significant errors by adopting a residual error analysis method, a correction value analysis method, a generalized likelihood ratio method, a Bayes method, an increment method and an principal component analysis method.

By adopting the control algorithm, the fluctuation of the temperature of the decomposing furnace under the normal working condition can be reduced to be within +/-5 ℃, even can be kept within +/-2 ℃ for a long time, and the fluctuation of the whole working condition is within +/-10 ℃. Compared with manual control, the fluctuation range is reduced by 75%.

The algorithm can be used for long-term stable and safe operation. The system is safe and reliable, can safely and stably operate for a long time under normal working conditions, and reduces the fluctuation of the production process to the maximum extent. The system operation rate can reach more than 99 percent, and the labor intensity of operators is greatly reduced.

Although the preferred embodiments of the present invention have been described above with reference to the accompanying drawings, the present invention is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and those skilled in the art can make many modifications without departing from the spirit and scope of the present invention as defined in the appended claims.

Claims

1. A full-working-condition self-adaptive temperature optimization control method for a cement burning decomposing furnace is characterized by comprising the following steps:

2. The method for controlling the full-working-condition self-adaptive temperature optimization of the cement burning decomposing furnace according to the claim 1, wherein in the step a, the specific steps of establishing the GPC controller are as follows:

A(z^-1)Y(k)＝B(z^-1)R(k-1)+ζ(k)/Δ （2）

A (z^{- 1}) = 1 + a_{1} z^{- 1} + a_{2} z^{- 2} + ... + a_{n_{a}} z^{- n_{a}}

B (z^{- 1}) = b_{0} + b_{1} z^{- 1} + b_{2} z^{- 2} + ... + b_{n_{b}} z^{- n_{b}}

wherein, y (k) is a control object output vector, r (k) is a control object input vector, and ζ (k) is a zero-mean noise sequence which is uncorrelated with each other; a (z)^-1)、B(z^-1) Respectively representing the operators z^-1A polynomial of (a); Δ =1-z^-1Is a difference operator; then A (z)^-1)、B(z^-1) R (k-1) corresponds to an input variable: the corresponding matrix of the kiln tail coal feeding amount, the kiln feeding amount, the tertiary air temperature and the smoke chamber temperature is as follows:

B (z^{- 1}) = {[\begin{matrix} B_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * B_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * B_{V_{2}} (z^{- 1}) * A_{V_{3}} (z^{- 1}) \\ A_{1} (z^{- 1}) * A_{V_{1}} (z^{- 1}) * A_{V_{2}} (z^{- 1}) * B_{V_{3}} (z^{- 1}) \end{matrix}]}^{T}

R (k - 1) = [\begin{matrix} u (k - 1) \\ v_{1} (k - 1) \\ v_{2} (k - 1) \\ v_{3} (k - 1) \end{matrix}]

wherein,

\begin{matrix} A_{i} (z^{- 1}) = 1 - e^{- \frac{T_{s}}{T i}} z^{- 1} & i = 1, v_{1}, v_{2}, v_{3} \end{matrix}

from the above, A (z)^-1) Of order n_a=4，B(z^-1) Of order of

In equation (3), i is the number corresponding to the different input variables: i =1 corresponds to the input variable of the coal feeding quantity of the kiln tail; i-v₁Inputting variables corresponding to the feeding amount of the kiln; i-v₂Inputting variables corresponding to the temperature of the tertiary air; i-v₃Inputting variables corresponding to the temperature of the smoke chamber; u (k-1) is an input variable value of the kiln tail coal feeding quantity at the k-1 moment; v. of₁(k-1) inputting a variable value for the kiln feeding amount at the time of k-1; v. of₂(k-1) inputting a variable value of the temperature of the tertiary air at the k-1 moment; v. of₃(k-1) inputting a variable value for the smoke chamber temperature at the moment of k-1; a. the_i、B_iIs an operator z corresponding to different input variables^-1Polynomial expression: tau is_iIs the time lag for different input variables: k_iIs the scaling factor for different input variables: t is_iIs the time constant for the different input variables: t is_sIs the sampling time of the system;

step a-3: definition E (z)^-1) And F (z)^-1) Is in the form of formula (4)From said step a-2, A (z) is obtained^-1) Calculate outFurther, an intermediate variable E (z) is obtained^-1) And F (z)^-1) Wherein E (z)^-1) And F (z)^-1) The expression of (a) is as follows:

E(z^-1)＝[E₁(z^-1),…E_P(z^-1)]^T （4）

F(z^-1)＝[F₁(z^-1),…F_P(z^-1)]^T

E_f(z^-1)=e₀+e₁z^-1+e₂z^-1+…+e_j-1z^-j+1

F_j(z^-1)=f_j,0+f_j,1z^-1+f_j,2z^-2+f_j,3z^-3+f_j,4z^-4

\begin{matrix} f_{j + 1, i} = f_{j, i + 1} - {\tilde{a}}_{i + 1} e_{j} & i = 1, 2, 3 \end{matrix}

\begin{matrix} f_{j + 1, 4} = - {\tilde{a}}_{5} e_{j} = - {\tilde{a}}_{5} f_{j, 0} & i = 4 \end{matrix}

E_j+1(z^-1)=E_j(z^-1)+e_jz^-j；

when the initial value j =1, E₁(z^-1)=1，e₀＝1，

\begin{matrix} G_{j} (z^{- 1}) = g_{0} + g_{1} z^{- 1} + g_{2} z^{- 2} + ... + g_{n_{b} + j - 1} z^{- n b - j + 1} & j = 1, 2, ..., P \end{matrix}

g_j＝e_jb₀+h_j,0

h_j+1,i-1＝e_jb₀+h_j,i 1≤i≤n_b

h_{j + 1, n_{b} - 1} = e_{j} b_{n_{b}}

when initial value j = 1:

G₁(z^-1)=g₀＝e₀b₀，

f(k)＝[f₁(k) f₂(k) … f_P(k)]^T

3. The method as claimed in claim 2, wherein in step d, the optimal control increment Δ r (k) ═ d [ Y ] is calculated from steps a-4 and a-5_s-f(k)-c*e(k)]Wherein d isFirst row of (2), matrixIs a matrix formed by the elements of the first row and the first column of each element in the matrix G in the step a-4, Y_s＝[y_s(k+1)^T,…,y_s(k+N)^T]^TIs the following control target curve:

y_s(k)＝y(k)

y_s(k+j)＝ay_s(k+j-1)+(1-a)W,j=1,…,P

the prediction error e (k) is expressed as: e (a)k)＝y(k)-y_p(k) Wherein y (k) is an actual value of the current decomposing furnace outlet temperature, y_p(k) Is the predicted value of the outlet temperature of the decomposing furnace at the last moment;

4. The method for controlling the optimal temperature of the cement burning decomposing furnace in the full-working-condition self-adaption mode according to the claim 2, wherein in the step d, the specific method for controlling the model parameters in the self-correction mode is as follows:

5. The method for controlling the full-condition adaptive temperature optimization of the cement burning decomposing furnace according to the claim 1, wherein in the step c, the method for preprocessing the data comprises the following steps: rejecting invalid data by setting an upper limit and a lower limit; for random noise, removing by high-pass filtering, low-pass filtering, data smoothing and Kalman filtering; and eliminating the significant errors by adopting a residual error analysis method, a correction value analysis method, a generalized likelihood ratio method, a Bayes method, an increment method and an principal component analysis method.