Hot continuous rolling strip steel plate shape control method and device and electronic equipment

Abstract

The invention relates to a hot continuous rolling strip steel plate shape control method and device and electronic equipment. The method comprises the steps that plate blank PDI data, equipment parameters and model parameters are read; calculating speed, temperature and thickness parameters of each finish rolling rack based on a finish rolling model; the roll bending force set value of each rack is calculated, and the roll bending force reference value, the convexity self-learning distribution value, the straightness self-learning distribution value and the operator intervention value of the SAMP model table are superposed in response to the situation that the rolling specification is not replaced, so that the roll bending force set value is generated; in response to replacement of steel types, thickness or width specifications, on the basis of superposition of non-replaced specifications, a roll bending force difference value of the SAPP model table and the SAMP model table is introduced to be multiplied by a dynamic adjustment coefficient, and a roll bending force set value is generated; a roll bending force set value is issued to an execution system at the finish rolling inlet moment; and the actual convexity and straightness of the finish rolling outlet strip steel are collected, target values are compared, and the set value of the next plate blank is corrected through self-learning. The method improves the target convexity and straightness hit rate of the hot-rolled strip steel.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automatic control technology for rolling steel, in particular to a hot continuous rolling strip shape control method and device and electronic equipment. BACKGROUND

[0002] As a core shape quality index of hot rolling strip, the strip shape directly determines the quality grade of the final product. Due to the coupling effect of multiple factors such as steel composition, rolling specification, temperature distribution, equipment stiffness, etc. on the strip shape, and the highly complex action law, the control problem of the strip shape has long plagued the field of hot rolling strip. In recent years, affected by the external economic environment and the overcapacity of the steel industry, users have continuously improved the quality requirements for the strip. Not only are they concerned about the material organization performance, but they have also significantly increased the strictness of the shape quality (especially the strip shape). According to statistics, the strip shape problem accounts for a significant proportion of quality disputes, and has become a key bottleneck restricting product competitiveness.

[0003] The conventional strip shape control model performs well in steady-state rolling, but has obvious limitations in dynamic conditions such as steel grade change and specification change (such as the rolling of the first strip). The insufficient control accuracy of the crown and flatness leads to a decrease in the target value hit rate, and frequent manual intervention of the bending roller force is required to maintain the stability of the strip shape. This not only increases the workload, but also makes it easier to cause strip instability and even steel jamming accidents due to the fluctuation of the strip shape between the racks. Therefore, it is necessary to develop a self-adaptive and high-precision strip shape control method to break through the technical bottleneck in dynamic conditions and meet the production requirements of high quality. SUMMARY

[0004] Therefore, it is necessary to provide a hot continuous rolling strip shape control method, device and electronic equipment to solve the problem of insufficient control accuracy of the crown and flatness of the strip during the change of the steel grade and the specification.

[0005] The present application provides a hot continuous rolling strip shape control method, which comprises:

[0006] reading the slab PDI data, equipment parameters and model parameters, the model parameters including the crown self-learning value, the flatness self-learning value, the operator intervention value self-learning value of the bending roller force of each rack in the rolling mill model table, the bending roller force set value of the previous strip, and the bending roller force set value of the previous strip in the product self-learning model with the same steel grade, thickness and width layer;

[0007] calculating the speed, temperature and thickness parameters of each rack in the finishing mill based on the finishing mill model;

[0008] The bending roll force setting value of each stand is calculated according to the plate shape model, wherein, in response to the non-replacement rolling specification, the bending roll force reference value, the crown self-learning distribution value, the flatness self-learning distribution value and the operator intervention value of the SAMP model table are superimposed to generate the bending roll force setting value; in response to the replacement of the steel grade, the thickness or the width specification, the bending roll force difference value of the SAPP model table and the SAMP model table is multiplied by a dynamic adjustment coefficient on the basis of the superposition of the non-replacement specification to generate the bending roll force setting value;

[0009] The bending roll force setting value is issued to the execution system at the finishing rolling entrance moment;

[0010] The actual crown and flatness of the strip steel at the finishing rolling exit moment are collected, the deviations are compared with the target values, and the next slab setting value is corrected through self-learning.

[0011] In one of the embodiments, the superimposition in response to the non-replacement rolling specification comprises:

[0012] The bending roll force reference value, the crown self-learning distribution value, the flatness self-learning distribution value and the operator intervention value of the SAMP model table are directly added to generate the bending roll force setting value; wherein, the crown self-learning distribution value is dynamically distributed based on the crown deviation of the previous strip steel, so that the first several stands have non-zero values and the last several stands have zero values, and the flatness self-learning distribution value is dynamically distributed based on the flatness deviation of the previous strip steel, so that the last stand has a non-zero value and the other stands have zero values.

[0013] In one of the embodiments, the generation of the bending roll force setting value comprises:

[0014] The bending roll force difference value of the SAPP model table and the SAMP model table is multiplied by a dynamic adjustment coefficient and superimposed on the basis of the superposition of the non-replacement specification, wherein, the dynamic adjustment coefficient is dynamically determined based on the change of the steel grade, the thickness or the width to adjust the bending roll force setting value in response to the specification change;

[0015] The dynamic adjustment coefficient coef[i] is dynamically adjusted according to different stand positions, wherein i represents the stand number, and the coefficient is generated by an expert system based on historical production data to optimize the crown and flatness hit rate of the first strip steel after the specification change.

[0016] In one of the embodiments, the self-learning correction comprises:

[0017] The crown self-learning value is dynamically updated based on the crown deviation, and the flatness self-learning value is dynamically updated based on the flatness deviation, wherein, the updated self-learning value is used for the calculation of the bending roll force setting value of the next slab to continuously optimize the plate shape control in response to the measured deviation.

[0018] In one of the embodiments, the updating of the crown self-learning value comprises:

[0019] allocating the crown deviation to each stand so that the crown self-learning allocation value of the first several stands is increased or decreased, and the value of the last several stands remains zero;

[0020] the updating the flatness self-learning value comprises:

[0021] allocating the flatness deviation to the last stand so that the flatness self-learning allocation value of the last stand is increased or decreased, and the value of the other stands remains zero.

[0022] In one of the embodiments, the operator intervention value self-learning value is dynamically adjusted based on operator input and directly superimposed into the bending force setting value calculated according to the plate shape model to respond to real-time operation intervention to optimize plate shape control, wherein the operator intervention value is included in the calculation regardless of whether the gauge is changed or not.

[0023] In one of the embodiments, the operator intervention value self-learning value is dynamically updated through the self-learning function of the plate shape setting model and adjusted based on historical operation data to reduce the operator workload and improve the plate shape stability.

[0024] In one of the embodiments, the crown self-learning value, the flatness self-learning value and the operator intervention value self-learning value are all dynamically generated based on the production data of the previous piece of strip steel and stored in the rolling mill model table or the product self-learning model and updated through the expert system learning to respond to the changes in the continuous rolling process.

[0025] The application also provides a hot continuous rolling strip steel plate shape control device, which comprises:

[0026] a data reading module for reading slab PDI data, equipment parameters and model parameters, wherein the model parameters comprise the crown self-learning value, the flatness self-learning value, the operator intervention value self-learning value of each stand bending force and the bending force setting value of the previous piece of strip steel in the rolling mill model table, and the bending force setting value of each stand of the previous piece of strip steel in the product self-learning model with the same steel grade, thickness and width layer;

[0027] a parameter calculation module for calculating the speed, temperature and thickness parameters of each stand in the finishing mill based on the finishing mill model;

[0028] a calculation response module for calculating the bending force setting value of each stand according to the plate shape model, wherein in response to the non-changing rolling gauge, the bending force reference value, the crown self-learning allocation value, the flatness self-learning allocation value and the operator intervention value of the SAMP model table are superimposed to generate the bending force setting value; in response to the changing steel grade, thickness or width gauge, the bending force difference value of the SAPP model table and the SAMP model table multiplied by a dynamic adjustment coefficient is introduced on the basis of the superposition of the non-changing gauge to generate the bending force setting value;

[0029] a data sending module configured to send the bending force setting value to an execution system at the finishing rolling entry moment;

[0030] a comparison correction module configured to collect the actual crown and flatness of the strip at the finishing rolling exit, compare the target value to obtain a deviation, and correct the setting value of the next slab through self-learning.

[0031] The application further provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the hot strip shape control method according to any one of the above embodiments when executing the computer program.

[0032] The hot strip shape control method, device and electronic device, in the stage of reading slab data, integrate the historical self-learning value of the rolling mill model and the setting value of the same specification strip, when detecting the change of the steel grade / thickness / width specification, introduce the bending force difference compensation term of the SAPP and SAMP model table into the basic superposition model, quantize the nonlinear interference of the specification mutation on the strip shape by using the dynamic adjustment coefficient, directly improve the bending force setting precision of the first strip of the changed specification, meanwhile, based on the position characteristic difference of the mill stand, differentiate the self-learning value, focus the crown self-learning value on the front mill stand to control the rolling stability, and apply the flatness self-learning value to the last mill stand to optimize the finished product quality, combine the continuous accumulation of the operator intervention value to form a closed-loop experience library, and accurately send the setting value at the finishing rolling entry; finally, through collecting the deviation of the actual strip shape data and the target value, trigger the self-learning mechanism to correct the subsequent slab setting in real time, thereby systematically solving the core defect of the insufficient crown / flatness hit rate when changing the specification in the background technology, and synchronously eliminating the strip wearing and steel clamping accidents caused by the frequent adjustment of the load and the suppression of the strip shape fluctuation between the mill stands. BRIEF DESCRIPTION OF DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without any creative effort.

[0034] Figure 1 A flow chart of the hot strip shape control method of one embodiment;

[0035] Figure 2 A flow chart of the hot strip shape control method of another embodiment;

[0036] Figure 3 A schematic diagram of the hot strip shape control device of one embodiment;

[0037] Figure 4An internal structure diagram of a computer device of an embodiment. DETAILED DESCRIPTION

[0038] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts should fall within the protection scope of the present application.

[0039] The embodiments of the present application will be described below with reference to the drawings. Figures 1-4 A hot strip shape control method, device and electronic equipment are described.

[0040] As shown in FIG. 1, in one embodiment, a hot strip shape control method comprises the following steps: Figure 1

[0041] In step S110, slab PDI data, equipment parameters and model parameters are read. The model parameters include crown self-learning values, flatness self-learning values, operator intervention value self-learning values of each stand bending force and the previous strip bending force set values in the rolling mill model table, and the previous strip bending force set values of each stand in the product self-learning model of the same steel grade, the same thickness and the same width layer.

[0042] The read crown self-learning values, flatness self-learning values and operator intervention value self-learning values are dynamically generated based on the production data of the previous strip and stored in the rolling mill model table or the product self-learning model. These values are updated by an expert system to respond to changes in the continuous rolling process. This data fusion mechanism enables the system to respond to multi-dimensional changes in the rolling process (such as roll gap offset and incoming material fluctuation) simultaneously, ensuring the consistency of the shape of long-period rolling and achieving the collaborative optimization of multi-source data.

[0043] The bending force set values of the previous strip in the product self-learning model of the same steel grade, the same thickness and the same width layer are dynamically matched with the current slab parameters (such as width layer) by the self-learning function of the shape setting model to optimize the set values in response to changes in the steel grade, thickness or width layer. When the steel grade, thickness or width layer changes, the system automatically calls the closest historical data optimization set value to avoid control errors caused by specification jumps and to accurately match the layer parameters.

[0044] In step S120, the speed, temperature and thickness parameters of each stand in the finishing mill are calculated based on the finishing mill model. The speed, temperature distribution, thickness and loop angle parameters of each stand are calculated based on the finishing mill model to provide input for shape control.

[0045] ​In step S130, the bending roll force setting value of each stand is calculated according to the plate shape model. In response to the non-replacement rolling specification, the bending roll force reference value, the crown self-learning distribution value, the flatness self-learning distribution value and the operator intervention value in the SAMP model table are superimposed to generate the bending roll force setting value. In response to the replacement of the steel grade, thickness or width specification, the bending roll force difference value between the SAPP model table and the SAMP model table is multiplied by a dynamic adjustment coefficient on the basis of the superposition of the non-replacement specification to generate the bending roll force setting value.

[0046] In response to the non-replacement rolling specification, the superposition includes directly adding the bending roll force reference value, the crown self-learning distribution value, the flatness self-learning distribution value and the operator intervention value in the SAMP model table to generate the bending roll force setting value, which simplifies the calculation process. The crown self-learning distribution value is dynamically distributed to the first few stands based on the crown deviation of the previous strip, so that the first few stands have non-zero values and the last few stands have zero values. The flatness self-learning distribution value is dynamically distributed based on the flatness deviation of the previous strip and is concentrated in the last stand, so that the last stand has a non-zero value and the other stands have zero values. This targeted distribution allows the front stands to prioritize correcting the crown defect during rolling and the last stand to accurately control the flatness, thereby avoiding redundant intervention of multiple stands, improving response speed and plate shape quality, and improving the bending roll force setting efficiency and optimizing the plate shape control precision.

[0047] In response to the replacement of the steel grade, thickness or width specification, the generation of the bending roll force setting value includes calculating the bending roll force difference value between the SAPP model table and the SAMP model table, multiplying it by a dynamic adjustment coefficient, and superimposing it on the basis of the superposition of the non-replacement specification. The dynamic adjustment coefficient is dynamically determined based on the changes in the steel grade, thickness or width to adjust the bending roll force setting value in response to the specification change. When the steel grade, thickness or width specification is replaced, the bending roll force difference value between the SAPP model table and the SAMP model table is multiplied by a dynamic adjustment coefficient (such as coef[i]) and superimposed on the basis of the setting value of the non-replacement specification. This coefficient is dynamically determined according to the type of specification change (such as steel grade / thickness / width) to quickly adapt the bending roll force setting value to the process requirements of the new specification, significantly reducing the crown and flatness deviation of the first strip after the specification change, and solving the problem of out-of-control plate shape of the first strip after the specification change.

[0048] The dynamic adjustment coefficient coef[i] is dynamically adjusted according to different stand positions, where i represents the stand number, and the coefficient is generated by an expert system based on historical production data to optimize the crown and flatness hit rate of the first strip after the specification change. The dynamic adjustment coefficient coef[i] is represented as:

[0049]

[0050] In the formula, ΔH is the thickness difference between the current slab and the previous slab, H refH is the thickness reference (e.g. 20mm), AH is the width difference between the current slab and the previous slab, B ref H is the width reference (e.g. 1000mm), SteelGrade is the steel grade change indicator (0 = no change, 1 = change), k1, k2, k3 are the stand weight coefficients (need to be learned by expert system, example: F1 stand k1 = 0.8, k2 = 0.5, k3 = 0.3). Front stands (F1-F3): focus on thickness change compensation (k1 is larger); middle stands (F4-F6): balance thickness / width influence; end stands (F7-F8): focus on width change compensation (k2 is larger).

[0051] The generation method of the dynamic adjustment coefficient coef[i] is as follows:

[0052] When the steel grade / thickness / width specification change is detected, coef[i] is calculated differently according to the stand position. Taking the F1 stand as an example:

[0053] The previous slab thickness H prev = 2.0mm, the current slab thickness H curr = 1.8mm, AH = |1.8-2.0| = 0.2mm; the thickness reference H ref = 20mm, the thickness weight k1 = 0.8; similarly, the width term contribution is: Steel grade does not change, SteelGrade = 0; the final coef[i] = 0.8x(0.2 / 20) + 0.005 + 0 = 0.013, k1, k2, k3 of each stand are generated by expert system learning historical optimization records, for example, F1 stand focuses on thickness compensation (k1 is larger), F8 stand focuses on width compensation (k2 is larger).

[0054] The coefficient coef[i] of different stand positions (i represents the stand number) is generated by expert system based on historical production data learning, realizing the differentiated adjustment of the bending force of each stand when changing specifications. For example, the front stands focus on convexity compensation, and the end stands strengthen flatness control, thereby targetedly improving the convexity and flatness hit rate of the first strip, reducing the trial rolling scrap rate, and optimizing the stand adaptability of the dynamic adjustment coefficient.

[0055] The operator intervention value self-learning value is dynamically adjusted based on operator input and directly superimposed on the bending roll force set value in step S130 to respond to real-time operation intervention to optimize the plate shape control; wherein the operator intervention value is included in the calculation when the gauge is changed or not changed. The operator intervention value self-learning value is directly superimposed on the bending roll force set value when the gauge is changed or not changed, and is dynamically adjusted based on operator input. By integrating human experience into the automatic control system in real time, the system can quickly respond to sudden working conditions (such as foreign matter pressing in), enhance the robustness of the system and reduce the risk of abnormal strip breaking, and strengthen the real-time integration capability of operation intervention.

[0056] The operator intervention value self-learning value is dynamically updated through the self-learning function of the plate shape setting model, and is adjusted based on historical operation data to reduce the workload of the operator and improve the stability of the plate shape. For example, the scene of frequent manual intervention is learned and converted into an automatic compensation value, gradually reducing the workload of the subsequent operator, reducing the intensity of manual intervention, and improving the stability of the plate shape.

[0057] Step S140, the bending roll force set value is issued to the execution system at the entry time of the finishing mill. At the entry time of the finishing mill, the bending roll force set value is issued to the first-level basic automation system for execution.

[0058] Step 150, the actual crown and flatness of the strip steel at the exit of the finishing mill are collected, the deviation from the target value is obtained, and the set value of the next slab is corrected through self-learning.

[0059] The self-learning correction includes: dynamically updating the crown self-learning value based on the crown deviation, and dynamically updating the flatness self-learning value based on the flatness deviation; wherein the updated self-learning value is used for the calculation of the bending roll force set value of the next slab to respond to the continuous optimization of the plate shape control based on the measured deviation. Based on the measured crown deviation and flatness deviation at the exit of the finishing mill, the crown self-learning value and the flatness self-learning value are dynamically updated, and the updated value is used for the set calculation of the next slab. Through the closed-loop feedback mechanism, the system responds to the rolling fluctuations (such as temperature changes and roll consumption) in real time, gradually eliminates the deviation accumulation, realizes the continuous stability of the plate shape control, and realizes the continuous self-optimization capability of the plate shape control.

[0060] The updating of the crown self-learning value includes: assigning a crown bias to each stand, so that the crown self-learning assignment value of the first few stands is increased or decreased, and the value of the last few stands remains zero; the updating of the flatness self-learning value includes: assigning a flatness bias to the last stand, so that the flatness self-learning assignment value of the last stand is increased or decreased, and the value of other stands remains zero. The crown bias is only assigned to the first few stands (the value of the rear stands remains zero), and the flatness bias is only assigned to the last stand (the value of other stands remains zero). This assignment mode conforms to the law that the front stands dominate the crown control and the last stand dominates the flatness control in the hot continuous rolling process, avoids the plate shape oscillation caused by the interference of multiple stands, improves the control efficiency, and refines the stand division strategy of bias assignment.

[0061] The hot continuous rolling strip plate shape control method of the embodiment, through a dynamic specification identification mechanism and a multi-source learning value cooperative calculation strategy, first integrates the historical self-learning values (including crown, flatness, and operator intervention values) of the rolling mill model and the set values of the same specification strip in the stage of reading the slab data. When the steel grade / thickness / width specification is detected to change, the difference compensation item of the bending force of the SAPP and SAMP model tables is introduced into the basic superposition model (SAMP reference value+crown / flatness assignment value+operator intervention value), the non-linear interference of the specification mutation on the plate shape is quantified by using the dynamic adjustment coefficient, and the bending force setting accuracy of the first strip of the changed specification is directly improved. At the same time, the self-learning values are differentially assigned based on the position characteristics of the stands--the crown self-learning value focuses on the front stands to control the rolling stability, and the flatness self-learning value acts on the last stand to optimize the finished product quality, and the continuous accumulation of the operator intervention value forms a closed-loop experience library, and the setting value is accurately issued at the entry of the finishing mill. Finally, by collecting the deviation of the actual plate shape data and the target value at the outlet, the self-learning mechanism is triggered to correct the subsequent slab setting in real time, thereby systematically solving the core defects of the insufficient crown / flatness hit rate when changing the specification in the background technology, and simultaneously eliminating the strip wearing and steel clamping accidents caused by the frequent adjustment of the load by the operator and the suppression of the plate shape fluctuation between the stands.

[0062] As shown in FIG. Figure 2 In one embodiment, the present application provides a hot continuous rolling strip plate shape control method based on an expert system, which specifically comprises the following steps:

[0063] Firstly, the slab PDI (Pramary Data Input, referred to as PDI data) data is read, and the device and the related parameters of the finishing mill setting model are set. When the strip reaches the entry and the outlet pyrometer of the finishing mill, the measured values are extracted, and data preparation is made for the finishing mill setting model calculation.

[0064] The slab PDI data includes slab number, steel grade, slab specification, product target thickness, target crown, final rolling target temperature, and coiling target temperature.

[0065] The model parameters mainly include the rigidity of each stand rolling mill, the device parameters such as size, the self-learning value of convexity read from the rolling mill model table, the self-learning value of flatness, the self-learning value of the operator intervention value of the bending roll force of each stand, and the set value of the bending roll force of the previous piece, the set value of the bending roll force of the previous piece of the same steel grade, the same thickness, and the same width layer from the product self-learning model, and the like.

[0066] In the second step, the speed, temperature, loop angle and other parameters of each stand of the finishing mill are calculated according to the finishing mill model.

[0067] In the third step, the bending roll force set value of each stand is calculated according to the plate shape model.

[0068] Compared with the previous piece, if the specification is not changed:

[0069] BendSetup[i] = BendSamp[i] + BendCrownVern[i] + BendFlat[i] + BendOperator[i],

[0070] Compared with the previous piece, if the specification is changed:

[0071] BendSetup[i] = BendSamp[i] + (BendSapp[i] - BendSamp[i]) * coef[i] + BendCrownVern[i]

[0072] + BendFlat[i] + BendOperator[i],

[0073] wherein,

[0074] BendSetup: the bending roll force set value;

[0075] i: the stand of the finishing mill;

[0076] BendSamp[i]: the bending roll force value saved in the SAMP model table;

[0077] BendSapp[i]: the bending roll force value saved in the SAPP model table;

[0078] BendCrownVern[i]: the bending roll force allocated to each stand after self-learning of convexity, the bending roll force of the first several stands has a value, and the bending roll force of the last several stands is 0;

[0079] BendFlatVern[i]: the bending roll force allocated to each stand after self-learning of flatness, the bending roll force of the last stand has a value, and the bending roll force of the other stands is 0;

[0080] BendOperator[i]: the bending roll force of the operator intervention;

[0081] Coef[i]: Change the steel, change the thickness, change the width of the bending roll force change coefficient.

[0082] Fourthly, at the time of the finishing rolling entrance, the finishing rolling model setting result is issued to the first level basic automation execution (the first level basic automation is to execute the model setting value issued by the second level, and is sent to the equipment operation).

[0083] Fifthly, through the finishing rolling exit multifunctional instrument, the actual value of the strip and flatness at the finishing rolling exit is collected, and is compared with the designed target value, so that the deviation of the strip crown and flatness at the finishing rolling exit is obtained, and through the self-learning function of the shape setting model, the setting value of the next slab is corrected, so that the target crown and flatness are good.

[0084] According to the actual production data, the self-learning of the crown, the self-learning of the flatness, the self-learning of the operator bending roll intervention value, and the determination of the bending roll force setting value of each stand are combined with different steel, thickness and width in the embodiment; the actual value of the strip crown and flatness at the finishing rolling exit is collected, and is compared with the designed target value, so that the deviation of the strip crown and flatness at the finishing rolling exit is obtained, and through the self-learning function of the shape setting model, the bending roll force of the next slab is corrected, so that the target crown and flatness are good.

[0085] Through the hot continuous rolling strip shape control based on the expert system, the target crown and flatness hitting rate of the hot rolling strip are improved, and especially the good strip shape is maintained. Compared with the traditional method, the embodiment has the following positive effects:

[0086] Under the existing equipment conditions, the self-learning of the bending roll force intervention value, the self-learning of the crown and the self-learning of the flatness can effectively eliminate the influence of various interferences in the embodiment. The hitting rate of the crown and the flatness is effectively improved, especially the hitting rate of the crown and the flatness of the first strip of the first steel and the first specification. The problem of unstable strip threading and steel clamping caused by unstable interstand shape is effectively solved, and the workload of the operator is reduced.

[0087] The present application takes a certain hot continuous rolling 1250mm 8-stand hot rolling production line as an example for illustration as follows:

[0088] Firstly, the slab PDI data is read, the slab number is ap202506140035, the steel type is SAE1010, the slab specification is 200*910*10000mm, the target thickness after rough rolling is 33mm, the product target thickness is 1.73mm, the final rolling target temperature is 860℃, the coiling target temperature is 650℃, the equipment constant and the coiling temperature model related parameters are read, the actual value measured by the high temperature meter before the slab reaches the finishing rolling machine front entrance is 1100℃, and the actual value measured by the high temperature meter at the finishing rolling exit is 1012℃.

[0089] Second, according to the finishing model, the speed, temperature, loop angle and other parameters of each stand are calculated.

[0090] Third, according to the plate shape model, the bending force setting value of each stand is calculated.

[0091] Compared with the previous block, if the specification is not changed:

[0092] BendSetup[i] = BendSamp[i] + BendCrownVern[i] + BendFlat[i] + BendOperator[i],

[0093] Compared with the previous block, if the specification is changed:

[0094] BendSetup[i] = BendSamp[i] + (BendSapp[i] - BendSamp[i]) * coef[i] + BendCrownVern[i]

[0095] + BendFlat[i] + BendOperator[i],

[0096] Wherein,

[0097] BendSetup: Bending force setting value;

[0098] i: Finishing mill stand;

[0099] BendSamp[i]: Bending force value saved in SAMP model table;

[0100] BendSapp[i]: Bending force value saved in SAPP model table;

[0101] BendCrownVern[i]: Bending force allocated to each stand after crown self-learning, the bending force of the first few stands has a value, and the bending force of the last few stands is 0;

[0102] BendFlatVern[i]: Bending force allocated to each stand after flatness self-learning, the bending force of the last stand has a value, and the bending force of other stands is 0;

[0103] BendOperator[i]: Operator intervention bending force;

[0104] Coef[i]: Steel grade change, thickness change, width change bending force change coefficient.

[0105] F1 F2 F3 F4 F5 F6 F7 F8 BendSamp 853.3 836.2 781.6 832.3 866.5 854.5 843.1 781.9 BendSapp 963.2 921.1 906.5 780.5 702.6 654.3 754.8 760.8 BendCrownVern 82.3 51.5 38.5 19.7 0 0 0 0 BendFlat 0 0 0 0 0 0 0 50.5 BendOperator 0 0 71.8 0 0 0 0 0 BendSetup 968.57 913.17 929.37 836.46 817.33 794.44 816.61 826.07

[0106] The fourth step is to send the finishing model setting results to the first-level basic automation for execution at the finishing entrance (the first-level basic automation is to execute the model setting values ​​sent by the second level and send them to the equipment for operation).

[0107] The fifth step is to collect the actual values ​​of the strip and flatness at the finishing rolling exit through the multifunctional instrument at the finishing rolling exit, compare them with the designed target values, and obtain the deviation of the convexity and flatness of the strip at the finishing rolling exit. Through the self-learning function of the plate shape setting model, correct the setting value of the next slab to ensure the target convexity and flatness.

[0108] The hot-rolled strip shape control device provided by the present invention is described below. The hot-rolled strip shape control device described below and the hot-rolled strip shape control method described above can be referenced to each other.

[0109] like Figure 3 As shown, in one embodiment, a hot rolled strip shape control device includes a data reading module 310 , a parameter calculation module 320 , a calculation response module 330 , a data sending module 340 and a comparison and correction module 350 .

[0110] The data reading module 310 is used to read the slab PDI data, equipment parameters and model parameters. The model parameters include the convexity self-learning value, the straightness self-learning value, the operator intervention value self-learning value of the bending roll force of each stand and the bending roll force setting value of the previous strip in the rolling mill model table, as well as the bending roll force setting value of each stand of the previous strip of the same steel grade, thickness and width layer in the product self-learning model.

[0111] The parameter calculation module 320 is used to calculate the speed, temperature and thickness parameters of each finishing rolling stand based on the finishing rolling model.

[0112] The calculation response module 330 is used to calculate the bending roll force setting value of each frame according to the plate shape model, wherein, in response to the rolling specifications not being changed, the bending roll force reference value, the convexity self-learning distribution value, the straightness self-learning distribution value and the operator intervention value of the SAMP model table are superimposed to generate the bending roll force setting value; in response to the change of steel type, thickness or width specification, on the basis of the superposition of the specifications not being changed, the bending roll force difference between the SAPP model table and the SAMP model table is introduced and multiplied by the dynamic adjustment coefficient to generate the bending roll force setting value.

[0113] The data sending module 340 is used to send the bending roll force setting value to the execution system at the finishing rolling entrance moment.

[0114] The comparison and correction module 350 is used to collect the actual convexity and flatness of the finished strip at the exit, compare the deviation with the target value, and correct the set value of the next slab through self-learning.

[0115] Figure 4An example is shown in the schematic diagram of the physical structure of an electronic device, which can be a smart terminal, and the internal structure diagram can be as shown in Figure 4 The electronic device includes a processor, a memory and a network interface connected by a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the electronic device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement a hot continuous rolling strip shape control method, which includes:

[0116] Read slab PDI data, equipment parameters and model parameters, the model parameters include convexity self-learning values, flatness self-learning values, operator intervention values self-learning values of bending roll force of each stand in the mill model table, and the bending roll force setting values of the previous strip of each stand in the product self-learning model, which are in the same steel grade, the same thickness and the same width layer;

[0117] Calculate the speed, temperature and thickness parameters of each stand in the finishing mill based on the finishing mill model;

[0118] Calculate the bending roll force setting value of each stand according to the strip shape model, wherein, in response to not changing the rolling specification, the bending roll force reference value, the convexity self-learning distribution value, the flatness self-learning distribution value and the operator intervention value of the SAMP model table are superimposed to generate the bending roll force setting value; in response to changing the steel grade, thickness or width specification, on the basis of the superposition of the unchanged specification, the bending roll force difference value of the SAPP model table and the SAMP model table is introduced multiplied by a dynamic adjustment coefficient to generate the bending roll force setting value;

[0119] Issue the bending roll force setting value to the execution system at the finishing mill inlet time;

[0120] Collect the actual convexity and flatness of the strip at the finishing mill outlet, compare the target value to obtain the deviation, and correct the next slab setting value through self-learning.

[0121] Those skilled in the art can understand that Figure 4 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0122] On the other hand, the present application also provides a computer storage medium storing a computer program, which is executed by a processor to implement a hot continuous rolling strip shape control method, which includes:

[0123] reading slab PDI data, equipment parameters and model parameters, the model parameters including convexity self-learning values, flatness self-learning values, operator intervention value self-learning values of bending roll force of each stand in the rolling mill model table, and bending roll force setting values of each stand of the previous slab in the product self-learning model;

[0124] calculating speed, temperature and thickness parameters of each stand of the finishing mill based on the finishing mill model;

[0125] calculating bending roll force setting values of each stand according to the plate shape model, wherein, in response to not changing the rolling specification, superimposing the bending roll force reference value, the convexity self-learning distribution value, the flatness self-learning distribution value and the operator intervention value of the SAMP model table to generate the bending roll force setting value; in response to changing the steel grade, thickness or width specification, introducing the bending roll force difference value of the SAPP model table and the SAMP model table multiplied by a dynamic adjustment coefficient on the basis of the superposition of the unchanged specification to generate the bending roll force setting value;

[0126] issuing the bending roll force setting value to the execution system at the moment of the finishing mill inlet;

[0127] collecting actual convexity and flatness of the strip at the finishing mill outlet, comparing the target values to obtain deviations, and correcting the next slab setting value through self-learning.

[0128] In another aspect, a computer program product or computer program is provided, which includes computer instructions stored in a computer readable storage medium. A processor of an electronic device reads the computer instructions from the computer readable storage medium, and the processor implements a hot rolling strip shape control method when executing the computer instructions, the method comprising:

[0129] reading slab PDI data, equipment parameters and model parameters, the model parameters including convexity self-learning values, flatness self-learning values, operator intervention value self-learning values of bending roll force of each stand in the rolling mill model table, and bending roll force setting values of each stand of the previous slab in the product self-learning model;

[0130] calculating speed, temperature and thickness parameters of each stand of the finishing mill based on the finishing mill model;

[0131] According to the plate shape model, the bending roll force setting value of each stand is calculated, wherein, in response to the non-replacement rolling specification, the bending roll force reference value, the crown self-learning distribution value, the flatness self-learning distribution value and the operator intervention value of the SAMP model table are superimposed to generate the bending roll force setting value; in response to the replacement of the steel grade, the thickness or the width specification, on the basis of the superposition of the non-replacement specification, the bending roll force difference value of the SAPP model table and the SAMP model table is introduced to generate the bending roll force setting value multiplied by the dynamic adjustment coefficient;

[0132] The bending roll force setting value is issued to the execution system at the finishing rolling entrance moment;

[0133] The actual crown and flatness of the strip steel at the finishing rolling exit are collected, the deviation is obtained by comparing the target value, and the next slab setting value is corrected through self-learning.

[0134] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware, and the computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments of each method. Any reference to memory, storage, database or other medium used in each embodiment provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory.

[0135] As an illustration but not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0136] Each technical feature of the above embodiments can be combined arbitrarily, and in order to make the description concise, not all possible combinations of each technical feature in the above embodiments are described, however, as long as the combination of these technical features does not exist contradictory, it should be considered as the scope of the present application.

[0137] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the present application. It should be noted that for ordinary skilled persons in the art, without departing from the concept of the present application, several modifications and improvements can be made, which are within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A method of shape control of hot strip, characterized by, The method comprises: reading slab PDI data, equipment parameters and model parameters, the model parameters including camber self-learning values in a rolling mill model table, flatness self-learning values, operator intervention value self-learning values of each stand bending force operation and previous strip bending force set values, and product self-learning model different steel, thickness, width layer of the previous strip each stand bending force set value; calculating the speed, temperature and thickness parameters of each stand of the finishing mill based on the finishing mill model; calculating the bending force set value of each stand according to the shape model, wherein, in response to no change of rolling specifications, the bending force reference value, camber self-learning distribution value, flatness self-learning distribution value and operator intervention value of the SAMP model table are superimposed to generate the bending force set value; in response to the change of steel, thickness or width specifications, the bending force difference value of the SAPP model table and the SAMP model table is introduced on the basis of the superposition of no change of specifications multiplied by a dynamic adjustment coefficient to generate the bending force set value; issuing the bending force set value to the execution system at the entry moment of the finishing mill; collecting the actual camber and flatness of the strip at the exit of the finishing mill, comparing the target value to obtain the deviation, and correcting the next slab set value by self-learning.

2. The hot strip shape control method according to claim 1, characterized by, The superposition in response to no change of rolling specifications comprises: directly adding the bending force reference value, camber self-learning distribution value, flatness self-learning distribution value and operator intervention value of the SAMP model table to generate the bending force set value; wherein the camber self-learning distribution value is dynamically distributed based on the camber deviation of the previous strip, so that the first few stands have non-zero values and the last few stands have zero values, and the flatness self-learning distribution value is dynamically distributed based on the flatness deviation of the previous strip, so that the last stand has a non-zero value and the other stands have zero values.

3. The method of shape control of hot rolled steel strip according to claim 1, characterized in that, The generation of the bending force set value comprises: calculating the bending force difference value of the SAPP model table and the SAMP model table, multiplying by a dynamic adjustment coefficient, and superimposing on the basis of the superposition of no change of specifications, wherein the dynamic adjustment coefficient is dynamically determined based on the change of steel, thickness or width to adjust the bending force set value in response to the change of specifications; The dynamic adjustment coefficient coef[i] is dynamically adjusted according to different stand positions, wherein i represents the stand number, and the coefficient is generated by an expert system based on historical production data to optimize the camber and flatness hit rate of the first strip after the change of specifications.

4. The method of shape control of hot rolled steel strip according to claim 1, characterized in that, The self-learning correction comprises: dynamically updating the camber self-learning value based on the camber deviation and dynamically updating the flatness self-learning value based on the flatness deviation, wherein the updated self-learning value is used for the calculation of the bending force set value of the next slab to continuously optimize the shape control in response to the measured deviation.

5. The hot strip shape control method according to claim 4, characterized by, The updating of the camber self-learning value comprises: distributing the camber deviation to each stand so that the camber self-learning distribution value of the first few stands increases or decreases, and the value of the last few stands remains zero; The updating of the flatness self-learning value comprises: distributing the flatness deviation to the last stand so that the flatness self-learning distribution value of the last stand increases or decreases, and the value of the other stands remains zero.

6. The method of shape control of hot strip steel according to claim 1, characterized in that, The self-learning value of the operator intervention value is dynamically adjusted based on the operator input and is directly superimposed on the bending roll force setting value calculated according to the plate shape model to optimize the plate shape control in response to real-time operation intervention, wherein the operator intervention value is included in the calculation when changing the specification or not.

7. The hot strip shape control method according to claim 6, characterized by, The operator intervention value self-learning value is dynamically updated through the self-learning function of the plate shape setting model and is adjusted based on historical operation data to reduce the operator's workload and improve plate shape stability.

8. The method of shape control of hot strip steel according to claim 1, characterized in that, The convexity self-learning value, flatness self-learning value and operator intervention value self-learning value are dynamically generated based on the production data of the previous strip, and stored in the rolling mill model table or product self-learning model, and are updated through expert system learning to respond to changes in the continuous rolling process.

9. A hot rolled strip shape control device, characterized in that: The device comprises: A data reading module is used to read slab PDI data, equipment parameters, and model parameters. The model parameters include the crown self-learning value, flatness self-learning value, operator intervention value self-learning value of each stand bending roll force, and the previous strip bending roll force setting value in the rolling mill model table, as well as the previous strip bending roll force setting value of each stand for the same steel grade, thickness, and width layer in the product self-learning model; Parameter calculation module, used to calculate the speed, temperature and thickness parameters of each finishing rolling stand based on the finishing rolling model; A calculation response module is used to calculate the bending roll force set value of each stand based on the plate shape model. In response to the unchanged rolling specifications, the bending roll force reference value, the crown self-learning distribution value, the flatness self-learning distribution value, and the operator intervention value in the SAMP model table are superimposed to generate the bending roll force set value. In response to the change of steel type, thickness, or width specifications, the bending roll force difference between the SAPP model table and the SAMP model table is introduced on the basis of the superposition of the unchanged specifications and multiplied by the dynamic adjustment coefficient to generate the bending roll force set value. A data sending module is used to send the bending roll force setting value to the execution system at the finishing rolling entrance; The comparison and correction module is used to collect the actual convexity and flatness of the finished rolled strip at the exit, compare the deviation with the target value, and correct the set value of the next slab through self-learning.

10. An electronic device comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the method for controlling the shape of a hot-rolled strip steel is implemented according to any one of claims 1 to 8.

Images