JPH082456B2 - Dimension control method for bar and wire rod rolling - Google Patents

Description

TECHNICAL FIELD The present invention relates to a dimensional control method in rolling a steel bar and a wire rod.

[Prior Art] Known examples of conventional dimension control methods for bar steel and wire rod rolling are as follows.

In the technique of Japanese Patent Application Laid-Open No. 62-173018 filed by the inventors of the present invention, a rolling equation of state consisting of deformation, load, and temperature is used to set the roll rotational speed and roll gap as manipulated variables, and the target rolling is performed. The gauge height of each stand ridge side gauge considering the tension between each stand and monitor compensation of the machine group is controlled by feedback of control deviation and state vector based on the optimal regulator theory.

Further, in the technique of Japanese Patent Laid-Open No. 62-244509 filed by the present inventors, the roll rotational speed and roll gap are used as manipulated variables by using the rolling equation of state consisting of deformation, load and temperature. The height of the gauge between each stand in the rolling mill group and the calculated height of the gauge on the mountain side of each stand in consideration of monitor compensation are controlled by feedback of control deviation and state vector based on the optimal regulator theory.

In any of the above-mentioned conventional techniques, the number of rolling stands is one system, and when the number of rolling stands increases, the precision decreases due to the linear rolling model, the precision of the control system decreases, and the dimension of the matrix increases, As the online calculation time becomes longer, the control cycle becomes longer and the control performance deteriorates.

[Problems to be Solved by the Invention] In the present invention, in rolling steel bar and wire rod, when the number of rolling stands is large, in one system control, there is a problem that the accuracy of control system design decreases and the online calculation time becomes long. The task is to solve.

[Means for Solving the Problems] The present invention divides all stands into subsystems for every several stands, and uses a rolling state equation consisting of deformation, load, and temperature for each subsystem to perform roll rotation. The control deviation and state are based on the optimal regulator theory based on the optimal regulator theory based on the optimal regulator theory. Optimal input of vector feedback controls the overall stand-to-stand tension or loop height in the rolling mill train within the subsystem and the height and width dimensions of the rolled stock discharged from the most downstream stand in the subsystem, thus It is a method of controlling the cross-sectional dimension of a rolled material in bar steel and wire rod rolling that implements multivariable control over a stand.

That is, the present inventors have disclosed in Japanese Patent Laid-Open Nos. Sho 62-173018 and Sho 62-244509, each stand stand-out based on the optimum regulator theory or the stand height in consideration of loop height and monitor compensation. In the method of controlling the side gauge meter calculation height, when the number of stands is several or more, there is a problem that the control performance deteriorates in the control system of one system. The above problem is solved by performing multi-variable control in the subsystem using the tension or loop height in the subsystem and the adjacent tension or loop height in the adjacent subsystem for each subsystem. It is a solution.

[Operation] In the present invention, in the case of continuous rolling of several or more stands, it is divided into subsystems every several stands, and a multivariable control system is configured for each subsystem. Since it is controlled by the subsystem of the upstream stand, the control system design is highly accurate, the online calculation time is short, and the control system with good control performance can be designed relatively easily.

[Examples] Examples of the present invention will be described in detail below. here,
A bar steel mill coarse row loopless 6 stand will be described as an example when it is carried out in one system and when it is divided into two subsystems of 2 stands upstream and 4 stands downstream.

First, the control system design of the tension control in which the operation amount is only the roll rotation number when the system is divided into two subsystems will be described. A list of symbols is shown in Table 1 below.

In this example, as the mill drive system dynamic characteristics, the roll gap control system is approximated by a primary delay system, and the roll speed control system is approximated by a secondary system.

Further, for each subsystem, the nonlinear rolling model is linearized and dimensionlessized around the reference state shown in Table 2 and displayed as a state space model, and further discretized with the control period Δt, the following state equation is obtained. obtain. Here, Δx is defined as a variation from the path schedule value x ₀ and is defined as Δx ^* ≡Δx / x ₀ (where ΔS ^* ≡Δs / h, Δtf ^* ≡Δtf / kfm). Also, Δ ^* is the deviation from the lock-on value ΔxL ^* in the gauge meter calculation.

That is, for each subsystem, the state belt x is expressed as xk = (Δik ^* , Δtfjk ^* , ΔNik ^* , Δik ^* ) ^T.
(1) (The superscript T means transposed matrix) yk = (Δik ^* , Δtfjk ^* ) ^T …… (2) Input uk = (ΔNref, ik ^* ) ^T …… (3) Disturbance wk = (ΔTmik ^* , ΔHik ^* , ΔBik ^* ) ^T …… (4) k = 1,2 (k: Subsystem No.) Number of subsystems = 2 ik: Stand No. included in the subsystem. If k = 1, ik = 1,2 If k = 2, ik = 3,4,5,6 jk: No.k Stand No. + No. (k-1) Downstream stand No. of subsystem . + No. (k + 1) subsystem's most upstream stand No., (No. (k-1) subsystem does not exist when k = 1, No. (k + 1) subsystem does not exist when k = m) ) If k = 1, jk = 1,2,3 If k = 2, jk = 2,3,4,5,6 nk: No.k Number of stands included in the subsystem If k = 1, nk = 2 If k = 2 When nk = 4, rolling equation of state xk (j + 1) = Ak xk (j) Bk uk (j) ...... (5) yk (j) = Ck xk (j) ...... (6) j: using a sample number (integer), the evaluation function (Here, Q and R are weighting matrices) Find the control input of the control deviation and state vector feedback that minimizes.

The control input u _K (k) is u (k) = − F ₁ e _K (k−1) −F ₂ [e _K (k) −e _K
(K-1)] - F _{3 [xs K (k) -xs K (} k-1) ] + u _K
(K-1) is given by (8). Here, the control deviation e _K, first floor difference partial state xs _K and input u _K v _K is _{e K (k) = r K} (k) -y K (k) ...... (9) (r K (k ) ≡ 0: Target value) xs _K (k) = (ΔNik ^* , Δik ^* ) ^T ... (10) v _K (k) = u _K (k) -u _K (k-1) ... (11) The optimal gain matrix defined by Is calculated from the following Riccati equation.

F = (R + G ^T PG) ⁻¹ G ^T PΦ …… (12) PΦ ^T PΦ−Φ ^T PG (R + G ^T PG) G ^T PΦ + H ^T QH …… (1
3) However, the matrices Φ, G, and H are error systems of the systems (5) and (6). X _K (k + 1) = Φ X _K (k) + Gv _K (k) (14) e _K (k) = _{HX K (k) ...... (15} ) X K (k) = _{^{[▲ e T K ▼ (k-}} 1), ▲ e T K ▼ (k) -
_{^{▲ e T K ▼ (k-}} 1), x ▲ s T K ▼ (k) -x ▲ s T K ▼ (k
−1)] ^T is a coefficient matrix of (16) and is obtained from matrices A to C. Although the optimum control system of the discrete time system was derived here, if the state equation of rolling (Equations (5) and (6)) is given in the continuous time system, the optimum control system of the continuous time system is constructed by the same method. it can.

By applying this optimal control law, an optimal control system as shown in FIG. 1 for wire rod and bar steel rolling is constructed. Fig. 1 shows No. 1 and 2 stands among 6 coarse row stands in a multi-stand mill, No. 1 subsystem, No. 3, 4, 5 and 6.
It is an example of tension control when the stand is the No. 2 subsystem. The control computer 8 samples the tension Δtfj ^* , the roll rotational speed ΔNj ^* and its derivative Δj, the roll gap Δi ^* , and the rolling load Δi ^*, and determines the optimum control input u _K (k) given by the equation (8). Calculated roll speed setting value ΔNref,
The ik ^*, (12), the weights Q, R of formula (13) was determined by repeating the simulation.

No.1 stand in the coarse row 6-stand mill The average temperature of the material on the inlet side, the height of the inlet side, and the width of the inlet side are ΔTm ₁ = -25 ° C, ΔH ₁ ^* = ΔB ₁
We examined the case where the change was ^* = 0.007 in steps or signs.

The step response without control is shown in FIGS. 2, 3 and 4. That is, FIGS. 2 to 4 show changes in the average temperature of the inlet material, the height of the inlet material, and the width of the inlet material ΔTm ₁ = −25 ° C., ΔH ₁ ^*
= ΔB ₁ ^* = 0.007, the step response of the height Δhi ^{* of the} cross section of the rolled material on the mountain side of each stand, the width Δbi ^* of the rolled material on the mountain side, and the tension Δtfj ^* is illustrated. Here, it is assumed that the time t = 2 seconds when the material temperature and the dimensional variation reach the No. 1 stand.

Step responses in the case of one system and in the case of being divided into two subsystems with control are shown in FIGS. 5 to 8 and 9 to 12, respectively. 5 to 12 show ΔTm ₁ = −25 ° C. and ΔH ₁ ^* = Δ when control is performed (tension control).
The step response of the mountain side height Δhi ^* , the mountain side width Δbi ^* , the tension Δtfj ^* , and the roll rotational speed ΔNj ^* according to B ₁ ^* = 0.007 is illustrated. It can be seen that the tension, peak height, and peak width are controlled.

Next, ΔTm ₁ = -25 ° C, ΔH ₁ ^* = ΔB ₁ ^* = 0.007, period
Consider the case where a cycle disturbance of 2.6 seconds is applied.
Figures 13 to 15 show the peak-side height Δhi ^* , without control.
This is an example of the response of Yamakawa width Δbi ^* and tension Δtfj ^* . Here, the time when the tip of the material reaches the No. 1 stand is t = 2.0 seconds.

Figures 16 to 23 show the mountain side height Δhi ^* and the mountain side width with control (tension control) for one system (Figs. 16 to 19) and when divided into two subsystems (Figs. 20 to 23). This is an example of the response of Δbi ^* and tension Δtfj ^* .

In this example, the number of target stands was as small as 6, so that good control performance was obtained even when the whole system was controlled by one system.

In the above example, the looperless continuous rolling was shown, but even when there is a looper between the stands, if the state equation is configured using the relational expression between the looper height and the rolled material tension, the same method as in the above example is used. You can control the dimensions with.

[Effects] As described above, according to the present invention, since the optimal regulator control is performed by dividing the subsystems, the following effects can be obtained.

(1) The tension control and the height control of each stand mountain side gauge meter can settling time is short and the offset can be controlled to zero.

(2) The offset of the finishing stand mountain side height and mountain side width can be controlled to zero.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of an apparatus for carrying out the present invention. FIG. 2, FIG. 3 and FIG. 4 are graphs showing examples of the step response of the peak-side height Δhi ^* , the peak-side width Δbi ^* , and the tension Δtfj ^* when there is no control. 5, 6, 6, 7, 8, 9, 10 and
Fig. 11 and Fig. 12 show the peak-side height Δhi ^{* with control} ,
It is a graph which shows the example of the step response of mountain side width (DELTA ⁾ bi ^* , tension (DELTA ⁾ tfj ^* , and roll speed (DELTA ⁾ Nj ^* . Figures 13, 14, and 15 show the peak-side height Δhi ^* , the peak-side width Δbi ^* , and the tension Δtfj ^{* in} the case of sine disturbance without control ^.
5 is a graph showing an example of the response of FIG. FIGS. 16, 17, 18, and 19 show the peak-side height Δhi ^* and the peak-side width Δb in the case of control in one system configuration.
Graph showing i ^* , tension Δtfj ^* , roll speed ΔNj ^* , second
0 Figure, Figure 21, Figure 22 and Figure 23, the mountain side height of the controlled case in the second sub-system configuration .DELTA.hi ^*, mountain width Derutabi ^*, tension Derutatfj ^*, an example of a roll rotation speed DerutaNj ^* It is a graph shown. 1,3,5: Horizontal rolls 2,4,6: Vertical rolls 7: Material 8: Control computer 9,10: Sub-system division stand 11, 12: Computer subsystem

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Masayuki Araki 1-1-1 Edamitsu, Yawatahigashi-ku, Kitakyushu-shi, Fukuoka (3) Technical Research Laboratory, Nippon Steel & Co., Ltd. (56) Reference JP-A 64-71512 ( JP, A)

Claims (1)

[Claims] 1. A roll gap deviation in each stand constituting the rolling mill train during continuous rolling of a wire rod and a steel bar by the rolling mill train. Is Si ^* , i = 1,2, ..., n Deviation of forward tension of rolled material Δtj ^* , j = 1,2, ..., n−1 Roll rotation speed deviation [(measured value Ni) / (reference value No, i) -1] and its differential amount are ΔNi ^* , Δi ^* , i = 1,2, ..., n Rolling load deviation [(actual load Pi) / (reference value Po, i) -1] is ΔPi ^* , i = 1,2, ..., n Of the height of the cross section of the rolled material [(measured value hi) / (reference value ho, i) -1] is Δhi ^* , i = 1,2, ..., n [where the measured value hi is the gauge meter type hi ＝ Hcal, i ＋ Si ＋ αiPi / Mi (Hcal, i: Caliber depth, Si: Roll gap, αi: Scale factor, Pi: Load, Mi: Mill rigidity)], Roll gap, Rolling load, Rolled section height The lock-on values are SL, i, PL, i, hL, i, and their deviations are ΔSL ^* ,
i, ΔPL ^* , i, ΔhL ^* , i (ΔSL ^* , i ＝ (SL, i−So, i) / ho, i, ΔPL ^* , i ＝ (PL, i−Po, i) / Po, i, ΔhL ^* , i ＝ (hL, i−ho, i) / ho, i, hL, i ＝ H cal, i ＋ SL, i ＋ αiPL, i / Mi), and the deviation Δh ^* i And the deviation ΔhL ^* , i of the lock-on value of the cross-section height of the rolled material on the exit side, plus the monitor compensation amount Δh ^* M, i, Δ ^* i is Δ ^* i = Δi ^* + αiΔi ^* / M ^* i + ΔhM ^* , i, i = 1,
2, ・ ・ ・, n (where Δi ^* = ΔSi ^* −ΔSL ^* , i, Δi ^* = ΔPi ^* −ΔPL ^* , i, ΔMi ^* = ho, i Mi / Po, i, and monitor compensation amount ΔhM ^* , i is
From the measured values of the deviations of the height and width dimensions of the rolled material measured by the rolled material cross-sectional dimension meter provided on the stand exit side, the amount of reduction at each stand is determined by the height and width dimensions of the rolled material cross section at the finish stand exit side. Based on the influence coefficient, it is distributed to each stand and obtained by integration.) Average temperature of the material on the entrance side of each stand, deviation of the height dimension and width dimension of the rolled material cross section [(measured value) / (reference value) −1] is ΔTmi ^* , ΔHi ^* , ΔBi ^* , i = 1,2, ..., n Deviation of loop height of rolled material [(measured value) / (reference value)-
1] is yhj ^* [j = 1,2, ..., n−1, (n: the total number of stands in the target rolling mill train)], all stands can be converted into subsystems of several stands. The state vector x is divided for each subsystem, and (a) when there is no loop between stands, xk = (Δik ^* , Δtfjk ^* , ΔNik ^* , Δik ^* ) ^T (The superscript T is the transposed matrix. (B) When there is a loop between stands, xk = (Δik ^* , Δyhjk ^* , ΔNik ^* , Δik ^* ) ^T, and when there is no loop between (a) stands, yk = (Δik ^* , Δtfjk ^* ) ^T (b) When there is no loop between stands, yk = (Δik ^* , Δyhjk ^* ) ^T , input uk = (Δref, ik ^* , ΔNref, ik ^* ) ^T , and disturbance wk = ( ΔTmik ^* , ΔHik ^* , ΔBik ^* ) ^T , k = 1,2, ..., m [k: Subsystem No., m: Number of subsystems, ik: Stand No. included in No.k subsystem. , jk: No.k sub Stand No. + No. (k-1) included in the system Downstream of stand No. + No. (k + 1) Uppermost stand No. of subsystem, No. (k-1) when k = 1 Subsystem does not exist, No. (k + 1) subsystem does not exist when k = m, nk: No.k number of stands included in the subsystem], and the rolling state equation xk (j + 1) = Ak xk (J) + Bk uk (j) yk (j) = Ck xk (j) j: Using the sample number (integer), the control input uk = (Δref, lk ^* , ΔNref, ik ^* ) ^T has the stand number lk at lk. = 0, n (k-1) to nk, under the condition that any combination can be taken, when lk = 0, the roll gap is excluded from the manipulated variable, and the evaluation function [Where Q and R are weighting matrices, ek (j) = rk (j) −yk (j) = − yk (j), (rk (j) = 0 is the target value) vk (j) = uk (j ) −uk (j−1)] by optimal input of control deviation and state vector feedback to minimize inter-stand tension or loop height in the rolling mill train in the subsystem and the most downstream stand in the subsystem ( No.n in No.k subsystem
A dimension control method in bar and wire rod rolling, characterized in that multi-variable control is performed over all stands by controlling the height dimension and width dimension of the rolled material discharged from the k-stand).