CN101881563B - Multi-area intelligent online optimizing control method for thermal efficiency of heating furnace - Google Patents

Abstract

The invention relates to a multi-area intelligent online optimizing control method for thermal efficiency of a heating furnace, which belongs to the technical field of the control of the thermal efficiency of heating furnaces. The method is characterized in that a thermal efficiency optimized off-line part carries out working area division on the working condition of a heating furnace according to a thermal load and obtains optimizing values of oxygen content and negative pressure in each working area by historical data excavation. An online part confirms the working area of the heating furnace according to process real-time data, can be maintained in a better working state in time when the load is changed by starting with the optimizing values of the oxygen content and the negative pressure obtained by the off-line part and carries out thermal efficiency self-optimizing control when operation is steady, thereby achieving the aims of rapid optimization and long-time working near an optimizing state. A smoke air system carries out area control on the oxygen content and the negative pressure of a hearth by adopting a control method based on dynamic feedforward and steady feedback, and can ensure that the heating furnace works near an optimized working point given by a thermal efficiency optimizing method so as to realize the optimal thermal efficiency.

Description

Multi-area intelligent online optimizing control method for thermal efficiency of heating furnace

Technical field

The present invention relates to the on-line optimizing and controlling method of thermal efficiency of heating furnace, belong to petrochemical industry tubular heater and production process automation field.

Background technology

Tubular heater (to call heating furnace in the following text) is the important technological equipment of production process extensive uses such as oil refining and petrochemical industry, almost in every suit commercial plant heating furnace is arranged all.Heating furnace is that device provides thermal source, is the capital equipment of device energy-wasting.For example, the atmospheric and vacuum distillation fuel consumption accounts for the device total energy consumption at 82-92%, and delayed coking is about 90%.Heating furnace fuel can produce CO in combustion process ₂, gases such as CO.Heating furnace to operation carries out thermal efficiency optimization, makes its long-term work near optimum combustion regime, and under the situation that does not change technology, realize saving energy and reduce the cost, and reduce the environmental pollution that causes because of imperfect combustion, be a urgent task.

The thermal efficiency of heating furnace is the percentage that the effective heat that utilizes of heated medium accounts for the aggregate supply heat.As far as heating furnace, if air quantity is too small, can cause fuel combustion insufficient, waste fuel also produces gaseous contamination environment such as CO; If air quantity is excessive, it is too much then to discharge the heat that flue gas takes away, and the thermal efficiency reduces.Therefore, there is the optimum thermal efficiency in corresponding certain operating mode.The adjustment means of thermal efficiency of heating furnace are air quantity, through the adjustment air quantity, keep rational excess air coefficient, reach optimum thermal efficiency.

Control and optimization about thermal efficiency of heating furnace have had some researchs and application result.The at present industrial method that more generally adopts is that oxygen content is controlled efficiency of combustion indirectly in the control flue gas, but load, when fuel changes, respective change also can take place in optimal oxygen content.The bibliographical information internal model control and the thermal efficiency are used to the real-time control and the optimization of thermal efficiency of heating furnace from methods such as optimizing.But the control to the thermal efficiency can not guarantee optimum thermal efficiency, also exists the controllability problem.Adopt the method for the thermal efficiency from optimizing, when the heating furnace thermic load changed greatly, optimizing speed was slower.

PID control is generally adopted in the control of industry upper fume system.Owing to the heating furnace air quantity is regulated reasons such as executing agency's sensitivity is relatively poor, the dead band is big, general discomfort is fit to do continuous adjusting, and above-mentioned control loop major part is in manual state for a long time.Owing to the heating furnace air quantity is regulated reasons such as executing agency's sensitivity is relatively poor, the dead band is big, general discomfort is fit to do continuous adjusting, and above-mentioned control loop is in manual state for a long time, let alone the optimization of the thermal efficiency.Therefore, need exploitation appropriate control strategy to realize automatic control, for the optimization that realizes the thermal efficiency lays the foundation to the heating furnace air and gas system.

Summary of the invention

The object of the invention: the multi-area intelligent online optimizing control method for thermal efficiency of heating furnace that provides a kind of practicality.Method comprises thermal efficiency optimal control and air and gas system control two parts.Thermal efficiency optimal control partly is divided into off-line and online two parts.The off-line part is carried out the working region division according to thermic load to the heating furnace operating mode, excavates through historical data, obtains the optimization work point in each working region.Online part is confirmed the working region of heating furnace according to the process real-time data, in each working region, carries out the thermal efficiency then from seeking optimum control, reaches near fast seeking and the long-term work purpose the optimization state.The control method based on " dynamic Feedforward, stable state feedback " is adopted in the control of air and gas system, and oxygen content and combustion chamber draft are controlled, and guarantees that heating furnace has stable fired state, for optimization lays the foundation.

The invention is characterized in: said method realizes in host computer successively according to the following steps:

Steps A: host computer initialization:

In said host computer, set up with lower module: on-line optimization module, air and gas system control module and real-time data base/OPC bitcom module, wherein:

The on-line optimization module; Through controling the real time data of gathering heating furnace from heating furnace controlled device and Distributed Control System, and be sent to said real-time data base to the historical optimization work point in the thermic load of heating furnace zone and the regional current optimization work point of thermic load and supply said air and gas system control module use through the OPC bitcom;

The air and gas system control module; Under the effect of said OPC bitcom; Gather oxygen content and combustion chamber draft numerical value in real time; And read the determined optimization work point of on-line optimization module in the said real-time data base; The combustion chamber draft in the air-fuel ratio when fuel is increased, thermic load zone oxygen content and thermic load zone is controlled, and oxygen content and combustion chamber draft are maintained near the optimization work point, and the control action that the combustion chamber draft control in the air-fuel ratio control when fuel increases, thermic load zone Control for Oxygen Content and thermic load zone is calculated is sent to said heating furnace controlled device and Distributed Control System through said OPC bitcom;

Step B: said on-line optimization module, carry out thermal efficiency offline optimization and thermal efficiency on-line optimization successively according to the following steps:

Step B1: thermal efficiency offline optimization, its step is following:

Step B1.1: step test obtains the steady-state response time of the thermal efficiency:

Under the steady situation of process, the oxygen content setting value is applied the step test signal, record thermal efficiency change curve, the steady-state response time T of the acquisition thermal efficiency _r

Step B1.2: set sampling period T, T ∈ [0.25T _r, T _r], and gather following heating furnace field data: heating furnace oxygen content, heating furnace burner hearth negative pressure, furnace outlet temperature, heating furnace inlet temperature, furnace charge flow, fuel flow rate and intake, and carry out the off-line modeling analysis according to the following steps;

Step B1.3: heated medium is not had the heating furnace of phase transformation, calculate the effective heat duty Q (k) of heating furnace:

Q(k)＝F(k)C _p[T _out(k)-T _in(k)]

Wherein: k is sampling instant,

Q (k) is the k effective heat duty of heating furnace constantly,

F (k) is the k flow of heated medium constantly,

T _Out(k) be the k outlet temperature of heated medium constantly,

T _In(k) be the k inlet temperature of heated medium constantly,

C _pSpecific heat for heated medium;

Step B1.4: utilize thermic load territorial classification device to be divided into N zone according to the residing working region of big young pathbreaker's heating furnace thermic load of thermic load, N=5, Ω is used in the zone _iExpression, i=1,2 ..., 5;

The working range of loading when the heating furnace normal heat is designing effective heat duty Q ₀0.75 times when between 1.25 times, changing, the zone in 5 zones limit is respectively:

[0.75Q ₀，0.85Q ₀)，[0.85Q ₀，0.95Q ₀)，[0.95Q ₀，1.05Q ₀)，

[1.05Q ₀，1.15Q ₀)，[1.15Q ₀，1.25Q ₀]

Use Ω _IL, Ω _IHRepresent regional Ω respectively _iThe lower limit and the upper limit;

Step B1.5: according to positive balance method calculating thermal efficiency:

$\mathrm{\η}\left(k\right)=\frac{Q\left(k\right)}{H_f{F}_f\left(k\right)}$

Wherein: η (k) is a k calculating thermal efficiency value constantly,

F _f(k) be k fuel flow rate constantly,

H _fBe the fuel combustion calorific value;

Step B1.6:, seek thermic load working region Ω according to the heating furnace field data that obtains among the step B1.2 _iThe historical optimization work point of internally heated oven makes object function j _iMaximum:

$J_i=\underset{O_{2i}\left(k\right),P_i\left(k\right)}{\max }{\mathrm{\η}}_i\left(k\right)$ i＝1，2，…，5

Wherein: η _i(k) be thermic load working region Ω _iInterior k is the calculating thermal efficiency value constantly,

O _2i(k) be the corresponding constantly measurement of oxygen content value of k,

P _i(k) be the corresponding constantly combustion chamber draft measured value of k;

Making maximum oxygen content of the thermal efficiency and combustion chamber draft value as said thermic load working region Ω _iThe historical optimization work point of internally heated oven is used Ω _{I, opt}Expression:

Ω _i，opt＝(O _2i，opt，P _i，opt) i＝1，2，…，5

Wherein: Ω _{I, opt}Be thermic load working region Ω _iHistorical optimization work point,

O _{2i, opt}Be thermic load working region Ω _iThe historical optimal value of interior oxygen content,

P _{I, opt}Be thermic load working region Ω _iThe historical optimal value of internal furnace negative pressure;

Step B2: thermal efficiency on-line optimization, its step is following:

Step B2.1: the real time data of gatherer process comprises: heating furnace oxygen content, heating furnace burner hearth negative pressure, furnace outlet temperature, heating furnace inlet temperature, furnace charge flow and fuel flow rate;

Step B2.2: utilize the thermic load territorial classification device deterministic process present located thermic load working region Ω among the step B1.4 _i, and obtain the historical optimization work point Ω of corresponding calculated off line gained _{I, opt}

Step B2.3: the said method of B1.5 is calculated current time thermal efficiency η (k) set by step;

Step B2.4: whether the decision criteria deterministic process by following is in stable state:

$\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}\left(\frac{1}{L}\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}\right|\frac{y_{\mathrm{jl}}-{\stackrel{\&OverBar;}{y}}_j}{{\stackrel{\&OverBar;}{y}}_j}\left|\right)<\mathrm{\&epsiv;}$

Wherein: y _Il, j=1,2,3 are respectively l value of characteristic variable furnace outlet temperature, fire box temperature and inlet amount,

L is whether deterministic process is in stable historical data length, and L*T=30min, T are the sampling periods,

j=1; 2; 3 are respectively the mean value of characteristic variable furnace outlet temperature, fire box temperature and inlet amount

ε is preassigned stable state decision threshold, and span is (0,0.1);

Step B2.5: confirm optimization work point under the current thermic load according to the following steps:

Step B2.5.1: if process is in stable state, then the online optimizing of the thermal efficiency is adopted from seeking method for optimally controlling, serves as to transfer excellent variable with the oxygen content setting value, with the historical optimal value O of oxygen content _{2i, opt}Be initial value, online searching makes the highest oxygen content setting value O of the thermal efficiency _2s(k), step is following:

Step B2.5.1.1: the changing value Δ η (k) that calculates the thermal efficiency of going up a moment k-1 relatively:

Δη(k)＝η(k)-η(k-1)；

Step B2.5.1.2: if | Δ η (k) |＜Δ η _Min, then stop optimizing and record oxygen content setting value at this moment, wherein Δ η _MinBe preset thermal efficiency adjustment dead band;

Step B2.5.1.3: if | Δ η (k) |>=Δ η _Min, then with Δ O _2s(k) be said oxygen content setting value O _2s(k) optimizing step-length automatic optimal:

[0061]? $\mathrm{\&Delta;}{O}_{2s}\left(k\right)=\mathrm{\&lambda;}\frac{\mathrm{\&Delta;\&eta;}\left(k\right)}{\mathrm{\&Delta;}{O}_{2s}(k-1)}{O}_{2s}(k-1)$

O _2s(k)＝O _2s(k-1)+ΔO _2s(k)

[0063]Wherein: λ for the adjustment coefficient, span be (0,1];

Step B2.5.2: if process is not in stable state, with the historical optimal value O of oxygen content _{2i, opt}Be oxygen content setting value O _2s(k), guarantee that heating furnace is in the duty of suboptimum;

Step B2.6: the oxygen content setting value O that calculates _2s(k) be sent to said real-time data base, supply the Control for Oxygen Content device in the said air and gas system control module to implement;

Step C: the control object of said air and gas system control module comprises the air-fuel ratio when oxygen content, combustion chamber draft and fuel increase; The control target is to be controlled at said oxygen content and combustion chamber draft in the optimization set point or given range that is drawn jointly by said offline optimization and on-line optimization, and step is following:

Step C1: step test obtains the steady-state response time of oxygen content and combustion chamber draft:

Under the steady situation of process, intake is applied the step test signal, record oxygen content change curve, the steady-state response time T of acquisition oxygen content _O2ss

Under the steady situation of process, the air-introduced machine inlet baffle is applied the step test signal, record combustion chamber draft change curve, the steady-state response time T of acquisition combustion chamber draft _Pss

Step C2: the control cycle of setting the air and gas system control module is T _c, T _c=min (T _O2ss, T _Pss)/40;

Step C3: when fuel increases air-fuel ratio is controlled according to the following steps:

Step C3.1: if fuel recruitment Δ F _f(k _c) last relatively intake adjustment moment k _c-1 fuel quantity F _F0Ratio surpasses predetermined threshold value β, and β ∈ (0,0.2], promptly

$\frac{\mathrm{\&Delta;}{F}_f\left(k_c\right)}{F_{f0}}>\mathrm{\&beta;}$

The feedforward variation delta F of air intake when then fuel increases _AF(k _c) do

ΔF _aF(k _c)＝α·AFR·ΔF _f(k _c)

Wherein: Δ F _f(k _c)=F _f(k _c)-F _F0Be fuel change amount, F _f(k _c) be fuel flow rate,

α is an excess air coefficient, is [1.05,1.15] to the fuel gas span;

AFR is a stoichiometric air/, and to fuel gas, it is calculated as:

$\mathrm{AFR}=0.01\&times;4.76\&times;[0.5\mathrm{CO}+0.5H_2+\mathrm{\&Sigma;}(m+\frac{n}{4})C_m{H}_n+1.5H_{2}S-O_2]$

Wherein: CO, H ₂, C _mH _n, H ₂S, O ₂Being each constituent content in the fuel gas, is unit with %;

Step C4: according to the following steps oxygen content is carried out Region control:

Step C4.1: if the oxygen content setting value is O _2s(k), the regional extent of its permission is:

[O _2s(k)-δ，O _2s(k)+δ]

Wherein: δ is that oxygen content departs from oxygen content setting value O _2s(k) zone limit, δ ∈ (0,0.5);

Step C4.2: at Control for Oxygen Content moment k _c,

If O ₂(k _c)＞O _2HPerhaps O ₂(k _c)＜O _2L, O wherein _2H=O _2s(k)+and δ, O _2L=O _2s(k)-δ, and said Control for Oxygen Content device is not in the stand-by period, then is calculated as follows required intake and changes:

$\mathrm{\&Delta;}{F}_a\left(k_c\right)=\frac{O_{2s}\left(k\right)-O_2\left(k_c\right)}{21-O_{2s}\left(k\right)}[F_a(k_c-1)+F_f\left(k_c\right)]$

Wherein: Δ F _a(k _c) be the variable quantity of intake,

O _2s(k) be the oxygen content setting value,

O ₂(k _c) be the measured value of oxygen content;

If O _2L≤O ₂(k _c)≤O _2H, Δ F then _a(k _c)=0;

According to Control for Oxygen Content moment k _cCombustion chamber draft P (k _c) by the variation delta F of following principle correction intake _a(k _c):

If P (k _c)＞P _HS, P _HSBe the upper safety limit of combustion chamber draft, and the variation delta F of intake _a(k _c)＞0 then keeps intake constant, makes Δ F _a(k _c)=0,

If P (k _c)＜P _LS, P _LSBe the lower safety limit of combustion chamber draft, and the variation delta F of intake _a(k _c)＜0 then keeps intake constant, makes Δ F _a(k _c)=0;

Step C4.3: be calculated as follows required intake:

F _a(k _c)＝F _a(k _c-1)+ΔF _a(k _c)

Wherein: F _a(k _c) be intake;

After process made feedback regulation, wait for the steady-state response time T of an oxygen content _O2ss

Step C5: according to the following steps combustion chamber draft is carried out Region control:

Step C5.1: if the setting value of combustion chamber draft is P _s, and be in thermic load working region Ω _i, P then _s=P _{I, opt}, i=1,2 ..., 5, it allows the regional extent of change to be:

[P _s-σ，P _s+σ]

Wherein: σ is the zone limit that combustion chamber draft departs from setting value, σ ∈ (0,10);

Step C5.2: at combustion chamber draft control moment k _c,

If P (k _c)＞P _HPerhaps P (k _c)＜P _L, P wherein _H=P _s+ σ, P _L=P _s-σ, and the combustion chamber draft controller is not in the stand-by period, and then the deviation of combustion chamber draft is e _P(k _c)=P _s-P (k _c), air-introduced machine inlet baffle variation delta MV then ₂(k _c) do

$\mathrm{\&Delta;}{\mathrm{MV}}_2\left(k_c\right)=\frac{e_P\left(k_c\right)}{K_1}$

Wherein: K ₁Proportionality coefficient for air-introduced machine inlet baffle and combustion chamber draft variation;

After process made feedback regulation, wait for the steady-state response time T of a combustion chamber draft _Pss

Under other situation, Δ MV ₂(k _c)=0;

Step C5.3: FEEDFORWARD CONTROL is carried out in the variation of blower variable frequency by following formula:

ΔMV _2F(k _c)＝K ₂ΔMV ₁(k _c)

Wherein: Δ MV _2F(k _c) be the feedforward variation of air-introduced machine inlet baffle,

Δ MV ₁(k _c)=MV ₁(k _c)-MV ₁(k _cThe variation of-1) exporting for blower variable frequency,

K ₂Be the feed-forward coefficients between baffle plate variation and the frequency conversion output.

Use proof: the heating furnace offline optimization carries out the working region according to thermic load to the heating furnace operating mode to be divided, and excavates through historical data, can access the optimization work point in each working region.On-line optimization is confirmed the working region of heating furnace according to the process real-time data, in each working region, is that starting point is carried out the thermal efficiency from seeking optimum control with historical optimization work point, reaches near fast seeking and the long-term work purpose the optimization state.The control of air and gas system can effectively be removed the dangerous influence that the intake continuous closed-loop is regulated based on the control method of " stable state feedback, dynamic Feedforward " thought, removes the stability influence of Dynamic Coupling to control.

Description of drawings

Fig. 1 is heating furnace flow process and thermal efficiency multi-area intelligent optimal control structural representation.

Fig. 2 is that the thermal efficiency is optimized off-line part steps sketch map.

Fig. 3 is that the thermal efficiency is optimized online optimizing and air and gas system control general steps sketch map.

Fig. 4 is that the thermal efficiency is optimized thermic load territorial classification device.

Fig. 5 is a thermal efficiency on-line optimization program flow diagram.

The specific embodiment

Below in conjunction with accompanying drawing and case study on implementation, specific embodiments of the invention describes in further detail.Wherein, Fig. 1 is heating furnace flow process and thermal efficiency multi-area intelligent optimal control structural representation; Fig. 2 is that the thermal efficiency is optimized off-line part steps sketch map; Fig. 3 is that the thermal efficiency is optimized online optimizing and air and gas system control general steps sketch map; Fig. 4 is that the thermal efficiency is optimized thermic load territorial classification device; Fig. 5 is a thermal efficiency on-line optimization program flow diagram.

Thermal efficiency offline optimization

Step B1.1: step test obtains the steady-state response time of the thermal efficiency:

Under the steady situation of process, the oxygen content setting value is applied the step test signal, record thermal efficiency change curve, the steady-state response time T of the acquisition thermal efficiency _r

Step B1.2: set sampling period T, T ∈ [0.25T _r, T _r], and gather following heating furnace field data: heating furnace oxygen content, heating furnace burner hearth negative pressure, furnace outlet temperature, heating furnace inlet temperature, furnace charge flow, fuel flow rate and intake;

Gathering above data a period of time (as one month) is used for the off-line modeling analysis;

Step B1.3: heated medium is not had the heating furnace of phase transformation, calculate the effective heat duty Q (k) of heating furnace:

Q(k)＝F(k)C _p[T _out(k)-T _in(k)]

Wherein: k is sampling instant,

Q (k) is the k effective heat duty of heating furnace constantly,

F (k) is the k flow of heated medium constantly,

T _Out(k) be the k outlet temperature of heated medium constantly,

T _In(k) be the k inlet temperature of heated medium constantly,

C _pSpecific heat for heated medium;

Step B1.4: utilize thermic load territorial classification device to be divided into N zone according to the residing working region of big young pathbreaker's heating furnace thermic load of thermic load, N=5, Ω is used in the zone _iExpression, i=1,2 ..., 5;

The working range of loading when the heating furnace normal heat is designing effective heat duty Q ₀0.75 times when between 1.25 times, changing, the zone in 5 zones limit is respectively:

[0.75Q ₀，0.85Q ₀)，[0.85Q ₀，0.95Q ₀)，[0.95Q ₀，1.05Q ₀)，

[1.05Q ₀，1.15Q ₀)，[1.15Q ₀，1.25Q ₀]

Use Ω _IL, Ω _IHRepresent regional Ω respectively _iThe lower limit and the upper limit;

Step B1.5: according to positive balance method calculating thermal efficiency:

$\mathrm{\&eta;}\left(k\right)=\frac{Q\left(k\right)}{H_f{F}_f\left(k\right)}$

Wherein: η (k) is a k calculating thermal efficiency value constantly,

F _f(k) be k fuel flow rate constantly,

H _fBe the fuel combustion calorific value;

Step B1.6:, seek thermic load working region Ω according to the heating furnace field data that obtains among the step B1.2 _iThe historical optimization work point of internally heated oven makes object function J _iMaximum:

$J_i=\underset{O_{2i}\left(k\right),P_i\left(k\right)}{\max }{\mathrm{\&eta;}}_i\left(k\right)$ i＝1，2，…，5

Wherein: η _i(k) be thermic load working region Ω _iInterior k is the calculating thermal efficiency value constantly,

O _2i(k) be the corresponding constantly measurement of oxygen content value of k,

P _i(k) be the corresponding constantly combustion chamber draft measured value of k;

Making maximum oxygen content of the thermal efficiency and combustion chamber draft value as said thermic load working region Ω _iThe historical optimization work point of internally heated oven is used Ω _{I, opt}Expression:

Ω _i，opt＝(O _2i，opt，P _i，opt) i＝1，2，…，5

Wherein: Ω _{I, opt}Be thermic load working region Ω _iHistorical optimization work point,

O _{2i, opt}Be thermic load working region Ω _iThe historical optimal value of interior oxygen content,

P _{I, opt}Be thermic load working region Ω _iThe historical optimal value of internal furnace negative pressure.

Thermal efficiency on-line optimization

In host computer, the optimization timer is set, the cycle is T.The thermal efficiency is optimized online part and is carried out the following step at each optimization cycle:

Step B2.1: the real time data of gatherer process comprises: heating furnace oxygen content, heating furnace burner hearth negative pressure, furnace outlet temperature, heating furnace inlet temperature, furnace charge flow and fuel flow rate;

Step B2.2: utilize the thermic load territorial classification device deterministic process present located thermic load working region Ω among the step B1.4 _i, and obtain the historical optimization work point Ω of corresponding calculated off line gained _{I, opt}

Step B2.3: the said method of B1.5 is calculated current time thermal efficiency η (k) set by step;

Step B2.4: whether the decision criteria deterministic process by following is in stable state:

$\underset{j=1}{\overset{3}{\mathrm{\&Sigma;}}}\left(\frac{1}{L}\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}\right|\frac{y_{\mathrm{jl}}-{\stackrel{\&OverBar;}{y}}_j}{{\stackrel{\&OverBar;}{y}}_j}\left|\right)<\mathrm{\&epsiv;}$

Wherein: y _Il, j=1,2,3 are respectively l value of characteristic variable furnace outlet temperature, fire box temperature and inlet amount,

L is whether deterministic process is in stable historical data length, and L*T=30min, T are the sampling periods,

j=1; 2; 3 are respectively the mean value of characteristic variable furnace outlet temperature, fire box temperature and inlet amount

ε is preassigned stable state decision threshold, and span is (0,0.1);

Step B2.5: confirm optimization work point under the current thermic load according to the following steps:

Step B2.5.1: if process is in stable state, then the online optimizing of the thermal efficiency is adopted from seeking method for optimally controlling, serves as to transfer excellent variable with the oxygen content setting value, with the historical optimal value O of oxygen content _{2i, opt}Be initial value, online searching makes the highest oxygen content setting value O of the thermal efficiency _2s(k), step is following:

Step B2.5.1.1: the changing value Δ η (k) that calculates the thermal efficiency of going up a moment k-1 relatively:

Δη(k)＝η(k)-η(k-1)；

Step B2.5.1.2: if | Δ η (k) |＜Δ η _Min, then stop optimizing and record oxygen content setting value at this moment, wherein Δ η _MinBe preset thermal efficiency adjustment dead band;

Step B2.5.1.3: if | Δ η (k) |>=Δ η _Min, then with Δ O _2s(k) be said oxygen content setting value O _2s(k) optimizing step-length automatic optimal:

$\mathrm{\&Delta;}{O}_{2s}\left(k\right)=\mathrm{\&lambda;}\frac{\mathrm{\&Delta;\&eta;}\left(k\right)}{\mathrm{\&Delta;}{O}_{2s}(k-1)}{O}_{2s}(k-1)$

O _2s(k)＝O _2s(k-1)+ΔO _2s(k)

Wherein: λ for the adjustment coefficient, span be (0,1];

Step B2.5.2: if process is not in stable state, with historical optimization work point O _{2i, opt}Be setting value O _2s(k), guarantee that heating furnace is in the duty of suboptimum;

Step B2.6: the oxygen content setting value O that calculates _2s(k) be sent to said real-time data base, supply the Control for Oxygen Content device in the said air and gas system control module to implement;

Through the classification of thermic load multizone, reach to specific operation fast seeking and near the purpose of the long-term work optimization state, especially be fit to the heating furnace that thermic load often changes.

Air and gas system control

The control of air and gas system comprises air-fuel ratio control, oxygen content Region control and the combustion chamber draft Region control when fuel increases.Its target is that oxygen content and combustion chamber draft are controlled in the optimization set point or given range that above-mentioned offline optimization and on-line optimization determine.

Step C1: step test obtains the steady-state response time of oxygen content and combustion chamber draft:

Under the steady situation of process, intake is applied the step test signal, record oxygen content change curve, the steady-state response time T of acquisition oxygen content _O2ss

Under the steady situation of process, the air-introduced machine inlet baffle is applied the step test signal, record combustion chamber draft change curve, the steady-state response time T of acquisition combustion chamber draft _Pss

Step C2: the control cycle of setting the air and gas system control module is T _c, T _c=min (T _O2ss, T _Pss)/40;

Step C3: the air-fuel ratio control when fuel increases.The purpose of air-fuel ratio control is when fuel increases, to carry out dynamic Feedforward, prevents that burning is not enough.

Step C3.1: if fuel recruitment Δ F _f(k _c) last relatively intake adjustment moment k _c-1 fuel quantity F _F0Ratio surpasses predetermined threshold value β, and β ∈ (0,0.2], promptly

$\frac{\mathrm{\&Delta;}{F}_f\left(k_c\right)}{F_{f0}}>\mathrm{\&beta;}$

Wherein: Δ F _f(k _c)=F _f(k _c)-F _F0Be fuel change amount, F _f(k _c) be fuel flow rate;

The feedforward variation delta F of air intake when then fuel increases _AF(k _c) do

ΔF _aF(k _c)＝α·AFR·ΔF _f(k _c)

Wherein: α is an excess air coefficient, is [1.05,1.15] to the fuel gas span;

AFR is a stoichiometric air/, and it is calculated as to fuel gas:

$\mathrm{AFR}=0.01\&times;4.76\&times;[0.5\mathrm{CO}+0.5H_2+\mathrm{\&Sigma;}(m+\frac{n}{4})C_m{H}_n+1.5H_{2}S-O_2]$

Wherein: CO, H ₂, C _mH _n, H ₂S, O ₂Being each constituent content in the fuel gas, is unit with %;

Step C4: oxygen content Region control:

The controlled variable of oxygen content Region control is an oxygen content, and performance variable is intake (or calculating intake).The target of oxygen content Region control is with near the zone of Control for Oxygen Content setting value.

Step C4.1: if the oxygen content setting value is O _2s(k), the regional extent of its permission is:

[O _2s(k)-δ，O _2s(k)+δ]

Wherein: δ is that oxygen content departs from oxygen content setting value O _2s(k) zone limit, δ ∈ (0,0.5);

Step C4.2: at Control for Oxygen Content moment k _c,

If O ₂( _Kc)＞O _2HPerhaps O ₂(k _c)＜O _2L, O wherein _2H=O _2s(k)+and δ, O _2L=O _2s(k)-δ, and said Control for Oxygen Content device is not in the stand-by period, then is calculated as follows required intake and changes:

$\mathrm{\&Delta;}{F}_a\left(k_c\right)=\frac{O_{2s}\left(k\right)-O_2\left(k_c\right)}{21-O_{2s}\left(k\right)}[F_a(k_c-1)+F_f\left(k_c\right)]$

Wherein: Δ F _a(k _c) be the variable quantity of intake,

O _2s(k) be the oxygen content setting value,

O ₂(k _c) be the measured value of oxygen content;

If O _2L≤O ₂(k _c)≤O _2H, Δ F then _a(k _c)=0;

According to Control for Oxygen Content moment k _cCombustion chamber draft P (k _c) by the variation delta F of following principle correction intake _a(k _c):

If P (k _c)＞P _HS, P _HSBe the upper safety limit of combustion chamber draft, and the variation delta F of intake _a(k _c)＞0 then keeps intake constant, makes Δ F _a(k _c)=0,

If P (k _c)＜P _LS, P _LSBe the lower safety limit of combustion chamber draft, and Δ F _a(k _c)＜0 then keeps intake constant, makes Δ F _a(k _c)=0;

Step C4.3: be calculated as follows required intake:

F _a(k _c)＝F _a(k _c-1)+ΔF _a(k _c)

Wherein: F _a(k _c) be intake;

After process made feedback regulation, wait for the steady-state response time T of an oxygen content _O2ss

Step C5: combustion chamber draft Region control:

The controlled variable of combustion chamber draft Region control is a combustion chamber draft, and performance variable is the blower inlet baffle plate.The target of combustion chamber draft Region control is that combustion chamber draft is controlled near the zone the setting value.

Step C5.1: if the setting value of combustion chamber draft is P _s, and be in thermic load working region Ω _i, P then _s=P _{I, opt}, i=1,2 ..., 5, it allows the regional extent of change to be:

[P _s-σ，P _s+σ]

Wherein: σ is the zone limit that combustion chamber draft departs from setting value, σ ∈ (0,10);

Step C5.2: at combustion chamber draft control moment k _c,

If P (k _c)＞P _HPerhaps P (k _c)＜P _L, P wherein _H=P _s+ σ, P _L=P _s-σ, and the combustion chamber draft controller is not in the stand-by period, and then the deviation of combustion chamber draft is e _P(k _c)=P _s-P (k _c), air-introduced machine inlet baffle variation delta MV then ₂(k _c) do

$\mathrm{\&Delta;}{\mathrm{MV}}_2\left(k_c\right)=\frac{e_P\left(k_c\right)}{K_1}$

Wherein: K ₁Proportionality coefficient for air-introduced machine inlet baffle and combustion chamber draft variation;

After process made feedback regulation, wait for the steady-state response time T of a combustion chamber draft _Pss

Under other situation, Δ MV ₂(k _c)=0;

Step C5.3: FEEDFORWARD CONTROL is carried out in the variation of blower variable frequency by following formula:

ΔMV _2F(k _c)＝K ₂ΔMV ₁(k _c)

Wherein: Δ MV _2F(k _c) be the feedforward variation of air-introduced machine inlet baffle,

Δ MV ₁(k _c)=MV ₁(k _c)-MV ₁(k _cThe variation of-1) exporting for blower variable frequency,

K ₂Be the feed-forward coefficients between baffle plate variation and the frequency conversion output.

Claims (1)

1. multi-area intelligent online optimizing control method for thermal efficiency of heating furnace is characterized in that, said method realizes in host computer successively according to the following steps:

Steps A: host computer initialization:

In said host computer, set up with lower module: on-line optimization module, air and gas system control module and real-time data base/OPC bitcom module, wherein:

The on-line optimization module; Through controling the real time data of gathering heating furnace from heating furnace controlled device and Distributed Control System, and be sent to said real-time data base to the historical optimization work point in the thermic load of heating furnace zone and the regional current optimization work point of thermic load and supply said air and gas system control module use through the OPC bitcom;

The air and gas system control module; Under the effect of said OPC bitcom; Gather oxygen content and combustion chamber draft numerical value in real time; And read the determined optimization work point of on-line optimization module in the said real-time data base; The combustion chamber draft in the air-fuel ratio when fuel is increased, thermic load zone oxygen content and thermic load zone is controlled, and oxygen content and combustion chamber draft are maintained near the optimization work point, and the control action that the combustion chamber draft control in the air-fuel ratio control when fuel increases, thermic load zone Control for Oxygen Content and thermic load zone is calculated is sent to said heating furnace controlled device and Distributed Control System through said OPC bitcom;

Step B: said on-line optimization module, carry out thermal efficiency offline optimization and thermal efficiency on-line optimization successively according to the following steps:

Step B1: thermal efficiency offline optimization, its step is following:

Step B1.1: step test obtains the steady-state response time of the thermal efficiency:

Under the steady situation of process, the oxygen content setting value is applied the step test signal, record thermal efficiency change curve, the steady-state response time T of the acquisition thermal efficiency _r

Step B1.2: set sampling period T, T ∈ [0.25T _r, T _r], and gather following heating furnace field data: heating furnace oxygen content, heating furnace burner hearth negative pressure, furnace outlet temperature, heating furnace inlet temperature, furnace charge flow, fuel flow rate and intake, and carry out the off-line modeling analysis according to the following steps;

Step B1.3: heated medium is not had the heating furnace of phase transformation, calculate the effective heat duty Q (k) of heating furnace:

Q(k)＝F(k)C _p[T _out(k)-T _in(k)]

Wherein: k is sampling instant,

Q (k) is the k effective heat duty of heating furnace constantly,

F (k) is the k flow of heated medium constantly,

T _Out(k) be the k outlet temperature of heated medium constantly,

T _In(k) be the k inlet temperature of heated medium constantly,

C _pSpecific heat for heated medium;

Step B1.4: utilize thermic load territorial classification device to be divided into N zone according to the residing working region of big young pathbreaker's heating furnace thermic load of thermic load, N=5, Ω is used in the zone _iExpression, i=1,2 ..., 5;

The working range of loading when the heating furnace normal heat is designing effective heat duty Q ₀0.75 times when between 1.25 times, changing, the zone in 5 zones limit is respectively:

[0.75Q ₀，0.85Q ₀)，[0.85Q ₀，0.95Q ₀)，[0.95Q ₀，1.05Q ₀)，

[1.05Q ₀，1.15Q ₀)，[1.15Q ₀，1.25Q ₀]

Use Ω _IL, Ω _IHRepresent regional Ω respectively _iThe lower limit and the upper limit;

Step B1.5: according to positive balance method calculating thermal efficiency:

Wherein: η (k) is a k calculating thermal efficiency value constantly,

F _f(k) be k fuel flow rate constantly,

H _fBe the fuel combustion calorific value;

Step B1.6:, seek thermic load working region Ω according to the heating furnace field data that obtains among the step B1.2 _iThe historical optimization work point of internally heated oven makes object function J _iMaximum:

i＝1，2，…，5

Wherein: η _i(k) be thermic load working region Ω _iInterior k is the calculating thermal efficiency value constantly,

O _2i(k) be the corresponding constantly measurement of oxygen content value of k,

P _i(k) be the corresponding constantly combustion chamber draft measured value of k;

Making maximum oxygen content of the thermal efficiency and combustion chamber draft value as said thermic load working region Ω _iThe historical optimization work point of internally heated oven is used Ω _{I, opt}Expression:

Ω _i，opt＝(O _2i，opt，P _i，opt) i＝1，2，…，5

Wherein: Ω _{I, opt}Be thermic load working region Ω _iHistorical optimization work point,

O _{2i, opt}Be thermic load working region Ω _iThe historical optimal value of interior oxygen content,

P _{I, opt}Be thermic load working region Ω _iThe historical optimal value of internal furnace negative pressure;

Step B2: thermal efficiency on-line optimization, its step is following:

Step B2.1: the real time data of gatherer process comprises: heating furnace oxygen content, heating furnace burner hearth negative pressure, furnace outlet temperature, heating furnace inlet temperature, furnace charge flow and fuel flow rate;

Step B2.2: utilize the thermic load territorial classification device deterministic process present located thermic load working region Ω among the step B1.4 _i, and obtain the historical optimization work point Ω of corresponding calculated off line gained _{I, opt}

Step B2.3: the said method of B1.5 is calculated current time thermal efficiency η (k) set by step;

Step B2.4: whether the decision criteria deterministic process by following is in stable state:

Wherein: y _Jl, j=1,2,3 are respectively l value of characteristic variable furnace outlet temperature, fire box temperature and inlet amount,

L is whether deterministic process is in stable historical data length, and L*T=30min, T are the sampling periods,

j=1; 2; 3 are respectively the mean value of characteristic variable furnace outlet temperature, fire box temperature and inlet amount

ε is preassigned stable state decision threshold, and span is (0,0.1);

Step B2.5: confirm optimization work point under the current thermic load according to the following steps:

Step B2.5.1: if process is in stable state, then the online optimizing of the thermal efficiency is adopted from seeking method for optimally controlling, serves as to transfer excellent variable with the oxygen content setting value, with the historical optimal value O of oxygen content _{2i, opt}Be initial value, online searching makes the highest oxygen content setting value O of the thermal efficiency _2s(k), step is following:

Step B2.5.1.1: the changing value Δ η (k) that calculates the thermal efficiency of going up a moment k-1 relatively:

Δη(k)＝η(k)-η(k-1)；

Step B2.5.1.2: if | Δ η (k) |＜Δ η _Min, then stop optimizing and record oxygen content setting value at this moment, wherein Δ η _MinBe preset thermal efficiency adjustment dead band;

Step B2.5.1.3: if | Δ η (k) |>=Δ η _Min, then with Δ O _2s(k) be said oxygen content setting value O _2s(k) optimizing step-length automatic optimal:

O _2s(k)＝O _2s(k-1)+ΔO _2s(k)

Wherein: λ for the adjustment coefficient, span be (0,1];

Step B2.5.2: if process is not in stable state, with the historical optimal value O of oxygen content _{2i, opt}Be oxygen content setting value O _2s(k), guarantee that heating furnace is in the duty of suboptimum;

Step B2.6: the oxygen content setting value O that calculates _2s(k) be sent to said real-time data base, supply the Control for Oxygen Content device in the said air and gas system control module to implement;

Step C: the control object of said air and gas system control module comprises the air-fuel ratio when oxygen content, combustion chamber draft and fuel increase; The control target is to be controlled at said oxygen content and combustion chamber draft in the optimization set point or given range that is drawn jointly by said offline optimization and on-line optimization, and step is following:

Step C1: step test obtains the steady-state response time of oxygen content and combustion chamber draft:

Under the steady situation of process, intake is applied the step test signal, record oxygen content change curve, the steady-state response time T of acquisition oxygen content _O2ss

Under the steady situation of process, the air-introduced machine inlet baffle is applied the step test signal, record combustion chamber draft change curve, the steady-state response time T of acquisition combustion chamber draft _Pss

Step C2: the control cycle of setting the air and gas system control module is T _c, T _c=min (T _O2ss, T _Pss)/40;

Step C3: when fuel increases air-fuel ratio is controlled according to the following steps:

Step C3.1: if fuel recruitment Δ F _f(k _c) last relatively intake adjustment moment k _c-1 fuel quantity F _F0Ratio surpasses predetermined threshold value β, and β ∈ (0,0.2], promptly

The feedforward variation delta F of air intake when then fuel increases _AF(k _c) do

ΔF _aF(k _c)＝α·AFR·ΔF _f(k _c)

Wherein: Δ F _f(k _c)=F _f(k _c)-F _F0Be fuel change amount, F _f(k _c) be fuel flow rate,

α is an excess air coefficient, is [1.05,1.15] to the fuel gas span;

AFR is a stoichiometric air/, and to fuel gas, it is calculated as:

Wherein: CO, H ₂, C _mH _n, H ₂S, O ₂Being each constituent content in the fuel gas, is unit with %;

Step C4: according to the following steps oxygen content is carried out Region control:

Step C4.1: if the oxygen content setting value is O _2s(k), the regional extent of its permission is:

[O _2s(k)-δ，O _2s(k)+δ]

Wherein: δ is that oxygen content departs from oxygen content setting value O _2s(k) zone limit, δ ∈ (0,0.5);

Step C4.2: at Control for Oxygen Content moment k _c,

If O ₂(k _c)＞O _2HPerhaps O ₂(k _c)＜O _2L, O wherein _2H=O _2s(k)+and δ, O _2L=O _2s(k)-δ, and said Control for Oxygen Content device is not in the stand-by period, then is calculated as follows required intake and changes:

Wherein: Δ F _a(k _c) be the variable quantity of intake,

O _2s(k) be the oxygen content setting value,

O ₂(k _c) be the measured value of oxygen content;

If O _2L≤O ₂(k _c)≤O _2H, Δ F then _a(k _c)=0;

According to Control for Oxygen Content moment k _cCombustion chamber draft P (k _c) by the variation delta F of following principle correction intake _a(k _c):

If P (k _c)＞P _HS, P _HSBe the upper safety limit of combustion chamber draft, and the variation delta F of intake _a(k _c)＞0 then keeps intake constant, makes Δ F _a(k _c)=0,

If P (k _c)＜P _LS, P _LSBe the lower safety limit of combustion chamber draft, and the variation delta F of intake _a(k _c)＜0 then keeps intake constant, makes Δ F _a(k _c)=0;

Step C4.3: be calculated as follows required intake:

F _a(k _c)＝F _a(k _c-1)+ΔF _a(k _c)

Wherein: F _a(k _c) be intake;

After process made feedback regulation, wait for the steady-state response time T of an oxygen content _O2ss

Step C5: according to the following steps combustion chamber draft is carried out Region control:

Step C5.1: if the setting value of combustion chamber draft is P _s, and be in thermic load working region Ω _i, P then _s=P _{I, opt}, i=1,2 ..., 5, it allows the regional extent of change to be:

[P _s-σ，P _s+σ]

Wherein: σ is the zone limit that combustion chamber draft departs from setting value, σ ∈ (0,10);

Step C5.2: at combustion chamber draft control moment k _c,

If P (k _c)＞P _HPerhaps P (k _c)＜P _L, P wherein _H=P _s+ σ, P _L=P _s-σ, and the combustion chamber draft controller is not in the stand-by period, and then the deviation of combustion chamber draft is e _P(k _c)=P _s-P (k _c), air-introduced machine inlet baffle variation delta MV then ₂(k _c) do

Wherein: K ₁Proportionality coefficient for air-introduced machine inlet baffle and combustion chamber draft variation;

After process made feedback regulation, wait for the steady-state response time T of a combustion chamber draft _Pss

Under other situation, Δ MV ₂(k _c)=0;

Step C5.3: FEEDFORWARD CONTROL is carried out in the variation of blower variable frequency by following formula:

ΔMV _2F(k _c)＝K ₂ΔMV ₁(k _c)

Wherein: Δ MV _2F(k _c) be the feedforward variation of air-introduced machine inlet baffle,

Δ MV ₁(k _c)=MV ₁(k _c)-MV ₁(k _cThe variation of-1) exporting for blower variable frequency,

K ₂Be the feed-forward coefficients between baffle plate variation and the frequency conversion output.

Images