CN104942251B - Cast-on time determining method of steel mill continuous casting machines - Google Patents

Abstract

The invention provides a cast-on time determining method of steel mill continuous casting machines. The method comprises the steps that a cast-on decision controller is connected with a manufacturing execution system (MES) database of a steel mill and an MES client side system; a target function of a cast-on decision is established, the to-be-cast-on frequency and left task restraint condition, the constraint condition about whether continuous casting is needed during the to-be-cast-on frequency, the constraint condition between the to-be-cast-on frequency time and the planned period starting and ending time, the amount of redundant metal besides the needed metal amount planned to need to be finished in the task period on a production line, and other constraint conditions are established, and the established known physical quantity and parameter values are obtained; the cast-on time shown in the specification of a jth cast-on time of an ith casting machine in a kth planned period is obtained through an optimal solver. Whether the to-be-cast-on frequency of the continuous casting machines is continuous or not can be judged, the cast-on time can be scientifically calculated, a foundation is laid for formulating a reasonable steel mill production dispatching operation plan, the production running among shifts can be stabilized, and the amount of accumulated metal is reduced.

Description

The casting time of steel plant continuous casting machine determines method

Technical field

The present invention relates to technical field of metallurgical control, be specifically related to the casting time side of determination of a kind of steel plant continuous casting machine Method.

Background technology

Conticaster casting time decision-making (Casting start time decision of continuous caster, CSDCC) the very corn of a subject task determines that whether conticaster respectively time even can water and time concrete opening is watered with front watering wait opening to water time Between.Conticaster waits to open that to water rationally determining of time casting time be the precondition of establishment steel smelting-continuous casting production schedules.Mesh Before, in actual production, steel-making continuous casting production schedules is typically to have worked out after manually being determined casting time by dispatcher Becoming, the science of casting time and reasonability become the key factor that can restriction steel mill production schedules be implemented smoothly.Cause This research CSDCC problem has important realistic meaning.

The metals resources amount balance of the molten iron/molten steel on steel mill production line is that the important bar watering decision-making effectiveness is opened in impact Part.In recent years, the planned dispatching problem that domestic and international experts and scholars produce for steel smelting-continuous casting is widely paid close attention to, and research obtains substantially Progress.In the research relevant to conticaster casting time, Hubert M etc. devises one and utilizes the circulation of known iron water amount repeatedly In generation, calculates the dispatching algorithm of casting time；The round grade of Tan have studied including watering and time opening and to water etc. from minimum power consumption angle Schedule job Time Calculation method；Wang Chuan etc. propose and solve other under assuming certain conticaster casting time known case even The method of the casting time of casting machine.But, at present research, can the big multipair conticaster that affects connect the factor watered according to plan and enter Go simplification process, research emphasis has been concentrated on the preparation method side of production schedules being analyzed time factor processing Face, is not directed to molten iron on the production line affecting casting time/molten steel metals resources be balanced each factor and opens the relation watering decision-making Research, thus constrain the real world applications effect of steel-making continuous casting Production Scheduling Study on Problems achievement.

Summary of the invention

In order to overcome defect present in above-mentioned prior art, it is an object of the invention to provide a kind of steel plant continuous casting machine Casting time determines that method, the method are capable of conticaster and water the scientific algorithm of time casting time.

In order to realize the above-mentioned purpose of the present invention, according to an aspect of the present invention, the invention provides a kind of steel mill The casting time of conticaster determines method, and it comprises the steps:

S1, controller is connected with manufacturing execution system (MES) data base and the MES FTP client FTP of steel mill respectively；

S2, sets up to open and waters objective function of decision-making, and described object function is:

$\max f=C_1{\mathrm{\Θ}}^k-C_2{Q}_o^k---\left(1\right)$

$\begin{array}{c}{\mathrm{\Θ}}^k=\frac{\underset{i=1}{\overset{H}{\Σ}}(\frac{{\mathrm{rq}}_{ij}^k}{{\mathrm{rs}}_{ij}^k\·\mathrm{\ρ}\·{\mathrm{ra}}_{ij}^k}+\underset{j=1}{\overset{N_i^k}{\Σ}}\left(min\left({\mathrm{\τ}}_e^k,x_{ij}^k+{\mathrm{p\τ}}_{ij}^k\right)-x_{ij}^k\right)}{({\mathrm{\τ}}_e^k-{\mathrm{\τ}}_s^k)\·H}\×100\%,\\ \∀i\∈\[1,H\],\∀j\∈\[1,N_i^k\]\end{array}---\left(2\right)$

$Q_o^k=Q_E^k+Q_I^k-Q_S^k-Q_L^k-Q_C^k---\left(3\right)$

$\begin{array}{c}{Q}_C^k=\underset{i=1}{\overset{H}{\Σ}}{\mathrm{rq}}_{ij}^k+\underset{i=1}{\overset{H}{\Σ}}\underset{j=1}{\overset{N_i^k}{\Σ}}\left(min\left({\mathrm{\τ}}_e^k,x_{ij}^k+{\mathrm{p\τ}}_{ij}^k\right)-x_{ij}^k\right)\·\mathrm{\ρ}\·{\mathrm{wa}}_{ij}^k\·{\mathrm{ws}}_{ij}^k,\\ \∀i\∈\[1,H\],\∀j\∈\[1,N_i^k\]\end{array}---\left(4\right)$

$Q_{ca}^k=\frac{Q_C^k}{{\mathrm{\τ}}_e^k-{\mathrm{\τ}}_s^k}---\left(5\right)$

$Q_L^k=Q_C^k\·\frac{\mathrm{\η}}{1-\mathrm{\η}}---\left(6\right)$

$Q_a^k=\frac{{\mathrm{\τ}}_A^k\·Q_{ca}^k}{1-\mathrm{\η}}---\left(7\right)$

$Q_S^k=Q_a^k+Q_{rcon}^k---\left(8\right)$

Wherein, k is plan phase sequence number, k ∈ 1,2 ..., and Ω }, Ω is plan phase sum；I is casting machine serial number, i ∈ 1, 2 ..., H}, H are casting machine sum；J for watering sequence number, It is that the i-th casting machine waters time sum；ρ is molten steel Density；η is metal loss factor；q₀Steel amount is averagely watered for each heat；Whole process average stream time kth plan phase；For Start time kth plan phase；For finish time kth plan phase；Cast for kth plan phase the i-th casting machine jth casting plan Stove number；Time minimum interval is watered for kth plan phase the i-th casting machine jth；Water secondary leaving over for kth plan phase the i-th casting machine jth to appoint Business finish time；Time legacy tasks average pull rate is watered for kth plan phase the i-th casting machine jth；Cast for the kth plan phase i-th Machine jth waters time casting cycle；Wait to open for kth plan phase the i-th casting machine jth and water time average pull rate；For the kth plan phase i-th Casting machine jth waters time legacy tasks basal area；Wait to open for kth plan phase the i-th casting machine jth and water subtask basal area；It is K plan phase the i-th casting machine jth is watered time legacy tasks and is watered steel amount；C1 is benefit coefficient；C2 is cost coefficient；For the kth plan phase Enter iron, for known quantity；For stock's amount of metal that kth plan is initial, for known quantity；For overstocking of the kth plan end of term Amount of metal, for Unknown Process variable；For the line safety inventory amount of metal of kth plan phase, for Unknown Process variable；For kth Plan phase metal loss amount, for Unknown Process variable；Steel amount is watered, for Unknown Process variable for the kth plan phase；For kth meter Draw phase the i-th casting machine jth to wait to open to water time to open and water the moment, for unknown number to be asked；

Formula (1) represents that object function is deducted the cost structure of overstocked amount of metal by serialization degree income；Formula (2) represents even Continuousization degree (Θ^k) water time casting time sum divided by H platform even equal to the legacy tasks casting time of each conticaster with waiting to open Time plan phase that casting machine is total；(3) overstocked amount of metal is representedIron is entered respectively by this plan phaseOpening inventory Amount of metalSafety on line stock's amount of metalMetal loss amountWater steel amountRelation constitute；Formula (4) the expression plan phase adds up to water steel amount and time waters steel amount sum equal to the legacy tasks amount of each conticaster with waiting to open to water；Formula (5) table Show that the H platform conticaster unit interval within the plan phase averagely waters steel amountEqual to accumulative steel amount of watering divided by plan phase length；Formula (6) represent when planning phase metal loss factor and being known, accumulative metal loss amount and the accumulative mutual relation watering steel amount；Formula (7) Represent when averagely watering the steel amount average line stock amount of metal of calculating with the unit intervalTime, it should utilize metal loss factor η It is converted to the iron water amount of respective numbers；Formula (8) represents that safety on line stock's amount of metal needs at average line stock metal A random fluctuation demand amount of metal is added on the basis of amount

S3, set up wait to open water time and legacy tasks constraints, wait to open whether connect between watering time water constraints, respectively treat Open time time of watering and plan beginning and ending time phase constraints and the constraints of overstocked amount of metal,

S4, controller obtains each known quantity parameter in step S2 by MES data storehouse and full factory each MES FTP client FTP Value；

S5, according to the exploitation intermediate variable of known quantity parameter in step S4, described intermediate variable includes respectively waiting to open watering Secondary casting cycle, each casting machine legacy tasks casting time, each casting machine single bit time water steel amount, collects metal loss factor, whole Overstock in amount of metal formula coefficient before real number decision variable after reason, arrange after coefficient before Real-valued decision variable；

S6, determines each coefficient of object function according to the intermediate variable that step S5 obtains, and waters secondary for the i-th conticaster jth, Order:

F (i, j)=f (i, j)+arrange after before Real-valued decision variable coefficient × conticaster i jth water time casting time x (i, J), (i, j), ((i, j), wherein, (i is j) that Real-valued defined in matlab becomes to sdpvar to described x for i, j)=sdpvar to make p=f The function of amount；

S7, determines the restriction relation in step S3 according to the intermediate variable that step S5 obtains；

S8, utilizes Optimization Solution device solvesdp (F ,-p), obtain kth plan phase the i-th casting machine jth wait open water time open water time CarveAnd whether connect, between watering time, the parameter wateredWherein, solvesdp () is matlab majorized function, and F is that total constraint is closed System, object function identifier total when being maximizing for-p；

S9, controller is according to obtainingData carry out the shop job scheduling of steel smelting-continuous casting production, and to life Produce operation system to implement effectively to control.

The angle of molten iron/molten steel metals resources balance that the present invention is to be processed from steel mill's production line, establishes with continuous casting Produce total benefit to be the conticaster of object function to the maximum and open and water the mixed integer programming algorithm of decision-making and solve.It is capable of continuous casting Machine waters the scientific algorithm of time casting time, lays the foundation for working out rational production schedule of steelmaking plant production plan, also contributes to Stablize the production run between each order of classes or grades at school, reduce the overstocked amount of metal on production line, opening in order to water and carry for steel plant continuous casting machine Supply technological means.

In the preferred embodiment of the present invention, in described step S3, constraints is:

1) respectively wait to open the company between watering time and water relation constraint:

$\begin{array}{c}{\mathrm{\τ}}^{\′}+{\mathrm{s\τ}}_{ij}^k\·y_{ij}^k\≤x_{ij}^k\≤{\mathrm{\τ}}^{\′}+({\mathrm{\τ}}_e^k-{\mathrm{\τ}}^{\′})\·y_{ij}^k,\\ \∀i\∈\[1,H\],\∀j\∈\[1,N_i^k\]\end{array}---\left(9\right)$

Wherein,For binary variable, 1 represents that kth plan phase the i-th casting machine jth is watered secondary company with legacy tasks and watered, and 0 represents Break and water,For unknown number to be asked；

F1: when judge wait open water time whether connect between legacy tasks water time, in formula (9)

${\mathrm{\τ}}^{\′}={\mathrm{r\τ}}_{ij}^k=\frac{{\mathrm{rq}}_{ij}^k}{\mathrm{\ρ}\·{\mathrm{rs}}_{ij}^k\·{\mathrm{ra}}_{ij}^k},\∀i\∈\[1,H\],\∀j\∈\[1,N_i^k\]---\left(10\right)$

$\begin{array}{c}{\mathrm{rq}}_{ij}^k=n_{i(j=N_i^{k-1})}^{k-1}\·q_o-({\mathrm{\τ}}_e^{k-1}-w_{i(j=N_i^{k-1})}^{k-1})\·\mathrm{\ρ}\·{\mathrm{wa}}_{i(j=N_i^{k-1})}^{k-1}\·{\mathrm{ws}}_{i(j=N_i^{k-1})}^{k-1},\\ \∀i\∈\[1,H\],\∀j\∈\[1,N_i^{k-1}\]\end{array}---\left(11\right)$

Formula (9-10) represent when wait open water time even water with legacy tasks time, casting time point takes legacy tasks finish time, Non-company takes legacy tasks finish time and terminates certain moment to the plan phase plus watering the minor tick time when watering；Formula (11) represents kth meter Draw i-th conticaster jth of phase and water the difference that time legacy tasks amount plans equal to this casting machine previous plan phase to water steel amount with casting amount Value；

F2: when decision-making remaining respectively wait open water time between whether connect water time, in formula (9)

$\begin{array}{cc}{\mathrm{\τ}}^{\′}=x_{i(j-1)}^k+{\mathrm{p\τ}}_{ij}^k,& \∀i\∈\[1,H\],\∀j\∈\[1,N_i^k\]\end{array}---\left(12\right)$

Formula (9), (12) represent when except remaining adjacent with legacy tasks respectively wait open water that time interconnection is watered mutually time, this casting time Point waters time finish time before taking, and non-company waters time finish time and terminates the some time to the plan phase plus watering the minor tick time before taking when watering Carve；

F3: respectively wait to open time time of watering and plan beginning and ending time phase relation constraint condition:

$\begin{array}{cc}{\mathrm{\τ}}_s^k\≤x_{ij}^k\≤{\mathrm{\τ}}_e^k,& \∀i\∈\[1,H\],\∀j\∈\[1,N_i^k\]\end{array}---\left(13\right)$

F4: overstocked amount of metal constraints:

$0\≤Q_E^k+Q_I^k-Q_S^k-Q_L^k-Q_C^k---\left(14\right).$

In the preferred embodiment of the present invention, the intermediate variable obtained in described step S5 is:

Respectively wait to open and water the stove number * average Metal Weight of each stove in time casting cycle=respectively wait to open waters time/(wait to open and water time pulling rate * Wait to open and water time basal area * molten steel density)；

Each casting machine legacy tasks casting time=each casting machine legacy tasks molten steel amount/(each casting machine legacy tasks pulling rate * respectively casts Machine legacy tasks basal area * molten steel density)；

Each casting machine single bit time waters steel amount=water time pulling rate * and waters time basal area * molten steel density；

The metal loss factor collected=(1+ metal loss factor/(1-metal loss factor)+whole process logistics time/ ((plan finish time phase-start time plan phase) * (1-metal loss factor))；

After arrangement, in overstocked amount of metal formula, before real number decision variable, coefficient=-each casting machine single bit time waters steel amount * and collects Metal loss factor；

Constant term in amount of metal constraint formulations=estimate into ferrum total amount+opening inventory iron-stochastic demand is overstock after arrangement The metal loss factor * of stock's amount of metal-collect (respectively wait to open and water time casting by each casting machine legacy tasks molten steel amount sum+each casting machine Cycle * each casting machine single bit time waters steel amount+plan phase finish time * each casting machine single bit time and waters steel amount sum)；

During metal loss factor * each casting machine unit that after arrangement, before Real-valued decision variable, coefficient=cost coefficient * collects Between water steel amount+income coefficient/((plan finish time phase-start time plan phase) * conticaster number).

In the another kind of preferred implementation of the present invention, described intermediate variable set-up procedure S3 obtained according to step S5 In the method for restriction relation be:

1) watering time for the i-th conticaster the 1st, make F1 (1,1)=0, i=1,2 ..., H, described H is positive integer；

F1 (i, j)=F1 (i, j)+set (casting machine i legacy tasks casting time+casting machine i water time between minimum time interval * casting First binary variable y of machine i (i, 1)≤first casting time x of casting machine i (i, 1)≤casting machine i legacy tasks casting time + (plan finish time phase-casting machine i legacy tasks first binary variable y of casting time * casting machine i (i, 1)), wherein, y (i, 1) corresponding in constraints(i, j) corresponding in constraints for yX (i, 1) is corresponding in constraints (i, j) corresponding in constraints for xSet () be MATLAB YALMIP optimization tool in constraints function is set；

2) the i-th conticaster jth is watered time, make F2 (1,2)=0, i=1,2 ..., H, $j=2,3,......,N_i^k;$

F2 (i, j)=F2 (i, j)+set ((casting machine i jth-1 casting time x (i, j-1)+casting machine i water time casting cycle)+ Casting machine i water time between minimum time interval * casting machine i jth binary variable y (i, j)≤casting machine i jth casting time x (and i, j) < =(casting machine i jth-1 casting time x (i, j-1)+casting machine i waters time casting cycle)+(plan finish time phase-(casting machine i jth-1 Casting time x (i, j-1)+casting machine i waters time casting cycle)+casting machine i jth binary variable y (i, j)))；

3) the i-th conticaster jth is watered time, make F3 (1,1)=0, i=1,2 ..., H, $j=1,2,......,N_i^k,$ X (i, j)≤plan start time phase)；

4) the i-th conticaster jth is watered time, make F4 (1,1)=0, i=1,2 ..., H,

((i, j)+set (overstocks in amount of metal formula coefficient * conticaster i before real number decision variable to F4 after arrangement for i, j)=F4 Jth waters time casting time x (i overstocks constant term in amount of metal constraint formulations after j)≤arrangement)；

5) F1=F1 (i, j), F2=F2 (i, j), F3=F3 (i, j), F4=F4 (i, j)；

Total restriction relation: F=F1+F2+F3+F4.

By arrange constraints F1 (i, j), F2 (i, j), F3 (i, j), it is possible to achieve under different task state accurately Determining the casting time of each steel plant continuous casting machine, by arranging constraints F4, (i j) may insure that between the adjacent plan phase not The fluctuation of stock's amount of metal occur, the continuous casting production for equalization stable creates favorable conditions.

The additional aspect of the present invention and advantage will part be given in the following description, and part will become from the following description Obtain substantially, or recognized by the practice of the present invention.

Accompanying drawing explanation

Above-mentioned and/or the additional aspect of the present invention and advantage are from combining the accompanying drawings below description to embodiment and will become Substantially with easy to understand, wherein:

Fig. 1 is steel smelting-continuous casting production process schematic diagram in prior art；

Fig. 2 is molten iron/molten steel equilibrium of stock relationship change schematic diagram；

Fig. 3 is the casting time Gantt chart of actual design decision-making；

Fig. 4 be the design for Fig. 3 utilize that the method for the present invention is optimized by the casting time Gantt chart of class's decision-making；

Fig. 5 is the casting time Gantt chart of the per diem decision-making that the design for Fig. 3 utilizes the method for the present invention to be optimized.

Detailed description of the invention

Embodiments of the invention are described below in detail, and the example of described embodiment is shown in the drawings, the most from start to finish Same or similar label represents same or similar element or has the element of same or like function.Below with reference to attached The embodiment that figure describes is exemplary, is only used for explaining the present invention, and is not considered as limiting the invention.

Fig. 1 is 3 production links that existing steel smelting-continuous casting process mainly comprises: steel-making, refine and continuous casting.Make steel and connect The casting each self-contained parallel unit of link, and refine link generally comprises multiple parallel unit, to realize different refinery practice Requirement.General steel smelting-continuous casting production process is as shown in Figure 1: the high temperature liquid iron shipped from blast furnace is through hot metal pre process procedures After be blended in converter smelting becomes the ladle that molten steel, molten steel are poured under converter on chassis, by overhead traveling crane and the hauling operation of chassis, Steel ladle is transported to refine link, and the refined molten steel on different refining equipments successively according to manufacturing technique requirent, refine completes After, then by overhead traveling crane and chassis, steel ladle be transported to continuous casting and implement casting, forming strand.

In steel mill's production scheduling, heat refers to the molten steel that certain converter produces in a smelting cycle, due to one The molten steel of heat is loaded in a ladle, so the object being scheduled before steel-making to continuous casting is heat, heat is steel mill Production unit minimum in production scheduling.Water the secondary heat set referring to direct casting on same conticaster, be that steel mill produces Production unit maximum in scheduling.Steel smelting-continuous casting scheduling scheme formulates flow process: first turned according to technical standard by user's contract Turn to production contract；Then according to the establishment charging plan such as steel capacity and technological requirement and casting plan, and heat is combined Roll the establishment hot rolling modular plan such as rolling power and technological requirement, form steel-making and produce, with hot rolling, the production lot meter coordinated mutually Draw.In Production Lot Planning, it has been determined that water secondary casting casting machine and water time processing sequence of interior heat and production work Skill.The research contents of the present invention is namely based on Production Lot Planning and steel mill MES data storehouse, opening of research steel plant continuous casting machine Water the time.

In practical situation, steel mill typically has some conticasters, carry out conticaster in time opening opening of watering time and water decision-making, The present invention turns to target with the comprehensive benefit maximum of decision-making, i.e. from the cost minimization of the overstocked amount of metal controlled production line with make The serialization degree Income Maximum of conticaster considers.

The metals resources equilibrium condition of the molten iron/molten steel on production line can affect the casting time of conticaster and run shape State.In order to quantify and the metals resources relevant influence factor of balance, carry out Unify legislation with overstocked amount of metal and constituent relation formula thereof: [overstocking amount of metal]=[initial line stock amount of metal]+[adding up into iron]-[adding up to water steel amount]-[accumulative metal loss Amount]-[end of term safety on line stock amount of metal].If it is excessive to overstock amount of metal, then waters decision objective disagree, if being less than with opening Zero, entail dangers to end of term safety on line stock amount of metal and the direct casting of next plan phase conticaster, ensureing enough peaces again In the case of full stock's amount of metal, the lower limit of overstocked amount of metal is controlled standard and is set to zero by the present invention.Add up in the plan phase to water steel Amount waters steel amount sum for each conticaster, and separate unit conticaster waters steel amount and watered steel amount and this meter by previous plan phase legacy tasks respectively Draw expectation to open to water and time water steel amount and constitute.

It is relevant that accumulative metal loss amount adds up to water steel amount with the plan phase.Calculate for simplifying, present invention assumes that plan phase metal Loss factor is equal to metal loss factor known to the previous plan phase, then this plan phase adds up metal loss amount and is represented by: tired Meter metal loss amount=loss factor × add up to water steel amount/(1-loss factor).

Safety on line stock's amount of metal is made up of average line stock amount of metal and random fluctuation demand two parts amount of metal, Wherein, average line stock amount of metal formula " average inventory=mean unit time output × average flow time " calculates； Random fluctuation demand amount of metal embodies and need to consider because the change at random factors such as casting cycle, pulling rate, flow sheet equipment fault cause The demand to metals resources more than the part of average computation value, calculate for simplifying, be empirically taken as constant.Fig. 2 is Molten iron/molten steel equilibrium of stock relationship change schematic diagram.

For the conticaster serialization extent description in decision objective be: serialization degree=each casting machine activity duration it With/(during opening casting machine quantity × plan).

Accordingly, content of policy decision and the optimization aim of CSDCC problem are specified.In the present embodiment, ask for beneficially problem Solve, make following basic assumption:

(1) within the plan phase, each conticaster waits to open that to water sub-quantity known, waters phenomenon with occurring without unplanned stopping in watering time；

(2) the plan phase enters ferrum total amount and can predict, and each conticaster waters order limit without preferentially connecting.

The casting time that the invention provides a kind of steel plant continuous casting machine determines method, and it comprises the steps:

S1, opens and waters decision controller and be connected with the MES data storehouse of steel mill and full factory each MES FTP client FTP respectively；

S2, sets up to open and waters objective function of decision-making, and object function is:

$\max f=C_1{\mathrm{\&Theta;}}^k-C_2{Q}_o^k---\left(1\right)$

$\begin{array}{c}{\mathrm{\&Theta;}}^k=\frac{\underset{i=1}{\overset{H}{\&Sigma;}}(\frac{{\mathrm{rq}}_{ij}^k}{{\mathrm{rs}}_{ij}^k\&CenterDot;\mathrm{\&rho;}\&CenterDot;{\mathrm{ra}}_{ij}^k}+\underset{j=1}{\overset{N_i^k}{\&Sigma;}}\left(min\left({\mathrm{\&tau;}}_e^k,x_{ij}^k+{\mathrm{p\&tau;}}_{ij}^k\right)-x_{ij}^k\right)}{({\mathrm{\&tau;}}_e^k-{\mathrm{\&tau;}}_s^k)\&CenterDot;H}\&times;100\%\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array},---\left(2\right)$

$Q_o^k=Q_E^k+Q_I^k-Q_S^k-Q_L^k-Q_C^k---\left(3\right)$

$\begin{array}{c}{Q}_C^k=\underset{i=1}{\overset{H}{\&Sigma;}}{\mathrm{rq}}_{ij}^k+\underset{i=1}{\overset{H}{\&Sigma;}}\underset{j=1}{\overset{N_i^k}{\&Sigma;}}\left(min\left({\mathrm{\&tau;}}_e^k,x_{ij}^k+{\mathrm{p\&tau;}}_{ij}^k\right)-x_{ij}^k\right)\&CenterDot;\mathrm{\&rho;}\&CenterDot;{\mathrm{wa}}_{ij}^k\&CenterDot;{\mathrm{ws}}_{ij}^k,\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array}---\left(4\right)$

$Q_{ca}^k=\frac{Q_C^k}{{\mathrm{\&tau;}}_e^k-{\mathrm{\&tau;}}_s^k}---\left(5\right)$

$Q_L^k=Q_C^k\&CenterDot;\frac{\mathrm{\&eta;}}{1-\mathrm{\&eta;}}---\left(6\right)$

$Q_a^k=\frac{{\mathrm{\&tau;}}_A^k\&CenterDot;Q_{ca}^k}{1-\mathrm{\&eta;}}---\left(7\right)$

$Q_S^k=Q_a^k+Q_{rcon}^k---\left(8\right)$

Wherein, k is plan phase sequence number, k ∈ 1,2 ..., and Ω }, Ω is plan phase sum；I is casting machine serial number, i ∈ 1, 2 ..., H}, H are casting machine sum；J for watering sequence number, It is that the i-th casting machine waters time sum；ρ is Molten steel density；η is metal loss factor；q₀Steel amount is averagely watered for each heat；Whole process average stream time kth plan phase；For start time kth plan phase；For finish time kth plan phase；For kth plan phase the i-th casting machine jth casting plan Casting furnace number；Time minimum interval is watered for kth plan phase the i-th casting machine jth；Time something lost is watered for kth plan phase the i-th casting machine jth Remain in office business finish time；Time legacy tasks average pull rate is watered for kth plan phase the i-th casting machine jth；For the kth plan phase I casting machine jth waters time casting cycle；Wait to open for kth plan phase the i-th casting machine jth and water time average pull rate；For the kth plan phase I-th casting machine jth waters time legacy tasks basal area；Wait to open for kth plan phase the i-th casting machine jth and water subtask basal area； Water time legacy tasks for kth plan phase the i-th casting machine jth and water steel amount；C1 is benefit coefficient；C2 is cost coefficient；For kth meter Draw the phase enter iron, for known quantity；For stock's amount of metal that kth plan is initial, for known quantity；For the kth plan end of term Overstock amount of metal, for Unknown Process variable；For the line safety inventory amount of metal of kth plan phase, for Unknown Process variable； For kth plan phase metal loss amount, for Unknown Process variable；Steel amount is watered, for Unknown Process variable for the kth plan phase；For Kth plan phase the i-th casting machine jth is waited to open to water time to open and is watered the moment, for unknown number to be asked.

Formula (1) represents that object function is deducted the cost structure of overstocked amount of metal by serialization degree income, wherein: benefit system Number C₁The continuous casting zone ton steel energy medium brought, middle Bao Naicai, auxiliary material etc. can be improved from serialization degree and consume minimizing, And labor productivity raising brings the factors such as cost of labor saving to be calculated reference value；Cost coefficient C₂Can unite from steel mill Count in the average ton steel production cost in storehouse because molten iron/molten steel overstocks the power consumption cost of the LF operation intensification that wait is caused Increase relation etc. calculates.

Formula (2) represents serialization degree Θ^kTime casting is watered with waiting to open equal to the legacy tasks casting time of each conticaster Time sum is divided by H platform conticaster total time plan phase.

Formula (3) represents overstocked amount of metalIron is entered respectively by this plan phaseOpening inventory amount of metalOnline Safety inventory amount of metalMetal loss amountWater steel amountRelation constitute, wherein For known parameters.

Formula (4) the expression plan phase add up to water steel amount equal to each conticaster legacy tasks amount with wait to open water time water steel amount it With.

Formula (5) represents that the H platform conticaster unit interval within the plan phase averagely waters steel amountEqual to accumulative water steel amount divided by Plan phase length.

Formula (6) represents that, when planning phase metal loss factor and being known, accumulative metal loss amount waters the mutual of steel amount with accumulative Relation.

Formula (7) represents when averagely watering the steel amount average line stock amount of metal of calculating with the unit intervalTime, it should utilize gold Belong to loss factor η and be converted to the iron water amount of respective numbers.

Formula (8) represents that safety on line stock's amount of metal needs to add one on the basis of average line stock amount of metal at random Variation demand amount of metal

S3, set up wait to open water time and legacy tasks constraints, wait to open whether connect between watering time water constraints, respectively treat Open time time of watering and plan beginning and ending time phase constraints and the constraints of overstocked amount of metal, in the present embodiment, about Bundle condition is:

1) respectively wait to open the company between watering time and water relation constraint:

$\begin{array}{c}{\mathrm{\&tau;}}^{\&prime;}+{\mathrm{s\&tau;}}_{ij}^k\&CenterDot;y_{ij}^k\&le;x_{ij}^k\&le;{\mathrm{\&tau;}}^{\&prime;}+({\mathrm{\&tau;}}_e^k-{\mathrm{\&tau;}}^{\&prime;})\&CenterDot;y_{ij}^k,\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array}---\left(9\right)$

Wherein,For binary variable, 1 represents that kth plan phase the i-th casting machine jth is watered secondary company with legacy tasks and watered, and 0 represents Break and water,For unknown number to be asked；

F1: when judge wait open water time whether connect between legacy tasks water time, in formula (9)

$\begin{array}{cc}{\mathrm{\&tau;}}^{\&prime;}={\mathrm{r\&tau;}}_{ij}^k=\frac{{\mathrm{rq}}_{ij}^k}{\mathrm{\&rho;}\&CenterDot;{\mathrm{rs}}_{ij}^k\&CenterDot;{\mathrm{ra}}_{ij}^k},& \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array}---\left(10\right)$

$\begin{array}{c}{\mathrm{rq}}_{ij}^k=n_{i(j=N_i^{k-1})}^{k-1}\&CenterDot;q_o-({\mathrm{\&tau;}}_e^{k-1}-w_{i(j=N_i^{k-1})}^{k-1})\&CenterDot;\mathrm{\&rho;}\&CenterDot;{\mathrm{wa}}_{i(j=N_i^{k-1})}^{k-1}\&CenterDot;{\mathrm{ws}}_{i(j=N_i^{k-1})}^{k-1},\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^{k-1}\&rsqb;\end{array}---\left(11\right)$

Formula (9-10) represent when wait open water time even water with legacy tasks time, casting time point takes legacy tasks finish time, Non-company takes legacy tasks finish time and terminates certain moment to the plan phase plus watering the minor tick time when watering；Formula (11) represents kth meter Draw i-th conticaster jth of phase and water the difference that time legacy tasks amount plans equal to this casting machine previous plan phase to water steel amount with casting amount Value；

F2: when decision-making remaining respectively wait open water time between whether connect water time, in formula (9)

$\begin{array}{cc}{\mathrm{\&tau;}}^{\&prime;}=x_{i(j-1)}^k+{\mathrm{p\&tau;}}_{ij}^k,& \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array}---\left(12\right)$

Formula (9), (12) represent when except remaining adjacent with legacy tasks respectively wait open water that time interconnection is watered mutually time, this casting time Point waters time finish time before taking, and non-company waters time finish time and terminates the some time to the plan phase plus watering the minor tick time before taking when watering Carve；

F3: respectively wait to open time time of watering and plan beginning and ending time phase relation constraint condition:

$\begin{array}{cc}{\mathrm{\&tau;}}_s^k\&le;x_{ij}^k\&le;{\mathrm{\&tau;}}_e^k,& \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array}---\left(13\right)$

F4: overstocked amount of metal constraints:

$0\&le;Q_E^k+Q_I^k-Q_S^k-Q_L^k-Q_C^k---\left(14\right).$

S4, opens and waters MES data storehouse that decision controller passes through and MES FTP client FTP obtains each known quantity ginseng in step S2 The value of number；

S5, according to the exploitation intermediate variable of known quantity parameter in step S4, described intermediate variable includes respectively waiting to open watering Secondary casting cycle, each casting machine legacy tasks casting time, each casting machine single bit time water steel amount, collects metal loss factor, whole Overstock in amount of metal formula coefficient before real number decision variable after reason, arrange after coefficient before Real-valued decision variable.

In the present embodiment, it is thus achieved that intermediate variable be:

Respectively wait to open and water the stove number * average Metal Weight of each stove in time casting cycle=respectively wait to open waters time/(wait to open and water time pulling rate * Wait to open and water time basal area * molten steel density)；

Each casting machine legacy tasks casting time=each casting machine legacy tasks molten steel amount/(each casting machine legacy tasks pulling rate * respectively casts Machine legacy tasks basal area * molten steel density)；

Each casting machine single bit time waters steel amount=water time pulling rate * and waters time basal area * molten steel density；

The metal loss factor collected=(1+ metal loss factor/(1-metal loss factor)+whole process logistics time/ ((plan finish time phase-start time plan phase) * (1-metal loss factor))；

After arrangement, in overstocked amount of metal formula, before real number decision variable, coefficient=-each casting machine single bit time waters steel amount * and collects Metal loss factor；

Constant term in amount of metal constraint formulations=estimate into ferrum total amount+opening inventory iron-stochastic demand is overstock after arrangement The metal loss factor * of stock's amount of metal-collect (respectively wait to open and water time casting by each casting machine legacy tasks molten steel amount sum+each casting machine Cycle * each casting machine single bit time waters steel amount+plan phase finish time * each casting machine single bit time and waters steel amount sum)；

During metal loss factor * each casting machine unit that after arrangement, before Real-valued decision variable, coefficient=cost coefficient * collects Between water steel amount+income coefficient/((plan finish time phase-start time plan phase) * conticaster number).

S6, determines each coefficient of object function according to the intermediate variable that step S5 obtains, and waters secondary for the i-th conticaster jth, Order:

F (i, j)=f (i, j)+arrange after before Real-valued decision variable coefficient × conticaster i jth water time casting time x (i, J), (i, j), ((i, j), wherein, (i is j) that Real-valued defined in matlab becomes to sdpvar to described x for i, j)=sdpvar to make p=f The function of amount；

S7, determines the restriction relation in step S3 according to the intermediate variable that step S5 obtains.Particularly as follows:

1) watering time for the i-th conticaster the 1st, make F1 (1,1)=0, i=1,2 ..., H, described H is positive integer；

F1 (i, j)=F1 (i, j)+set (casting machine i legacy tasks casting time+casting machine i water time between minimum time interval * casting First binary variable y of machine i (i, 1)≤first casting time x of casting machine i (i, 1)≤casting machine i legacy tasks casting time + (plan finish time phase-casting machine i legacy tasks first binary variable y of casting time * casting machine i (i, 1)), wherein, y (i, 1) corresponding in constraints(i, j) corresponding in constraints for yX (i, 1) is corresponding in constraints(i, j) corresponding in constraints for xSet () be MATLAB YALMIP optimization tool in constraints letter is set Number；

2) the i-th conticaster jth is watered time, make F2 (1,2)=0, i=1,2 ..., H, $j=2,3,......,N_i^k;$

F2 (i, j)=F2 (i, j)+set ((casting machine i jth-1 casting time x (i, j-1)+casting machine i water time casting cycle)+ Casting machine i water time between minimum time interval * casting machine i jth binary variable y (i, j)≤casting machine i jth casting time x (and i, j) < =(casting machine i jth-1 casting time x (i, j-1)+casting machine i waters time casting cycle)+(plan finish time phase-(casting machine i jth-1 Casting time x (i, j-1)+casting machine i waters time casting cycle)+casting machine i jth binary variable y (i, j)))；

3) the i-th conticaster jth is watered time, make F3 (1,1)=0, i=1,2 ..., H, $j=1,2,......,N_i^k,$

F3 (i, j)=F3 (i, j)+set (plan finish time phase≤conticaster i jth water time casting time x (i, j)≤ Start time plan phase)；

4) the i-th conticaster jth is watered time, make F4 (1,1)=0, i=1,2 ..., H, $j=1,2,......,N_i^k,$

((i, j)+set (overstocks in amount of metal formula coefficient * conticaster i before real number decision variable to F4 after arrangement for i, j)=F4 Jth waters time casting time x (i overstocks constant term in amount of metal constraint formulations after j)≤arrangement)；

5) F1=F1 (i, j), F2=F2 (i, j), F3=F3 (i, j), F4=F4 (i, j)；Total restriction relation: F=F1+ F2+F3+F4。

S8, utilizes Optimization Solution device solvesdp (F ,-p), obtain kth plan phase the i-th casting machine jth wait open water time open water time CarveAnd whether connect, between watering time, the parameter wateredWherein, solvesdp () is matlab majorized function, and F is total restriction relation Identifier, object function identifier total when being maximizing for-p；

S9, controller is according to obtainingData carry out the shop job scheduling of steel smelting-continuous casting production, and to life Produce operation system to implement effectively to control.

Owing to CSDCC problem comprises real number decision variable simultaneouslyWith 0-1 decision variableBelong to mixed integer programming to ask Topic, and solvesdp is exactly the function solving above decision problem.Solvesdp function can be by input model target letter Number, constraints, variable quantity and classification, the identification of the feature such as independence of variable, solve MIXED INTEGER in auto-call library Special solved function BNB20 of planning is consistent with the method for solving needed for CSDCC problem.

It is as follows that its branch and bound method used (function BNB20) solves CSDCC problem principle:

The first step. arranging initial continuous casting and producing total benefit is f=-∞, by the 0-1 decision variable of former CSDCC problemPine The continuous variable that relaxation is 0-1 scope, solves the optimal solution of the CSDCC* problem after relaxing, if each solvedComponent is all 0 Or 1 value, then in this solutionBe exactly CSDCC problem optimum respectively wait to open water time casting time and water time between connect the state of watering. Target function value is both to have obtained continuous casting to produce total benefit.

If second step. in solution at least oneValue b of variable (i, j)^kFor non-zero or 1, then to b (i, j)^kRound nearby [b(i,j)]^k, and to CSDCC problem addition of constraints respectivelyNew to form two CSDCC* ', CSDCC* " problem, the most respectively by continuous variable to CSDCC* ', CSDCC* " seek optimal solution.

3rd step. by current solution CSDCC* ', CSDCC* " continuous casting that obtains produce total benefit with both continuous casting produces total Benefit comparison, finds out the maximum and produces, as continuous casting, the upper bound f*. that total benefit is new

If the 4th step. CSDCC* ', CSDCC* " continuous casting produce in total benefit and have less than f*, then cut this branch；If More than or equal to f* butDo not meet 0 or 1 value, then repeat the continuation of second step method rightVariable branch. until all ofIt is all 0 or 1 value, nowBe exactly CSDCC problem respectively wait to open water time casting time and water time between connect the state of watering, f* is Excellent continuous casting produces total benefit.As can be seen here, those skilled in the art utilize the method for the present invention just can accurately obtain kth plan Phase the i-th casting machine jth is waited to open to water time to open and is watered the moment

The angle of molten iron/molten steel metals resources balance that the present invention is to be processed from steel mill's production line, establishes with continuous casting Produce total benefit to be the conticaster of object function to the maximum and open and water the mixed integer programming algorithm of decision-making and solve.It is capable of continuous casting Machine waters the scientific algorithm of time casting time, contributes to working out rational production schedule of steelmaking plant production plan, stablize each order of classes or grades at school it Between working condition, reduce the overstocked amount of metal on production line, opening in order to water and provide technological means for steel plant continuous casting machine.

In a preferred embodiment of the present invention, open with certain steel plant continuous casting machine domestic and water decision-making real data as reality Example, carries out the application validity check of algorithm. and compare analysis with producing actual decision-making (by manually determining in production) result.

Example enterprises has 5 conticasters, and in production, CSDCC is by manually determining.Take the creation data of 3 continuous teams and groups such as Under: $Q_{I1}^k=5909t,Q_{I2}^k=5575t,Q_{I3}^k=5323t,Q_E^k=2815t,Q_{rcon}^k=290t,{\mathrm{\&tau;}}_A^k=270\min ,$ ρ=7.1t/ m³、 ${\mathrm{s\&tau;}}_{1j}^k=80\min ,{\mathrm{s\&tau;}}_{2j}^k=80\min ,{\mathrm{s\&tau;}}_{3j}^k=90\min ,{\mathrm{s\&tau;}}_{4j}^k=90\min ,{\mathrm{s\&tau;}}_{5j}^k=100\min ,$ C₁=43.5 yuan/1%, C₂ =1.8 yuan/t, each class conticaster parameter is as shown in table 1, wherein, and N_oFor casting machine number, gr₁For teams and groups number.When the opening of actual decision-making is watered Between Gantt chart as shown in Figure 3, using this as the basis of inventive algorithm results contrast.

Experiment point two groups: one group is with teams and groups' activity duration for the plan phase, and each conticaster of each teams and groups waters steel and always water number of times, enter Iron is identical with producing reality, is called for short by class's decision-making；Another group is with the whole day activity duration for the plan phase, and each conticaster of whole day is opened Water always water number of times, enter iron with produce reality identical, be called for short per diem decision-making.

Table 1. each teams and groups conticaster parameter

Utilizing computational methods of the present invention to calculate, obtain such as the data of Fig. 4, Fig. 5 and Biao 2, Fig. 4 is opening by class's decision-making Water time Gantt chart.Fig. 5 is the casting time Gantt chart of per diem decision-making. table 2 is with casting time as foundation, by above-mentioned relevant general Read calculate actual and open the index contrast tables such as desired value and metal balance that the method watering decision-making and utilize the present invention calculates, table Middle band " * " numerical value has been deducted and has been unsatisfactory for (14) formula and overstocks the value of calculation of amount of metal nonnegativity restrictions part.

Table 2 decision objective and the contrast of metals resources balance index

For being analyzed as follows of Fig. 3-Fig. 5 and Biao 2: walkthrough

1) result of decision calculated in the present inventive method is substantially better than actual decision-making in terms of producing stability contorting. and actual Decision-making occurs in that in 201min waters steel while 5 conticasters, the situation of only 1 conticaster casting in 66min；And the present invention The result of calculation of method has remained that within the casting plan phase 2 and above conticaster even water, and produces relative equilibrium stable. from In table 2 actual decision-making with also have embodiment by the standard deviation data of class's decision-making. its reason is: relative to the method for the present invention, for Reach the casting sequence of each casting plan, carry out quantifying meter to metals resources balance on production line owing to lacking during actual decision-making The means calculated, teams and groups 1,2 all occur in that because of unlatching conticaster in advance and cause-138t ,-1946t respectively and overstock amount of metal, leading The safety on line stock's amount of metal having caused respective numbers is not enough, and causes teams and groups 3 to postpone to open conticaster, makes again the plan phase terminate Time overstocked amount of metal increase. and during the casting time of the conticaster utilizing the method for the present invention to calculate, because having considered ferrum The influence factor of steel equilibrium of stock has also quantified the constituent relation between each factor, effectively prevent because of the unreasonable of casting time The production instability problem produced.

2) method of the present invention is better than actual decision-making in terms of total benefit. and essentially identical in whole day average serialization degree In the case of, the cost that the overstocked amount of metal difference that main reason is that under different decision mode of total benefit difference is caused occurs Change is caused, and wherein, actual decision-making have impact on total benefit because its overstocked amount of metal significantly deviate from its control standard.

3) when be intended to phase time range from a class be loosened to one day (three classes) and only whole day scope limit each company Casting machine wait open water time sum time, the present invention can find in bigger time range and meet each conticaster of new constraints, respectively waters Secondary optimization casting time, and it is simultaneously achieved decision objective and the further optimization of the metals resources each index of balance, embody Stronger Parameters variation adapts to ability. and in actual applications, the increase plan phase has reduction to predict the risk into iron accuracy, also The effectiveness of the result of decision may be affected, therefore, it should in the decision basis with the daily planning phase, with teams and groups' activity duration be The plan phase carries out opening calculation and check and the adjustment watering decision-making.

Conticaster of the present invention is opened and watered the purpose of decision-making is on the basis of guaranteeing to produce stable row, is overstock by minimizing and is giving birth to Produce overstocked amount of metal on line to reduce cost, to improve the serialization degree of conticaster to increase benefit, finally realize continuous casting and give birth to Produce the maximization of total benefit.Accumulative steel amount and opening inventory amount of metal, safety on line stock's metal is watered considering the plan phase Measure, add up, on the basis of iron, the overstocked factor such as amount of metal mutual relation and process time constraint, to establish with CSDCC total Benefit is the Integer programming of target to the maximum.In producing according to reality, each order of classes or grades at school company alternately waters state and plans even to water The feature that stove number is relevant, devise using safety on line stock's amount of metal of previous plan phase with overstocked amount of metal sum as rear one The processing method of the opening inventory amount of metal of plan phase, and apply YALMIP optimization toolbox based on MATLAB to solve. Application example shows, the present invention can solve the casting time decision problem of steel mill multiple stage conticaster. under the same conditions with Actual decision-making compares, and can improve the total benefit of decision-making, reduce overstock on production line metals resources amount, stablize adjacent teams and groups Between working condition, opening in order to water and provide technological means for steel plant continuous casting machine.

In the description of this specification, reference term " embodiment ", " some embodiments ", " example ", " specifically show Example " or the description of " some examples " etc. means to combine this embodiment or example describes specific features, structure, material or spy Point is contained at least one embodiment or the example of the present invention.In this manual, to the schematic representation of above-mentioned term not Necessarily refer to identical embodiment or example.And, the specific features of description, structure, material or feature can be any One or more embodiments or example in combine in an appropriate manner.

Although an embodiment of the present invention has been shown and described, it will be understood by those skilled in the art that: not These embodiments can be carried out multiple change in the case of departing from the principle of the present invention and objective, revise, replace and modification, this The scope of invention is limited by claim and equivalent thereof.

Claims (4)

1. the casting time of a steel plant continuous casting machine determines method, it is characterised in that comprise the steps:

S1, opens and waters decision controller and be connected with MES data storehouse and the MES FTP client FTP of steel mill respectively；

S2, sets up to open and waters objective function of decision-making, and described object function is:

$\max f=C_1{\mathrm{\&Theta;}}^k-C_2{Q}_o^k---\left(1\right)$

$\begin{array}{c}{\mathrm{\&Theta;}}^k=\frac{\underset{i=1}{\overset{H}{\mathrm{\&Sigma;}}}(\frac{{\mathrm{rq}}_{ij}^k}{{\mathrm{rs}}_{ij}^k\&CenterDot;\mathrm{\&rho;}\&CenterDot;{\mathrm{ra}}_{ij}^k}+\underset{j=1}{\overset{N_i^k}{\mathrm{\&Sigma;}}}(\min ({\mathrm{\&tau;}}_e^k,x_{ij}^k+{\mathrm{p\&tau;}}_{ij}^k)-x_{ij}^k\left)\right)}{({\mathrm{\&tau;}}_e^k-{\mathrm{\&tau;}}_s^k)\&CenterDot;H}\&times;100\%\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array},---\left(2\right)$

$Q_o^k=Q_E^k+Q_I^k-Q_S^k-Q_L^k-Q_C^k---\left(3\right)$

$\begin{array}{c}{Q}_C^k=\underset{i=1}{\overset{H}{\mathrm{\&Sigma;}}}{\mathrm{rq}}_{ij}^k+\underset{i=1}{\overset{H}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{N_i^k}{\mathrm{\&Sigma;}}}\left(\min \right({\mathrm{\&tau;}}_e^k,x_{ij}^k+{\mathrm{p\&tau;}}_{ij}^k)-x_{ij}^k)\&CenterDot;\mathrm{\&rho;}\&CenterDot;{\mathrm{wa}}_{ij}^k\&CenterDot;{\mathrm{ws}}_{ij}^k,\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array}---\left(4\right)$

$Q_{ca}^k=\frac{Q_C^k}{{\mathrm{\&tau;}}_e^k-{\mathrm{\&tau;}}_s^k}---\left(5\right)$

$Q_L^k=Q_C^k\&CenterDot;\frac{\mathrm{\&eta;}}{1-\mathrm{\&eta;}}---\left(6\right)$

$Q_a^k=\frac{{\mathrm{\&tau;}}_A^k\&CenterDot;Q_{ca}^k}{1-\mathrm{\&eta;}}---\left(7\right)$

$Q_S^k=Q_a^k+Q_{rcon}^k---\left(8\right)$

Wherein, k is plan phase sequence number, k ∈ 1,2 ..., and Ω }, Ω is plan phase sum；I is casting machine sequence number, i ∈ 1,2 ..., H}, H are casting machine sum；J for watering sequence number, It is that the i-th casting machine waters time sum；ρ is molten steel density；η For metal loss factor；q₀Steel amount is averagely watered for each heat；Whole process average stream time kth plan phase；For kth plan Start time phase；For finish time kth plan phase；For kth plan phase the i-th casting machine jth casting plan casting furnace number； Time minimum interval is watered for kth plan phase the i-th casting machine jth；At the end of watering time legacy tasks for kth plan phase the i-th casting machine jth Carve；Time legacy tasks average pull rate is watered for kth plan phase the i-th casting machine jth；Water secondary for kth plan phase the i-th casting machine jth Casting cycle；Wait to open for kth plan phase the i-th casting machine jth and water time average pull rate；Water for kth plan phase the i-th casting machine jth Secondary legacy tasks basal area；Wait to open for kth plan phase the i-th casting machine jth and water subtask basal area；For the kth plan phase i-th Casting machine jth is watered time legacy tasks and is watered steel amount；C1 is benefit coefficient；C2 is cost coefficient；Iron is entered for the kth plan phase, for Known quantity；For stock's amount of metal that kth plan is initial, for known quantity；For the overstocked amount of metal in the kth plan end of term, for not Know process variable；For the line safety inventory amount of metal of kth plan phase, for Unknown Process variable；Damage for kth plan phase metal Consumption, for Unknown Process variable；Steel amount is watered, for Unknown Process variable for the kth plan phase；For kth plan phase the i-th casting machine J waits to open to water time to open and waters the moment, for unknown number to be asked；

Formula (1) represents that object function is deducted the cost structure of overstocked amount of metal by serialization degree income；

Formula (2) represents serialization degree (Θ^k) water time casting time equal to the legacy tasks casting time of each conticaster with waiting to open Sum is divided by H platform conticaster total time plan phase；

Formula (3) represents overstocked amount of metalIron is entered respectively by this plan phaseOpening inventory amount of metalOnline Safety inventory amount of metalMetal loss amountWater steel amountRelation constitute；

Formula (4) the expression plan phase adds up to water steel amount and time waters steel amount sum equal to the legacy tasks amount of each conticaster with waiting to open to water；

Formula (5) represents that the H platform conticaster unit interval within the plan phase averagely waters steel amountSteel amount is watered divided by meter equal to accumulative Draw phase length；

Formula (6) represents when planning phase metal loss factor and being known, accumulative metal loss amount and the accumulative mutual relation watering steel amount；

Formula (7) represents when averagely watering the steel amount average line stock amount of metal of calculating with the unit intervalTime, it should utilize metal Loss factor η is converted to the iron water amount of respective numbers；

Formula (8) represents that safety on line stock's amount of metal needs to add a random fluctuation on the basis of average line stock amount of metal Demand amount of metal

S3, set up wait to open water time and legacy tasks constraints, wait to open whether connect between watering time water constraints, respectively wait to open to water Secondary time and plan beginning and ending time phase constraints and the constraints of overstocked amount of metal；

S4, controller obtains taking of each known quantity parameter in step S2 by the MES data storehouse of steel mill and MES FTP client FTP Value；

S5, according to physical quantity known in step S4 and the exploitation intermediate variable of parameter, described intermediate variable includes respectively waiting out Water time casting cycle, each casting machine legacy tasks casting time, each casting machine single bit time water steel amount, the metal loss factor that collects, Overstock in amount of metal formula coefficient before real number decision variable after arrangement, arrange after coefficient before Real-valued decision variable；

S6, determines each coefficient of object function according to the intermediate variable that step S5 obtains, and waters secondary for the i-th conticaster jth, order:

F (i, j)=f (i, j)+arrange after before Real-valued decision variable coefficient × conticaster i jth water time casting time x (i, j), (i, j), ((i, j), wherein, (i j) is Real-valued variable defined in matlab to sdpvar to described x for i, j)=sdpvar to make p=f Function；

S7, determines the restriction relation in step S3 according to the intermediate variable that step S5 obtains；

S8, utilizes Optimization Solution device solvesdp (F ,-p), obtains kth plan phase the i-th casting machine jth and waits to open to water time to open and water the moment And whether connect between watering time and to waterWherein, solvesdp () is matlab majorized function, and F is total restriction relation identifier ,-p For object function identifier total during maximizing；

S9, controller is according to obtainingData carry out the shop job scheduling of steel smelting-continuous casting production, and to production run System is implemented effectively to control.

2. the casting time of steel plant continuous casting machine as claimed in claim 1 determines method, it is characterised in that: in described step S3 Constraints is:

1) respectively wait to open the company between watering time and water relation constraint:

$\begin{array}{c}{\mathrm{\&tau;}}^{\&prime;}+{\mathrm{s\&tau;}}_{ij}^k\&CenterDot;y_{ij}^k\&le;x_{ij}^k\&le;{\mathrm{\&tau;}}^{\&prime;}+({\mathrm{\&tau;}}_e^k-{\mathrm{\&tau;}}^{\&prime;})\&CenterDot;y_{ij}^k,\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;\end{array}---\left(9\right)$

Wherein,For binary variable, 1 represents that kth plan phase the i-th casting machine jth is watered secondary company with legacy tasks and watered, and 0 represents disconnected waters,For unknown number to be asked；

F1: when judge wait open water time whether connect between legacy tasks water time, in formula (9)

${\mathrm{\&tau;}}^{\&prime;}={\mathrm{r\&tau;}}_{ij}^k=\frac{{\mathrm{rq}}_{ij}^k}{\mathrm{\&rho;}\&CenterDot;{\mathrm{rs}}_{ij}^k\&CenterDot;{\mathrm{ra}}_{ij}^k},\&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;---\left(10\right)$

$\begin{array}{c}{\mathrm{rq}}_{ij}^k=n_{i(j=N_i^{k-1})}^{k-1}\&CenterDot;q_o-({\mathrm{\&tau;}}_e^{k-1}-{\mathrm{w\&tau;}}_{i(j=N_i^{k-1})}^{k-1})\&CenterDot;\mathrm{\&rho;}\&CenterDot;{\mathrm{wa}}_{i(j=N_i^{k-1})}^{k-1}\&CenterDot;{\mathrm{ws}}_{i(j=N_i^{k-1})}^{k-1},\\ \&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^{k-1}\&rsqb;\end{array}---\left(11\right)$

Formula (9-10) represent when wait open water time even water with legacy tasks time, casting time point takes legacy tasks finish time, non-company Taking legacy tasks finish time when watering terminates certain moment to the plan phase plus watering the minor tick time；

Formula (11) represents that i-th conticaster jth of kth plan phase is watered time legacy tasks amount and watered equal to the plan of this casting machine previous plan phase Steel amount and the difference of casting amount；

F2: when decision-making remaining respectively wait open water time between whether connect water time, in formula (9)

${\mathrm{\&tau;}}^{\&prime;}=x_{i(j-1)}^k+{\mathrm{p\&tau;}}_{ij}^k,\&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;---\left(12\right)$

Formula (9), (12) represent when except remaining adjacent with legacy tasks respectively wait open water that time interconnection is watered mutually time, this casting time point takes Before water time finish time, non-company waters time finish time and terminates certain moment to the plan phase plus watering the minor tick time before taking when watering；

F3: respectively wait to open time time of watering and plan beginning and ending time phase relation constraint condition:

${\mathrm{\&tau;}}_s^k\&le;x_{ij}^k\&le;{\mathrm{\&tau;}}_e^k,\&ForAll;i\&Element;\&lsqb;1,H\&rsqb;,\&ForAll;j\&Element;\&lsqb;1,N_i^k\&rsqb;---\left(13\right)$

F4: overstocked amount of metal constraints:

$0\&le;Q_E^k+Q_I^k-Q_S^k-Q_L^k-Q_C^k---\left(14\right).$

3. the casting time of steel plant continuous casting machine as claimed in claim 1 determines method, it is characterised in that: in described step S5 The intermediate variable obtained is:

Respectively wait to open and water the stove number * average Metal Weight of each stove in time casting cycle=respectively wait to open waters time/(wait to open and water time pulling rate * and wait out Water time basal area * molten steel density)；

Each casting machine legacy tasks casting time=each casting machine legacy tasks molten steel amount/(each casting machine legacy tasks each casting machine of pulling rate * is lost Remain in office business basal area * molten steel density)；

Each casting machine single bit time waters steel amount=water time pulling rate * and waters time basal area * molten steel density；

The metal loss factor collected=(1+ metal loss factor/(1-metal loss factor)+whole process logistics time/((meter Draw finish time phase-start time plan phase) * (1-metal loss factor))；

After arrangement, in overstocked amount of metal formula, before real number decision variable, coefficient=-each casting machine single bit time waters the gold that steel amount * collects Belong to loss factor；

Constant term in amount of metal constraint formulations=estimate into ferrum total amount+opening inventory iron-stochastic demand stock is overstock after arrangement The metal loss factor * of amount of metal-collect (respectively wait to open and water time casting cycle * by each casting machine legacy tasks molten steel amount sum+each casting machine Each casting machine single bit time waters steel amount+plan phase finish time * each casting machine single bit time and waters steel amount sum)；

Metal loss factor * each casting machine single bit time that after arrangement, before Real-valued decision variable, coefficient=cost coefficient * collects waters Steel amount+income coefficient/((plan finish time phase-start time plan phase) * conticaster number).

4. the casting time of steel plant continuous casting machine as claimed in claim 1 determines method, it is characterised in that: described according to step The method of the restriction relation in intermediate variable set-up procedure S3 that S5 obtains is:

1) watering time for the i-th conticaster the 1st, make F1 (1,1)=0, i=1,2 ..., H, described H is positive integer；

F1 (i, j)=F1 (i, j)+set (casting machine i legacy tasks casting time+casting machine i water time between minimum time interval * casting machine i First binary variable y (i, 1)≤first casting time x of casting machine i (i, 1)≤casting machine i legacy tasks casting time+ (plan finish time phase-casting machine i legacy tasks first binary variable y of casting time * casting machine i (i, 1)), wherein, y (i, 1) corresponding in constraints(i, j) corresponding in constraints for yX (i, 1) is corresponding in constraints (i, j) corresponding in constraints for xSet () be MATLAB YALMIP optimization tool in constraints function is set；

2) the i-th conticaster jth is watered time, makes F2 (1,2)=0, i=1,2 ..., H, j=2,3 ...,

F2 (i, j)=F2 (i, j)+set ((casting machine i jth-1 casting time x (i, j-1)+casting machine i waters time casting cycle)+casting machine i Minimum time interval * casting machine i jth binary variable y (i, j)≤casting machine i jth casting time x (i, j)≤(casting between watering time Machine i jth-1 casting time x (i, j-1)+casting machine i waters time casting cycle)+(finish time phase-(casting machine i jth-1 is opened when watering in plan Between x (i, j-1)+casting machine i water time casting cycle)+casting machine i jth binary variable y (i, j)))；

3) the i-th conticaster jth is watered time, makes F3 (1,1)=0, i=1,2 ..., H, j=1,2 ...,

((i, j) (plan finish time phase≤conticaster i jth waters time casting time x (i, j)≤plan to+set to F3 for i, j)=F3 Start time phase)；

4) the i-th conticaster jth is watered time, makes F4 (1,1)=0, i=1,2 ..., H, j=1,2 ...,

((i, j)+set (overstocks in amount of metal formula coefficient * conticaster i jth before real number decision variable to F4 after arrangement for i, j)=F4 Water time casting time x (i overstocks constant term in amount of metal constraint formulations after j)≤arrangement)；

5) make F1=F1 (i, j), F2=F2 (i, j), F3=F3 (i, j), F4=F4 (i, j)；

Total restriction relation: F=(F1+F2+F3+F4).