JPH04361810A - Controlling method for hot rolling - Google Patents

Abstract

PURPOSE:To contrive cooperation of tension components and to reduce the fluctuation of looper angle by using a frequency weight filter and controlling by using H infinity control logic. CONSTITUTION:A signal abtd. by subjecting a tensile deviation signal (a) to pass through a frequency weight filter W1 26, signal abtd. by subjecting at angular deviation signal (b) to pass, through a frequency weight filter W 2 27 and signal abtd. by subjecting the tension (c) of sheet and looper angle (d) to pass through a frequency weight filter W3 28 are taken as new state variables and signals that the respective state variables that are newly added are multiplied by state feedback gains 29, 30, 21 and 31, 32, 25 are added to conventional state feedback signals.

Description

Description: FIELD OF THE INVENTION The present invention relates to a control method for keeping tension constant during hot rolling by changing roll circumferential speed and looper motor torque. [0002] In hot rolling control equipment, automatic plate thickness deviation control system (AG) is used to realize highly accurate plate thickness.
C) is adopted (Japan Iron and Steel Institute, "Theory and Practice of Plate Rolling", p. 223 to p. 256). The automatic inner plate thickness deviation control system will be explained in detail with reference to FIG. In FIG. 8, M: Rolling mill rigidity coefficient (KgW/mm) Q: Rolling material plasticity coefficient (KgW/mm) Δu: Rolling system mechanism operation command amount (mm) ΔS: Rolling mechanism operation amount (mm) ΔP : Rolling load variation (KgW) Δh: Output side plate thickness deviation (mm) ΔH: Inlet side plate thickness deviation (mm) ΔRr: Rolling reference (mm) α: Tuning factor (---) (0≦
α≦1) s: Plus operator (1/sec) G: Integral constant (1/sec) T1: Time constant (sec), provided that T1 <<1.0. In this case, the transfer function G1 from the inlet plate thickness deviation (ΔH) to the outlet plate thickness deviation (Δh) is expressed by Equation 2 using W expressed by Equation 1. [Equation 1] [Equation 2] Now considering the characteristics in a steady state (s=0.0), G1 is expressed by Equation 3. ##EQU00003## Therefore, in order to completely eliminate the entrance plate thickness deviation (ΔH), the tuning factor must be α=1.0. [0011] Furthermore, when using the automatic inner plate thickness deviation control system, the tension of the rolled material, the thickness of the rolled material, the speed of the rolled material on the exit and entry sides of the rolling mill, and the roll of the upstream stand from the looper angle. A method of changing the circumferential speed has been adopted, for example, Japanese Patent Application Laid-Open No. 59-110410, Japanese Patent Application Laid-open No. 62-72411.
In these methods, the usual 15 to 2
A DC mill motor with a crossover frequency of 5 rad/sec is used. [0012] However, the conventional method has the following problems, which will be explained in detail with reference to FIG. FIG. 9 is a diagram showing a conventional hot rolling control device. In FIG. 9, 1 is a rolled material, 2 is an i-1 rolling mill (hereinafter referred to as an upstream rolling mill), 3 is an i-th rolling mill (hereinafter referred to as a downstream rolling mill), and 4 is a looper. roll, 5 is the looper arm, 6 is the looper drive motor,
7 is a CPU, 10 is an upstream mill motor system, and 11 is a downstream mill motor system, and the rolled material flows from the upstream rolling mill to the downstream rolling mill. In FIG. 9, Hi: Thickness of the plate at the entrance of the i-th rolling mill (mm) Vi: Speed at the entrance of the i-th rolling mill (mm/sec) hi:
I-th rolling mill outlet side plate thickness (mm) vi : I-th rolling mill outlet speed (mm/sec) Hi-1
: i-1 No. 1 rolling mill entrance plate thickness (mm) Vi-1 : i-
No. 1 rolling mill entry speed (mm/sec) hi-1: i-
No. 1 rolling mill outlet side plate thickness (mm) vi-1 : i-1 rolling mill exit speed (mm/sec) ΔVRi-1 : i-1 rolling mill circumferential speed change amount (mm/sec). [0015] In steady state, the i-th stand and i+
Between the No. 1 stands, Vi-1 is expressed by Equation 4. [Formula 4] vi-1 = Vi [0017] At the i-th stand, the relationship shown in Equation 5 holds true. [Equation 5] Hi ・Vi = hi ・vi 0019
For example, when the deformation resistance of the rolled material changes and the plate thickness at the exit side of the rolling machine changes as hi → hi +Δhi, Vi changes accordingly as shown in Equation 6. ##EQU6## Furthermore, as shown in the equation 7, the velocity equilibrium is broken. [Equation 7] vi ≠ Vi + 1 If left as it is, the rolled material may form a loop and fold into the rolling mill, causing the rolled material to break. By changing the roll circumferential speed of the side stand, the balance in Equation 7 can be restored. [0024] In this way, it is attempted to suppress changes in tension caused by changes in plate speed by controlling the torque of the looper motor and the circumferential speed of the rolls of the upstream or downstream stand. In other words, this is a redundant control system that attempts to control one controlled variable using two manipulated variables. In this way, conventional control includes non-interference control and LQ control.
In non-interference control, the looper height is controlled by the looper motor torque, and the tension is controlled by the roll peripheral speed, and only the manipulated variable of the roll peripheral speed was used for tension control. Therefore, disturbance suppression using non-interference control is as shown in FIG. FIG. 6 shows how the steel plate tension changes when the entry side plate speed Vi changes due to changes in deformation resistance, etc., and the influence of AGC. The vertical axis of this figure is the steel plate tension (K
gf), the horizontal axis represents time (seconds). Reference numeral 1 in the figure corresponds to tension component 1 in the looper control block diagram shown in FIG. 3, which will be described later.
corresponds to tension component 2, and the sign T corresponds to the tension that is the sum thereof. Note that tension component 1 represents the amount of tension affected by manipulating the roll circumferential speed, and tension component 2 represents the amount of tension affected by manipulating the looper angle. In addition, as shown in FIG. 6, in non-interference control, tension component 1 caused by manipulating the roll circumferential speed is equal to the total tension fluctuation, that is, if the tension influence due to the looper angle change is 0, the tension component 2 Since no suppression is performed, when the advance rate fluctuates by about 10% due to the hydraulic pressure reduction of the AGC, it is greatly affected by plate speed fluctuation disturbance, and tension fluctuation with an amplitude of about 700 Kgf occurs. Further, the period corresponds to a response of 120 rad under hydraulic pressure. FIG. 7 shows a case where LQ control is used. In this case, tension components 1 and 2, which are the tension effects of the two manipulated variables, cancel each other out to some extent, and the total tension fluctuation is suppressed to a swing width of about 100 kgf. However, since tension component 2 is used too much,
Looper angle variation [Tension component 2 x (L/EhB) KLS
] is large, resulting in unstable operation. Therefore, a design is required that allows the two manipulated variables to cooperate quantitatively. In other words, it is necessary to appropriately separate and coordinate the sharing of the two tension components in terms of frequency. in particular,
Large, slow movements are handled by the roll circumferential speed, or tension component 1. Moving the looper too much can impede operational stability, so fine, fast movements are handled by the looper's up and down movements, or tension component 2. It is considered necessary to take appropriate measures. The present invention provides a hot rolling control method that advantageously solves the above problems. [Means for Solving the Problems] The present invention provides for measuring the tension of the rolled material detected by a tension meter for measuring the tension of the rolled material, an angle meter and an angular velocity meter of a looper that moves the rolled material up and down during rolling of a steel plate. , looper angle, and looper angular velocity, when the central processing unit calculates the mill motor rotational speed correction amount and looper torque or looper motor rotational speed correction amount, it calculates the tension control amount due to the roll peripheral speed change and the tension control amount due to the looper angle change. A hot rolling control method characterized by using a frequency weighting filter to divide the control portion into appropriate frequency ranges and determining a state feedback gain based on H∞ control theory to change the roll circumferential speed and looper motor torque. It is. [Function] The present invention will be explained in detail below along with its function. FIG. 1 shows an example of a hot rolling control device suitable for carrying out the present invention and a hot rolling control method using the same. In FIG. 1, 1 is the rolled material, 2 is the i-1th
No. 1 rolling mill (hereinafter referred to as the upstream rolling mill), 3 is the i-th rolling mill (hereinafter referred to as the downstream rolling mill), 4 is the looper roll, 5 is the looper arm, 6 is the looper drive motor, 7 8 is a CPU, 8 is an upstream mill motor system, and 9 is a downstream mill motor system, and the rolled material 1 is flowing from the upstream rolling mill to the downstream rolling mill. In FIGS. 1 and 2, Hi :i
Thickness at the entrance of the No. rolling mill (mm) Vi : Speed at the entrance of the No. i rolling mill (mm/sec) hi : Thickness of the sheet at the exit side of the No. i rolling mill (mm)
c) Hi-1: No. i-1 rolling mill entrance plate thickness (mm) Vi
-1: i-1 rolling mill entry speed (mm/sec) hi
-1: No. i-1 rolling mill exit plate thickness (mm) vi-1:
i-1 rolling mill exit speed (mm/sec) ΔVRi-1
: I-1 rolling mill circumferential speed change amount (mm/sec). Further, in FIG. 2, W1, W2, W3
: Frequency weighting filter ▶ mark : Changeover switch (soft switch), ON is LQ (ILQ) control mode, OFF is robust control mode. KI: Integral gain KP: State feedback gain KS: Gain adjustment parameter Further, the CPU 7 and upstream mill motor system 8 are equipped with an AC mill motor system of the present invention having a cross frequency of 40 rad/sec or more. There is. In hot rolling,
In order to achieve highly accurate plate thickness, it is necessary to set the tuning factor to 0.85 to 1.0, which increases the swing angle of the looper and the tension fluctuation of the steel plate, but in order to stabilize the operation, the tuning factor of the looper needs to be set to 0.85 to 1.0. It is necessary to suppress the angle and tension variation of the steel plate to below a certain level. FIG. 2 is a block diagram of a looper multivariable optimal control system implementing the present invention. First, a tension target and an angle target are set from the outside, and compared with the actual tension (c) and actual angle (d), respectively, and the difference is input to the integrators 10 and 11. Then, the outputs of the integrators 10 and 11 are given integral gains of 12, 13, and 1.
Multiplyed by 4 and 15. To this, state feedback signals (a) and (b), which will be explained below, are added, and finally, the values multiplied by adjustment gains 16 and 17 become the mill circumferential speed and the command value to the looper. The calculation of the state feedback signals (a) and (b) is a process state quantity. State feedback gains 18, 19, 20, and 22 for steel plate tension (c), looper angle (d), and looper angular velocity (e), respectively.
Multiply by 23 and 24. Further, the following features of the present invention are that the tension deviation signal (F) is a signal passed through a frequency weighting filter W1 26, and the angular deviation signal (G) is a signal passed through a frequency weighting filter W2 27. The signals obtained by passing the frequency weighting filter W3 28 to the steel plate tension (c) and looper angle (d) signals are used as new state variables, and each newly added state variable is given a state feedback gain of 29, 30,
21, and the signal multiplied by 31, 32, and 25 is added to the conventional state feedback signal. Frequency weighting filter W1 26, W2 2
7, W3 28 is the key point of the present invention, and the process + operation end 33 will be explained below with reference to FIG. 4 and FIG. 3 as well. FIG. 2B is a block diagram of the process operating end 33 of the looper optimum control system implementing the present invention. Reference numeral 100 is a quadratic system approximation of the response of the rolling roll speed control system, and the output of this rolling roll speed control system 100 corresponds to the speed difference in the circumferential speed of the rolls between the stands, and the output of the rolling roll speed control system 100 corresponds to the speed difference in the circumferential speed of the rolls between the stands and Vd, and the signal obtained by subtracting the feedback term (a) to the plate speed due to the tension change, (b) is integrated by 10
1 becomes the amount of deflection between the loops (c), which is multiplied by Young's modulus and divided by the length 104 between the stands becomes the tension component 1 (d). [0047] Also, tension component 1(d) is obtained from the looper torque.
A signal obtained by multiplying by the influence coefficient KgT105 from tension to torque, a signal obtained by multiplying the looper angular velocity (f) by the damping constant of the looper 107, and a looper angle signal (g)
Influence coefficient KgS from looper angle to looper torque
The signal obtained by subtracting the signal multiplied by 108 is divided by the moment of inertia of the looper, and the signal obtained by integrating 106 becomes the looper angular velocity (f), and the signal further integrated by 109 becomes the looper angle signal (g). Note that each coefficient is as follows. E: Young's modulus of the steel plate, I: moment of inertia of the looper system, L: distance between stands, f: damping constant of the looper, KVT: influence term from tension to plate speed, KgT: torque from tension (from steel plate to KLS: Influence term from looper angle to plate elongation, KgS: Influence term from looper angle to looper load torque, KVT: i to (i+1) tension between stands Since the change corresponds to a change in the forward tension of the i-stand and the rearward tension of the i+1 stand, this changes the forward movement rate fi of the i stand and the backward movement rate bi+1 of the i+1 stand. Therefore, even if the circumferential speed of i, i+1 stand rolls is constant, i
The plate speeds on the exit side of the stand and the entrance side of the i+1 stand change. Advance rate: f=(VO−Vr)/Vr
Backward movement rate: b=(Vr - Vi)/Vr Since the steel plate slips on the exit and entry sides of the roll, the roll circumferential speed and the steel plate speed do not match. Vo: Outgoing plate speed, Vi: Incoming plate speed, Vr: Roll circumferential speed Vo>Vr>Vi, KVT is as shown in Equation 8. [Formula 8] KVT=(dfi/dT)・Vr,i+(db
i+1 /dT)・Vr,i+1 0053]KgT
: Torque Gt exerted on the looper by steel plate tension T (kg) (
kg·m) is expressed as shown in Equation 9. [Equation 9] Gt = R1 ・T ・(sin(θ+β)−s
in(θ-α))(kg·m), therefore, KgT becomes as shown in Equation 10, and KgS becomes as shown in Equation 11. [Formula 10] KgT(θ)=R1 ・(sin(θ+β)−si
n(θ−α))(kg・m) [Formula 11] KLS: Plate elongation 1=11 +12 −L=(
L1 +R1 ・cosθ)/cosα+(L−L1
-R1 ・cos θ)/cos β-L where α=tan-1{(R1 ・sin θ+R2 −H1
)/(L1 +R1 ・cosθ)}, β=tan-1
{(R1 ・sin θ+R2 −H1 )/(L−L1
−R1 ・cos θ)}. [Equation 12] Therefore, KLS=dl/dθ 0
[059] KgS: The load torque that changes as the looper angle changes includes torque due to the weight of the steel plate, torque due to the weight of the looper, torque due to bending of the steel plate, and, for example, due to the restoring force of an elastic body. Torque due to steel plate weight GM = WL
・cos θ・R1, WL: Torque due to the weight of the weight looper of the steel plate between the stands GS = WS ・co
sθ・R1/2, WS: Looper weight, torque due to looper bending GE =k・(L・sinθ−H1)c
osθ Therefore, KgS becomes as shown in Equation 13. [Formula 13] KgS=d(GM +GS +GE)/d
θ0062 Note that L, R1, R2, H1, L1
, E, I, θ are as follows. L: Distance between stands
5.486m R1: Looper arm length 0
．． 612m R2: Looper roll radius
0.092m H1:
Distance between looper shaft and pass line: 0.184m
L1: Stand ~ looper shaft
2.345m E: Young's modulus of steel plate
1.8×1
010kg/m2 I: Looper total moment of inertia 0.74kg・m・s2
θ: looper angle FIG. 4 is an explanation of the frequency weighting filters W1 111, W2 112, and W3 113, which are the key points of the present invention, and correspond to 26, 27, and 28 in FIG. W1 111 indicates the tension control sharing range by tension component 1, which is made larger in the low frequency range so that the low frequency range can be mainly shared. Further, W2 112 is a frequency weighting filter that indicates how far into the frequency domain the tension components 1 and 2, that is, the total tension control is to be performed, and increases the gain up to the response range to be controlled. Also W3 11
3 is a filter that indicates a frequency range that is not controlled by the tension component 2; In other words, the tension component 2 is the looper angle signal (
This corresponds to g), and if it fluctuates too much, the operation will become unstable, so the amount of operation should be limited as much as possible. Therefore, the tension control by the looper should be performed finely at high frequencies, and W3 represents the frequency range in which the looper does not move. The input to W1 is the difference between tension component 1 and the tension target, and the input to W2 is the tension deviation itself, W
The input to 3 is tension component 2. These inputs are W
Using the known H∞ design theory, the state feedback gains (18, 19, 20, 21,
22, 23, 24, 25, 29, 30, 31, 32) are determined. [Example] L: Distance between stands
5.486m R1: Looper arm length 0
．． 612m R2: Looper roll radius
0.092m H1:
Distance between looper shaft and pass line: 0.184m
L1: Stand ~ looper shaft
2.345m E: Young's modulus of steel plate
1.8×1
010kg/m2 I: Looper total moment of inertia 0.74kg・m・s2
θ: In the case of looper angle, plate speed disturbance Vd in Fig. 3 (hydraulic pressure: 120
FIG. 5 shows the tension component 1, tension component 2, and total tension when the tension fluctuation is suppressed by the control of the present invention in the case where the advance rate changes by 10% on a sine wave due to rad. Effects of the Invention As explained above, the conventional
Compared to the non-interference control and LQ control shown in FIG.
5 It can be seen that the amplitude of tension fluctuation is suppressed to a large extent (about 15 kgf). It can also be seen that the amplitude of tension component 2 is smaller than in LQ control, and the variation in the looper angle is also small. This is because the roles of the two tension components in the frequency range are sufficiently separated by the frequency weighting filter, so there is good coordination.For example, as set in W3 113 in Fig. 4, the looper angle fluctuation is It is kept small.

[Brief explanation of drawings]

FIG. 1 is a drawing showing an example of a hot rolling control device suitable for implementing the present invention.

FIG. 2 is a block diagram showing a looper multivariable optimal control system implementing the present invention.

FIG. 3 is a block diagram of a process operation end 33 of a looper optimum control system implementing the present invention, and is a drawing obtained by equivalently converting FIG. 1.

FIG. 4: Frequency weighting filter W1 1 implementing the present invention
11, W2112, and W3 113. FIG.

FIG. 5 is a waveform diagram showing the suppression characteristics against disturbance when controlled by the method of the present invention.

FIG. 6 is a waveform diagram showing a state of disturbance control under non-interference control.

FIG. 7 is a waveform diagram showing the state of disturbance control in LQ control.

FIG. 8 is a block diagram showing an automatic inner plate thickness deviation control system in the hot rolling control device.

FIG. 9 is a drawing showing an example of a conventional hot rolling control device.

[Explanation of symbols]

1 Rolled material 2 Upstream rolling mill 3 Downstream rolling mill 4 Looper roll 5 Looper arm 6 Looper drive motor 7 CPU 8 Upstream mill motor system 9 Downstream mill motor system

Claims (1)

[Claims] Claim 1: During rolling of a steel plate, the tension of the rolled material detected by a tension meter that measures the tension of the rolled material, an angle meter and an angular velocity meter of a looper that moves the rolled material up and down, the looper angle,
When calculating the mill motor rotational speed correction amount and looper torque or looper motor rotational speed correction amount using the central processing unit from the looper angular velocity, the tension control amount by changing the roll circumferential speed and the tension control amount by changing the looper angle are appropriately calculated. A hot rolling control method characterized by using a frequency weighting filter for dividing into a frequency range, determining a state feedback gain according to H∞ control theory, and changing the roll circumferential speed and looper motor torque.