Search

CN-122007173-A - Plate defect characteristic identification and quantification control method based on 26-zone sensor

CN122007173ACN 122007173 ACN122007173 ACN 122007173ACN-122007173-A

Abstract

The invention relates to the technical field of strip shape control of reversible cold-rolled sheet strips of a 20-roller multi-roller cold-rolling mill, in particular to a strip shape defect characteristic identification and quantification control method based on a 26-zone sensor, which divides the 26-zone strip shape sensor into five functional zones: the method comprises the steps of setting acquisition periods at the left part, namely 1-3 areas, the left 1/4 area, namely 6-9 areas, the middle area, namely 12-15 areas, the right 1/4 area, namely 18-21 areas, and the right part, namely 24-26 areas, acquiring sensor data of each area in real time in each acquisition period, calculating a middle wave characteristic value delta I_center, a double-side wave characteristic value delta I_edge, a single-side wave characteristic value delta I_asym, a left 1/4 wave characteristic value delta I_Q1 and a right 1/4 wave characteristic value delta I_Q3, and comparing each characteristic value with a corresponding preset threshold value to judge whether defects exist.

Inventors

  • REN YAN
  • ZHAO HE
  • SHI XIAOCHEN

Assignees

  • 山西太钢不锈钢精密带钢有限公司

Dates

Publication Date
20260512
Application Date
20260115

Claims (5)

  1. 1. A plate defect characteristic identification and quantification control method based on a 26-area sensor is characterized in that the 26-area plate sensor is divided into five functional areas, namely a left-side area, namely a 1-3 area, a left-side area, namely a 6-9 area, a middle-side area, namely a 12-15 area, a right-side area, namely an 18-21 area, a right-side area, namely a 24-26 area, an acquisition period is set, sensor data of each area are acquired in real time in each acquisition period, an intermediate wave characteristic value delta I_center, a double-side wave characteristic value delta I_edge, a single-side wave characteristic value delta I_asym, a left-side area 1/4 wave characteristic value delta I_Q1 and a right-side area 1/4 wave characteristic value delta I_Q3 are calculated, each characteristic value is compared with a corresponding preset threshold value, whether defects exist or not is judged, if the defects do not exist, the defects enter the next acquisition period to conduct the same judgment, if the defects exist, the types of the defects are judged, the corresponding adjustment modes and adjustment formulas are selected according to the types of the defects, adjustment formulas are calculated, adjustment quantity is controlled according to the adjustment quantity, an output control instruction is carried out, adjustment is carried out, the adjustment command is carried out, the adjustment is carried out, the output, the adjustment command is carried out, the preset value is calculated, the adjustment time is calculated, the preset value and the preset value is not to have the same value, the preset characteristic value, and the preset value is compared with the preset value, and the threshold value is not to the corresponding adjustment mode to the defect type to the corresponding to the type to the defect type, and the defect type to be corrected to the defect type, and the defect type.
  2. 2. The control method according to claim 1, wherein when ΔI_center >3 (I-units), which is a median section average value ,ΔI_center=I_center-I_edge,I_center=(I_12+I_13+I_14+I_15)/4,I_edge=(I_1+I_2+I_3+I_24+I_25+I_26)/6,I_center, I_edge is a side section average value, I_n is a value of an nth section, n is a sensor number, a natural number of 1 to 26, and I-units is a plate defect quantization index of 10 5 , is determined as a median wave defect; When Δi_edge >3 (I-units) and Δi_symmetry <4 (I-units), determining that the double-sided wave defect is detected, wherein ,ΔI_edge=I_edge_avg-I_center,ΔI_symmetry=|I_left-I_right|,I_edge_avg=(I_left+I_right)/2,I_left=(I_1+I_2+I_3)/3,I_right=(I_24+I_25+I_26)/3,ΔI_symmetry is a side symmetry determination value, i_edge_avg is a side integrated average value, i_left is a left average value, and i_right is a right average value; When- Δi_asym > +4, determining as a left unilateral wave defect, wherein Δi_asym=i_left-i_right; When-delta I_asym < -4, judging that the right single-side wave defect exists; When- Δi_q1>4, determining as left 1/4 wave defect, wherein Δi_q1=i_q1-i_ref_left, i_q1= (i_6+i_7+i_8+i_9)/4, i_ref_left= (i_left+i_center)/2, wherein i_q1 is a left 1/4 zone average value, i_ref_left is a left 1/4 zone reference value; When- Δi_q3>4, it is determined as a right 1/4 wave defect, wherein Δi_q3=i_q3-i_ref_right, i_q3= (i_18+i_19+i_20+i_21)/4, i_ref_right= (i_right+i_center)/2, wherein i_q3 is a right 1/4 zone average value, and i_ref_right is a right 1/4 zone reference value; when two or more defect judging conditions are satisfied, judging that the defect is a composite defect, wherein the composite defect comprises a middle wave and a single-side wave defect, a double-side wave and a single-side wave defect, a 1/4 wave and a middle wave defect, and a 1/4 wave and a single-side wave defect.
  3. 3. The control method according to claim 2, characterized in that: When the intermediate wave defect exists, the adjustment mode is roll force adjustment, wherein the roll force adjustment formula is delta F_band= -K_band multiplied by K_step multiplied by delta I_center multiplied by B multiplied by h, delta F_band is roll force increment, K_band is roll force influence coefficient, K_band unit is KN/[ (I-unit) m.mm) ], the value range is 8-12, K_step is steel grade correction coefficient, B is twice the width of strip steel, and h is the thickness of strip steel; When the double-sided wave defect exists, the roll bending force is adjusted firstly, then the roll bending force is adjusted, wherein the roll bending force is adjusted according to a formula of delta F_band=K_band×K_step×delta I_edge×B×h, the roll bending force is adjusted according to a formula of delta F_roll= -K_roll×K_step×delta I_edge×B, the unit of K_roll is kN/[ (I-unit) m ], the value range of K_roll is 50-80, wherein K_step is a steel grade correction coefficient, B is twice the width of strip steel, and h is the thickness of strip steel; When the single-side wave defect is the left single-side wave defect or the right single-side wave defect, the adjustment mode is that firstly, the inclination angle adjustment is carried out, then the string roll displacement adjustment is carried out, the inclination angle adjustment formula is delta theta-titl =K_ titl ×K_step×delta I_asym/B, the string roll displacement adjustment formula is delta S_shift= -K_shift×K_step×delta I_asym, wherein the delta theta-titl inclination angle is K_ titl, the unit is mrad.m/(I-unit), the value range is 0.015-0.025, the delta S_shift is the string roll displacement, the K_shift unit is mm/(I-unit), and the value range is 0.8-1.5; For the left 1/4 wave defect, the middle roller displacement adjustment formula is DeltaS_IR= -K_IR×K_step×DeltaI_Q1, the local cooling control formula is DeltaQ_cool=K_cool×K_step×DeltaI_Q1, wherein DeltaS_IR is the left 1/4 wave middle roller displacement amount, K_IR is the left 1/4 wave middle roller influence coefficient, K_IR unit is mm/(I-unit) and takes the value range of 0.6-1.0, and for the right 1/4 wave defect, the middle roller displacement adjustment formula is DeltaS_IL=K_IL×K_step×DeltaI_Q3, wherein DeltaS_IL is the right 1/4 wave middle roller displacement amount, K_IL is the right 1/4 middle roller influence coefficient, K_IL unit is the/(I-unit) and takes the value range of 0.6-1.0; For the composite defects, determining each defect, simultaneously adjusting different adjustment types, sequentially adjusting the same adjustment type according to the size of the characteristic value, and multiplying each adjustment quantity by a coupling correction coefficient K_coupling, wherein the coupling correction coefficient K_coupling takes a value in a range of 0.80-0.95.
  4. 4. A control method according to claim 3, wherein the modification factor of the ordinary austenitic steel grade is 1.00, the modification factor of the molybdenum-containing austenitic steel grade is 1.05-1.08, and the modification factor of the ferritic steel grade is 0.88-0.92.
  5. 5. The control method according to claim 3, wherein the ΔF_band single adjustment range is (-50 kN,50 kN), the cumulative adjustment range is (-300 kN,300 kN), the ΔF_roll single adjustment range is (-200 kN,200 kN), the cumulative adjustment range is (-15% set point, 15% set point), the Δθ_ titl single adjustment range is (-0.05 mrad,0.05 mrad), the cumulative adjustment range is (-0.20 mrad,0.20 mrad), the ΔS_shift single adjustment range is (-5 mm,5 mm), the cumulative adjustment range is (-20 mm,20 mm), the ΔS_IR single adjustment range is (-3 mm,3 mm), and the cumulative adjustment range is (-15 mm,15 mm).

Description

Plate defect characteristic identification and quantification control method based on 26-zone sensor Technical Field The invention relates to the technical field of strip shape control of reversible cold-rolled sheet strips of a 20-roller multi-roller cold-rolling mill, in particular to a strip shape defect characteristic identification and quantification control method based on a 26-zone sensor. Background The prior 20-roller multi-roller cold rolling mill plate shape control system has a plurality of technical problems in the ultra-thin precise strip steel rolling process. Firstly, multi-objective conflict optimization is difficult, a traditional PID control system can only process single objective control, and effective balance among a plurality of mutually-influenced objectives such as plate shape precision control, rolling force stability, surface quality assurance, energy consumption optimization and the like is difficult to realize. The accuracy of the plate shape is required to be controlled within a range of +/-2-3I, meanwhile, fluctuation of rolling force values is ensured to be within a reasonable range, the surface quality roughness Ra value is stable, and energy consumption optimization is realized on the premise of ensuring quality. Currently commercially available plate meters typically employ 20-30 measurement areas, with 26-area configurations being most common. However, the prior art lacks an explicit solution for how to use 26-zone sensor data for systematic defect identification and quantitative control. Operators face complex plate curves, and it is often difficult to quickly and accurately judge the defect types and adjustment amounts, so that the control effect is uneven. And secondly, the special control of the small-diameter working roller is difficult. The diameter of the working roll of the 20-roller rolling mill is only 21-45mm, and compared with the diameter of the working roll of the traditional rolling mill, the diameter of the working roll is smaller, so that the elastic flattening value of the roller is relatively larger, and the plate shape control precision is directly influenced. The small-diameter working roller is worn relatively quickly, the control parameters need to be frequently adjusted and are sensitive to pressure change, and the control stability of the system needs to be improved. Thirdly, the technological difficulty of plate defect identification and control. The defect identification system based on the 26-point plate shape detector has the difficulty in handling four typical defects that the middle wave is shown as abnormal areas of the sensors 12-15, rolling force needs to be properly adjusted, the double-side wave is shown as abnormal areas of the sensors 1-3 and 24-26, bending force needs to be properly adjusted, the single-side wave is shown as exceeding the standard of asymmetry and needs to be matched with the adjustment of the tilting roller and the tandem roller, and 1/4 wave is shown as abnormal areas of the sensors 6-9 and 18-21, and intermediate roller displacement adjustment is needed. Limitations of prior art systems are mainly manifested in: defect identification lacks quantitative criteria, wherein the prior art mostly adopts qualitative descriptions (such as 'middle waves', 'side waves'), and lacks quantitative identification formulas based on sensor data. Operators need to judge the defect type by visually observing the plate-shaped curve, the identification accuracy is about 80 percent, and the response time is 3-5 seconds. Control adjustments lack explicit formulas-the prior art fails to provide specific control formulas and parameters for identified plate defects. For example, there is no clear guidance on how much kN the roll force should be adjusted, and how much kN the rolling force should be adjusted after the intermediate wave is found. The adjustment amplitude difference of different operators can reach +/-50%. The sensor zoning function is ambiguous, which areas are used for monitoring side waves, which areas are used for monitoring middle waves and which areas are used for monitoring 1/4 waves in the 26-zone sensor, and the prior art lacks system definition. The composite defect is difficult to treat, when the composite defects such as middle waves and single-side waves occur at the same time, the prior art lacks a system identification method and a coordination control strategy, repeated trial and error is needed, and the once correction success rate is only about 40%. The parameter self-optimization mechanism is lacking, namely the influence coefficient and the control threshold value are usually fixed and cannot be automatically optimized according to the actual control effect. Disclosure of Invention The invention aims to solve the technical problems of constructing a multi-target collaborative optimization control system suitable for the characteristics of a small-diameter working roll of a 20-roll mill, realizing intelligen