CN103160626B - Method for determining cold blast furnace hearth - Google Patents

Abstract

The invention discloses a method for determining cold blast furnace hearth. The method is as below: respectively establishing an RBF neural network calculation model of Si content in molten iron of blast furnace, a blast furnace material speed calculation model and a theoretical combustion temperature calculation model; and using cold determinations, Index1, Index2, Index3, for blast furnace hearth by the three models to comprehensively determine whether the blast furnace hearth is too cold, if delta 1.*Index1 + delta 2*Index2 + delta 3*Index3 is no less than 1, wherein delta 1 belongs to 0-1, delta 2 belongs to 0-1, and delta 3 belongs to 0-1, the blast furnace hearth is too cold. The invention comprehensively considers physical factors and chemical factors of blast furnace to determine whether the blast furnace hearth is too cold; compared with the prior art, the method can determine cold blast furnace hearth earlier and more accurately, so as to take measures to control the temperature of blast furnace in a reasonable range, and avoid deviation or error on operation condition of the blast furnace different operators; therefore, the method is very beneficial to the efficient operation of blast furnace.

Description

A kind of method that judges that blast furnace hearth is excessively cool

Technical field

The present invention relates to blast furnace ironmaking monitoring field, especially a kind of method that judges that blast furnace hearth is excessively cool.

Background technology

Blast furnace ironmaking is the raw ferriferous major way of modern steel enterprise, its production process is accompanied by the physical and chemical reaction of a large amount of tight couplings, non-linear, large dead time, and the feature of this complex process of high furnace interior has caused the conventional test set of very difficult use and detection means directly to obtain the physics and chemistry state of high furnace interior.And the hot state of high furnace interior is the most directly embodying of blast furnace stable smooth operation, blast furnace is excessively cool to be unfavorable for tapping a blast furnace smoothly separated with slag iron, overheatedly not only causes blast furnace to be difficult to walk but also causes a large amount of energy dissipations.Therefore; the hot state that obtains ahead of time blast furnace hearth can be earlier to blast furnace Intervention; can guarantee to reduce energy consumption under the prerequisite of smooth operation of furnace; reduce coke ratio; reduce the discharge of obnoxious flavour, for iron and steel enterprise, realize green low-carbon production and adapt to national environmental protection policy important in inhibiting.In view of heat state of blast furnace is the important indicator of reflection blast furnace working order, judge that as early as possible blast furnace hearth crosses cool extremely importantly for blast furnace operating personnel, but high furnace interior complicated and changeable determined blast furnace hearth, cross cool situation and can not directly by blast furnace operating personnel, be obtained.Conventional judges that it is all indirectly to judge that by means of the silicone content in blast-melted whether blast furnace hearth is excessively cool that blast furnace hearth is crossed the means of cool situation, but this method need to solve the various process variables of blast furnace and blast-melted in silicone content between mathematical model, and mathematical model is series model or the model such as end user's artificial neural networks judges that blast furnace hearth all exists the undesirable situation of judged result duration of service.

Summary of the invention

The object of this invention is to provide a kind of method that judges that blast furnace hearth is excessively cool, being intended to treat with a certain discrimination affects the influence factor that blast furnace hearth is excessively cool, reality from blast furnace iron-making process, consider blast furnace material distribution, blast furnace blast, historical molten iron silicon content to the excessively cool impact of blast furnace hearth, in conjunction with RBF neural network and blast furnace expertise advantage separately, form comprehensive mathematical model efficiently, thereby judge as early as possible the situation that blast furnace hearth is excessively cool, result is implemented to control to blast furnace accordingly.

Realizing object of the present invention and design altogether three technical schemes, is respectively that judgement scheme, high furnace charge speed computation model calculation result and the blast furnace hearth that Si content in blast-melted and blast furnace hearth are excessively cool crossed the judgement scheme that cool judgement scheme, theoretical combustion temperature computation model calculation result and blast furnace hearth are excessively cool.

1, the judgement scheme that the Si content in blast-melted and blast furnace hearth are excessively cool:

This scheme comprises the following steps:

1) choose the input variable of RBF calculating model of neural networks.Because the parameter relating in blast furnace actually operating is numerous, therefore for the structure of simplifying model, improve arithmetic speed and the generalization ability of model, Si content-Si (i-1) when definite input variable comprises cold flow, hot blast temperature, blast, pressure reduction, ventilation property, top pressure, top temperature, hour coal powder blowing amount, front tapping a blast furnace for twice after deleting in molten iron, Si (i).

2) input variable is implemented to be normalized.Because the order of magnitude between input variable differs the large and requirement of RBF neural network to input data, determine and adopt Mean Method to process input variable:

$\stackrel{\‾}{X}=\frac{X-X_{\min }}{X_{\max }-X_{\min }}$

Wherein: for the data after normalization method, x is input variable, X _minfor the minimum value of input variable, X _maxmaximum value for input variable.

3) determine RBF neural network structure.The determining of RBF neural network structure mainly comprises the number of input layer, the choosing of the number of hidden layer neuron, the neuronic number of output layer, RBF center, RBF center width.The number of input layer is actual is exactly the number of network input variable, and the neuronic number of output layer can be chosen according to the prediction ability of network and actual demand.RBF center width r can first elect 1 as, and value can suitably relax or reduce, and its impact of prediction ability on network reality is not very large, and its impact can also compensate by the optimization at network weight HeRBF center.The center of RBF neural network can learn to obtain by nearest neighbor classifier method.

The detailed process of nearest neighbor classifier method is as follows:

First select a suitable Gaussian function width r, define a vector A (l) for depositing the output vector sum that belongs to all kinds of, define a counter B (l) for adding up the number of samples that belongs to all kinds of, wherein l is classification number.From first data to (X ₁, Y ₁) start, on X1, set up a cluster centre, make C ₁=X ₁, A (1)=Y ₁, B (1)=1.The RBF network of setting up like this only has an implicit unit, and the center of this implicit unit is C ₁, this implicit unit is W to the weight vector of output layer ₁=A (1)/B (1).Consider that second sampled data is to (X ₂, Y ₂), obtain X ₂to C ₁the distance of this cluster centre || X ₂-C ₁||.If || X ₂-C ₂||≤r, C ₁for X ₂the most contiguous cluster.Make A (1)=Y ₁+ Y ₂, B (1)=2, W ₁=A (1)/B (1); If || X ₂-C ₂|| > r, by X ₂as a new cluster centre.Make C ₂=X ₂, A (2)=Y ₂, B (2)=1 is adding a hidden unit in the RBF of above-mentioned foundation neural network, and this hidden unit is W to the weight vector of output layer ₂=A (2)/B (2).Suppose that we consider that k sampled data is to (X _k, Yk) (k=3,4 ..., in the time of N), there is H cluster centre, its central point is respectively C ₁, C ₂..., C _h, existing H hidden unit in the RBF of above-mentioned foundation network.Obtain respectively again X _kdistance to this H cluster centre || X _k-C _i||, i=1,2 ..., H, establishes || X _k-C _j|| be the minor increment in these distances, i.e. C _jfor X _knearest neighbor classifier: if || X _k-C _j|| > r, by X _kas a new cluster centre.C (H+1)=X _k, A (H+1)=Y _k, B (H+1)=1, and keep A (i), the value of B (i) is constant, i=1, and 2 ..., H adds H+1 hidden unit in the RBF of above-mentioned foundation network again, and this hidden unit is W to the weight vector of output layer _h+1=A (H+1)/B (H+1).If || X _k-C _j|| < r, order: A (j)=A (j)+Y _k, B (j)=B (j)+1.When i ≠ j, i=1,2 ..., H, keeps A (i), and the value of B (i) is constant.Hidden unit is W to the weight vector of output layer _i=A (i)/B (i), i=1,2 ..., H.

The RBF neural network of setting up is like this output as:

$f\left(X_k\right)=\frac{{\mathrm{\&Sigma;}}_i^H{W}_i\exp (-\frac{{\left|\right|X_k-C_i\left|\right|}^2}{r^2})}{{\mathrm{\&Sigma;}}_i^H\exp (-\frac{{\left|\right|X_k-C_i\left|\right|}^2}{r^2})}$

Wherein, f (X _k) silicone content in molten iron while being required-next blast furnace casting.

4) collect the input variable after the normalization method of unit in the sampling period, be input in RBF neural network, forecast obtains the silicone content Si (i+1) that next time taps a blast furnace blast-melted.The blast-melted silicone content that once taps a blast furnace before Si (i-1) representative, Si (i) represents this blast-melted silicone content that taps a blast furnace:

The 1:Si if satisfied condition (i-1)≤a1, Si (i) <a2, Si (i+1) <a3

: Index1=1

If do not satisfy condition 1, and the 2:Si that satisfies condition (i)≤a4, Si (i+1)≤a5

: Index1=0.5

If do not satisfy condition 1, do not satisfy condition 2 again

: Index1=0

Wherein, the threshold value of Si while once tapping a blast furnace before a1 representative, the threshold value 1 of Si when a2 represents current tapping a blast furnace, the threshold value 1 of Si when a3 representative is tapped a blast furnace next time, the threshold value 2 of Si when a4 represents current tapping a blast furnace, the threshold value 2 of Si when a5 representative is tapped a blast furnace next time, a1<a2<a3;

The result of index Index1 is directly connected to final to the excessively cool judgement of blast furnace hearth.

2, high furnace charge speed computation model calculation result and blast furnace hearth are crossed cool judgement scheme

The input variable of high furnace charge speed computation model comprises the degree of depth that the stock rod of blast furnace declines, and output is this model to the excessively cool judgement of blast furnace hearth.According to the principle of blast furnace blanking and blast furnace hearth temperature variation relation, the algorithm of model is as follows: from identification stock rod shape, stock rod degree of depth time series data, extract eigenwert, do discharging chi translational speed, acceleration judges the abnormal working of a furnace and evaluates blanking state.Eigenwert is for representing the data of stock rod shape facility, identification different time sequence data.First calculate the data such as momentary velocity, top speed, minimum velocity, speed variation, transient acceleration, peak acceleration, minimum acceleration and acceleration bias, data all derive from take the stock rod depth data that Δ T collects as the cycle, then, data value and lowering speed are pressed to the different stock rod degree of depth-time series array datas.

$v_i\left(j\right)=\frac{l_i\left(j\right)}{T}$

$a_i\left(j\right)=\frac{\begin{array}{cc}{v}_i(j+1)& v_i\left(j\right)\end{array}}{T}$

v _max＝max(v _i(j))

v _min＝min(v _i(j))

$v_{\mathrm{aver}}=\frac{\begin{array}{cc}{v}_{\max }& v_{\min }\end{array}}{T}$

$v_{\mathrm{std}}=\frac{\sqrt{(v_i\left(j\right)}{v_{\mathrm{aver}})}^2}{\begin{array}{cc}n& 1\end{array}}$

a _max＝max(a _i(j))

a _min＝min(a _i(j))

$a_{\mathrm{aver}}=\frac{\begin{array}{cc}{a}_{\max }& a_{\min }\end{array}}{T}$

$a_{\mathrm{std}}=\frac{\sqrt{(a_i\left(j\right)}{a_{\mathrm{aver}})}^2}{\begin{array}{cc}n& 1\end{array}}$

In formula: i=1,2,3; J=1,2,3 ... n.V, v _max, v _min, v _aver, v _stdrepresent respectively blanking velocity, speed maximum value, speed minimum value, speed average, velocity standard deviation; A, a _max, a _min, a _aver, a _stdrepresent respectively blanking acceleration, acceleration maximum value, acceleration minimum value, accelerate mean value, acceleration standard deviation, wherein speed variation and acceleration bias are got the value of one-period.

1:a (j) > (the b1 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.3

2:a (j) > (the b2 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.5

3:a (j) > (the b3 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.7

If above three conditions do not meet,

: Index2=0

Wherein: a (j) represents that stock rod is at j blanking acceleration constantly, and b1, b2, b3 represent respectively threshold value 1, threshold value 2, the threshold value 3 of blanking acceleration, a _stdrepresent acceleration bias data, j=1,2,3 ... n.

The result of index Index2 is directly connected to final to the excessively cool judgement of blast furnace hearth.

3, the judgement scheme that theoretical combustion temperature computation model calculation result and blast furnace hearth are excessively cool

Carbon element (comprising other combustiblesubstance) is incomplete combustion in the hot blast that is containing hygroscopic water in air port zonal combustion, products of combustion is CO, H2 and N2, theoretical tuyere combustion temperature is that the heat that measurement burning heat release and air blast are brought into is all passed to the level that coal gas can reach, and the hot state of theoretical combustion temperature and blast furnace is closely related.Theoretical combustion temperature is the transient function under the various operating parameters effects of blast furnace, lifetime hysteresis quality not, it can be used as a kind of reference of grasping at any time heat state of blast furnace, but be subject to the impact of various high furnace parameters comparatively frequent just because of its variation, therefore the index of heat state of blast furnace is weighed in the conduct that theoretical combustion temperature can not be single, needs to coordinate other means jointly to obtain the hot state of blast furnace.According to blast furnace theoretical combustion temperature calculation formula:

$T_L=\frac{Q_C+Q_F+Q_R-Q_X}{V_g\&CenterDot;c_p^t}$

Wherein: Q _cfor burning before air port, carbon element produces CO liberated heat, kJ/t;

Q _ffor the physical thermal that air blast and coal powder injection carrier gas are brought into, kJ/t;

Q _rthe sensible heat of bringing into while entering zone of combustion for coke, kJ/t;

Q _xfor minute heat of desorption of moisture decomposition and fuel injection in air blast, kJ/t;

V _g, for burning production coal gas volume and at T _lspecific heat capacity during temperature, m ³/ t and kJ/ (m ^{3 ℃}).

If satisfied condition: c1<T _lc2

: Index3=0.3

If satisfied condition: T _lc1

: Index3=0.5

Other situation: Index3=0

Wherein, c1 and c2 be threshold value 1 and the threshold value 2 of representation theory temperature of combustion respectively, and the result of index Index3 is directly connected to final to the excessively cool judgement of blast furnace hearth.

4, comprehensively judge that blast furnace hearth is excessively cool

By three computation models such as RBF calculating model of neural networks, high furnace charge speed computation model, theoretical combustion temperature computation model, can be obtained respectively separately the excessively cool judgement of blast furnace hearth, be respectively Index1, Index2, Index3.Therefore can be defined as follows rule:

If: δ 1Index1+ δ 2Index2+ δ 3Index3 >=1, wherein δ 1 ∈ [0,1], δ 2 ∈ [0,1], δ 3 ∈ [0,1]

: blast furnace hearth is excessively cool

Enforcement of the present invention can judge that blast furnace hearth is excessively cool early, the blast furnace temperature of can taking measures to improve accordingly, control blast furnace temperature in rational scope, deviation or the erroneous judgement of having avoided different operating personnel to produce operation of blast furnace situation, be highly profitable to efficient blast furnace operating.

Accompanying drawing explanation

Fig. 1 is that blast furnace hearth is crossed cool judgment models schematic diagram;

Fig. 2 is RBF neural network training schema in the present invention.

Embodiment

Below in conjunction with specific embodiment, the excessively cool method of judgement blast furnace hearth of the present invention is further described:

The object of this invention is to provide a kind of method that judges that early blast furnace hearth is excessively cool.Realizing object of the present invention and design altogether three technical schemes, is respectively that judgement scheme, high furnace charge speed computation model calculation result and the blast furnace hearth that Si content in blast-melted and blast furnace hearth are excessively cool crossed the judgement scheme that cool judgement scheme, theoretical combustion temperature computation model calculation result and blast furnace hearth are excessively cool.According to the Computing Principle of scheme, as shown in Figure 1, need altogether three computation models, be respectively RBF calculating model of neural networks, high furnace charge speed computation model, theoretical combustion temperature computation model.The high low degree of RBF calculating model of neural networks can calculate the blast furnace next one while tapping a blast furnace blast-melted middle silicone content, thus draw the judgement that this model is too low to blast furnace hearth temperature.It is too low that high furnace charge speed computation model can show that this model is judged blast furnace hearth temperature according to the decline situation of high furnace charge chi.Theoretical combustion temperature computation model can obtain the physical thermal that burning production heat and fuel brings into according to products of combustion burning and show that this model is to the excessively cool judgement of blast furnace hearth.Because three models consider to cause the too low reason of blast furnace hearth temperature from physics heat transfer, chemistry heat transfer, historical data equal angles respectively, substantially all factors that affect blast furnace hearth furnace temperature have been included, therefore the judged result of comprehensive three models, just can judge that blast furnace hearth furnace temperature is excessively cool more accurately.

1, the judgement scheme that the Si content in blast-melted and blast furnace hearth are excessively cool:

Utilize this nonlinear approximation capability of RBF to forecast the height of the blast-melted middle silicone content of next smelting cycle.Comprise following steps:

1) choose the input variable of RBF calculating model of neural networks.Because the parameter relating in blast furnace actually operating is numerous, therefore for the structure of simplifying model, improve arithmetic speed and the generalization ability of model, Si content-Si (i-1) when definite input variable comprises cold flow, hot blast temperature, blast, pressure reduction, ventilation property, top pressure, top temperature, hour coal powder blowing amount, front tapping a blast furnace for twice after deleting in molten iron, Si (i).

2) input variable is implemented to be normalized.Determine and adopt Mean Method to process input variable:

$\stackrel{\&OverBar;}{X}=\frac{X-X_{\min }}{X_{\max }-X_{\min }}$

Wherein: for the data after normalization method, x is input variable, X _minfor the minimum value of input variable, X _maxmaximum value for input variable.Si content-Si (i-1) when cold flow, hot blast temperature, blast, pressure reduction, ventilation property, top pressure, top temperature, hour coal powder blowing amount, front tapping a blast furnace for twice in molten iron, the input such as Si (i) data are after normalized, its numerical value can all fall between [0,1].

3) determine RBF neural network structure.Because the sample size that network training needs is many, in this example, select altogether 300 stove data, wherein 200 blast furnace data are for training network structure, 100 stove data are for verifying network, sampled data is input variable and the output variable after normalization method, with vector (X, Y), represent:

(X, Y)=[cold flow hot blast temperature blast pressure reduction ventilation property is pushed up temperature hour coal powder blowing amount Si (i-1) Si (i) Si (i+1) that bears down on one]

Use the detailed process of nearest neighbor classifier method Training RBF Neural Network as shown in Figure 2:

First Gaussian function width r is 0.06, defines a vector A (l) for depositing the output vector sum that belongs to all kinds of, defines a counter B (l) for adding up the number of samples that belongs to all kinds of, and wherein l is classification number.From first data to (X ₁, Y ₁) start, on X1, set up a cluster centre, make C ₁=X ₁, A (1)=Y ₁, B (1)=1.The RBF network of setting up like this only has an implicit unit, and the center of this implicit unit is C ₁, this implicit unit is W to the weight vector of output layer ₁=A (1)/B (1).Consider that second sampled data is to (X ₂, Y ₂), obtain X ₂to C ₁the distance of this cluster centre || X ₂-C ₁||.If || X ₂-C ₁||≤r, C ₁for X ₂the most contiguous cluster.Make A (1)=Y ₁+ Y ₂, B (1)=2, W ₁=A (1)/B (1); If || X ₂-C ₂|| > r, by X ₂as a new cluster centre.Make C ₂=X ₂, A (2)=Y2, B (2)=1 is adding a hidden unit in the RBF of above-mentioned foundation neural network, and this hidden unit is W to the weight vector of output layer ₂=A (2)/B (2).Suppose that we consider that k sampled data is to (X _k, Yk) (k=3,4 ..., in the time of N), there is H cluster centre, its central point is respectively C ₁, C ₂..., C _h, existing H hidden unit in the RBF of above-mentioned foundation network.Obtain respectively again X _kdistance to this H cluster centre || X _k-C _i||, i=1,2 ..., H, establishes || X _k-C _j|| be the minor increment in these distances, i.e. C _jfor X _knearest neighbor classifier: if || X _k-C _j|| > r, by X _kas a new cluster centre.C (H+1)=X _k, A (H+1)=Y _k, B (H+1)=1, and keep A (i), the value of B (i) is constant, i=1, and 2 ..., H adds H+1 hidden unit in the RBF of above-mentioned foundation network again, and this hidden unit is W to the weight vector of output layer _h+1=A (H+1)/B (H+1).If || X _k-C _j||≤r, order: A (j)=A (j)+Y _k, B (j)=B (j)+1.When i ≠ j, i=1,2 ..., H, keeps A (i), and the value of B (i) is constant.Hidden unit is W to the weight vector of output layer _i=A (i)/B (i), i=1,2 ..., H.

The RBF neural network of setting up is like this output as:

$f\left(X_k\right)=\frac{{\mathrm{\&Sigma;}}_i^H{W}_i\exp (-\frac{{\left|\right|X_k-C_i\left|\right|}^2}{r^2})}{{\mathrm{\&Sigma;}}_i^H\exp (-\frac{{\left|\right|X_k-C_i\left|\right|}^2}{r^2})}$

Wherein, f (X _k) be required-silicone content in molten iron during blast furnace casting next time.

In the present embodiment, the neural network structure after training is that input layer number is 10, and middle layer neuron number is 14, and output layer neuron number is 1.

4) the silicone content Si (i+1) that next time taps a blast furnace blast-melted obtaining according to RBF neural network prediction.Simultaneously according to following rule, show that RBF neural network is to the excessively cool judgement of blast furnace hearth:

The 1:Si if satisfied condition (i-1)≤a1, Si (i) <a2, Si (i+1) <a3

: Index1=1

If do not satisfy condition 1, and the 2:Si that satisfies condition (i)≤a4, Si (i+1)≤a5

: Index1=0.5

If do not satisfy condition 1, do not satisfy condition 2 again

: Index1=0

In this example, a1=0.2, a2=0.25, a3=0.3, a4=0.2, a5=0.2.

2, the excessively cool judgement scheme of high furnace charge speed computation model calculation result and blast furnace hearth

According to the principle of blast furnace blanking and blast furnace hearth temperature relation, the algorithm of model is as follows: from identification stock rod shape, stock rod degree of depth time series data, extract eigenwert, do discharging chi translational speed, acceleration judges the abnormal working of a furnace and evaluates blanking state.Eigenwert is for representing the data of stock rod shape facility, identification different time sequence data.First calculate the data such as momentary velocity, top speed, minimum velocity, speed variation, transient acceleration, peak acceleration, minimum acceleration and acceleration bias, data all derive from take the stock rod depth data that Δ T collects as the cycle, then, data value and lowering speed are pressed to the different stock rod degree of depth-time series array datas.

Suppose that blast furnace has 3 stock rods, the degree of depth that definition stock rod declines is respectively l ₁, l ₂, l ₃, in unit period, the variation of 3 stock rod descending depths is respectively Δ l ₁(j), Δ l ₂(j), Δ l ₃(j), a computation period is Δ T, and blast furnace hearth temperature is too low judgment result is that Index2.Can calculate the speed V that in the unit time, stock rod declines _iand acceleration alpha (j) _i(j).

$v_i\left(j\right)=\frac{l_i\left(j\right)}{T}$

$a_i\left(j\right)=\frac{\begin{array}{cc}{v}_i(j+1)& v_i\left(j\right)\end{array}}{T}$

v _max＝max(v _i(j))

v _min＝min(v _i(j))

$v_{\mathrm{aver}}=\frac{\begin{array}{cc}{v}_{\max }& v_{\min }\end{array}}{T}$

$v_{\mathrm{std}}=\frac{\sqrt{(v_i\left(j\right)}{v_{\mathrm{aver}})}^2}{\begin{array}{cc}n& 1\end{array}}$

a _max＝max(a _i(j))

a _min＝min(a _i(j))

$a_{\mathrm{aver}}=\frac{\begin{array}{cc}{a}_{\max }& a_{\min }\end{array}}{T}$

$a_{\mathrm{std}}=\frac{\sqrt{(a_i\left(j\right)}{a_{\mathrm{aver}})}^2}{\begin{array}{cc}n& 1\end{array}}$

In formula: i=1,2,3; J=1,2,3 ... n.V, v _max, v _min, v _aver, v _stdrepresent respectively blanking velocity, speed maximum value, speed minimum value, velocity standard deviation; A, a _max, a _min, a _aver, a _stdrepresent respectively blanking acceleration, acceleration maximum value, acceleration minimum value, acceleration standard deviation.Wherein speed variation and acceleration bias are got one-period.

Impact according to blast furnace blanking velocity on heat state of blast furnace:

1:a (j) > (the b1 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.3

2:a (j) > (the b2 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.5

3:a (j) > (the b3 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.7

If above three conditions do not meet,

: Index2=0

Wherein: a (j) represents that stock rod is at j blanking acceleration constantly, and b1, b2, b3 represent respectively threshold value 1, threshold value 2, the threshold value 3 of blanking acceleration, a _stdrepresent acceleration bias data, j=1,2,3 ... n.

In this example, b1=0.01, b2=0.02, b3=0.03, both can obtain high furnace charge speed computation model to the excessively cool judgement of blast furnace hearth by above calculating.

3, theoretical combustion temperature computation model calculation result and blast furnace blast furnace hearth are crossed cool judgement scheme

According to blast furnace theoretical combustion temperature calculation formula:

$T_L=\frac{Q_C+Q_F+Q_R-Q_X}{V_g\&CenterDot;c_p^t}$

Wherein: Q _cfor burning before air port, carbon element produces CO liberated heat, kJ/t;

Q _ffor the physical thermal that air blast and coal powder injection carrier gas are brought into, kJ/t;

Q _rthe sensible heat of bringing into while entering zone of combustion for coke, kJ/t;

Q _xfor minute heat of desorption of moisture decomposition and fuel injection in air blast, kJ/t;

V _g, for burning, produce coal gas volume and the specific heat capacity when TL temperature thereof, m ³/ t and kJ/ (m ^{3 ℃}).

If satisfied condition: c1<T _lc2

: Index3=0.3

If satisfied condition: T _lc1

: Index3=0.5

Other situation: Index3=0

Wherein, c1 and c2 be threshold value 1 and the threshold value 2 of representation theory temperature of combustion respectively, in this example, and c1=1950, c2=2100, the result of index Index3 is directly connected to final to the excessively cool judgement of blast furnace hearth.

4, comprehensively judge that blast furnace hearth is excessively cool

By three computation models such as RBF calculating model of neural networks, high furnace charge speed computation model, theoretical combustion temperature computation model, can be obtained respectively separately the excessively cool judgement of blast furnace hearth, be respectively Index1, Index2, Index3.

If: δ 1Index1+ δ 2Index2+ δ 3Index3 >=1, wherein δ 1 ∈ [0,1], δ 2 ∈ [0,1], δ 3 ∈ [0,1]

: blast furnace hearth is excessively cool

In blast furnace actual production, the burden structure of blast furnace, operating duty etc. all can change, so the self-teaching capability for correcting of model and the judgement no less important of model.So degree of agreement of the meeting periodical survey judged result of the computation model in the present invention and actual result.If computation model is after having moved for some time, when its judgement precision can not meet blast furnace Production requirement, RBF calculating model of neural networks can be upgraded and revise network structure from new collection sample, forms new RBF calculating model of neural networks.Theoretical combustion temperature computation model also can recalculate to meet blast furnace operating mode according to the situation of blast furnace crude fuel.The model using in present method has good adaptability and robustness, can meet the requirement that now large blast furnace is produced.

Claims (4)

1. a method that judges that blast furnace hearth is excessively cool, it is characterized in that setting up respectively RBF calculating model of neural networks, high furnace charge speed computation model, the theoretical combustion temperature computation model of the Si content in blast-melted, then judgement-Index1, the Index2, the Index3 that utilize 3 models to cross cool result to blast furnace hearth comprehensively judge that whether blast furnace hearth is excessively cool, if δ 1Index1+ δ 2Index2+ δ 3Index3 >=1, δ 1 ∈ [0 wherein, 1], δ 2 ∈ [0,1], δ 3 ∈ [0,1], can judge that blast furnace hearth is excessively cool.

2. a kind of method that judges that blast furnace hearth is excessively cool according to claim 1, is characterized in that the judgement scheme that Si content in blast-melted and blast furnace hearth are excessively cool comprises the following steps:

1) choose the input variable of RBF calculating model of neural networks;

2) input variable is implemented to be normalized;

3) determine RBF neural network structure, mainly comprise the number of input layer, the determining of the number of hidden layer neuron, the neuronic number of output layer, also comprise and use nearest neighbor classifier method to determine the choosing of RBF center, RBF center width;

4) collect the input variable after the normalization method of unit in the sampling period, be input in RBF neural network, forecast obtains the silicone content Si (i+1) that next time taps a blast furnace blast-melted, the blast-melted silicone content that once taps a blast furnace before Si (i-1) representative, Si (i) represents this blast-melted silicone content that taps a blast furnace:

The 1:Si if satisfied condition (i-1)≤a1, Si (i) <a2, Si (i+1) <a3

: Index1=1

If do not satisfy condition 1, and the 2:Si that satisfies condition (i)≤a4, Si (i+1)≤a5

: Index1=0.5

If do not satisfy condition 1, do not satisfy condition 2 again

: Index1=0

Wherein, the threshold value of Si while once tapping a blast furnace before a1 representative, the threshold value 1 of Si when a2 represents current tapping a blast furnace, the threshold value 1 of Si when a3 representative is tapped a blast furnace next time, the threshold value 2 of Si when a4 represents current tapping a blast furnace, the threshold value 2 of Si when a5 representative is tapped a blast furnace next time, a1<a2<a3;

The result of index Index1 is directly connected to final to the excessively cool judgement of blast furnace hearth.

3. a kind of method that judges that blast furnace hearth is excessively cool according to claim 1, the input variable that it is characterized in that set up high furnace charge speed computation model comprises the degree of depth that stock rod declines, output is this model to the excessively cool judgement of blast furnace hearth, according to blast furnace blanking velocity, the impact of the hot cupola well state of blast furnace is carried out to index access Index2, the result of Index2 is directly connected to final to the excessively cool judgement of blast furnace hearth, the impact according to blast furnace blanking velocity on the hot state of blast furnace hearth:

1:a (j) > (the b1 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.3

2:a (j) > (the b2 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.5

3:a (j) > (the b3 a if each stock rod of blast furnace all satisfies condition _std)

: Index2=0.7

If above three conditions do not meet,

: Index2=0

Wherein: a (j) represents that stock rod is at j blanking acceleration constantly, and b1, b2, b3 represent respectively threshold value 1, threshold value 2, the threshold value 3 of blanking acceleration, a _stdrepresent acceleration bias data, j=1,2,3 ... n;

The result of index Index2 is directly connected to final to the excessively cool judgement of blast furnace hearth.

4. a kind of method that judges that blast furnace hearth is excessively cool according to claim 1, it is characterized in that set up theoretical combustion temperature computation model is that the hot state relation degree of cupola well based on theoretical combustion temperature and blast furnace designs, according to blast furnace theoretical combustion temperature calculation formula:

Wherein: T _lrepresentation theory temperature of combustion, ℃;

Q _cfor burning before air port, carbon element produces CO liberated heat, kJ/t;

Q _ffor the physical thermal that air blast and coal powder injection carrier gas are brought into, kJ/t;

Q _rthe sensible heat of bringing into while entering zone of combustion for coke, kJ/t;

Q _xfor minute heat of desorption of moisture decomposition and fuel injection in air blast, kJ/t;

V _g, for burning production coal gas volume and at T _lspecific heat capacity during temperature, m ³/ t and kJ/ (m ³℃);

If satisfied condition: c1<T _lc2

: Index3=0.3

If satisfied condition: T _lc1

: Index3=0.5

Other situation: Index3=0

Wherein, c1 and c2 be threshold value 1 and the threshold value 2 of representation theory temperature of combustion respectively, and the result of index Index3 is directly connected to final to the excessively cool judgement of blast furnace hearth.