CN110373510A - A kind of narrow window control method and system of steel quality based on bimodal switching - Google Patents
A kind of narrow window control method and system of steel quality based on bimodal switching Download PDFInfo
- Publication number
- CN110373510A CN110373510A CN201910642694.1A CN201910642694A CN110373510A CN 110373510 A CN110373510 A CN 110373510A CN 201910642694 A CN201910642694 A CN 201910642694A CN 110373510 A CN110373510 A CN 110373510A
- Authority
- CN
- China
- Prior art keywords
- control
- time
- representing
- narrow window
- molten steel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 94
- 229910000831 Steel Inorganic materials 0.000 title claims abstract description 78
- 239000010959 steel Substances 0.000 title claims abstract description 78
- 230000002902 bimodal effect Effects 0.000 title claims abstract description 29
- 238000009628 steelmaking Methods 0.000 claims abstract description 65
- 238000003723 Smelting Methods 0.000 claims abstract description 36
- 230000007246 mechanism Effects 0.000 claims abstract description 34
- 238000005457 optimization Methods 0.000 claims abstract description 20
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 19
- 238000012706 support-vector machine Methods 0.000 claims abstract description 12
- 241000255581 Drosophila <fruit fly, genus> Species 0.000 claims abstract description 9
- 238000007664 blowing Methods 0.000 claims description 55
- 239000002826 coolant Substances 0.000 claims description 52
- 230000008569 process Effects 0.000 claims description 36
- 238000010168 coupling process Methods 0.000 claims description 32
- 230000008878 coupling Effects 0.000 claims description 31
- 238000005859 coupling reaction Methods 0.000 claims description 31
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 claims description 30
- 229910052799 carbon Inorganic materials 0.000 claims description 30
- 229910052760 oxygen Inorganic materials 0.000 claims description 30
- 239000001301 oxygen Substances 0.000 claims description 30
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 29
- 230000006870 function Effects 0.000 claims description 15
- 230000003044 adaptive effect Effects 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 9
- URWAJWIAIPFPJE-YFMIWBNJSA-N sisomycin Chemical compound O1C[C@@](O)(C)[C@H](NC)[C@@H](O)[C@H]1O[C@@H]1[C@@H](O)[C@H](O[C@@H]2[C@@H](CC=C(CN)O2)N)[C@@H](N)C[C@H]1N URWAJWIAIPFPJE-YFMIWBNJSA-N 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 230000000670 limiting effect Effects 0.000 claims description 5
- 229910052770 Uranium Inorganic materials 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 7
- 238000005265 energy consumption Methods 0.000 abstract description 5
- 238000003908 quality control method Methods 0.000 abstract description 5
- 239000013589 supplement Substances 0.000 description 12
- 241000255588 Tephritidae Species 0.000 description 11
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 8
- 239000000047 product Substances 0.000 description 8
- 238000005261 decarburization Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 7
- 238000001514 detection method Methods 0.000 description 6
- 238000013528 artificial neural network Methods 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 230000001502 supplementing effect Effects 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 229910052742 iron Inorganic materials 0.000 description 4
- 230000009467 reduction Effects 0.000 description 4
- 239000000523 sample Substances 0.000 description 4
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000013145 classification model Methods 0.000 description 3
- 238000006477 desulfuration reaction Methods 0.000 description 3
- 238000004134 energy conservation Methods 0.000 description 3
- 239000003546 flue gas Substances 0.000 description 3
- 238000011478 gradient descent method Methods 0.000 description 3
- 230000036961 partial effect Effects 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 241000894007 species Species 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 2
- NINIDFKCEFEMDL-UHFFFAOYSA-N Sulfur Chemical compound [S] NINIDFKCEFEMDL-UHFFFAOYSA-N 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000011217 control strategy Methods 0.000 description 2
- 230000001808 coupling effect Effects 0.000 description 2
- 230000023556 desulfurization Effects 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 239000007789 gas Substances 0.000 description 2
- 238000010438 heat treatment Methods 0.000 description 2
- 238000012886 linear function Methods 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 229910052698 phosphorus Inorganic materials 0.000 description 2
- 239000011574 phosphorus Substances 0.000 description 2
- 230000002829 reductive effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000004904 shortening Methods 0.000 description 2
- 239000002893 slag Substances 0.000 description 2
- 239000002436 steel type Substances 0.000 description 2
- 229910052717 sulfur Inorganic materials 0.000 description 2
- 239000011593 sulfur Substances 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 108010046685 Rho Factor Proteins 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000005275 alloying Methods 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000007872 degassing Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000004836 empirical method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000017525 heat dissipation Effects 0.000 description 1
- 238000000265 homogenisation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000005304 joining Methods 0.000 description 1
- 238000012933 kinetic analysis Methods 0.000 description 1
- PYLLWONICXJARP-UHFFFAOYSA-N manganese silicon Chemical compound [Si].[Mn] PYLLWONICXJARP-UHFFFAOYSA-N 0.000 description 1
- 229910052751 metal Inorganic materials 0.000 description 1
- 239000002184 metal Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003647 oxidation Effects 0.000 description 1
- 238000007254 oxidation reaction Methods 0.000 description 1
- 238000004886 process control Methods 0.000 description 1
- 238000011112 process operation Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000010079 rubber tapping Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000012932 thermodynamic analysis Methods 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Classifications
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C5/00—Manufacture of carbon-steel, e.g. plain mild steel, medium carbon steel or cast steel or stainless steel
- C21C5/28—Manufacture of steel in the converter
- C21C5/30—Regulating or controlling the blowing
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21C—PROCESSING OF PIG-IRON, e.g. REFINING, MANUFACTURE OF WROUGHT-IRON OR STEEL; TREATMENT IN MOLTEN STATE OF FERROUS ALLOYS
- C21C2300/00—Process aspects
- C21C2300/06—Modeling of the process, e.g. for control purposes; CII
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P10/00—Technologies related to metal processing
- Y02P10/25—Process efficiency
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Materials Engineering (AREA)
- Metallurgy (AREA)
- Organic Chemistry (AREA)
- Carbon Steel Or Casting Steel Manufacturing (AREA)
Abstract
The invention discloses a kind of narrow window control method and system of steel quality based on bimodal switching, by according to execute-in-place system and smelting steel grade requirement, obtain the narrow window section of converter terminal technique, determine that pneumatic steelmaking two blows the modality decision mechanism of stage control and based on modality decision mechanism based on the twin support vector machine classification method of drosophila optimization algorithm, stage progress MFA control is blown to pneumatic steelmaking two, solves prior art technical problem bad to steel quality control effect in convertor steelmaking process, it realizes according to different mode, using different control programs, and narrow window concept is introduced into steelmaking process for the first time, and it ensure that low energy consumption and high quality in converter smelting based on narrow window, to improve product quality, realize energy-saving and emission-reduction.
Description
Technical Field
The invention relates to the technical field of converter steelmaking, in particular to a molten steel quality narrow window control method and system based on bimodal switching.
Background
The iron and steel industry is an important basic raw material industry and belongs to the national economic prop industry. At present, the steel industry faces important common problems of heavy load of environment, low resource utilization rate, product homogenization, low value and the like, so that the research of a molten steel quality narrow window control method based on bimodal switching is a necessary way for realizing the intellectualization of steel process.
Converter steelmaking is the mainstream steelmaking method in the world at present, and 60% of steel in the world and 90% of steel in China are smelted by converters at present. The converter steelmaking in China mainly adopts a top-bottom combined blowing method, namely a converter smelting method for blowing oxygen from the top of a converter and blowing different gases from the bottom of the converter simultaneously. The main tasks can be summarized as four-step removal (desulfuration, deoxidation, dephosphorization and decarburization), two-step removal (degassing and impurity removal) and two-step adjustment (component adjustment and temperature adjustment), and the main adopted technical means are as follows: oxygen supply, temperature rise, slag formation, addition of a deoxidizer and alloying.
The molten steel quality end point control mainly refers to end point component and end point temperature control, and because the processes of desulfurization and dephosphorization are more complicated than the decarburization process, sulfur and phosphorus are always removed to the range required by the end point as early as possible. Therefore, the molten steel quality end point control is simplified into carbon content and temperature control. Inaccuracy in endpoint control can cause a number of hazards: the metal consumption is increased, the blowing time is prolonged, the service life of a furnace lining is reduced, the normal production order of a workshop is disturbed, and the quality of molten steel is further influenced. Meanwhile, a considerable number of domestic small and medium-sized converters are still at the experience smelting and single static control level, so that the control level for improving the quality of molten steel is urgently needed, and the stable control of the narrow window has important significance for improving the product quality, shortening the smelting time, improving the smelting rate and improving the economic benefit of enterprises.
The existing converter later-stage dynamic control technology can be roughly divided into a mechanism control model, an experience control model and an intelligent control model. The mechanism control model has two starting points, namely heat conservation and mass conservation at a macroscopic angle, thermodynamic analysis and kinetic analysis at a microscopic angle, but the derivation process of the model is simplified and assumed to different degrees, so that the model precision is low. Most of the theoretical bases of the empirical control model are decarburization rate curves, which are considered from the same empirical trend, and the universality is poor. The intelligent control model is established based on statistical data of an actual steelmaking process, adopts a black box principle, only considers the behavior of converter steelmaking input and output parameters, has unique advantages on the premise of incomplete mechanism research, but ignores process factors of converter smelting.
CN103882176A on-line dynamic optimal control method for converter steelmaking process based on data driving
Application No. 201410114943.7 application No. 2014.03.25
Application publication No. CN103882176A application publication No. 2014.06.25
The patent provides an online dynamic optimal control method for a converter steelmaking process based on data driving. Establishing an elite heat database, selecting a heat data set matched with the current smelting process information, determining a reference curve, determining an optimized set value set of each operation variable by optimally controlling the current prediction model to have minimum deviation with the reference curve, realizing real-time online control, facilitating an operator to select an operation set value according to actual working conditions, and improving the production efficiency of a steelmaking factory.
The converter steelmaking on-line dynamic control system adopted by the patent comprises a sublance, a spectrum analyzer, a flue gas analyzer and a throwing probe, but the hardware conditions of a domestic equivalent number of small and medium-sized converter steelmaking systems are immature.
CN108570528A control method for improving steelmaking converter blowing end point temperature
Application No. 201810355320.7 application No. 2018.04.19
Application publication No. CN108570528A application publication No. 2018.09.25
The patent proposes a control method for increasing the converting end point temperature of a steel converter. According to the metallurgical dynamics, thermodynamics and reaction engineering principles, different stages of converter blowing are calculated in a distributed mode, and the corresponding gun positions, the corresponding slag charge adding amount and the corresponding oxygen supply strength of a silicon-manganese oxidation period and a decarburization period are determined, so that the consistency of the slagging reaction and the molten iron reaction process in the converter steelmaking process is ensured, the heat dissipation coefficient and the heating rate in the smelting process are stabilized, and the blowing end point temperature control capability is improved.
The method mainly adjusts a charging system, an oxygen supply system, a slagging system and a temperature system in the smelting process of the converter, is based on a mechanism model and empirical operation, has certain simplification and assumption and has weak disturbance resistance.
CN108647407A method for analyzing and determining carbon in converter steelmaking flue gas
Application No. 201810372680.8 application No. 2018.04.24
Application publication No. CN108647407A application publication No. 2018.10.12
The patent provides a method for analyzing and determining carbon in converter steelmaking flue gas. The molten pool uniformity concept is fused on the basis of exponential carbon determination, the behavior characterization of decarburization in the converter steelmaking process is perfected, the calculation precision is improved, the problem that the initial carbon content is difficult to directly determine based on a carbon element mass conservation model is solved, meanwhile, an end point carbon content curve fitting model is improved, and the influence of process operation parameters such as oxygen lance position, bottom blowing gas flow, top blowing oxygen flow and the like on decarburization rate in the smelting process is considered.
The influence of complex interference factors such as furnace age and splashing is not considered, the model precision is still limited, and the use is limited to a certain extent.
CN105974896A converter steelmaking optimization control system and method based on information physical fusion
Application No. 201610394984.5 application No. 2016.06.07
Application publication No. CN105974896A application publication No. 2016.09.28
The patent provides a converter steelmaking optimal control system and method based on information physical fusion. The control and optimization of converter smelting are organically combined together, the stable operation of converter steelmaking is easily optimized in real time by using a complex optimization algorithm, the sampling time of an upper computer process control module and the cloud server forms optimization positive feedback, the optimization effect has self-adaptability, the optimization efficiency is improved, the stability of a control system is enhanced, and meanwhile, a monitoring layer is added, so that the result can be automatically adjusted, and the production efficiency of the iron and steel industry is improved.
The upper computer sampling time, the operation optimization sampling time, the optimization data transmission time and the historical data transmission time are set, and when interference exists, the time quantum cannot be dynamically adjusted in time.
In summary, the prior art has the corresponding defects, and therefore the invention is provided.
Disclosure of Invention
The invention provides a molten steel quality narrow window control method and system based on bimodal switching, and solves the technical problem that the molten steel quality control effect is poor in the converter steelmaking process in the prior art.
In order to solve the technical problem, the molten steel quality narrow window control method based on bimodal switching provided by the invention comprises the following steps:
acquiring a narrow window interval of a converter end point process according to a field operation system and smelting steel seed requirements;
determining a modal decision mechanism of converter steelmaking second blowing stage control by a twin support vector machine classification method based on a drosophila optimization algorithm;
and performing self-adaptive control on the converter steelmaking second blowing stage based on a modal decision mechanism.
Further, the classification decision function of the modal decision mechanism is specifically:
Class i＝min|K(xT,CT)ωj+bj|，
wherein i ═ 1, -1, Class +1 denotes the addition of coolants, Class-1 denotes the absence of coolants, K denotes the gaussian kernel function, x denotesTRepresenting input samples, CTTwo sample sets representing the addition of coolant and the absence of coolant, j ═ 1,2, w1Weight, w, representing hyperplane 1 solved for the set of added coolant class samples2Representing the weight of the hyperplane 2 solved for the set of samples without the addition of a coolant class, b1Represents the deviation of the hyperplane 1, b2Indicating the deviation of the hyperplane 2.
Further, based on a modal decision mechanism, the self-adaptive control of the converter steelmaking second blowing stage comprises the following steps:
based on a modal decision mechanism, classifying control types of a converter steelmaking second blowing stage, and specifically classifying the control types into a single-variable control type and a double-variable coupling control type, wherein the single-variable control type specifically means that only oxygen quantity supplementary blowing control needs to be carried out on the converter steelmaking second blowing stage, and the double-variable coupling control type specifically means that oxygen quantity supplementary blowing control and coolant addition control need to be carried out on the converter steelmaking second blowing stage;
aiming at the single variable control type, SISO model-free self-adaptive control is adopted;
and aiming at the bivariate coupling control type, MIMO model-free self-adaptive decoupling control is adopted.
Further, for a single variable control type, the controller adopting SISO model-free adaptive control is specifically designed as follows:
wherein u (t) represents the input of the system at the time t, u (t-1) represents the input of the system at the time t-1, Deltau (t-1) represents the input difference value of the system at the time t and the time t-1, and y*(t +1) represents a true value at the t +1 moment, y (t) represents the output of the system at the t moment, Δ y (t) represents the output difference value between the system at the t moment and the t-1 moment, α represents a forgetting factor, and α represents the output difference value1A value, alpha, representing a forgetting factor when the control system enters a steady state2The forgetting factor representing when the control system is in motion calculates an intermediate value,an online estimate representing the pseudo-partial derivative of the system at time t,an online estimate representing the pseudo-partial derivative of the system at time t-1,represents the on-line estimated value of the pseudo-partial derivative of the system at time t ═ 1, p represents the sequence of steps in u (t) calculation, and 0 < p ≦ 1, and η represents the parameterEstimating a time step sequence, wherein eta is more than 0 and less than or equal to 1, c represents the adjustment rate of a forgetting factor, U represents a judgment condition, xi represents a threshold value of the judgment condition U, lambda represents a weight coefficient and is used for limiting the change of a control input quantity, lambda is more than or equal to 0, and mu represents a parameterThe weight coefficient when estimating, and mu is more than or equal to 0, epsilon represents a small enough positive number.
Further, for a bivariate coupling control type, a controller adopting MIMO model-free adaptive decoupling control is specifically designed as follows:
wherein i is 1,2, ui(t) represents the value of the ith input at time t, ui(t-1) represents the value of the ith input at time t-1, Δ ui(t-1) represents the difference between the ith input at time t and time t-1, yi(t) represents the value of the ith output at time t, Δ yi(t) represents the difference between the ith output at time t and time t-1,which represents the true value at time t +1,an estimate representing the coupling between the control systems at time t,an estimate representing the coupling between the control systems at time t-1,an estimate of the time-varying parameter representing the ith input to the ith output at time t,a time-varying parameter estimate representing the ith input to the ith output at time t-1,showing the time-varying parameter estimated value of the ith input corresponding to the ith output at the moment 1, alpha is a forgetting factor, alpha1A value, alpha, representing a forgetting factor when the control system enters a steady state2Representing the calculation intermediate value of the forgetting factor when the control system is in a dynamic state, U representing a judgment condition, xi representing the threshold value of the judgment condition U, c representing the adjustment rate of the forgetting factor, rho representing the step length sequence of U (t) in the calculation, and 0 < rho ≦ 1, eta representing a parameterEstimating a time step length sequence, wherein eta is more than 0 and less than or equal to 1, lambda represents a weight coefficient and is used for limiting the change of the control input quantity, lambda is more than or equal to 0, and mu represents a parameterThe weight coefficient when estimating, and mu is more than or equal to 0, epsilon represents a small enough positive number.
Further, the narrow window area of the converter end point process specifically comprises the following steps:
the narrow window interval of the molten steel end point temperature is 1630-1660 ℃, and the interval of the molten steel end point carbon content is 0.06-0.08%.
The invention provides a molten steel quality narrow window control system based on bimodal switching, which comprises:
the device comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein the processor realizes the steps of the molten steel quality narrow-window control method based on the bimodal switching when executing the computer program.
Compared with the prior art, the invention has the advantages that:
according to the molten steel quality narrow window control method and system based on bimodal switching, the narrow window interval of the converter end point process is obtained according to the requirements of a field operation system and a smelting steel type, the mode decision mechanism of the converter steelmaking second blowing stage control is determined based on the twin support vector machine classification method of the drosophila optimization algorithm, model-free adaptive control is carried out on the converter steelmaking second blowing stage based on the mode decision mechanism, the technical problem that the molten steel quality control effect is poor in the converter steelmaking process in the prior art is solved, different control schemes are adopted according to different modes, the narrow window concept is introduced into the steelmaking process for the first time, and low energy consumption and high quality in converter smelting are guaranteed based on the narrow window, so that the product quality is improved, and energy conservation and emission reduction are realized.
Drawings
FIG. 1 is a flow chart of a molten steel quality narrow window control method based on bimodal switching according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a molten steel quality narrow window control method based on bimodal switching according to a second embodiment of the present invention;
FIG. 3 is a coolant addition and non-coolant profile for different heats of example two of the present invention;
FIG. 4 is a block diagram of the overall flow of the modal decision mechanism according to the second embodiment of the present invention;
FIG. 5 is a block diagram of a converter endpoint coupling control system according to a second embodiment of the present invention;
FIG. 6 is a graph showing the control result of the carbon content at the end point of the converter in the second embodiment of the present invention;
FIG. 7 is a graph showing the results of controlling the endpoint temperature of the converter in the second embodiment of the present invention;
FIG. 8 is a flowchart of a molten steel quality narrow window control method based on bimodal switching according to a third embodiment of the present invention;
FIG. 9 is a block diagram of a molten steel quality narrow window control system based on bimodal switching according to an embodiment of the present invention.
Reference numerals:
10. a memory; 20. a processor.
Detailed Description
In order to facilitate an understanding of the invention, the invention will be described more fully and in detail below with reference to the accompanying drawings and preferred embodiments, but the scope of the invention is not limited to the specific embodiments below.
The embodiments of the invention will be described in detail below with reference to the drawings, but the invention can be implemented in many different ways as defined and covered by the claims.
Example one
Referring to fig. 1, a molten steel quality narrow window control method based on bimodal switching according to an embodiment of the present invention includes:
s101, acquiring a narrow window interval of a converter terminal process according to a field operation system and smelting steel seed requirements;
s102, determining a modal decision mechanism for controlling a converter steelmaking second blowing stage by a twin support vector machine classification method based on a drosophila optimization algorithm;
and S103, performing self-adaptive control on the converter steelmaking second blowing stage based on a modal decision mechanism.
According to the molten steel quality narrow window control method based on bimodal switching, provided by the embodiment of the invention, the narrow window interval of the converter terminal process is obtained according to the field operation system and the smelting steel variety requirement, the twin support vector machine classification method based on the drosophila optimization algorithm determines the mode decision mechanism of the converter steelmaking second blowing stage control and performs model-free adaptive control on the converter steelmaking second blowing stage based on the mode decision mechanism, the technical problem that the molten steel quality control effect in the converter steelmaking process is poor in the prior art is solved, different control schemes are adopted according to different modes, the narrow window concept is introduced into the steelmaking process for the first time, and the low energy consumption and high quality in the converter smelting are ensured based on the narrow window, so that the product quality is improved, and the energy conservation and emission reduction are realized.
Example two
Referring to fig. 2, a molten steel quality narrow window control method based on bimodal switching according to a second embodiment of the present invention includes:
and step S201, acquiring a narrow window interval of the converter terminal process according to the field operation system and the smelting steel seed requirement.
Because the current steel industry is more and more competitive, narrow window control needs to be carried out on molten steel components at the end point of converter steelmaking in order to optimize the product structure and improve the product quality, and because the processes of desulfurization and dephosphorization are complex in the smelting process of the converter compared with decarburization, sulfur and phosphorus are always removed to the range required by the end point as far as possible in advance, so that the narrow window control of the molten steel components at the end point is simplified into narrow window control of the molten steel end point temperature and narrow window control of the molten steel end point carbon content. In the scheme, by consulting the converter steelmaking operation system and the smelting steel species requirement of field operation workers, narrow window regions for obtaining the HRB400E steel species are respectively as follows: the narrow window interval of the molten steel end point temperature is 1630-1660 ℃, and the interval of the molten steel end point carbon content is 0.06-0.08%.
And S202, determining a mode decision mechanism for controlling the converter steelmaking second blowing stage based on a twin support vector machine classification method of a drosophila optimization algorithm.
In the converter steelmaking process, a first detection point of a sublance is taken as a demarcation point, the whole smelting process is divided into two stages, and a main blowing stage is adopted before sublance detection; and determining the amount of added coolant and the amount of oxygen supplement blowing in the second blowing stage according to the information of the molten steel temperature and the carbon content detected by the sublance for the first time, so as to adjust the temperature and the carbon content in a converter molten pool and enable the temperature and the carbon content to reach the narrow process window interval.
The coolant added in the second blowing stage in the converter steelmaking process is mainly used for adjusting the temperature, when the temperature is higher and the carbon content is not low in the converter smelting process, a certain amount of coolant needs to be added, and the temperature of the molten steel at the end point is reduced by utilizing the physical heat consumed by heating the coolant and the chemical heat consumed by the chemical reaction generated by the coolant. However, when data of 2018 of a certain steel plant from 1 month to 4 months are counted, it is found that the coolant is not required to be added for all the heats, and there are cases of adding the coolant and not adding the coolant, as shown in table 1. Therefore, before dynamic control of the second blowing phase is performed, a mode decision mechanism needs to be determined to determine whether coolant needs to be added.
TABLE 1
| Categories | Adding a coolant | Without addition of coolant |
| Frequency of occurrence | 668 | 332 |
FIG. 3 is a diagram showing the coolant adding and coolant not adding profiles of the sublance for the first detection of different carbon contents and temperatures in different heats in the actual smelting of the converter. As can be seen from fig. 3, there are distinct regions in the cases of adding coolant and not adding coolant, and since the boundary between the regions is not distinct and cannot be simply divided by a straight line, the embodiment of the present invention determines the mode decision mechanism by using the twin support vector machine classification method based on the drosophila optimization algorithm.
Under the nonlinear condition of the upper graph, the converter steelmaking data is classified twice by using a twin support vector machine, the data is mapped from a low-dimensional space to a high-dimensional space by introducing a kernel function, and the original problem of the nonlinear twin support vector machine is represented as follows:
wherein, c1,c2> 0 are two penalty parameters, e1And e2Is the unit column vector of when dimension, ξ is the relaxation vector, C ═ A B]T，CTDenotes the transpose of C, K denotes the kernel function, w1Weight, w, representing hyperplane 1 solved for the set of added coolant class samples2Representing the weight of the hyperplane 2 solved for the set of samples without the addition of a coolant class, b1Represents the deviation of the hyperplane 1, b2Indicating the deviation of the hyperplane 2. The two hyperplanes are represented as:
f1(x):K(xT,CT)ω1+b1＝0 (3)
f2(x):K(xT,CT)ω2+b2＝0 (4)
the classification decision function is:
Class i＝min|K(xT,CT)ωj+bj| (5)
wherein i ═ 1, -1, Class +1 denotes the addition of coolants, Class-1 denotes the absence of coolants, K denotes the gaussian kernel function, x denotesTRepresenting input samples, CTTwo sample sets representing the addition of coolant and the absence of coolant, j ═ 1,2, w1Weight, w, representing hyperplane 1 solved for the set of added coolant class samples2Representing the weight of the hyperplane 2 solved for the set of samples without the addition of a coolant class, b1Representing the deviation of said hyperplane 1, b2Representing the deviation of said hyperplane 2.
A common method for TWSVM parameter selection is an empirical method, certain limitations exist, classification is inaccurate, an FOA (fruit fly optimization algorithm) is adopted in the scheme to find an optimal parameter combination, and the specific classification steps are as follows:
step 1: initializing population scale G, iteration number N, fruit fly population position range LR and fruit fly single flight range FR, and each individual initial position of fruit fly (X)axis，Yaxis)
Xaxis＝rand(LR) (6)
Yaxis＝rand(LR) (7)
Step 2: endowing each fruit fly with random flight direction and distance when searching food by smell
Xi＝Xaxis+rand(FR) (8)
Yi＝Yaxis+rand(FR) (9)
Step3 calculating the distance Dist from the individual fruit fly to the origin because the food position is unknowniThen, the taste concentration judgment value S is calculatedi
Si＝1/Disti (11)
Step 4: selecting the fruit fly with the best taste concentration value in the current fruit fly group, and recording the corresponding concentration value and position
Smelli＝fitness(Si) (12)
[bestSmell,bestIndex]＝min(Smell) (13)
The fitness represents a taste concentration judgment function, and the scheme represents the accuracy of classification of whether the coolant is added or not when solving by using FOA;
step 5: retaining the optimal taste concentration value and the position of the fruit fly to which other fruit flies fly by vision
Smellbest＝bestSmell (14)
Xaxis＝X(bestIndex) (15)
Yaxis＝Y(bestIndex) (16)
Step 6: performing iterative optimization within the maximum iteration number, repeating Step2-4, judging whether the taste concentration is better than the taste concentration retained in the previous time, and if so, executing Step5
The scheme specifically selects the Gaussian kernel function as follows:
K(x,y)＝exp(-||x-y||2/2σ2) (17)
and S203, classifying control types of the converter steelmaking second blowing stage based on a modal decision mechanism, wherein the control types are specifically divided into a single-variable control type and a double-variable coupling control type, the single-variable control type specifically means that only oxygen supplement blowing control needs to be carried out on the converter steelmaking second blowing stage, and the double-variable coupling control type specifically means that oxygen supplement blowing control and coolant addition control need to be carried out on the converter steelmaking second blowing stage.
Twin support vector machine classification model input variables based on a drosophila optimization algorithm are respectively the molten pool carbon content at the first detection point of the sublance, the molten pool temperature at the first detection point of the sublance, the carbon content at the target end point of the molten steel process and the target end point temperature of the molten steel process; the output variables are respectively coolant added and coolant not added, corresponding to the two modalities of the modality decision mechanism: and controlling the oxygen supplementing blowing amount, and controlling the oxygen supplementing blowing amount and adding a coolant in a coupling manner. As shown in fig. 4.
And step S204, aiming at the single variable control type, adopting SISO model-free self-adaptive control.
With the increase of complexity in the industrial production process, a mathematical model of a controlled object generally cannot be accurately obtained, and the problems of low accuracy and the like are often caused by a plurality of simplifications in the actual engineering application of a model-based control method and a model-based control theory. On the other hand, in the standard model-free control method, the PID controller is widely applied to system control, but the PID control has a not ideal effect and does not have learning capability in the system control such as strong nonlinearity, time-varying property, uncertainty and the like, and cannot adapt to the change of the system structure in time. Therefore, the embodiment of the invention adopts a model-free self-adaptive control method with variable forgetting factors.
1) Description of converter end point dynamic oxygen quantity compensation blowing control problem
Through the division of the mode decision mechanism, the dynamic control process at the last stage of the converter has a mode: the Single-loop control mode of the oxygen-supplementing-blowing amount control, namely the Single-loop control mode that the end point temperature and the end point carbon content of the converter respectively change along with the oxygen-supplementing-blowing amount, can be used as the following general nonlinear Single-Input Single-Output (Single-Input Single-Output) systems:
y(t+1)＝f(y(t),y(t-1),…,y(t-ny),u(t),u(t-1),…,u(t-nu)) (18)
in the above formula, y (t) and u (t) are respectively the output and input of the system at the time t, i.e. the end point temperature, carbon content and oxygen supplement blowing amount of the converter, nyAnd nuRepresenting the unknown order of the system, and f (…) is a non-linear function of the system.
Before designing the controller, the conditions satisfied by equation (18) are analyzed:
introduction 5.1: the system equation (18) outputs a partial derivative to the input that is present and continuous;
and (5.2) introduction: the system equation (18) satisfies the generalized Lipschitz condition, i.e., there is a normal b, such that
|Δy(t+1)|≤b|Δu(t)| (19)
Where Δ y (t +1) ═ y (t +1) -y (t), and Δ u (t +1) ═ u (t +1) -u (t).
Theorem 5.1: for system equation (18), satisfying the arguments 5.1 and 5.2, then when Δ u (t) ≠ 0, there must be a "pseudo-partial derivative" quantity φ (t) such that
Δy(t+1)＝φ(t)Δu(t),φ(t)≤b (20)
Equation (20) of the present embodiment is a generic model of system equation (18).
2) Controller design
In order to improve the control performance of the system, the model-free adaptive control has stronger tracking capability and convergence capability in the initial control stage and stronger anti-interference capability in the later control stage, the scheme introduces a forgetting factor into the design of the controller on the basis of the generic model, and provides a model-free adaptive control method with a variable forgetting factor. To this end, the control input criteria function is:
J(u(t))＝|e(t+1)|2+λ|u(t)-αu(t-1)|2 (21)
wherein, lambda is more than or equal to 0 and is a weight coefficient, and alpha is more than 0 and is a forgetting factor.
Substituting the general model formula (20) into the above criterion function, deriving u (t), and making it be 0, so as to obtain the control rate algorithm as follows:
in the formula, rho (rho is more than 0 and less than or equal to 1) is a step sequence, and the forgetting factor variation formula is as follows:
wherein alpha is1To control the value of the forgetting factor when the system enters the steady state, alpha2Calculating an intermediate value for a forgetting factor when a control system is in a dynamic state, generally weakening the influence of the previous control input on the current control input in order to enable the system to have stronger tracking capability and convergence capability in the initial control stage, and enabling the value of the forgetting factor in the dynamic state to be smaller, namely alpha2When the current control input is smaller, in order to inhibit noise in the process in a steady state, a larger forgetting factor is generally adopted to enhance the dependence capacity of the current control input of the system on the previous time, so that the current control input has 1 ≧ alpha1＞α2Is greater than 0; c is the forgetting factor adjusting rate which is a controllable quantity; ζ is a threshold value of the determination condition U.
In the control rate algorithm (22), phi (t) is the pseudo-partial derivative of the system, and belongs to an unknown quantity, so the embodiment of the invention adopts an online estimation value(pseudo-partial derivatives) instead of phi (t), the estimation criterion function of the pseudo-partial derivatives is as follows:
in the same way, pairA partial derivative of 0 gives:
wherein mu is more than or equal to 0 and is a weight coefficient, and eta (eta is more than 0 and less than or equal to 1) is a step length sequence.
In order to make equations (22) and (25) true, it is necessary that the partial derivative denominator is not 0, that is:
the threshold reset algorithm is given as follows:
where ε is a sufficiently small positive number.
Therefore, the control rate of the system obtained by the joint type (22), the formula (23), the formula (25), and the formula (27) is as follows:
wherein u (t) represents the input of the system at the time t, u (t-1) represents the input of the system at the time t-1, Deltau (t-1) represents the input difference value of the system at the time t and the time t-1, and y*(t +1) represents a true value at the t +1 moment, y (t) represents the output of the system at the t moment, Δ y (t) represents the output difference value between the system at the t moment and the t-1 moment, α represents a forgetting factor, and α represents the output difference value1A value, alpha, representing a forgetting factor when the control system enters a steady state2The forgetting factor representing when the control system is in motion calculates an intermediate value,an online estimate representing the pseudo-partial derivative of the system at time t,an online estimate representing the pseudo-partial derivative of the system at time t-1,represents the on-line estimated value of the pseudo-partial derivative of the system at time t ═ 1, p represents the sequence of steps in u (t) calculation, and 0 < p ≦ 1, and η represents the parameterEstimating a time step sequence, wherein eta is more than 0 and less than or equal to 1, c represents the adjustment rate of a forgetting factor, U represents a judgment condition, xi represents a threshold value of the judgment condition U, and lambda represents a weight coefficient for limiting controlThe system input quantity is changed, lambda is more than or equal to 0, mu represents a parameterThe weight coefficient when estimating, and mu is more than or equal to 0, epsilon represents a small enough positive number.
The control quantity of the model-free self-adaptive control method for the forgetting factor is the oxygen supplement blowing quantity, the control quantity is the end point temperature and the carbon content of the converter, relevant parameters in a control algorithm are set off line by utilizing different heat data provided on site, and then the parameters are applied to smelting in the actual process.
And S205, aiming at the bivariate coupling control type, adopting MIMO model-free self-adaptive decoupling control.
Through the division of the above mode mechanisms, the other mode is double-input and double-Output coupling control, namely a coupling control mode of oxygen supplement and additive cooling agent amount, converter end point temperature and carbon content, aiming at a Multi-input and Multi-Output (Multi-input Multi-Output) coupling system, on the basis of the above SISO system general model, the scheme regards the coupling influence of other inputs on a certain loop as measurable interference, and estimates the measurable interference through an RBF neural network, so as to realize decoupling, and specifically refers to the structural block diagram of the converter end point coupling control system shown in FIG. 5.
1) Description of control problems of dynamic oxygen quantity supplement and cooling agent addition at converter end point
The coupling control system can be written as follows:
y(t+1)＝f(y(t),y(t-1),…,y(t-ny),u(t),u(t-1),…,u(t-nu)) (29)
in the above formula, y (t) ═ y1(t),y2(t))T,u(t)＝(u1(t),u2(t))TThe output and input of the system at the time t, namely the end temperature of the converter, the carbon content, the oxygen supplementing amount, the added coolant amount, nyAnd nuRepresenting the unknown order of the system, and f (…) is a non-linear function of the system.
2) Controller design
As above, a generic model concept is introduced, i.e.
y(t+1)＝y(t)+φ(t)Δu(t) (30)
Wherein,equation (30) is further written as follows:
wherein,for the coupling effect between converter steelmaking end point control systems, the neural network estimation is used for compensation, and further the decoupling of a controlled system is realized.
The control rate of the converter steelmaking terminal MIMO coupling system is as follows:
wherein i is 1, 2.
3) RBF network prediction model-free self-adaptive decoupling control algorithm based on gradient descent method
By analyzing the coupling effect between converter steelmaking end point control systems, the scheme adopts RBF neural network estimation based on a gradient descent method for compensation, thereby realizing the decoupling of a controlled system. The overall algorithm flow is as follows:
step 1: initializing network clustering centers, and selecting a certain number of clustering centers c from the training seti(i＝1,2,…,h)；
Step 2: push buttonCalculating the width σ of the centeri；
Step3: in [0,1 ]]Randomly initializing weights w from the hidden layer to the output layer of the networki；
Step 4: n-dimensional input sample vector X ═ X (X) for joining network1,x2,…,xn) According to the formulaOutput of a computing network, wherein
Step 5: according to respective parameter correction formula based on gradient descent method
Updating parameters of the network, wherein the index function
Step 6: and calculating the performance index of the network, if the performance index meets the requirement, finishing training, outputting the network, otherwise, turning to Step4 to continue training when the iteration time t is t + 1.
Coupling quantity omega of trained RBF network to above systemi(t) estimated, then, the input vector in the RBF network is X ═ Δ y (t), Δ y (t-1), Δ u (t-1), Δ u (t-2)]TOutput isWherein y (t) ═ y1(t),y2(t))T，u(t)＝(u1(t),u2(t))T，The following formula is adopted for calculation:
the actual coupling phase in the converter steelmaking end point MIMO model-free self-adaptive decoupling control system is as follows:
the RBF index function is then:
after selecting proper network structure and network parameters, the method can be realizedApproximation to ω (t) with arbitrary precision, so that there is a small positive number γ, so that the estimation error of the coupled phase is always smaller than γ, i.e.
In summary, the control rate of the whole decoupling control system is:
wherein i is 1, 2.
As can be seen from the control result diagram of the carbon content at the end point of the converter shown in FIG. 6, after a certain number of iterations, the embodiment of the present invention can control the carbon content to be between 0.06% and 0.08% in the narrow window region. As can be seen from the converter end point temperature control result diagram shown in fig. 7, after a certain number of iterations, the controllable molten steel end point temperature of the embodiment of the present invention is 1630 ℃ -1660 ℃ between the narrow window regions, so that it is not difficult to see that the method of the embodiment of the present invention greatly improves the molten steel quality control effect, and in addition, the narrow window regions of the embodiment of the present invention can be used for guiding the on-site smelting, improving the product quality, realizing energy saving and emission reduction, and reducing energy consumption; the two modes of the dynamic process at the final stage of converter steelmaking are divided, so that the control process is more definite, the smelting state is in an expected range, and only oxygen needs to be blown up, so that the end point target narrow window interval is reached; when the smelting state exceeds the expected range, oxygen is additionally blown and a coolant is added at the same time to reach the narrow window interval of the end point target; the variable forgetting factor model-free self-adaptive control method adopted by the scheme gets rid of the dilemma that a mathematical model of a controlled object cannot be accurately obtained, so that the control system has stronger tracking capability and convergence capability in the initial stage and stronger process interference resistance capability in the later control stage, and the control performance of the system is improved; the decoupling control strategy based on RBF neural network estimation can more accurately control the carbon content and temperature of the steelmaking end point, and has important significance in the aspects of ensuring the molten steel quality, shortening the smelting time, reducing the smelting cost, reducing the production rework and the like.
The embodiment of the invention aims to provide a narrow window control method based on bimodal switching molten steel quality, which improves the molten steel quality of a converter terminal point; acquiring the narrow window interval of the end point carbon content and the end point temperature of the converter steelmaking according to the requirements of a field operation system and a smelting steel type; establishing a two-classification model according to historical data of different steelmaking times of the converter in the past, and determining a modal decision mechanism of a control system; according to different modes, different control schemes are adopted, double variables adopt MIMO model-free adaptive decoupling control, and single variables adopt SISO model-free adaptive control.
The key technical points of the embodiment of the invention are as follows:
(1) the embodiment of the invention introduces the concept of narrow window into the steelmaking process for the first time, and the acquisition of the narrow window interval ensures low energy consumption and high quality in converter smelting, thereby improving the product quality and realizing energy conservation and emission reduction;
(2) the embodiment of the invention adopts a twin support vector machine mode decision mechanism based on a fruit fly optimization algorithm, and divides the later dynamic process of converter steelmaking into two modes, wherein one mode is a single-variable smelting process of oxygen supplement blowing, and the other mode is a coupling process of oxygen supplement blowing and coolant addition;
(3) in the embodiment of the invention, under the condition that the accuracy of a mechanism model is not high, a model-free control method is adopted, a variable forgetting factor is added to design a controller on the basis of the existing model-free self-adaptive control, so that the system has stronger tracking capability and convergence capability at the initial control stage, the influence of the previous control on the current control is generally weakened, the value of the forgetting factor in the dynamic state is smaller, and the larger forgetting factor is generally taken to enhance the dependence capability of the current control input of the system on the previous control in the steady state so as to inhibit the noise in the process, so that the value of the forgetting factor in the steady state is gradually increased;
(4) according to the embodiment of the invention, the coupling problem between the oxygen supplementing blowing amount and the cooling agent adding amount is converted into the measurable disturbance of the system, and the RBF neural network is adopted to estimate the measurable disturbance on line, so that decoupling is realized.
EXAMPLE III
Referring to fig. 8, in the embodiment, a converter of 150t in a certain iron and steel plant is selected to be installed and operated, and a narrow window interval of a converter end point process is obtained according to a field operation system and smelting steel species requirements; then establishing a two-classification model by using historical data, and determining a modal decision mechanism of the control system; then aiming at the problem of single variable control (oxygen supplement and blowing), SISO model-free adaptive control is adopted, and a variable forgetting factor is introduced into a model-free adaptive control algorithm; and finally, aiming at the bivariate coupling control problems (oxygen supplement and cooling agent addition), adopting MIMO model-free self-adaptive decoupling control, and regarding the coupling problem in the control system as system disturbance for decoupling control.
Specifically, the method comprises the following steps:
and S301, acquiring a target molten steel quality narrow window interval by combining the type of the field smelting steel and the experience of operators.
And step S302, judging whether the carbon content and the temperature information which are detected by the secondary lance for the first time are in the narrow window interval or not on the basis of the narrow window interval of the acquired target molten steel quality.
And S303, utilizing the trained FOA _ TWSVM model to make an online real-time decision on the first sublance detection information which does not hit the narrow window interval, and dividing two modes.
Step S304, finally, respectively controlling according to the two divided modes, and aiming at the single variable mode, combining the established variable forgetting factor model-free self-adaptive control strategy and the data in the converter production process to control the oxygen supplement blowing amount; aiming at a bivariate coupling mode, the established RBFNN is utilized to estimate the coupling problem of the oxygen supplementing amount and the coolant adding amount on line, and the decoupling is carried out and then the control is carried out according to the univariate mode. And when the secondary blowing is finished, the quality of the molten steel at the end point of the converter meets the tapping requirement.
Referring to fig. 9, a molten steel quality narrow window control system based on bimodal switching according to an embodiment of the present invention includes:
the control method comprises a memory 10, a processor 20 and a computer program stored on the memory 10 and capable of running on the processor 20, wherein the processor 20 realizes the steps of the molten steel quality narrow-window control method based on the bimodal switching when executing the computer program.
The specific working process and working principle of the molten steel quality narrow window control system based on the bimodal switching in this embodiment may refer to the working process and working principle of the molten steel quality narrow window control method based on the bimodal switching in this embodiment.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (7)
1. A molten steel quality narrow window control method based on bimodal switching is characterized by comprising the following steps:
acquiring a narrow window interval of a converter end point process according to a field operation system and smelting steel seed requirements;
determining a modal decision mechanism of converter steelmaking second blowing stage control by a twin support vector machine classification method based on a drosophila optimization algorithm;
and carrying out self-adaptive control on the converter steelmaking second blowing stage based on the mode decision mechanism.
2. The molten steel quality narrow window control method based on bimodal switching as claimed in claim 1, wherein the classification decision function of the modal decision mechanism is specifically:
Class i＝min|K(xT,CT)ωj+bj|，
wherein i ═ 1, -1, Class +1 denotes the addition of coolants, Class-1 denotes the absence of coolants, K denotes the gaussian kernel function, x denotesTRepresenting input samples, CTTwo sample sets representing the addition of coolant and the absence of coolant, j ═ 1,2, w1Weight, w, representing hyperplane 1 solved for the set of added coolant class samples2Representing the weight of the hyperplane 2 solved for the set of samples without the addition of a coolant class, b1Representing the deviation of said hyperplane 1, b2Representing the deviation of said hyperplane 2.
3. The molten steel quality narrow window control method based on bimodal switching as claimed in claim 1 or 2, wherein based on the modal decision mechanism, the adaptive control of the converter steelmaking second blowing stage comprises:
based on the modal decision mechanism, classifying control types of a converter steelmaking second blowing stage, and specifically classifying the control types into a single variable control type and a bivariate coupling control type, wherein the single variable control type specifically means that only oxygen quantity supplementary blowing control needs to be carried out on the converter steelmaking second blowing stage, and the bivariate coupling control type specifically means that oxygen quantity supplementary blowing control and coolant addition control need to be carried out on the converter steelmaking second blowing stage;
aiming at the single variable control type, SISO model-free self-adaptive control is adopted;
and aiming at the bivariate coupling control type, adopting MIMO model-free self-adaptive decoupling control.
4. The molten steel quality narrow window control method based on bimodal switching as claimed in claim 3, characterized in that, for single variable control type, the controller adopting SISO model-free adaptive control is specifically designed as:
wherein u (t) represents the input of the system at the time t, u (t-1) represents the input of the system at the time t-1, Deltau (t-1) represents the input difference value of the system at the time t and the time t-1, y*(t +1) represents a true value at the t +1 moment, y (t) represents the output of the system at the t moment, delta y (t) represents the output difference value between the system at the t moment and the t-1 moment, alpha represents a forgetting factor, and alpha represents the output of the system at the t moment1A value, alpha, representing a forgetting factor when the control system enters a steady state2The forgetting factor representing when the control system is in motion calculates an intermediate value,an online estimate representing the pseudo-partial derivative of the system at time t,an online estimate representing the pseudo-partial derivative of the system at time t-1,denotes the on-line estimate of the pseudo-partial derivative of the system at time t 1, p denotes the sequence of u (t) calculation time steps, and 0<Rho is less than or equal to 1, eta represents a parameterEstimate the time step sequence, and 0<Eta is less than or equal to 1, c represents forgetting factor adjustment rate, U represents judgment condition, xi represents threshold of judgment condition U, lambda represents weight coefficient for limiting control input amount variation, and lambda is greater than or equal to 0, mu represents parameterThe weight coefficient when estimating, and mu is more than or equal to 0, epsilon represents a small enough positive number.
5. The molten steel quality narrow window control method based on bimodal switching as claimed in claim 4, wherein for bivariate coupling control type, the controller adopting MIMO model-free adaptive decoupling control is specifically designed as:
wherein i is 1,2, ui(t) represents the value of the ith input at time t, ui(t-1) represents the value of the ith input at time t-1,. DELTA.ui(t-1) represents the difference between the ith input at time t and time t-1, yi(t) represents the value of the ith output at time t,. DELTA.yi(t) represents the difference between the ith output at time t and time t-1,which represents the true value at time t +1,an estimate representing the coupling between the control systems at time t,an estimate representing the coupling between the control systems at time t-1,an estimate of the time-varying parameter representing the ith input to the ith output at time t,a time-varying parameter estimate representing the ith input to the ith output at time t-1,showing the time-varying parameter estimated value of the ith input corresponding to the ith output at the moment 1, alpha is a forgetting factor, alpha1A value, alpha, representing a forgetting factor when the control system enters a steady state2Representing the forgetting factor calculation intermediate value when the control system is in a dynamic state, U represents a judgment condition, xi represents a threshold value of the judgment condition U, and c representsThe forgetting factor adjusts the rate, p represents the step length sequence of u (t) in calculation, and 0<Rho is less than or equal to 1, eta represents a parameterEstimate the time step sequence, and 0<Eta is less than or equal to 1, lambda represents weight coefficient for limiting control input variable, and lambda is greater than or equal to 0, mu represents parameterThe weight coefficient when estimating, and mu is more than or equal to 0, epsilon represents a small enough positive number.
6. The molten steel quality narrow window control method based on bimodal switching as claimed in claim 5, wherein the narrow window area of converter terminal process specifically includes:
the narrow window interval of the molten steel end point temperature is 1630-1660 ℃, and the interval of the molten steel end point carbon content is 0.06-0.08%.
7. A molten steel quality narrow window control system based on bimodal switching, the system comprises:
memory (10), processor (20) and computer program stored on the memory (10) and executable on the processor (20), characterized in that the steps of the method according to any of the preceding claims 1 to 6 are implemented when the computer program is executed by the processor (20).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910642694.1A CN110373510B (en) | 2019-07-16 | 2019-07-16 | Molten steel quality narrow window control method and system based on bimodal switching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910642694.1A CN110373510B (en) | 2019-07-16 | 2019-07-16 | Molten steel quality narrow window control method and system based on bimodal switching |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110373510A true CN110373510A (en) | 2019-10-25 |
| CN110373510B CN110373510B (en) | 2020-04-24 |
Family
ID=68253473
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910642694.1A Active CN110373510B (en) | 2019-07-16 | 2019-07-16 | Molten steel quality narrow window control method and system based on bimodal switching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110373510B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110991089A (en) * | 2019-12-26 | 2020-04-10 | 北京科技大学 | Method for forecasting carbon content in later stage of converter steelmaking |
| CN113930573A (en) * | 2021-09-08 | 2022-01-14 | 山东钢铁股份有限公司 | Bottom blowing instantaneous flow dynamic control method of top-bottom combined blown converter |
| CN114692091A (en) * | 2022-04-24 | 2022-07-01 | 江苏四象软件有限公司 | Data governance task operation system based on big data resources |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1603424A (en) * | 2003-09-29 | 2005-04-06 | 宝山钢铁股份有限公司 | Bessemerizing control method based on intelligent compound dynamic model with sublance converter |
| US20090193936A1 (en) * | 2008-02-01 | 2009-08-06 | Honeywell International, Inc. | Methods and apparatus for an oxygen furnace quality control system |
| CN202390466U (en) * | 2011-12-05 | 2012-08-22 | 济南钢铁股份有限公司 | Self-adapted oxygen flow control system for automatic steel smelting converter |
| CN103882176A (en) * | 2014-03-25 | 2014-06-25 | 东北大学 | On-line dynamic optimization control method for converter steelmaking process based on data driving |
| CN108153146A (en) * | 2017-12-12 | 2018-06-12 | 东北大学 | A kind of polynary molten steel quality MFA control system and method for blast furnace |
| CN108676955A (en) * | 2018-05-02 | 2018-10-19 | 中南大学 | A kind of BOF Steelmaking Endpoint carbon content and temprature control method |
-
2019
- 2019-07-16 CN CN201910642694.1A patent/CN110373510B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1603424A (en) * | 2003-09-29 | 2005-04-06 | 宝山钢铁股份有限公司 | Bessemerizing control method based on intelligent compound dynamic model with sublance converter |
| US20090193936A1 (en) * | 2008-02-01 | 2009-08-06 | Honeywell International, Inc. | Methods and apparatus for an oxygen furnace quality control system |
| CN202390466U (en) * | 2011-12-05 | 2012-08-22 | 济南钢铁股份有限公司 | Self-adapted oxygen flow control system for automatic steel smelting converter |
| CN103882176A (en) * | 2014-03-25 | 2014-06-25 | 东北大学 | On-line dynamic optimization control method for converter steelmaking process based on data driving |
| CN108153146A (en) * | 2017-12-12 | 2018-06-12 | 东北大学 | A kind of polynary molten steel quality MFA control system and method for blast furnace |
| CN108676955A (en) * | 2018-05-02 | 2018-10-19 | 中南大学 | A kind of BOF Steelmaking Endpoint carbon content and temprature control method |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110991089A (en) * | 2019-12-26 | 2020-04-10 | 北京科技大学 | Method for forecasting carbon content in later stage of converter steelmaking |
| CN113930573A (en) * | 2021-09-08 | 2022-01-14 | 山东钢铁股份有限公司 | Bottom blowing instantaneous flow dynamic control method of top-bottom combined blown converter |
| CN114692091A (en) * | 2022-04-24 | 2022-07-01 | 江苏四象软件有限公司 | Data governance task operation system based on big data resources |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110373510B (en) | 2020-04-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110373510B (en) | Molten steel quality narrow window control method and system based on bimodal switching | |
| CN108676955B (en) | A kind of BOF Steelmaking Endpoint carbon content and temprature control method | |
| Lucia et al. | Handling uncertainty in economic nonlinear model predictive control: A comparative case study | |
| Wang et al. | Applying input variables selection technique on input weighted support vector machine modeling for BOF endpoint prediction | |
| CN104375475B (en) | The optimal control method of Batch reaction processes in batch reactor | |
| CN109447346B (en) | Converter oxygen consumption prediction method based on gray prediction and neural network combined model | |
| CN106096788B (en) | Converter steelmaking process cost control method and system based on PSO _ ELM neural network | |
| Rashid et al. | Multi-rate modeling and economic model predictive control of the electric arc furnace | |
| Liu et al. | A dynamic analytics method based on multistage modeling for a BOF steelmaking process | |
| Zeng et al. | Data-driven predictive control for blast furnace ironmaking process | |
| An et al. | A multi-time-scale fusion prediction model for the gas utilization rate in a blast furnace | |
| CN108153146B (en) | Model-free self-adaptive control system and method for quality of blast furnace multi-element molten iron | |
| CN107368125B (en) | A kind of blast furnace temperature control system and method based on CBR Yu the parallel mixed inference of RBR | |
| Xie et al. | A hybrid control strategy for real-time control of the iron removal process of the zinc hydrometallurgy plants | |
| CN107918368B (en) | The dynamic prediction method and equipment of iron and steel enterprise's coal gas yield and consumption | |
| Zhao et al. | A nonlinear industrial model predictive controller using integrated PLS and neural net state-space model | |
| CN106054836B (en) | Converter steelmaking process cost control method and system based on GRNN | |
| JP4093934B2 (en) | Model parameter determination method and program thereof, model prediction method and program thereof | |
| CN111522233A (en) | Parameter self-tuning MIMO different-factor full-format model-free control method | |
| Sun et al. | Process knowledge-based random forest regression for model predictive control on a nonlinear production process with multiple working conditions | |
| Han et al. | Hybrid intelligent control of BOF oxygen volume and coolant addition | |
| Cardoso et al. | Artificial neural networks for modelling and controlling the variables of a blast furnace | |
| Wu et al. | Design and application of generalized predictive control strategy with closed-loop identification for burn-through point in sintering process | |
| CN114107653A (en) | Annealing furnace temperature control method, device and system | |
| Tang et al. | An estimation of distribution algorithm with resampling and local improvement for an operation optimization problem in steelmaking process |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |