CN100493748C - Control method for dynamically correcting strip steel thickness deviation by forward slip adaption - Google Patents

Abstract

The present invention relates to a control method for dynamically correcting thickness deviation of steel strip by utilizing forward creep adaptation. It is characterized by that said invention adopts automatically-controlled feedback principle to optimize forward creep model, and utilizes collection and treatment of actual thickness deviation to regulate objective actual forward creep value, then utilizes forward creep adaptation method to correct forward creep set value, so that the rolling speed can be indirectly regulated so as to attain the goal of improving thickness deviation of steel strip.

Description

Utilize the control method of forward slip adaptive dynamic correction strip steel thickness deviation

(1) technical field

The present invention relates to a kind of control method of utilizing forward slip adaptive dynamic correction strip steel thickness deviation.

(2) background technology

In the cold continuous rolling production process, the cold rolling parameter all is to come out by the relevant calculated with mathematical model of L2 system utilization, will calculate good parameter then and pass to the L1 control system before the band steel enters rolling-mill housing.When this band steel was rolled, if there are not abnormal conditions to occur, the L2 initialization system was just no longer intervened the operation of rolling.The thickness and precision of band steel is to revise the quality that guarantees to export the band steel automatically by AGC (the automatic gauge control) control system of L1, eliminates the deviation that band steel supplied materials or setting model are brought by closed loop control methods such as pre-control, monitoring and the controls of second flow amount.But, all there is effect in the AGC system to institute's organic frame, but according to the operating characteristic of himself certain several frame are wherein directly had an effect, as first frame, last frame and the second last frame, other frame then is to keep its poised state.Like this, the band steel does not guarantee at the thickness and precision when not having the frame of AGC control.

The method of existing belt steel thickness control accuracy has: as roll eccentricity control (patent No. US4531392A, patent No. US4580224A, patent No. JP01162509A etc.), calculate roll eccentricities or utilize fluctuations to wait to determine the roll eccentricities value by model pre-estimating each frame actual load and roll-force, thereby it is compensated on the control corresponding value according to period of change, be used for keeping the stable of each frame outlet belt steel thickness.And for example roll gap Position Control (patent No. US4125004A, patent No. US4244025A) by the on-line measurement to band steel inlet thickness and exit thickness deviation, is carried out pre-control and FEEDBACK CONTROL to the roll gap position, promotes whole thickness control accuracy.The characteristics of these methods all are to allow control strategy implement at Process Control System L1, if not just this method being taken into account at the beginning of Control System Design, implement relatively difficulty in the later stage.

Utilizing on-the-spot real data to carry out corrected Calculation at the rolling model initialization system improves the patent of thickness deviation of band steel and have: patent No. JP62130710A at first calculates setting values such as roll gap position according to the housing rigidity of estimating, revise housing rigidity by in the operation of rolling actual load and thickness data being carried out acquisition process then, recomputate rolling parameter then and set.This patent is a kind of method that improves the setting model precision, and this method need cooperate its proprietary Mathematical Modeling.Patent publication No. CN1589986 is that the technological parameter that adopts the on-line automatic measurement of rolling technological parameter that plate strip rolling process is controlled automatically is optimized, be characterized in the back calculated value of resistance of deformation K in the rolling metal plate and tape process and coefficientoffriction being carried out accurate Calculation according to actual parameter, with the rolling force F and the advancing slip value fs substitution of actual measurement is the binary high order Nonlinear System of Equations of unknown quantity with resistance of deformation K and coefficientoffriction, find the solution the back calculated value of two unknown quantitys simultaneously, and utilize them that resistance of deformation K and coefficientoffriction computation model are carried out adaptive learning optimization, and then improve the setting accuracy of rolling metal plate and tape power F model.

(3) summary of the invention

The object of the present invention is to provide a kind of control method of utilizing forward slip adaptive dynamic correction strip steel thickness deviation, this control method is under the situation of not disturbing cold rolling AGC control system duty, do not carry out the rolling-mill housing of THICKNESS CONTROL for the AGC system, eliminate the thickness deviation of this frame outlet band steel by increasing the new adaptive approach of cold rolling advancing slip model as far as possible, thereby guarantee the belt steel thickness precision in exit.

The present invention is achieved in that a kind of control method of utilizing forward slip adaptive dynamic correction strip steel thickness deviation:

At first, by calibrator and tachymeter roll linear velocity and belt steel thickness in the cold continuous rolling process are accurately measured;

And it is constant to suppose to keep the second flow amount total amount of band steel under certain rolling power, then has:

vRoll _act*fslip _act*h _act＝vRoll _set*fslip _set*h _set (1)

In the formula, vRoll is the roll linear velocity, and fslip is advancing slip, and h is a thickness, and subscript act represents actual value, and subscript set represents setting value, and Δ h represents thickness deviation;

Setting speed and actual speed are equal, i.e. vRoll _Act=vRoll _Set

Actual belt steel thickness and the pass of setting between the belt steel thickness are: h _Act=h _Set+ Δ h;

Arrangement formula (1) can get,

${\mathrm{fslip}}_{\mathrm{act}}=\frac{{\mathrm{fslip}}_{\mathrm{set}}*h_{\mathrm{set}}}{h_{\mathrm{set}}+\mathrm{\Δh}}$

(2)

$={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}}{h_{\mathrm{act}}})$

Consider thickness deviation contributive rate factor, then formula (2) becomes:

${\mathrm{fslip}}_{\mathrm{act}}={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}*\mathrm{factor}}{h_{\mathrm{act}}})---\left(3\right)$

Note, y '=fslip _Act

Consider feedback control loop simultaneously, comprise ratio amplification coefficient pPara that error is considered and the integration amplification coefficient IPara that error is considered, then formula (3) becomes:

$y=y\′+(\mathrm{pPara}+\mathrm{IPara})*(y\′-y0)$

$={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}*\mathrm{factor}}{h_{\mathrm{act}}})+(\mathrm{pPara}+\mathrm{IPara})*(y\′-y0)---\left(4\right)$

In the formula, y is actual advancing slip output, also is the advancing slip output valve that algorithm calculates;

Y0 is a last computing cycle actual advancing slip output valve y constantly;

PPara is the ratio amplification coefficient that error is considered, its value is 0～0.2;

IPara is the integration amplification coefficient that error is considered, its value is 0～0.2;

Factor is the thickness deviation contributive rate, and its value is 0.2～0.8;

Three coefficient sums are no more than 0.75 simultaneously;

Advancing slip result of calculation Y limit value is 1.00～1.07;

Through type (4) inputs to model adaptation system after calculating actual advancing slip value y, model adaptation system is exported revised model coefficient and is given preset model, preset model calculates the advancing slip value that makes new advances and carries out the speed adjusting to control system, thus correction strip steel thickness deviation.

The present invention adopts the feedback principle of control automatically that advancing slip model is optimized, and existing AGC control system is not carried out the intermediate stand of THICKNESS CONTROL adjusting and controls.On the frame that does not have the AGC control and regulation, control system is to control according to the parameter that preset model provides fully, because error, the interference of environmental factor and the influences such as characteristic of equipment executing agency of model itself, the output of Process Control System is certainly existing certain error.The present invention is by carrying out acquisition process to the actual (real) thickness deviation, adjust the actual advancing slip value of target, utilize the advancing slip setting value of forward slip adaptive method correction then, so just adjusted the speed of rolls indirectly, thereby reach the purpose of improving strip thickness deviation.The present invention is characterized in realizing that simply reaction speed is fast compared with prior art.

(4) description of drawings

Fig. 1 utilizes the control schematic diagram of the control method of forward slip adaptive dynamic correction strip steel thickness deviation for the present invention;

Fig. 2 is the 2nd frame exit thickness deviation profile figure among the embodiment;

Fig. 3 is the 3rd frame exit thickness deviation profile figure among the embodiment.

(5) specific embodiment

The invention will be further described below in conjunction with the drawings and specific embodiments.

Referring to Fig. 1, a kind of control method of utilizing forward slip adaptive dynamic correction strip steel thickness deviation:

At first, by calibrator and tachymeter roll linear velocity and belt steel thickness in the cold continuous rolling process are accurately measured;

And it is constant to suppose to keep the second flow amount total amount of band steel under certain rolling power, then has:

vRoll _act*fslip _act*h _act＝vRoll _set*fslip _set*h _set (1)

In the formula, vRoll is the roll linear velocity, and fslip is advancing slip, and h is a thickness, and subscript act represents actual value, and subscript set represents setting value, and Δ h represents thickness deviation.

Advancing slip is exactly with the value of steel to front slide.

Owing to do not have the AGC control loop, so setting speed is equal with actual speed, i.e. vRoll _Act=vRoll _Set

Actual belt steel thickness and the pass of setting between the belt steel thickness are: h _Act=h _Set+ Δ h;

Arrangement formula (1) can get,

${\mathrm{fslip}}_{\mathrm{act}}=\frac{{\mathrm{fslip}}_{\mathrm{set}}*h_{\mathrm{set}}}{h_{\mathrm{set}}+\mathrm{\Δh}}---\left(2\right)$

$={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}}{h_{\mathrm{act}}})$

Consider two factors: promptly

1, thickness deviation contributive rate factor, then formula (2) becomes:

${\mathrm{fslip}}_{\mathrm{act}}={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}*\mathrm{factor}}{h_{\mathrm{act}}})---\left(3\right)$

Note, y '=fslip _Act

2, feedback control loop is arranged, promptly comprise ratio amplification coefficient pPara that error is considered and the integration amplification coefficient IPara that error is considered, then formula (3) becomes:

$y=y\′+(\mathrm{pPara}+\mathrm{IPara})*(y\′-y0)$

$={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}*\mathrm{factor}}{h_{\mathrm{act}}})+(\mathrm{pPara}+\mathrm{IPara})*(y\′-y0)---\left(4\right)$

In the formula, y is actual advancing slip output, also is the advancing slip output valve that algorithm calculates, and its limit value is 1.00～1.07.

Y0 is a last computing cycle actual advancing slip output valve y constantly;

PPara is the ratio amplification coefficient that error is considered, its value is 0～0.2;

IPara is the integration amplification coefficient that error is considered, its value is 0～0.2;

Factor is the thickness deviation contributive rate, and its value is 0.2～0.8;

Three coefficient sums are no more than 0.75 simultaneously; These coefficient limit values have been kept the stable of result of calculation.Advancing slip result of calculation y limit value is 1.00～1.07.

Through type (4) inputs to model adaptation system after calculating actual advancing slip value y, model adaptation system is exported revised model coefficient and is given preset model, preset model calculates the advancing slip value that makes new advances and carries out the speed adjusting to control system, thus correction strip steel thickness deviation.

Can obtain from formula (4): the size of dynamically adjusting the actual advancing slip value of operation of rolling needs according to the thickness deviation of band steel, if current thickness deviation is an overgauge, the calculating of through type (4) will make actual advancing slip value descend on the basis of original setting value so.Formula (4) is adjusted the calculating of actual advancing slip value by two parts, first is calculated according to thickness deviation, second portion is to compare with current thickness deviation result of calculation according to a preceding result calculated to calculate, if this moment, deviation between the two was bigger, the result of calculation that shows advancing slip value is in the dynamic adjustment, can shorten the adjustment time by certain compensating coefficient, the thickness deviation of band steel is eliminated as early as possible.If the thickness deviation of band steel is eliminated, the result of calculation of second portion can think 0, the calculating by first as can be known: y=fslip _VetIn the calculating of next computing cycle, still can obtain y=fslip like this _VetThis moment is advancing slip no longer to change, and strip thickness deviation also no longer changes, thereby makes the forward slip adaptive system be in steady operation.

As can be seen from Figure 1, after the actual advancing slip calculating, this value passes to model adaptation system as input, advancing slip model adaptation will be revised relevant model coefficient according to calculating good advancing slip actual value, like this when model specification next time advancing slip calculated value more closing to reality need, the change of each advancing slip value all can be reacted directly into the speed of control system and regulate, and control system receives after the new advancing slip value value of regulating the speed automatically, thereby reaches the purpose of elimination thickness deviation.

Embodiment

Band steel inlet thickness 2.26mm, exit thickness 0.974mm, width 916mm.The advancing slip setting value of band steel that calculates under certain belt steel thickness specification is distributed is: the 2nd frame the 1.0136, the 3rd frame is 1.0216, rolling down in the speed of 600m/min, the actual exit thickness of the 2nd frame is 1.75mm, thickness deviation is that the actual exit thickness of 0.102mm, the 3rd frame is 1.386mm, and thickness deviation is 87um.

Embodiment one: get coefficient factor=0.4, pPara=0.1, IPara=0.0, for the first time the y0=fslip that calculates _Set, as calculated, the advancing slip real output value of the 2nd frame is:

The first step: the y1 when calculating factor=0.4,

$y1=((1.0136-1)*(1-\frac{0.102}{1.75}*0.4)+1)=1.013283$

Second step: the y2 when calculating pPara=0.1,

$y2=0.1*(((1.0136-1)*(1-\frac{0.102}{1.75}*0.4)+1)-1.0136)=-0.000032$

The 3rd step: the y3 when calculating IPara=0.0,

y3＝0

The 4th step: calculate the output of actual advancing slip y,

y＝y1+y2+y3＝1.013283-0.000032+0＝1.01325

The advancing slip real output value of the 3rd frame is:

The first step: the y1 when calculating factor=0.4,

$y1=((1.0216-1)*(1-\frac{0.087}{1.386}*0.4)+1)=1.021058$

Second step: the y2 when calculating pPara=0.1,

$y2=0.1*(((1.0216-1)*(1-\frac{0.087}{1.386}*0.4)+1)-1.0216)=-0.00005$

The 3rd step: the y3 when calculating IPara=0.0,

y3＝0

The 4th step: calculate the output of actual advancing slip y,

y＝y1+y2+y3＝1.021058-0.00005+0＝1.02101

Embodiment two: get coefficient factor=0.2, pPara=0.05, IPara=0.05, for the first time the y0=fslip that calculates _Set, as calculated, the advancing slip real output value of the 2nd frame is:

The first step: the y1 when calculating factor=0.2,

$y1=((1.0136-1)*(1-\frac{0.102}{1.75}*0.2)+1)=1.013441$

Second step: the y2 when calculating pPara=0.05,

$y2=0.05*(((1.0136-1)*(1-\frac{0.102}{1.75}*0.2)+1)-1.0136)=-0.000008$

The 3rd step: the y3 when calculating IPara=0.05,

$y3=0.05*(((1.0136-1)*(1-\frac{0.102}{1.75}*0.2)+1)-1.0136)=-0.000008$

The 4th step: calculate the output of actual advancing slip y,

y＝y1+y2+y3＝1.013441-0.000008-0.000008＝1.013425

The advancing slip real output value of the 3rd frame is:

The first step: the y1 when calculating factor=0.2,

$y1=((1.0216-1)*(1-\frac{0.087}{1.386}*0.2)+1)=1.021329$

Second step: the y2 when calculating pPara=0.05,

$y2=0.05*(((1.0216-1)*(1-\frac{0.087}{1.386}*0.2)+1)-1.0216)=-0.000014$

The 3rd step: the y3 when calculating IPara=0.05,

$y2=0.05*(((1.0216-1)*(1-\frac{0.087}{1.386}*0.2)+1)-1.0216)=-0.000014$

The 4th step: calculate the output of actual advancing slip y,

y＝y1+y2+y3＝1.021329-0.000014-0.000014＝1.021301

Embodiment three: get coefficient factor=0.6, pPara=0.05, IPara=0.1, for the first time the y0=fslip that calculates _Set, as calculated, the advancing slip real output value of the 2nd frame is:

The first step: the y1 when calculating factor=0.6,

$y1=((1.0136-1)*(1-\frac{0.102}{1.75}*0.6)+1)=1.013124$

Second step: the y2 when calculating pPara=0.05,

$y2=0.05*(((1.0136-1)*(1-\frac{0.102}{1.75}*0.6)+1)-1.0136)=-0.000024$

The 3rd step: the y3 when calculating IPara=0.1,

$y3=0.1*(((1.0136-1)*(1-\frac{0.102}{1.75}*0.6)+1)-1.0136)=-0.000048$

The 4th step: calculate the output of actual advancing slip y,

The advancing slip real output value of y=y1+y2+y3=1.013124-0.000024-0.000048=1.01305 the 3rd frame is:

The first step: the y1 when calculating factor=0.6,

$y1=((1.0216-1)*(1-\frac{0.087}{1.386}*0.6)+1)=1.02079$

Second step: the y2 when calculating pPara=0.05,

$y2=0.05*(((1.0216-1)*(1-\frac{0.087}{1.386}*0.6)+1)-1.0216)=-0.000041$

The 3rd step: the y3 when calculating IPara=0.1,

$y2=0.1*(((1.0216-1)*(1-\frac{0.087}{1.386}*0.6)+1)-1.0216)=-0.000082$

The 4th step: calculate the output of actual advancing slip y,

y＝y1+y2+y3＝1.02079-0.000041-0.000082＝1.02067

Through type (4) inputs to model adaptation system after calculating actual advancing slip value y, model adaptation system is exported revised model coefficient and is given preset model, preset model calculates the advancing slip value that makes new advances and carries out the speed adjusting to control system, thus correction strip steel thickness deviation.

Referring to Fig. 2, Fig. 3, be the 2nd, the 3rd frame exit thickness deviation profile figure among the embodiment.As can be seen, the band steel is reducing gradually by the exit thickness deviation after these two frames that do not have AGC control from the figure, and therefore, the thickness and precision after this two frame is by the direct acting result of the revised setting value of forward slip adaptive.

The present invention is characterized in realizing that simply reaction speed is fast compared with prior art.

Claims (1)

1. control method of utilizing forward slip adaptive dynamic correction strip steel thickness deviation is characterized in that:

At first, by calibrator and tachymeter roll linear velocity and belt steel thickness in the cold continuous rolling process are accurately measured;

And it is constant to suppose to keep the second flow amount total amount of band steel under certain rolling power, then has:

vRoll _act*fslip _act*h _act＝vRoll _set*fslip _set*h _set (1)

In the formula, vRoll is the roll linear velocity, and fslip is advancing slip, and h is a thickness, and subscript act represents actual value, and subscript set represents setting value, and Δ h represents thickness deviation;

Setting speed and actual speed are equal, i.e. vRoll _AcT=vRoll _Set

Actual belt steel thickness and the pass of setting between the belt steel thickness are: h _Act=h _Set+ Δ h;

Arrangement formula (1) can get,

${\mathrm{fslip}}_{\mathrm{act}}=\frac{{\mathrm{fslip}}_{\mathrm{set}}*h_{\mathrm{set}}}{h_{\mathrm{set}}+\mathrm{\Δh}}$

(2)

$={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}}{h_{\mathrm{act}}})$

Consider thickness deviation contributive rate factor, then formula (2) becomes:

${\mathrm{fslip}}_{\mathrm{act}}={\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}*\mathrm{factor}}{h_{\mathrm{act}}})---\left(3\right)$

Note, y '=fslip _Act

Consider feedback control loop simultaneously, comprise ratio amplification coefficient pPara that error is considered and the integration amplification coefficient IPara that error is considered, then formula (3) becomes:

$y=y\′+(\mathrm{pPara}+\mathrm{IPara})*(y\′-y0)$

(4)

${=\mathrm{fslip}}_{\mathrm{set}}*(1-\frac{\mathrm{\Δh}*\mathrm{factor}}{h_{\mathrm{act}}})+(\mathrm{pPara}+\mathrm{IPara})*(y\′-y0)$

In the formula, y is actual advancing slip output, also is the advancing slip output valve that algorithm calculates;

Y0 is a last computing cycle actual advancing slip output valve y constantly;

PPara is the ratio amplification coefficient that error is considered, its value is 0～0.2;

IPara is the integration amplification coefficient that error is considered, its value is 0～0.2;

Factor is the thickness deviation contributive rate, and its value is 0.2～0.8;

Three coefficient sums are no more than 0.75 simultaneously;

Advancing slip result of calculation y limit value is 1.00～1.07;

Through type (4) inputs to model adaptation system after calculating actual advancing slip value y, model adaptation system is exported revised model coefficient and is given preset model, preset model calculates the advancing slip value that makes new advances and carries out the speed adjusting to control system, thus correction strip steel thickness deviation.

Images