WO2017088674A1 - 一种面向全流程生产的炼钢组批与排产方法 - Google Patents

一种面向全流程生产的炼钢组批与排产方法 Download PDF

Info

Publication number
WO2017088674A1
WO2017088674A1 PCT/CN2016/105581 CN2016105581W WO2017088674A1 WO 2017088674 A1 WO2017088674 A1 WO 2017088674A1 CN 2016105581 W CN2016105581 W CN 2016105581W WO 2017088674 A1 WO2017088674 A1 WO 2017088674A1
Authority
WO
WIPO (PCT)
Prior art keywords
batch
slab
continuous casting
production
contract
Prior art date
Application number
PCT/CN2016/105581
Other languages
English (en)
French (fr)
Inventor
唐立新
汪恭书
徐文杰
Original Assignee
东北大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 东北大学 filed Critical 东北大学
Priority to US15/529,745 priority Critical patent/US10713399B2/en
Publication of WO2017088674A1 publication Critical patent/WO2017088674A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • 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
    • 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
    • C21C5/42—Constructional features of converters
    • C21C5/46—Details or accessories
    • C21C5/4673—Measuring and sampling devices
    • 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
    • C21C5/52—Manufacture of steel in electric furnaces
    • C21C5/5294—General arrangement or layout of the electric melt shop
    • C—CHEMISTRY; METALLURGY
    • C21—METALLURGY OF IRON
    • C21D—MODIFYING THE PHYSICAL STRUCTURE OF FERROUS METALS; GENERAL DEVICES FOR HEAT TREATMENT OF FERROUS OR NON-FERROUS METALS OR ALLOYS; MAKING METAL MALLEABLE, e.g. BY DECARBURISATION OR TEMPERING
    • C21D11/00—Process control or regulation for heat treatments
    • C—CHEMISTRY; METALLURGY
    • C23—COATING METALLIC MATERIAL; COATING MATERIAL WITH METALLIC MATERIAL; CHEMICAL SURFACE TREATMENT; DIFFUSION TREATMENT OF METALLIC MATERIAL; COATING BY VACUUM EVAPORATION, BY SPUTTERING, BY ION IMPLANTATION OR BY CHEMICAL VAPOUR DEPOSITION, IN GENERAL; INHIBITING CORROSION OF METALLIC MATERIAL OR INCRUSTATION IN GENERAL
    • C23G—CLEANING OR DE-GREASING OF METALLIC MATERIAL BY CHEMICAL METHODS OTHER THAN ELECTROLYSIS
    • C23G1/00—Cleaning or pickling metallic material with solutions or molten salts
    • C23G1/02—Cleaning or pickling metallic material with solutions or molten salts with acid solutions
    • C23G1/08—Iron or steel
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/04—Manufacturing
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
    • G16Z99/00—Subject matter not provided for in other main groups of this subclass
    • 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
    • C21C5/52—Manufacture of steel in electric furnaces
    • C21C2005/5288—Measuring or sampling devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • 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
    • Y02P10/20—Recycling
    • 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
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Definitions

  • the invention belongs to the technical field of metallurgical automatic control, and particularly relates to a batching and scheduling method for a steelmaking group for whole process production.
  • a prominent feature of the steel production process is the continuous physical and chemical reactions of the material flow in the process, which are constantly changing in terms of state, nature and shape.
  • the scale of the devices for realizing the physical and chemical reactions of the material flow in the steel production is very large, and the devices are
  • the operation mode includes continuous, quasi-continuous and batch production; in the continuous production process, the raw materials are continuously passed through the same equipment, each equipment is in a relatively stable state and only performs a specific processing task, semi-finished or finished product. Output in a continuous flow; in the batch process, the raw materials are processed according to the set processing order and operating conditions, and the products are output in batch mode; quasi-continuous production is between continuous and intermittent production.
  • the blast furnace production is a continuous operation process, which is always in a production state except for maintenance, and the molten iron is continuously outputted at each tapping port. Therefore, the iron making process is a typical continuous production process.
  • the converter or electric arc furnace can smelt 150-300 tons of molten steel each time, which is called a heat; after the molten steel smelting of each heat is completed and poured into the ladle, the converter or electric arc furnace can be cleaned after the next step.
  • the steelmaking process corresponds to one batch processing process per smelting-heating steel; the molten steel of multiple heats can be continuously cast on the continuous casting machine, but due to the service life of the crystallizer, the tundish and the steel grade And the influence of the diversity of the billet specifications, the continuous casting machine will stop after a certain number of furnaces, and the continuous casting machine will be cleaned, and the key equipment - tundish and crystallizer will need to be replaced, and this process often needs 2-3 hours; therefore, the continuous casting process belongs to the batch production process, and each production batch of the continuous casting process is called a pouring batch, which is defined as the number of times of continuous casting using the same tundish and crystallizer on the same continuous casting machine.
  • Continuous production process the production of a single product, in addition to the total capacity of the macro-control, can not be controlled from the micro-level of the physical and chemical properties of the product; but the batch production process is different, before the production of each batch of products It is usually necessary to batch-produce products with physicochemical properties with similar process requirements, thereby increasing production efficiency and yield, and reducing resource consumption and energy consumption.
  • the solid semi-finished products and finished products such as slab, hot coil and cold coil are closely related to customer needs, and the production organization process is order-oriented; the steelmaking process is oriented to orders.
  • the source process of production, the number of downstream units is numerous, and the layout of the production line is complicated. Therefore, when determining the production organization arrangement in the steelmaking stage, including determining the composition decision of the batch, it is necessary to pay close attention to the existing material inventory and material demand of the downstream unit, and equalize the unit.
  • This technical problem is also called the optimization of the steelmaking batch planning and batch scheduling for the whole process production. It is necessary to decide the batch composition of the steelmaking and continuous casting processes under the qualification conditions of the production process including the whole process of the whole process. Batch allocation and sorting, as well as material flow distribution between downstream processes in steelmaking.
  • the quantitative scientific calculation method is used to determine the batch planning and batch scheduling of the steelmaking group for the whole process production, on the basis of ensuring the normal operation of the equipment, reasonable Arrange the production organization of customer demand products to improve equipment production efficiency, improve inventory structure, streamline production logistics, optimize the production process level of steel whole process, improve the control level of production process, reduce process energy consumption and low cost manufacturing. Significance.
  • the published patent (“ZL200810011659.1”, a steelmaking-continuous casting furnace batch plan automatic preparation method and system) mainly realizes the mass production of the steelmaking process by combining the slabs into the heat generation;
  • the published patent (“ZL200610046981.9”, a steelmaking-continuous casting tundish batch planning method and system) mainly realizes batching of the furnace group to the tundish to realize mass production of the continuous casting machine.
  • Patent ZL200810011659.1 and patent ZL200610046981.9 are mainly to solve the problem of batch production technology in the single process of steel making and continuous casting.
  • the production load balance between multiple workshops and units is not considered, and the materials between the front and back processes are not considered.
  • the flow supply balance, and there is no technical problem such as integrating batch decision making with the optimization of the scheduling decision.
  • the present invention proposes a steelmaking batching and scheduling method for full-process production, in order to achieve the balance and punctual distribution of the steel material flow in the whole process equipment and time dimension.
  • a steelmaking batch and scheduling method for full-process production including the following steps:
  • Step 1 Describe the production environment by constructing a directed network topology diagram
  • each node on the directed network topology diagram represents a specific production unit or inventory equipment, including: converter, refining furnace, continuous casting machine, slab library, hot rolling unit, leveling unit, hot coil library, pickling Unit and acid rolling unit; each arc on the directed network topology shows a specific material transfer process from one unit or inventory to another or inventory equipment, including: molten steel, slab, hot coil and cold volume;
  • Step 2 Set the process parameters of the product according to the quality requirements of the final product according to different customer contracts, including: determining the mapping of the manufacturing process of the product on the directed network topology map, and calculating the casting width of different products on the continuous casting machine according to the steel type. Scope, determine the superior replacement relationship between steel grades, determine the mixing relationship and cost of different steel grades in the tundish;
  • Step 3 According to the steel type, variety attribute, optional manufacturing process and width range of the product, the product order group is judged. If the total demand of the customer is greater than or equal to the maximum number of continuous furnaces allowed by the tundish, then Belong to the large contract group, perform step 6; if the total product under-required by the customer is less than the maximum number of continuous furnaces allowed by the intermediate package, then belong to the small contract group, and perform steps 4 to 5;
  • Step 4 Describe the batch production decision of the multi-product in the steelmaking process by constructing a mathematical model
  • Step 4-1 Mapping the multi-product group batch plan in the steelmaking production process to a mathematical model decision variable
  • Step 4-2. Map the process limits of the steelmaking production process to mathematical model constraints, as follows:
  • Step 4-2-1 establishing process constraints for the replacement relationship of product steel grades
  • Step 4-2-2 establishing a process constraint of the casting width range of the product on the continuous casting equipment
  • Step 4-2-3 Establish a process constraint for limiting the smelting capacity of each batch of the converter, that is, smelting in the same batch.
  • the total weight of the slab and un-commissioned slabs required by the customer contract shall be close to the standard smelting capacity of the converter.
  • the weight of the part of the smelting capacity exceeding the standard smelting capacity of the converter and the weight of the standard smelting capacity of the insufficient converter shall be less than the weight of one slab;
  • the un-commissioned slab refers to the remaining materials that are not managed by the customer contract to meet the requirements of the full batch production process in the converter smelting process;
  • Step 4-2-4 Establishing the furnace flow balance process of the two streams of molten steel in the casting process of each furnace molten steel, that is, the casting time of the same furnace molten steel in the continuous casting machine is equal, in the model The number of slabs mapped onto the two streams is equal;
  • Step 4-2-5 establish the process constraint of cutting the length range of the slab on the continuous casting equipment, that is, the cutting process of the continuous casting machine and the length of the customer ordering, the length of any slab cast in a furnace steel water is required to be Within the prescribed range;
  • Step 4-2-6 construct a flexible management constraint on the customer's order quantity, that is, the part that is insufficient or exceeds the customer's order quantity is less than the weight of one slab;
  • Step 4-3 Mapping the optimized process index in the steelmaking production process to the objective function of the mathematical model, minimizing the total weight of the untrusted slabs of all batches of output, minimizing the total amount of replacement between the steel grades, Minimize the total deviation of the slab weight in all batches from the standard smelting capacity of the converter, and minimize the total deviation of the order quantity of all customer contracts;
  • Step 5 Construct a mutual mapping relationship between a real number matrix and a group batch scheme, and use the established real matrix as the controlled object to implement a multi-object parallel iterative improvement strategy to obtain a final optimized group batch scheme, and then obtain a small
  • the pre-grouping plan of the contract group in the continuous casting process is as follows:
  • Step 5-1 Construct a mutual mapping relationship between a real number matrix and a batch batch scheme, as follows:
  • Step 5-1-1 Construct a real matrix whose dimension is the product of the total number of products, the steel grade and the width.
  • the elements in the matrix are all batches assigned to a certain steel type and a certain width.
  • Step 5-1-2 Obtain the weight of the slab in all batches assigned to the target steel grade and the target width and the weight of all the contract slabs in all batches assigned to the target steel grade and the target width, For all steel grades and width combinations, sort the weight values of all contract slabs in all batches from large to small, and repeat steps 5-1-3 to 5-1-9 in this order;
  • Step 5-1-3 determine the slab weight vector assigned to all the steel grade and width combination batches, and construct an empty batch, and set the slab weight already included in the batch to 0;
  • Step 5-1-4 selecting the contract of the first slab weight greater than 0 in the slab weight vector, and comparing the remaining capacity of the empty batch with the weight of the first slab, if the remaining capacity is greater than or equal to the first For the slab weight, perform steps 5-1-5; otherwise, perform steps 5-1-6;
  • Step 5-1-5 replacing the production under-consumption of the corresponding product in the flexible management constraint of the customer order quantity with the product slab Weight, and obtain the slab of the integer block according to the process conditions defined in steps 4-2-5 to 4-2-6, put the slab into the empty batch, update the slab weight of the batch and set it at The slab weight of the product in the slab weight vector is 0;
  • Step 5-1-6 replacing the production under-consumption of the corresponding product in the flexible management constraint condition of the customer order quantity with the remaining capacity, and obtaining the integer block according to the process conditions defined in steps 4-2-5 to 4-2-6.
  • the slab put it into the empty batch, update the slab weight of the batch and set the slab weight in the slab weight vector to 0;
  • Step 5-1-7 without adding a non-commissioned slab, determine whether the slab in the empty batch meets the process constraints of the limit of each batch of smelting capacity of the converter, and if so, perform step 5- 1-8, otherwise, perform steps 5-1-9;
  • Step 5-1-8 judging whether the slab containing in the empty batch meets the furnace flow balance process constraint condition of the two-flow steel water consumption, and if so, directly creating the next empty batch without any contract, and setting the The weight of the slab already contained in the batch is 0. Otherwise, the batch is repaired by adding or removing a slab from the empty batch to satisfy the furnace flow balance constraint, and then the next one does not contain any contract. Empty batch, and set the weight of the slab already included in the batch to 0;
  • Step 5-1-9 determine whether the slab weight vector is equal to 0, and if so, in the last batch that is not empty, according to the process constraints of each batch of smelting capacity of the converter and the two-flow steel consumption
  • the furnace flow balance process constraint condition is added without the commissioned slab, otherwise, return to step 5-1-4;
  • Steps 5-1-10 the weights of all contract slabs in all batches are performed in steps 5-1-3 to 5-1-9 to obtain a batch plan for all types of steel and width combinations;
  • Step 5-2 Using the established real matrix as the controlled object, implementing a multi-object parallel iterative improvement strategy to obtain a final optimized group batch solution, specifically including:
  • Step 5-2 randomly generate NP real number matrices with the same real matrix structure as described in step 5-1-1, and put all the constructed real matrices into the set, each target matrix meets the target steel grade and the target The width of the element is set to 1, and the unsatisfied element is set to 0;
  • NP is a population size parameter preset based on a multi-object parallel iterative improvement strategy algorithm
  • Step 5-2-2 return all the generated real matrix to perform step 5-1-1 to step 5-1-10, establish a correspondence between each real matrix and the group batch scheme, and obtain a decision variable according to the group batch scheme. Take the value, substitute it into the objective function, and obtain the objective function corresponding to each real number matrix;
  • Step 5-2-3 sorting the obtained objective functions from small to large, and dividing the real-number matrix of the first one-half of the ranking into one group, and dividing the real-number matrix of the second-half of the ranking into one group;
  • Step 5-2-4 Perform a mutation operation and a cross operation on each real matrix according to the grouping of the objective function corresponding to each real matrix, obtain the real matrix after the operation, and return all the real matrix after the operation to perform step 5-1.
  • -1 to step 5-1-10 establish the correspondence between the real number matrix and the group batch scheme after each operation, obtain the value of the decision variable according to the group batch scheme, substitute it into the objective function, and obtain the real matrix after each operation.
  • Corresponding objective function ;
  • Step 5-2-5 judging the size of the objective function corresponding to the real matrix before and after the operation, selecting the real matrix corresponding to the smaller objective function as the updated real matrix, obtaining the updated matrix set, and returning to step 5-2- 2 to step 5-2-4 until the matrix set is no longer updated, obtaining the final matrix set;
  • Step 5-2-6 selecting a real matrix with the smallest objective function value in the final matrix set, and returning the matrix to perform step 5-1-1 to step 5-1-10 to obtain a final optimized group batch scheme;
  • Step 5-3 the obtained steelmaking batches are combined according to the steel type and the width, that is, the steelmaking batches having the same steel type and width are combined into one continuous pouring batch group, and the small contract group is completed in the continuous casting process. Designation of a batch plan;
  • Step 6 Formulate a batch plan for the large contract group in the steelmaking process and a pre-batch plan for the continuous casting process;
  • Step 7 Determine the scheduling decision of the continuous casting batch on the continuous casting equipment by constructing a quantitative mathematical model; specifically: selecting the decision variables of the continuous pouring batch scheduling; quantitatively describing the continuous pouring batch scheduling The goal pursued; quantitative description of the process constraints and management requirements to be followed in the development of continuous casting batch scheduling; the specific steps are as follows:
  • Step 7-1 selecting a decision variable for scheduling the continuous batching batch
  • Step 7-2 quantitatively describing the goal pursued by the continuous pouring batch group scheduling
  • tundish includes: maximizing the utilization of the tundish, minimizing the number of slabs of different steel grades, minimizing the number of slabs, minimizing the deviation of the amount of hot rolled stock, minimizing the deviation of the inventory of difficult-to-roll materials, minimizing heat Deviation of demand for rolling and cold rolling and minimizing customer contract delay time;
  • Step 7-3 Quantitatively describe the process constraints and management requirements to be followed in the formulation of the continuous pouring batch scheduling
  • the method includes: a distribution relationship constraint of the continuous casting batch on the continuous casting equipment and a feasible distribution rule constraint;
  • Step 8 The mathematical model established in step 7 is used as a quantitative calculation basis, and a mutual mapping relationship between the real number vector and the continuous casting device on the continuous casting equipment is established, and the established real vector is used as the basis.
  • the controlled object obtains a multi-object parallel iterative improvement strategy to obtain a scheduling scheme for continuous casting equipment on the continuous casting equipment;
  • Step 9 The group batch plan and the scheduling integration plan are adjusted, issued and executed.
  • the product set N g produced by the steel grade is determined
  • N represents the total product set for a given small contract group
  • g i represents the steel grade of product i
  • G ⁇ i ⁇ N
  • g i represents all steel species contained in product set N
  • s gig represents steel of product i
  • the process constraints for establishing the limit of each batch of smelting capacity of the converter as described in Step 4-2-3, that is, the total weight of the slab and un-committed slabs required to meet the customer contract requirements of the smelting in the same batch should be close to the converter standard.
  • the smelting capacity, the weight of the part exceeding the standard smelting capacity of the converter and the weight of the standard smelting capacity of the insufficient converter are less than the weight of one slab;
  • C is the standard smelting capacity of the converter and Q i is the production under-production of the product i.
  • l std indicates the standard length of the un-slab in the continuous casting production
  • h std indicates the standard thickness of the un-slab in the continuous casting production
  • ⁇ indicates the density of the molten steel
  • Step 4-2-4 in the process of casting each furnace molten steel in the continuous casting machine, the flow balance of the two streams of steel is limited, that is, the casting time of the same furnace molten steel in the continuous casting machine is equal.
  • the number of slabs cast on the model mapped to two streams is equal;
  • n gwk represents an auxiliary integer variable, indicating that the steel grade is the number of slab blocks produced by the kth batch of odd flow of g width w;
  • Step 4-2-5 the process constraint of establishing the length range of the slab on the continuous casting equipment, that is, the cutting process of the continuous casting machine and the length of the customer ordering, requiring any slab cast in a furnace steel water
  • the length is within the specified range
  • h i represents the slab thickness required for product i, with Indicates the maximum and minimum length of the slab required for product i;
  • the flexible management constraint for constructing the customer order quantity described in step 4-2-6, that is, the part that is insufficient or exceeds the customer order quantity is less than the weight of one slab;
  • step 4-3 The objective function described in step 4-3 is as follows:
  • F 0 represents the total cost of the production group of the small contract group
  • the total cost of the production group of the small contract group is the total amount of un-committed slabs of all batches, the total replacement quantity of steel grades, and all customer contracts.
  • A represents a
  • N represents the total product set for a given small contract group
  • W represents the cast product set
  • indicates that the weight of the slab in all batches corresponding to the contract N assigned to the steel grade G and the width W is less than the contract N production deficit. Ratio relationship.
  • the slab weights obtained in step 5-1-2 for all batches assigned to the target steel grade and the target width and all contract slabs in all batches assigned to the target steel grade and target width The weight is calculated as follows:
  • igw denotes the contract slab weight assigned to all batches of steel grade g and width w
  • B gw denotes all contract slabs assigned to all batches of steel grade g and width w
  • the weight, a igw represents the ratio of the slab weight in all batches corresponding to the contract i assigned to the steel grade g and width w to the under-production of the contract i
  • Q i represents the production deficit of the product i.
  • the production under-consumption of the corresponding product in the flexible management constraint condition of the customer order quantity is replaced by the weight of the product slab as described in step 5-1-5, that is, the formula
  • the production under-production of product i is replaced by the product slab weight b igw ;
  • step 5-1-6 the production under-consumption of the corresponding product in the flexible management constraint condition of the customer order quantity is replaced with the remaining capacity, that is, the formula
  • the production under-producing of product i is replaced by the remaining capacity CE k
  • C represents the standard smelting capacity of the converter
  • E k represents the weight of the slab already contained in the batch.
  • Step 5-2-4 according to the grouping of the objective function corresponding to each real matrix, perform a mutation operation and a cross operation on each real matrix, and the specific steps are as follows:
  • Step 5-2-4-1 selecting three real numbers A r1 , A different from the target real matrix A j and different from each other in the set ⁇ A 1 , A 2 , ..., A NP ⁇ composed of real numbers matrix R2 , A r3 , ie j ⁇ r1 ⁇ r2 ⁇ r3;
  • Step 5-2-4-2 randomly generating a variable asynchronous long factor F j from the uniformly distributed real interval [j/NP, 1];
  • Step 5-2-4-3 performing a difference operation on the real numbers A j , A r1 , A r2 , and A r3 to obtain a real matrix V j after the mutation operation;
  • j 1, 2, . . . , NP ⁇ ;
  • S′ represents the obtained objective function Sorting from small to large, ranking the first half of the real matrix group, I means sorting the obtained objective functions from small to large, and ranking the second half of the real matrix grouping;
  • Step 5-2-4-4 randomly generating a crossover probability factor CR igw from each element of the matrix from the uniformly distributed real interval [j/NP, 1];
  • Step 5-2-4-5 performing cross operation on each pair of real numbers A j and V j to generate a real matrix U j ;
  • Represents the elements inside the real matrix U j , Represents the elements inside the real matrix V j , Represents the elements inside the real matrix A j , j 1, 2, ..., NP, i ⁇ N, w ⁇ W, k ⁇ ⁇ 1, 2, ..., K gw ⁇ ; W represents the continuous casting machine for the casting product set N The set of all possible widths required for the crystallizer; K gw represents the upper limit of the number of batches of steel grade g and width w; Representing a random number between (0, 1) obeying a normal distribution;
  • Step 5-2-4-6 determine whether the element or If yes, the boundary condition processing is performed on the real number matrix U j , otherwise, the cross operation is completed;
  • L represents an element The lower limit of the range of values
  • U represents the element The upper limit of the value range.
  • the decision variables for selecting the continuous pouring batch group scheduling described in Step 7-1 are as follows:
  • Setting the 0-1 decision variable u rls indicates whether the continuous batch r is allocated at the lth position of the continuous casting device s; setting the auxiliary variable Q ls indicates the continuous pouring batch at the lth position of the continuous casting device s Group casting completion time; setting auxiliary variable T r indicates casting completion time of continuous pouring batch r; setting auxiliary variable t i indicating casting completion time of contract i; setting auxiliary variable Indicates the amount of hot bar stock planned in the hot rolling mill h before the day d; set the auxiliary variables Indicates the planned inventory of difficult-to-roll materials on the d-day of the hot-rolling mill h; the set auxiliary variable ⁇ fd indicates the amount of slab that the steel mill produces for the planned flow on day d;
  • F I represents the total cost of the tundish updated by all the consecutive batches
  • S represents the collection of continuous casting equipment
  • R represents the collection of all consecutive batches
  • a r represents the number of furnaces contained in the continuous batch r
  • Max represents the maximum number of furnaces that can be cast in a tundish
  • b s represents the number of batches that cannot be connected to the continuous casting equipment s
  • l ⁇ Indicates the position of the batch of the ⁇ th non-mixable
  • F 2 represents the total cost of continuous casting of all the different steel types in the continuous pouring batch group; Represents the continuous casting batch assigned at the l position on the continuous casting equipment s, Representing continuous casting batch Steel grade
  • F 3 represents the total cost of all the widening of the continuous batching batch; Representing continuous casting batch Width, the function h(.) takes 0 when the two parameters are the same before, otherwise it takes 1;
  • the deviation between the planned inventory of the hot bar before the hot rolling and the target inventory is required to ensure the smoothness of the hot rolling production
  • F 4 represents the risk of inventory of all hot rolled stocks
  • H represents the collection of hot rolling mills
  • D represents the set of days in the plan period, Indicates the target stock of hot bar stock required by hot rolling mill h on day d;
  • the planned inventory of the difficult-to-roll material in the pre-rolling warehouse is the smallest deviation from the maximum and minimum allowable inventory of the difficult-to-roll material to reduce the clogging of the logistics caused by the excessively difficult-to-roll material;
  • F 5 represents the risk cost of all difficult-to-roll materials inventory; Indicates the largest inventory of difficult-to-roll materials in hot rolling mill h on day d; Indicates the minimum allowable inventory of hot rolled mills on day d;
  • F 6 represents the customer contract satisfaction income
  • N R represents the contract set with strict delivery time requirements
  • Ear i represents the earliest delivery date of contract i
  • Due i represents the latest delivery date of contract i
  • Step 7-3 describes the process constraints and management requirements to be followed in the production of the continuous casting batch, as follows:
  • v rs represents the contract production process and continuous casting equipment s compatibility parameters contained in the continuous batch r, v rs ⁇ ⁇ 0, 1 ⁇ .
  • Step 8-1 setting a 2
  • dimensional real number vector PP [a 1 , a 2 ... a
  • a r and b r are dimensionless real parameters, the value range is [0, 1]; 1 ⁇ r ⁇
  • Step 8-2 determining the continuous pouring batch set Rs assigned to any continuous casting equipment s according to the b r value
  • Step 8-3 determining the order of all consecutive batches assigned to any continuous casting equipment s according to the value of a r ;
  • the continuous batch batch set Rs is sorted according to the a r value from small to large, and the order of all the consecutive batches assigned to the continuous casting equipment s is determined.
  • the invention provides a steelmaking batching and scheduling method for whole process production, and the invention passes the process technology level Finely characterize the diversified specifications of the product and the batch properties of the equipment and the suitability between the product and the equipment, and establish a mathematical model that quantitatively describes the decision-making problem of the product in the steelmaking and continuous casting process, and based on
  • the model design is based on the multi-object parallel iterative improvement method to obtain the batching scheme on the steelmaking process; at the production organization level, from the perspective of the productivity balance between the parallel equipment of the same process and the logistics connection between the upstream and downstream processes, the batch
  • the model of distribution and ordering in continuous casting equipment and time dimension is established.
  • a serial iterative improvement method based on neighborhood search is designed to obtain batch production scheduling scheme in continuous casting equipment and time dimension.
  • the plan is integrated with the production scheduling scheme, and further fine-tuned and then delivered to the production and manufacturing units in the steelmaking stage.
  • Each production unit is prepared and executed according to the plan to achieve the steel material flow in the whole process equipment and time dimension.
  • Equilibrium and punctual distribution; the invention improves product quality, improves yield and resource utilization Operating efficiency of the equipment, materials to achieve a smooth engagement between the serial device and load balancing on parallel devices, reducing traffic congestion stream, latency and downstream equipment inventory, inventory to achieve reasonable control.
  • FIG. 1 is a flow chart of a method for batching and scheduling a steelmaking group for full-process production according to an embodiment of the present invention
  • FIG. 2 is a topological view of a network to an embodiment of the present invention
  • FIG. 3 is a schematic diagram showing a coding structure of a manufacturing process according to an embodiment of the present invention.
  • FIG. 4 is a flow chart for describing a batch production decision of a multi-product in a steelmaking process by constructing a mathematical model according to an embodiment of the present invention
  • FIG. 5 is a flow chart of a pre-grouping method for obtaining a small contract group in a continuous casting process according to an embodiment of the present invention
  • FIG. 6 is a flow chart of a method for determining a scheduling decision of a continuous casting batch on a continuous casting apparatus according to an embodiment of the present invention.
  • This embodiment is a large steel enterprise; the steel enterprise has two steel mills and two hot rolling mills, one of which includes pickling, acid rolling, continuous retreating, hot dip galvanizing, hot dip galvanizing, electroplating tin, cross cutting And re-rolling the cold rolling production line of 8 units; the first steelmaking plant is equipped with 3 converters, 2 sets of RH, LF and Ar refining equipment, and 2 continuous casting machines with a casting width ranging from 750 to 1320 mm, each furnace The standard smelting capacity of molten steel is 150 tons, the maximum number of castable furnaces in the tundish is 8 furnaces; the second steelmaking plant is equipped with 2 converters, one RH, LF and Ar refining equipment, and two casting widths ranging from 900 to 1650 mm. Continuous casting machine, the standard smelting capacity of each furnace steel is 250 tons, and the maximum number of castable furnaces in the tundish is 8 furnaces;
  • the steelmaking group batching and scheduling method for the whole process production includes the following steps:
  • Step 1 Describe the production environment by constructing a directed network topology diagram
  • each node on the directed network topology diagram represents a specific production unit or inventory equipment, including: converter, refining furnace, continuous casting machine, slab library, hot rolling unit , leveling unit, hot coil library, pickling unit and acid rolling unit; each arc on the directed network topology shows a specific material transfer process from one unit or inventory unit to another unit or inventory equipment, including : molten steel, slab, hot coil and cold coil;
  • Step 2 Set the process parameters of the product according to the quality requirements of the final product according to different customer contracts, including: determining the mapping of the manufacturing process of the product on the directed network topology map, and calculating the casting width of different products on the continuous casting machine according to the steel type. Scope, determine the superior replacement relationship between steel grades, determine the mixing relationship and cost of different steel grades in the tundish;
  • the mapping of the manufacturing process of the determined product described in step 2 on the directed network topology map is implemented by coding a unified manufacturing process for any product;
  • the coding structure is as shown in FIG. A total of 17 digits, each corresponding to a process on the whole process of steel, each digit is not 0, indicating that the product can be produced on the specific equipment of the corresponding process, each digit of 0 indicates that the product does not need to go through the process. produce;
  • Whether the product can be produced on the specific equipment of the corresponding process is determined by the physical and chemical properties of the product and the specific equipment process parameters, wherein the process parameters of the continuous casting process include the thickness of the mold, the maximum casting width allowed by the mold, and the minimum Casting width, maximum width and minimum width of casting machine allowed for casting, drawing speed of continuous casting machine and service life of tundish; process parameters of hot rolling process include maximum rolling thickness and minimum rolling thickness, maximum rolling width and minimum Rolling width, the hardness group allowed for rolling; for the cold rolling process, the equipment process parameters include: the maximum number of kilometers allowed for rolling mill work rolls (and support rolls), and the variation of the mill's allowable different specification properties (including forward width) Maximum jump, reverse width maximum jump), the mill allows the production of materials of different specifications and properties per unit time (the mill specification properties generally include: width, thickness, roughness);
  • the casting width range of different products on the continuous casting machine is calculated according to the steel type described in step 2, and the calculation formula is as follows:
  • the determining the superior replacement relationship between the steels described in step 2 is achieved by calculating the difference in the contents of carbon, phosphorus, sulfur, manganese and silicon in different steel grades; Equation (30) determines the intersection of any steel species g with any chemical element contained in any steel species h to determine:
  • determining the mixing relationship and cost of different steel grades in the tundish according to step 2 is realized by calculating the consistent value of the corresponding mixed code and the index code of different steel grades; specifically, by formula ( 32) comparing the relationship between the random casting code and the index code corresponding to any steel type g and any steel type h to determine the type of mixing;
  • ⁇ (a,b) is a custom comparison function
  • m g and n g respectively represent the mixed casting code and index code corresponding to steel type g ;
  • Step 3 According to the steel type, variety attribute, optional manufacturing process and width range of the product, the product order group is judged. If the total demand of the customer is greater than or equal to the maximum number of continuous furnaces allowed by the tundish, then Belong to the large contract group, perform step 6; if the total product under-required by the customer is less than the maximum number of continuous furnaces allowed by the intermediate package, then belong to the small contract group, and perform steps 4 to 5;
  • Step 4 Describe the batch production decision of the multi-product in the steelmaking process by constructing a mathematical model
  • the method steps are shown in FIG. 4, and specifically include the following steps:
  • Step 4-1 Mapping the multi-product group batch plan in the steelmaking production process to a mathematical model decision variable
  • Step 4-2. Map the process limits of the steelmaking production process to mathematical model constraints, as follows:
  • Step 4-2-1 establishing process constraints for the replacement relationship of product steel grades
  • the product set N g produced by the steel grade is determined
  • N represents the total product set for a given small contract group
  • g i represents the steel grade of product i
  • G ⁇ i ⁇ N
  • g i represents all steel species contained in product set N
  • s gig represents steel of product i
  • Step 4-2-2 establishing a process constraint of the casting width range of the product on the continuous casting equipment
  • Step 4-2-3 Establish the process constraints of the limit of each batch of smelting capacity of the converter. That is, the total weight of the slab and un-contracted slabs required to meet the customer contract requirements of the smelting in the same batch should be close to the standard smelting capacity of the converter. The weight of the part exceeding the standard smelting capacity of the converter and the weight of the standard smelting capacity of the insufficient converter are less than the weight of a slab; the untrusted slab refers to the output that meets the requirements of the full batch production process in the converter smelting process. No remaining materials for contract management with customers
  • C is the standard smelting capacity of the converter and Q i is the production under-production of the product i.
  • l std indicates the standard length of the un-slab in the continuous casting production
  • h std indicates the standard thickness of the un-slab in the continuous casting production
  • ⁇ indicates the density of the molten steel
  • Step 4-2-4 Establishing the furnace flow balance process of the two streams of molten steel in the casting process of each furnace molten steel, that is, the casting time of the same furnace molten steel in the continuous casting machine is equal, in the model The number of slabs mapped onto the two streams is equal;
  • n gwk represents an auxiliary integer variable, indicating that the steel grade is the number of slab blocks produced by the kth batch of odd flow of g width w;
  • Step 4-2-5 establish the process constraint of cutting the length range of the slab on the continuous casting equipment, that is, the cutting process of the continuous casting machine and the length of the customer ordering, the length of any slab cast in a furnace steel water is required to be Within the prescribed range;
  • h i represents the slab thickness required for product i, with Indicates the maximum and minimum length of the slab required for product i;
  • Step 4-2-6 construct a flexible management constraint on the customer's order quantity, that is, the part that is insufficient or exceeds the customer's order quantity is less than the weight of one slab;
  • Step 4-3 Mapping the optimized process index in the steelmaking production process to the mathematical model objective function to minimize all The total weight of the unsponsored slabs produced by the batch, the minimum replacement amount between the steel grades, the total slab weight in all batches, and the total deviation of the standard smelting capacity of the converter, and the minimization of all customer contracts.
  • the inventory cost index of the untrusted slab is mapped to the objective function of the formula (33), that is, the total weight of the untrusted slabs of all batches is minimized;
  • the excellent replacement replacement cost index between the steel types is mapped to the objective function of the formula (34), that is, the optimal replacement amount between the steel types is minimized;
  • the operating efficiency index of the converter batch production is mapped to the objective function of the formula (35), that is, the deviation of the produced slab weight in all batches and the standard smelting capacity deviation of the converter is minimized;
  • the management index of the customer's satisfaction with the order weight is mapped to the objective function of the formula (36), that is, the order quantity deviation of all the customer contracts is minimized;
  • F 0 represents the total cost of the production group of the small contract group
  • the total cost of the production group of the small contract group is the total amount of un-committed slabs of all batches, the total replacement quantity of steel grades, and all customer contracts.
  • Step 5 Construct a mutual mapping relationship between a real number matrix and a group batch scheme, and use the established real matrix as the controlled object to implement a multi-object parallel iterative improvement strategy to obtain a final optimized group batch scheme, and then obtain a small
  • the pre-grouping plan of the contract group in the continuous casting process is shown in Figure 5, as follows:
  • Step 5-1 Construct a mutual mapping relationship between a real number matrix and a batch batch scheme, as follows:
  • Step 5-1-1 Construct a real matrix whose dimension is the product of the total number of products, the steel grade and the width.
  • the elements in the matrix are all batches assigned to a certain steel type and a certain width.
  • ) dimension expressed by the formula (12) is designed:
  • A represents a
  • N represents the total product set for a given small contract group
  • W represents the cast product set
  • A represents a
  • N represents the total product set for a given small contract group
  • W represents the cast product set
  • the N-time continuous casting machine crystallizer The set of all possible widths to be set, a
  • Step 5-1-2 Obtain the weight of the slab in all batches assigned to the target steel grade and the target width and the weight of all the contract slabs in all batches assigned to the target steel grade and the target width, For all steel grades and width combinations, sort the weight values of all contract slabs in all batches from large to small, and repeat steps 5-1-3 to 5-1-9 in this order to obtain Group batch plan for all types of steel and width combinations:
  • igw denotes the contract slab weight assigned to all batches of steel grade g and width w
  • B gw denotes all contract slabs assigned to all batches of steel grade g and width w
  • the weight, a igw represents the ratio of the slab weight in all batches corresponding to the contract i assigned to the steel grade g and width w to the under-production of the contract i
  • Q i represents the production deficit of the product i.
  • Step 5-1-3 for any combination of steel and a width (g, w), all contracts assigned to statistical weight in the slab and the width of the steel batch composition, referred to as (b 1gw b 2gw ... b
  • gw ) T , create an empty batch k that does not contain any contract, and set the slab weight E k 0 already included in the batch;
  • Step 5-1-4 in the slab weight vector (b 1gw b 2gw ... b
  • the size of the first slab weight b igw if the remaining capacity CE k is greater than or equal to the first slab weight b igw , then perform steps 5-1-5, otherwise, perform steps 5-1-6;
  • Step 5-1-5 replacing the production under-producing Q i of the corresponding product in the flexible management constraint of the customer order quantity with the product slab weight b igw , and following steps 4-2-5 to 4-2-6
  • Step 5-1-7 without adding a non-commissioned slab, determine whether the slab in the empty batch k satisfies the process constraints of the limit of each batch of smelting capacity of the converter, and if so, step 5 is performed. -1-8, otherwise, perform steps 5-1-9;
  • Step 5-1-9 it is determined slab weight vector (b 1gw b 2gw ... b
  • Steps 5-1-10 the weights of all contract slabs in all batches are performed in steps 5-1-3 to 5-1-9 to obtain a batch plan for all types of steel and width combinations;
  • Step 5-2 Using the established real matrix as the controlled object, implementing a multi-object parallel iterative improvement strategy to obtain a final optimized group batch solution, specifically including:
  • Step 5-2 randomly generate NP real number matrices having the same real matrix structure as described in step 5-1-1, and put all constructed real matrices into the set ⁇ A 1 , A 2 , ..., A NP ⁇
  • the triplet (i, g, w) satisfying the condition i ⁇ N g ⁇ P w , a igw is set to 1, and is not satisfied.
  • the element is set to 0;
  • a igw is a random number generated from a uniformly distributed real interval [L, U], L and U are respectively a lower bound and an upper bound of the interval;
  • NP is a population size parameter preset based on a multi-object parallel iterative improvement strategy algorithm;
  • Step 5-2-3 sort the obtained objective function f(A j ) from small to large, and divide the real number matrix of the first one of the ranking into a group, denoted as S, and rank the second two.
  • a real number matrix is divided into a group, denoted as I, that is, S and I satisfy max ⁇ f(A j )
  • Step 5-2-4-1 selecting three real numbers A r1 , A different from the target real matrix A j and different from each other in the set ⁇ A 1 , A 2 , ..., A NP ⁇ composed of real numbers matrix R2 , A r3 , ie j ⁇ r1 ⁇ r2 ⁇ r3;
  • Step 5-2-4-2 randomly generating a variable asynchronous long factor F j from the uniformly distributed real interval [j/NP, 1];
  • Step 5-2-4-3 performing a difference operation on the real numbers A j , A r1 , A r2 , and A r3 to obtain a real matrix V j after the mutation operation;
  • j 1, 2, . . . , NP ⁇ ;
  • S′ represents the obtained objective function Sorting from small to large, ranking the first half of the real matrix group, I means sorting the obtained objective functions from small to large, and ranking the second half of the real matrix grouping;
  • Step 5-2-4-4 randomly generating a crossover probability factor CR igw from each element of the matrix by uniformly distributed real interval [i/NP, 1];
  • Step 5-2-4-5 performing cross operation on each pair of real numbers A j and V j to generate a real matrix U j ;
  • Represents the elements inside the real matrix U j , Represents the elements inside the real matrix V j , Represents the elements inside the real matrix A j , j 1, 2, ..., NP, i ⁇ N, w ⁇ W, k ⁇ ⁇ 1, 2, ..., K gw ⁇ ; W represents the continuous casting machine for the casting product set N The set of all possible widths required for the crystallizer; K gw represents the upper limit of the number of batches of steel grade g and width w; Representing a random number between (0, 1) obeying a normal distribution;
  • Step 5-2-4-6 determine whether the element or If yes, the boundary condition processing is performed on the real number matrix U j , otherwise, the cross operation is completed;
  • L represents an element The lower limit of the range of values
  • U represents the element The upper limit of the value range.
  • the value of the batch plan corresponding decision variable (x, z, y) is substituted into formula (19), and the weighted objective function under the group batch scheme is obtained, denoted as f(U j ); compare f(U j ) and f(A) j ), the matrix is updated according to formula (26), and step 5-2-2 to step 5-2-4 are repeated for the updated matrix set ⁇ A 1 , A 2 , ..., A NP ⁇ until the set ⁇ A 1 , A 2 ,...,A NP ⁇ is no longer updated;
  • Step 5-2-6 selecting the real matrix A j* with the smallest objective function value f(A j ) in the final matrix set ⁇ A 1 , A 2 , . . . , A NP ⁇ , and returning the matrix to step 5 -1-1 to step 5-1-10 to obtain the final optimized batch plan;
  • Step 5-3 the obtained steelmaking batches are combined according to the steel type and the width, that is, the steelmaking batches having the same steel type and width are combined into one continuous pouring batch group, and the small contract group is completed in the continuous casting process. Designation of a batch plan;
  • Step 6 Formulate a batch plan for the large contract group in the steelmaking process and a pre-batch plan for the continuous casting process;
  • the production underrun of all products in the group is calculated.
  • the number of steelmaking batches that need to be produced by the large contract group according to the requirements of the full batch production process of the converter smelting
  • T max of the tundish the K steelmaking batches are decomposed into the continuous casting process.
  • the unit capacity is calculated, and the inventory structure is counted as follows: by calculating the difference between the standard production capacity of the steel making, refining, continuous casting and hot rolling equipment and the equipment maintenance plan, the difference in downtime is obtained.
  • Step 7 Determine the scheduling decision of the continuous casting batch on the continuous casting equipment by constructing a quantitative mathematical model; specifically: selecting the decision variables of the continuous pouring batch scheduling; quantitatively describing the continuous pouring batch scheduling The goal pursued; quantitative description of the process constraints and management requirements to be followed in the development of continuous casting batch scheduling; the process flow is shown in Figure 6, the specific steps are as follows:
  • Step 7-1 Select the decision variables for the continuous pouring batch group scheduling; the details are as follows:
  • Setting the 0-1 decision variable u rls indicates whether the continuous batch r is allocated at the lth position of the continuous casting device s; setting the auxiliary variable Q ls indicates the continuous pouring batch at the lth position of the continuous casting device s Group casting completion time; setting auxiliary variable T r indicates casting completion time of continuous pouring batch r; setting auxiliary variable t i indicating casting completion time of contract i; setting auxiliary variable Indicates the amount of hot bar stock planned in the hot rolling mill h before the day d; set the auxiliary variables Indicates the planned inventory of the difficult-to-roll material in the hot rolling mill h before the d-day; the auxiliary variable ⁇ fd indicates the amount of slab that the steel mill produces for the planned flow on day d;
  • Step 7-2 quantitatively describing the goal pursued by the continuous pouring batch group scheduling
  • tundish includes: maximizing the utilization of the tundish, minimizing the number of slabs of different steel grades, minimizing the number of slabs, minimizing the deviation of the amount of hot rolled stock, minimizing the deviation of the inventory of difficult-to-roll materials, minimizing heat Deviation of demand for rolling and cold rolling and minimizing customer contract delay time;
  • F I represents the total cost of the tundish updated by all the consecutive batches
  • S represents the collection of continuous casting equipment
  • R represents the collection of all consecutive batches
  • a r represents the number of furnaces contained in the continuous batch r
  • Max represents the maximum number of furnaces that can be cast in a tundish
  • b s represents the number of batches that cannot be connected to the continuous casting equipment s
  • l ⁇ Indicates the position of the batch of the ⁇ th non-mixable pour, ie, the continuous batch with For the steel type, it is strictly forbidden to mix or pour the batch with The difference in width exceeds the maximum allowable online widening of the caster.
  • F 2 represents the total cost of continuous casting of all the different steel types in the continuous pouring batch group; Represents the continuous casting batch assigned at the l position on the continuous casting equipment s, Representing continuous casting batch Steel grade
  • F 3 represents the total cost of all the widening of the continuous batching batch; Representing continuous casting batch Width, the function h(.) takes 0 when the two parameters are the same before, otherwise it takes 1;
  • the deviation between the planned inventory of the hot bar before the hot rolling and the target inventory is required to ensure the smoothness of the hot rolling production
  • F 4 represents the risk of inventory of all hot rolled stocks
  • H represents the collection of hot rolling mills
  • D represents the set of days in the plan period, Indicates the target stock of hot bar stock required by hot rolling mill h on day d;
  • the planned inventory of the difficult-to-roll material in the pre-rolling warehouse is the smallest deviation from the maximum and minimum allowable inventory of the difficult-to-roll material to reduce the clogging of the logistics caused by the excessively difficult-to-roll material;
  • F 5 represents the risk cost of all difficult-to-roll materials inventory; Indicates the largest inventory of difficult-to-roll materials in hot rolling mill h on day d; Indicates the minimum allowable inventory of hot rolled mills on day d;
  • F 6 represents the customer contract satisfaction income
  • N R represents the contract set with strict delivery time requirements
  • Ear i represents the earliest delivery date of contract i
  • Due i represents the latest delivery date of contract i
  • Step 7-3 Quantitatively describe the process constraints and management requirements to be followed in the development of the continuous casting batch group; the details are as follows:
  • v rs represents the contract production process and continuous casting equipment s compatibility parameters contained in the continuous batch r, v rs ⁇ ⁇ 0, 1 ⁇ .
  • the method includes: a distribution relationship constraint of the continuous casting batch on the continuous casting equipment and a feasible distribution rule constraint;
  • Step 8 The mathematical model established in step 7 is used as a quantitative calculation basis, and a mutual mapping relationship between the real number vector and the continuous casting device on the continuous casting equipment is established, and the established real vector is used as the basis.
  • the controlled object obtains a multi-object parallel iterative improvement strategy to obtain the scheduling scheme of the continuous casting batch on the continuous casting equipment; that is, the distribution and sequence of the continuous casting equipment for the continuous casting equipment;
  • Step 8-1 setting a 2
  • dimensional real number vector PP [a 1 , a 2 ... a R
  • a r and b r are dimensionless real parameters, the value range is [0, 1]; 1 ⁇ r ⁇
  • Step 8-2 determining the continuous pouring batch set Rs assigned to any continuous casting equipment s according to the b r value
  • Step 8-3 determining the order of all consecutive batches assigned to any continuous casting equipment s according to the value of a r ;
  • the order of all consecutive pouring batches allocated to the equipment is determined according to the a r value, that is, the continuous pouring batch group arg min ⁇ a r with the smallest a r value in the set R s
  • r ⁇ R s ⁇ is placed in the first position of the continuous casting equipment s, and the second smallest batch of ar min ⁇ a r
  • Step 9 The group batch plan and the scheduling integration plan are adjusted, issued and executed.
  • the batch planning and the production scheduling scheme are integrated to obtain the final production organization plan, according to the special characteristics of the beginning and end of the month, the actual supply of molten iron, the key contract delivery date information, but the RH refined cold steel contract.
  • the production units After the reservation and the whole process logistics connection, and further fine-tuning the production organization plan in combination with the actual on-site production fluctuations, the production units will be given to the steelmaking stage, and each production unit will prepare and execute the production according to the plan to achieve Equilibrium and punctual distribution of steel material flow in the whole process equipment and time dimension.

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Organic Chemistry (AREA)
  • Materials Engineering (AREA)
  • Metallurgy (AREA)
  • Human Resources & Organizations (AREA)
  • Manufacturing & Machinery (AREA)
  • Mathematical Physics (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Mechanical Engineering (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Tourism & Hospitality (AREA)
  • General Engineering & Computer Science (AREA)
  • Operations Research (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Algebra (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • General Chemical & Material Sciences (AREA)
  • Thermal Sciences (AREA)
  • Educational Administration (AREA)
  • Game Theory and Decision Science (AREA)
  • Health & Medical Sciences (AREA)
  • Development Economics (AREA)
  • Quality & Reliability (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)

Abstract

一种面向全流程生产的炼钢组批与排产方法，属于冶金自动控制技术领域，通过对产品的多样化规格属性和设备的批生产属性以及产品和设备之间适配性进行刻画，建立定量化描述产品在炼钢和连铸工序上组批决策问题的模型，并从同工序并行设备之间产能均衡性、上下游工序之间物流衔接性的角度出发，对批在连铸设备和时间维度上的分配与排序建模，将组批计划与生产调度方案进行集成下达给炼钢阶段各生产制造单元，以达到钢铁物质流在全流程工序设备和时间维度上的均衡与准时分布。

Description

一种面向全流程生产的炼钢组批与排产方法 技术领域
本发明属于冶金自动控制技术领域，具体涉及一种面向全流程生产的炼钢组批与排产方法。
背景技术
钢铁生产流程一个突出的特点就是物质流在工艺过程中不断的发生物理化学反应，在状态、性质和形状方面不断变化；实现钢铁生产物质流物理化学反应的装置规模都非常大，并且这些装置的作业方式包含了连续化、准连续化和批生产等形式；在连续化生产过程中，原材料持续不断的经过相同设备，各设备处于相对稳定状态并且只执行一个特定的加工任务，半成品或产成品以连续流动的方式输出；在批生产过程中，原材料按设定的加工顺序和操作条件进行加工，产品以有批量方式输出；准连续化生产则是介于连续化和间歇化生产之间的一种混合模式；在炼铁工序，高炉生产是一个连续化作业过程，除检修外一直处于生产状态，并且铁水在各出铁口轮换持续输出。因此，炼铁工序是典型的连续化生产过程。在炼钢工序，转炉或电弧炉每次可冶炼150-300吨的钢水，称为一个炉次；每个炉次的钢水冶炼完成倒入钢包后，转炉或电弧炉经清理后可以进行下一炉次钢水的冶炼；显然，炼钢工序每冶炼-炉次钢水对应一次批加工过程；多个炉次的钢水可以在连铸机上连续浇铸，但由于受到结晶器、中间包使用寿命以及钢级和铸坯规格的多样性影响，连铸机连续生产到一定炉次数后会停止，并对连铸机进行清理，同时需要对关键设备-中间包和结晶器进行更换，而这一过程往往需要2-3个小时；因此，连铸工序属于批生产过程，连铸工序的每个生产批次称为浇次，定义为在同一台连铸机上使用相同中间包和结晶器连续浇铸的炉次序列；在热轧工序，板坯加热后在连轧机上也是连续加工，但同样受轧辊磨损等工艺因素影响，连轧机连续加工一定块数的板坯后也需要停机，设备清理后更换轧辊；热轧工序也属于批生产过程，热轧工序的每个生产批次称为轧制单元，定义为使用同一轧辊连续轧制的板坯序列；冷轧工序和热轧工序相似，也是批生产过程。连续化生产过程，生产出来的产品单一，除了总产能总量上进行宏观调控外，无法从产品物理化学属性的微观层面进行控制；但批生产过程则不同，在生产每一个批次的产品前，通常需要将物理化学属性具有工艺要求相似度的产品进行组批生产，由此提高生产效率和成材率，降低资源损耗和能源消耗。
除了生产流程长和批生产模式的特征外，现代钢铁企业的规模逐渐增大，产线布局呈现同工序多机组并行结构、工序间串行网状交叉结构的特征；复杂的内部产线结构和日趋激烈的外部市场竞争，使得钢铁企业对其生产效率、能耗水平、产品质量和生产成本等综合生产指标提出了更高的要求；因此，有必要从全流程的视角出发，合理安排工序间的物流关系， 在生产计划与调度的层次上优化工序间的物料衔接，从而提高企业的整体生产效率。在铁水之前，高炉的生产是一个连续化过程，生产组织过程也是面向库存的。在炼钢之后，以批生产过程为主，生产出来的固态半成品和成品如板坯、热卷、冷卷都和客户需求密切相关，生产组织过程则是面向订单的；炼钢工序是面向订单生产的源头工序，下游机组众多，生产线布局复杂；因此在确定炼钢阶段的生产组织安排包括确定批的组成决策的时候，需要高度关注下游机组的现有物料库存量和物料需求量，均衡机组的产能和库存，防止有些机组涨库而有些机组断料停机；由此可知，如何科学定量的确定炼钢工序的批组成、批在不同炼钢车间和连铸设备上的分配与排序、以及在一定生产期内不同班次上为下游各机组提供半成品的物料流量，是钢铁企业面临的一个非常重要并且十分复杂的技术问题。该技术问题也称为面向全流程生产的炼钢组批计划与批调度集成优化问题，需要在考虑包括全流程各个工序的生产工艺的限定条件下，决策炼钢和连铸工序的批组成、批分配与排序，以及炼钢下游工序间物料流量分配。
当前，为解决面向全流程生产的炼钢组批计划与批调度集成优化问题，企业通常采用依赖于计划员主观经验的人工排产方法；该方法通常是通过计划员对各机组的产能和库存情况进行的定性分析后，先按经验估计法来分配炼钢到下游工序间物料流量，然后按照合并同类项的方法简单确定批组成，最后按照人为设定的规则将批分配到各机组并排序；由于实际生产中涉及到的大量的生产数据，需要考虑多而复杂的生产工艺因素，采用这种以人工排产方法存在以下问题：
(1)一些复杂的工艺约束由于没有进行定量化描述和建模，通常被简化或忽略，这样将导致最终确定出来的批在机组上可能无法生产，这就需要在生产现场重新调整批的组成，进而破坏了生产过程的顺畅性；
(2)人工排产时大量的生产数据信息导致无法全部被考虑，通常采用按数据属性分类的方法进行粗略统计后再部分考虑，简单的数据分类统计方法导致大部分数据信息的细节被掩盖，信息考虑不完备将直接降低计划与调度编制的全局优化性；
(3)没有对所编制出的计划与调度的技术指标和经济指标进行定量化计算，也没有对多个技术指标和经济指标进行折衷化权衡，因此导致所编制的计划与调度的效果在很大程度依赖于计划编制人员的业务水平，对该技术问题的解决缺乏定量性和科学性。
因此，通过对钢铁全流程的生产工艺过程的深层次分析，采用定量化科学计算的方法来决策面向全流程生产的炼钢组批计划与批调度，在保证设备正常运行的基础上，合理的安排客户需求产品的生产组织，以提升设备生产效率，改善库存结构，理顺生产物流，对优化钢铁全流程的生产工艺水平、提升生产过程的管控水平、降低工序能耗和低成本制造具有十分 重要的意义。
目前，已公开的专利(“ZL200810011659.1”，一种炼钢-连铸炉次批量计划自动编制方法及系统)主要实现了将板坯组合到炉次来实现炼钢工序的批量生产；已公开的专利(“ZL200610046981.9”，一种炼钢-连铸中间包批量计划方法及系统)主要实现了将炉次组批到中间包以实现连铸机的批量生产。专利ZL200810011659.1和专利ZL200610046981.9主要是解决炼钢和连铸单工序上的作业组批生产技术问题，没有考虑多个车间和机组之间的生产负荷均衡，也没有考虑前后工序间的物料流供给平衡，更没有将组批决策同排产决策集成优化等技术问题。
发明内容
针对现有技术的不足，本发明提出一种面向全流程生产的炼钢组批与排产方法，以达到钢铁物质流在全流程工序设备和时间维度上的均衡与准时分布的目的。
一种面向全流程生产的炼钢组批与排产方法，包括以下步骤：
步骤1、采用构建有向网络拓扑图的方式描述生产环境；
其中，有向网络拓扑图上的每个节点表示一个具体的生产机组或库存设备，包括：转炉、精炼炉、连铸机、板坯库、热轧机组、平整机组、热卷库、酸洗机组和酸轧机组；有向网络拓扑图上的每条弧表示从一个机组或库存设备到另一个机组或库存设备之间存在具体的物料转移过程，包括：钢水、板坯、热卷和冷卷；
步骤2、根据不同客户合同对最终产品的质量要求，设置产品的工艺参数，包括：确定产品的制造流程在有向网络拓扑图上的映射、按照钢种计算不同产品在连铸机上的浇铸宽度范围、确定钢种间的优充替代关系、确定不同钢种在中间包中的混浇关系及成本；
步骤3、根据客户合同需求产品的钢种、品种属性、可选制造流程和宽度范围，判断产品订单所属分组，若客户需求总欠量大于或等于中间包允许的最大工艺连浇炉数，则属于大合同组，执行步骤6；若客户需求的产品总欠量小于中间包允许的最大工艺连浇炉数，则属于小合同组，执行步骤4至步骤5；
步骤4、采用构建数学模型的方式描述多产品在炼钢工序上的组批生产决策；
具体包括以下步骤：
步骤4-1、将炼钢生产过程中多产品组批方案映射为数学模型决策变量；
步骤4-2、将炼钢生产过程的工艺限制映射为数学模型约束条件，具体如下：
步骤4-2-1、建立产品钢种替代关系的工艺约束；
步骤4-2-2、建立产品在连铸设备上的浇铸宽度范围的工艺约束；
步骤4-2-3、建立转炉每一批次冶炼容量的限制的工艺约束，即限定在同一批次内冶炼的 客户合同需求的板坯和无委托板坯的总重量应接近转炉标准冶炼容量，超出转炉标准冶炼容量部分的重量和不足转炉标准冶炼容量部分的重量都要小于一块板坯的重量；
所述的无委托板坯是指为满足转炉冶炼过程中要求满批生产工艺而产出的没有和客户合同管理的剩余材料；
步骤4-2-4、建立每一炉钢水在连铸机上浇铸过程中，两流钢水消耗量的炉流平衡工艺约束，即要求同一炉钢水在连铸机两流的浇铸时间相等，在模型上映射为两流铸出的板坯块数相等；
步骤4-2-5、建立板坯在连铸设备上切割长度范围的工艺约束，即受连铸机切割工艺和客户订货长度的限制，要求一炉钢水中铸造出的任意板坯的长度在规定范围内；
步骤4-2-6、构建客户订货量的柔性管理约束，即不足或超出客户订货量的部分要小于一块板坯的重量；
步骤4-3、将炼钢生产过程中优化的工艺指标映射为数学模型目标函数，实现最小化所有批次产出的无委托板坯总重量、最小化钢种之间优充替代总量、最小化所有批次中产出板坯重量同转炉标准冶炼容量偏差总量、最小化所有客户合同的订货量偏差总量；
步骤5、构建一种实数矩阵与组批方案之间的相互映射关系，并以所建立的实数矩阵作为被控对象，实现基于多对象并行迭代改进策略获取最终优化的组批方案，进而获得小合同组在连铸工序的预组批方案，具体如下：
步骤5-1、构建一种实数矩阵与组批方案之间的相互映射关系，具体如下：
步骤5-1-1、构建一个实数矩阵，该矩阵的维数为全部产品数、钢种和宽度的乘积，矩阵中的元素为某一合同分配到某一钢种且某一宽度的所有批次内的板坯重量占上述合同生产欠量的比率；
步骤5-1-2、获得某一合同分配到目标钢种且目标宽度的所有批次内的板坯重量和分配到目标钢种且目标宽度的所有批次内的所有合同板坯的重量，并对所有钢种和宽度组合，按所有批次内的所有合同板坯的重量值从大到小进行排序，并按该顺序重复执行步骤5-1-3至步骤5-1-9；
步骤5-1-3、确定所有合同分配到任意钢种和宽度组合批次内的板坯重量向量，并构建空批次，设置该批次内已经包含的板坯重量为0；
步骤5-1-4、在板坯重量向量选取第一个板坯重量大于0的合同，并比较空批次的剩余容量与上述第一个板坯重量的大小，若剩余容量大于等于第一个板坯重量，则执行步骤5-1-5，否则，执行步骤5-1-6；
步骤5-1-5、将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为该产品板坯 重量，并按照步骤4-2-5至步骤4-2-6限定的工艺条件获取整数块的板坯，将上述板坯放到空批次中，更新该批次的板坯重量并设置在板坯重量向量中该产品板坯重量为0；
步骤5-1-6、将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为剩余容量，并按照步骤4-2-5至步骤4-2-6限定的工艺条件获取整数块的板坯，将其放到空批次中，更新该批次的板坯重量并设置在板坯重量向量中该产品板坯重量为0；
步骤5-1-7、在不添加无委托板坯的情况下，判断空批次内的包含板坯是否满足转炉每一批次冶炼容量的限制的工艺约束条件，若是，则执行步骤5-1-8，否则，执行步骤5-1-9；
步骤5-1-8、判断空批次内的包含板坯是否满足两流钢水消耗量的炉流平衡工艺约束条件，若是，则直接创建下一个不包含任何合同的空批次，并设置该批次内已经包含的板坯重量为0，否则，通过从该空批次中增加或者移除一块板坯来修复该批次使其满足炉流平衡约束，再创建下一个不包含任何合同的空批次，并设置该批次内已经包含的板坯重量为0；
步骤5-1-9、判断板坯重量向量是否等于0，若是，则在最后一个不为空的批次内，按转炉每一批次冶炼容量的限制的工艺约束条件和两流钢水消耗量的炉流平衡工艺约束条件添加无委托板坯，否则，返回执行步骤5-1-4；
步骤5-1-10、所有批次内的所有合同板坯的重量均执行完步骤5-1-3至步骤5-1-9，获得所有钢种和宽度组合内合同的组批方案；
步骤5-2、以所建立的实数矩阵作为被控对象，实现基于多对象并行迭代改进策略获取最终优化的组批方案，具体包括：
步骤5-2-1、随机生成NP个与步骤5-1-1所述实数矩阵结构相同的实数矩阵，并将所有构建的实数矩阵放入集合中，每个矩阵中满足目标钢种且目标宽度的元素设置为1，不满足的元素设置为0；
其中，NP为预先设定基于多对象并行迭代改进策略算法的种群规模参数；
步骤5-2-2、将生成的所有实数矩阵返回执行步骤5-1-1至步骤5-1-10，建立每个实数矩阵与组批方案的对应关系，根据组批方案获得决策变量的取值，将其代入目标函数中，获得每个实数矩阵对应的目标函数；
步骤5-2-3、将所获得的目标函数从小到大进行排序，并将排名前二分之一的实数矩阵分为一组，将排名后二分之一的实数矩阵分为一组；
步骤5-2-4、根据每个实数矩阵对应的目标函数所在分组，对每个实数矩阵进行变异操作和交叉操作，获得操作后实数矩阵，再将操作后所有实数矩阵返回执行步骤5-1-1至步骤5-1-10，建立每个操作后实数矩阵与组批方案的对应关系，根据组批方案获得决策变量的取值，将其代入目标函数中，获得每个操作后实数矩阵对应的目标函数；
步骤5-2-5、判断操作前后实数矩阵对应的目标函数的大小，选择对应目标函数较小的实数矩阵为更新后的实数矩阵，获得更新后的矩阵集合，并返回执行步骤5-2-2至步骤5-2-4，直至矩阵集合不再更新，获得最终的矩阵集合；
步骤5-2-6、在最终的矩阵集合中选择目标函数值最小的实数矩阵，将该矩阵返回执行步骤5-1-1至步骤5-1-10，获得最终的优化组批方案；
步骤5-3、将获得的炼钢批次按照钢种和宽度进行合并，即具有相同钢种和宽度的炼钢批次合并为一个连浇批组，完成小合同组在连铸工序的预组批方案的指定；
步骤6、制定大合同组在炼钢工序的分批方案和连铸工序的预分批方案；
步骤7、采用构建定量化的数学模型的方式确定连浇批组在连铸设备上排产决策；具体包括：选取连浇批组排产的决策变量；定量化描述连浇批组排产所追求的目标；定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求；具体步骤如下：
步骤7-1、选取连浇批组排产的决策变量；
步骤7-2、定量化描述连浇批组排产所追求的目标；
具体包括：最大化中间包利用率、最小化异钢种连浇板坯数、最小化调宽板坯数、最小化烫辊材库存量偏差、最小化难轧材库存量偏差、最小化热轧和冷轧各流向需求量的偏差和最小化客户合同拖期时间；
步骤7-3、定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求；
具体包括：连浇批组在连铸设备上的分配关系约束和可行分配规则约束；
步骤8、将步骤7所建立的数学模型为定量计算依据，通过建立一种实数向量与连浇批组在连铸设备上排产方案之间的相互映射关系，并以所建立的实数向量作为被控对象获得基于多对象并行迭代改进策略来获得连浇批组在连铸设备上排产方案；
即获得连浇批组对于连铸设备的分配和顺序；
步骤9、组批计划与排产集成方案调整，下发和执行。
步骤4-1所述的将炼钢生产过程中多产品组批方案映射为数学模型决策变量；具体如下：
设定连续决策变量xigwk，表示合同i在钢种为g宽度为w的第k个批次的生产板坯重量；设定整数决策变量zigwk，表示合同i在钢种为g宽度为w的第k个批次的生产板坯块数；设定整数决策变量z0gwk，表示钢种为g宽度为w的第k个批次内无委托板坯块数；设定0-1决策变量ygwk，当钢种为g宽度为w的第k个批次决定生产时，ygwk取值为1；否则ygwk取值为0；
步骤4-2-1所述的建立产品钢种替代关系的工艺约束；
即对任意钢种g，确定由该钢种生产的产品集合Ng；
Figure PCTCN2016105581-appb-000001
其中，N表示给定小合同组的全部产品集，gi表示产品i的钢种，G＝∪i∈Ngi表示产品集N中包含的全部钢种集，sgig表示产品i的钢种gi与任意钢种g的替代关系；
步骤4-2-2所述的建立产品在连铸设备上的浇铸宽度范围的工艺约束；
即对连铸机结晶器设定的任意宽度w，确定板坯浇铸为该宽度的产品集合Pw；
其中，
Figure PCTCN2016105581-appb-000003
表示浇铸产品集N时连铸机的结晶器所需设定的全部宽度集合；
Figure PCTCN2016105581-appb-000004
分别表示产品i允许的最大浇铸宽度以及最小浇铸宽度；
步骤4-2-3所述的建立转炉每一批次冶炼容量的限制的工艺约束，即限定在同一批次内冶炼的客户合同需求的板坯和无委托板坯的总重量应接近转炉标准冶炼容量，超出转炉标准冶炼容量部分的重量和不足转炉标准冶炼容量部分的重量都要小于一块板坯的重量；
具体公式如下：
Figure PCTCN2016105581-appb-000005
Figure PCTCN2016105581-appb-000006
Figure PCTCN2016105581-appb-000007
其中，C表示转炉标准冶炼容量，Qi表示产品i的生产欠量，
Figure PCTCN2016105581-appb-000008
表示钢种为g且宽度为w的批次数上限，
Figure PCTCN2016105581-appb-000009
表示一个批次中不足转炉标准冶炼容量的部分，
Figure PCTCN2016105581-appb-000010
表示一个批次中超出转炉标准冶炼容量的部分，lstd表示连铸生产中无委托板坯的标准长度，hstd表示连铸生产中无委托板坯的标准厚度，ρ表示钢水的密度；
步骤4-2-4所述的建立每一炉钢水在连铸机上浇铸过程中，两流钢水消耗量的炉流平衡工艺约束，即要求同一炉钢水在连铸机两流的浇铸时间相等，在模型上映射为两流铸出的板坯块数相等；
具体公式如下：
Figure PCTCN2016105581-appb-000011
其中，ngwk表示辅助整数变量，表示钢种为g宽度为w的第k个批次奇流生产的板坯块数；
步骤4-2-5所述的建立板坯在连铸设备上切割长度范围的工艺约束，即受连铸机切割工艺和客户订货长度的限制，要求一炉钢水中铸造出的任意板坯的长度在规定范围内；
具体公式如下：
Figure PCTCN2016105581-appb-000012
其中，hi表示产品i所需的板坯厚度，
Figure PCTCN2016105581-appb-000013
和
Figure PCTCN2016105581-appb-000014
表示产品i所需的板坯最大和最小长度；
步骤4-2-6所述的构建客户订货量的柔性管理约束，即不足或超出客户订货量的部分要小于一块板坯的重量；
Figure PCTCN2016105581-appb-000015
Figure PCTCN2016105581-appb-000016
Figure PCTCN2016105581-appb-000017
其中，
Figure PCTCN2016105581-appb-000018
表示欠量不足部分，
Figure PCTCN2016105581-appb-000019
表示欠量超出部分；
步骤4-3所述的目标函数，具体如下：
Figure PCTCN2016105581-appb-000020
其中，F0表示小合同组生产组批总成本，所述的小合同组生产组批总成本即为所有批次无委托板坯总量、钢种之间优充替代总量、所有客户合同的订货量偏差和转炉标准冶炼容量偏差的线性加权总和，λ1，λ2，λ3，λ4∈[0，1]，表示不同目标的权重系数，且λ1+λ2+λ3+λ4＝1。
步骤5-1-1所述的实数矩阵，具体公式如下：
Figure PCTCN2016105581-appb-000021
其中，A表示一个|N|×(|G|×|W|)维实数矩阵，N表示给定的小合同组的全部产品集合，W表示浇铸产品集，N时连铸机的结晶器所需设定的全部可能宽度集合，a|N|，G|，|W|表示对应合同N分配到钢种为G且宽度为W的所有批次内的板坯重量占合同N生产欠量的比率关系。
步骤5-1-2所述的获得某一合同分配到目标钢种且目标宽度的所有批次内的板坯重量和分配到目标钢种且目标宽度的所有批次内的所有合同板坯的重量，具体计算公式如下：
Figure PCTCN2016105581-appb-000022
Figure PCTCN2016105581-appb-000023
其中，bigw表示合同i分配到钢种为g且宽度为w的所有批次内的板坯重量，Bgw表示分配到钢种为g且宽度为w的所有批次内的所有合同板坯的重量，aigw表示对应合同i分配到钢种为g且宽度为w的所有批次内的板坯重量占合同i生产欠量的比率关系，Qi表示产品i的生产欠量。
步骤5-1-5所述的将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为该产品板坯重量，即将公式
Figure PCTCN2016105581-appb-000024
中的产品i的生产欠量替换为产品板坯重量bigw；
步骤5-1-6所述的将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为剩余容量，即将公式
Figure PCTCN2016105581-appb-000025
中的产品i的生产欠量替换为剩余容量C-Ek，C表示转炉标准冶炼容量，Ek表示批次内已经包含的板坯重量。
步骤5-2-4所述的根据每个实数矩阵对应的目标函数所在分组，对每个实数矩阵进行变异操作和交叉操作，具体步骤如下：
步骤5-2-4-1、在由实数矩阵构成的集合{A1，A2，…，ANP}中选取三个不同于目标实数矩阵Aj且互不相同的实数矩阵Ar1、Ar2、Ar3，即j≠r1≠r2≠r3；
步骤5-2-4-2、从均匀分布的实数区间[j/NP，1]随机生成一个变异步长因子Fj；
步骤5-2-4-3、将实数矩阵Aj、Ar1、Ar2、Ar3进行差分运算，获得变异操作后的实数矩阵Vj；
Figure PCTCN2016105581-appb-000026
其中，Ar*当前目标函数值最小的实数矩阵，满足f(Ar*)＝min{f(Aj)|j＝1，2，…，NP}；S′表示将所获得的目标函数从小到大进行排序，排名前二分之一的实数矩阵分组，I表示将所获得的目标函数从小到大进行排序，排名后二分之一的实数矩阵分组；
步骤5-2-4-4、从均匀分布的实数区间[j/NP，1]为矩阵的每个元素随机生成一个交叉概率因子CRigw；
步骤5-2-4-5、将每对实数矩阵Aj和Vj执行交叉操作，生成实数矩阵Uj；
Figure PCTCN2016105581-appb-000027
其中，
Figure PCTCN2016105581-appb-000028
表示实数矩阵Uj内部的元素，
Figure PCTCN2016105581-appb-000029
表示实数矩阵Vj内部的元素，
Figure PCTCN2016105581-appb-000030
表示实数矩阵Aj内部的元素，j＝1，2，…，NP，i∈N，w∈W，k∈{1，2，…，Kgw}；W表示浇铸产品集N时连铸机的结晶器所需设定的全部可能宽度集合；Kgw表示钢种为g且宽度为w的批次数上限；
Figure PCTCN2016105581-appb-000031
表示服从正态分布的(0，1)之间随机数；
步骤5-2-4-6、判断是否元素
Figure PCTCN2016105581-appb-000032
或者
Figure PCTCN2016105581-appb-000033
若是，则对实数矩阵Uj进行边界条件处理，否则，完成交叉操作；
具体公式如下：
Figure PCTCN2016105581-appb-000034
其中，L表示元素
Figure PCTCN2016105581-appb-000035
的取值范围下限，U表示元素
Figure PCTCN2016105581-appb-000036
的取值范围上限。
步骤7-1所述的选取连浇批组排产的决策变量，具体如下：
设定0-1决策变量urls表示连浇批组r是否被分配在连铸设备s的第l个位置上；设定辅助变量Qls表示连铸设备s的第l个位置的连浇批组浇铸完成时间；设定辅助变量Tr表示连浇批组r的浇铸完成时间；设定辅助变量ti表示合同i的浇铸完成时间；设定辅助变量
Figure PCTCN2016105581-appb-000037
表示热轧厂h前库在第d天的烫棍材计划库存量；设定辅助变量
Figure PCTCN2016105581-appb-000038
表示热轧厂h前库在第d天的难轧材计划库存量；设定辅助变量Δfd表示炼钢厂在第d天为f流向计划生产的板坯量；
步骤7-2所述的最大化中间包利用率；
即要求浇铸完所有连浇批组所使用的中间包个数最小，具体公式为：
Figure PCTCN2016105581-appb-000039
其中，FI表示所有连浇批组排产所更新中间包总成本；S表示连铸设备集合；R表示所有连浇批组集合；ar表示连浇批组r中包含的炉数；Tunmax表示一个中间包最大可浇铸的炉数；bs表示分配到连铸设备s上的不可连浇的批组个数；lτ，
Figure PCTCN2016105581-appb-000040
表示第τ个不可混浇的批组位置；
步骤7-2所述的最小化异钢种连浇板坯数；
具体公式为：
Figure PCTCN2016105581-appb-000041
其中，F2表示连浇批组排产所有异钢种连浇总成本；
Figure PCTCN2016105581-appb-000042
表示连铸设备s上第l个位置上分配的连铸批组，
Figure PCTCN2016105581-appb-000043
表示连铸批组
Figure PCTCN2016105581-appb-000044
的钢种；
步骤7-2所述的最小化调宽板坯数；
具体公式为：
Figure PCTCN2016105581-appb-000045
其中，F3表示连浇批组排产所有调宽总成本；
Figure PCTCN2016105581-appb-000046
表示连铸批组
Figure PCTCN2016105581-appb-000047
的宽度，函数h(.)在前后两个参数相同时取0，否则取1；
步骤7-2所述的最小化烫辊材库存量偏差；
即要求热轧前库的烫棍材计划库存量同目标库存量的偏差最小以保证热轧生产的顺畅性；
具体公式如下：
Figure PCTCN2016105581-appb-000048
其中，F4表示所有烫辊材库存风险成本；H表示热轧厂集合，D表示计划期内的天数集合，
Figure PCTCN2016105581-appb-000049
表示热轧厂h在第d天需求的烫棍材目标库存量；
步骤7-2所述的最小化难轧材库存量偏差；
即要求热轧前库的难轧材计划库存量同最大和最小允许的难轧材库存量的偏差最小以减少难轧材过多导致物流堵塞；
具体公式如下：
Figure PCTCN2016105581-appb-000050
其中，F5表示所有难轧材库存风险成本；
Figure PCTCN2016105581-appb-000051
表示热轧厂h在第d天最大的难轧材库存量；
Figure PCTCN2016105581-appb-000052
表示热轧厂h在第d天最小允许的难轧材库存量；
步骤7-2所述的最小化客户合同拖期时间；
具体公式如下：
Figure PCTCN2016105581-appb-000053
其中，F6表示客户合同满意度收益；NR表示有严格交货期要求的合同集合，Eari表示合同i的最早交货期，Duei表示合同i的最晚交货期；
步骤7-3所述的定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求，具体如下：
制定连浇批组在连铸设备上的分配关系约束，即要求每个连浇批组只能分配到一个连铸设备上的一个位置，每个连铸设备上的每个位置最多只能分配一个连浇批组，每个连铸设备上未分配连浇批组的位置一定是在已分配连浇批组的位置之后，具体公式如下：
Figure PCTCN2016105581-appb-000054
Figure PCTCN2016105581-appb-000055
Figure PCTCN2016105581-appb-000056
制定可行分配规则约束，即只有连浇批组内所包含的合同生产制程与给定的连铸设备兼容时才允许将连浇批组分配到该连铸设备上，具体公式如下；
urls≤vrs        (27)
式中，vrs表示连浇批组r内所包含的合同生产制程与连铸设备s兼容性参数，vrs∈{0，1}。
步骤8所述的建立一种实数向量与连浇批组在连铸设备上排产方案之间的相互映射关系，具体如下：
步骤8-1、设定一个2|R|维实数向量PP＝[a1，a2…a|R||b1，b2…b|R|]，并确定实数向量PP内部数值；
其中，ar和br为无量纲实参数，取值范围为[0，1]；1≤r≤|R|，R表示所有连浇批组集合；
步骤8-2、按br值确定分配到任意连铸设备s上的连浇批组集合Rs；
即
Figure PCTCN2016105581-appb-000057
步骤8-3、按ar值确定分配到任意连铸设备s上所有连浇批组的排序；
即对连浇批组集合Rs按照ar值由小到大进行排序，确定分配到连铸设备s上所有连浇批组的排序。
本发明优点：
本发明提出一种面向全流程生产的炼钢组批与排产方法，本发明在工艺技术层面，通过 对产品的多样化规格属性和设备的批生产属性以及产品和设备之间适配性进行精细刻画，建立定量化描述产品在炼钢和连铸工序上组批决策问题的数学模型，并基于该模型设计一种基于多对象并行迭代改进方法得到炼钢工序上组批方案；在生产组织层面，从同工序并行设备之间产能均衡性、上下游工序之间物流衔接性的角度出发，对批在连铸设备和时间维度上的分配与排序建立模型，并基于该模型设计一种基于邻域搜索的串行迭代改进方法得到批在连铸设备和时间维度上的生产调度方案；将组批计划与生产调度方案进行集成，并做进一步的微调后下达给炼钢阶段各生产制造单元，各生产制造单元按该方案进行备料和执行生产，以达到钢铁物质流在全流程工序设备和时间维度上的均衡与准时分布；本发明改善了产品质量，提高了成材率、资源利用率、设备作业效率，实现并行设备上的负荷均衡和串行设备间物料衔接顺畅，降低物料流交通拥堵、下游设备等待时间和库存，实现合理控制库存量。
附图说明
图1为本发明一种实施例的面向全流程生产的炼钢组批与排产方法流程图；
图2为本发明一种实施例的向网络拓扑图；
图3为本发明一种实施例的制造流程编码构成示意图；
图4为本发明一种实施例的采用构建数学模型的方式描述多产品在炼钢工序上的组批生产决策流程图；
图5为本发明一种实施例的获得小合同组在连铸工序的预组批方案方法流程图；
图6为本发明一种实施例的确定连浇批组在连铸设备上排产决策方法流程图。
具体实施方式
下面结合实施例和附图对本发明方法做进一步说明。
本实施例为一个大型钢铁企业；该钢铁企业有两个炼钢厂、两个热轧厂，一条包含酸洗、酸轧、连退、热镀锌、热镀铝锌、电镀锡、横切、重卷8个机组的冷轧产线；第一炼钢厂配备3个转炉、RH、LF和Ar精炼设备各2台、2台浇铸宽度范围为750-1320mm的连铸机，每个炉钢水标准冶炼容量为150吨、中间包最大可浇铸炉数为8炉；第二炼钢厂配备2个转炉，RH、LF和Ar精炼设备各1台、2台浇铸宽度范围为900-1650mm的连铸机，每个炉钢水的标准冶炼容量为250吨、中间包最大可浇铸炉数为8炉；
采用某钢铁企业实际生产中一周的合同数据，其中合同对应不同阶段及制造流程总生产欠量如下表所示：
Figure PCTCN2016105581-appb-000058
Figure PCTCN2016105581-appb-000059
本发明实施例中，面向全流程生产的炼钢组批与排产方法，方法流程图如图1所示，包括以下步骤：
步骤1、采用构建有向网络拓扑图的方式描述生产环境；
本发明实施例中，如图2所示，有向网络拓扑图上的每个节点表示一个具体的生产机组或库存设备，包括：转炉、精炼炉、连铸机、板坯库、热轧机组、平整机组、热卷库、酸洗机组和酸轧机组；有向网络拓扑图上的每条弧表示从一个机组或库存设备到另一个机组或库存设备之间存在具体的物料转移过程，包括：钢水、板坯、热卷和冷卷；
步骤2、根据不同客户合同对最终产品的质量要求，设置产品的工艺参数，包括：确定产品的制造流程在有向网络拓扑图上的映射、按照钢种计算不同产品在连铸机上的浇铸宽度范围、确定钢种间的优充替代关系、确定不同钢种在中间包中的混浇关系及成本；
本发明实施例中，步骤2所述的确定产品的制造流程在有向网络拓扑图上的映射是通过对任意产品建立一类统一的制造流程的编码来实现的；编码构成如图3所示，总共包含17位，每一位对应钢铁全流程上的一道工序，每一位不为0数值表示产品可以在对应工序的具体设备上生产，每一位为0数值表示产品不需要经过该工序生产；
产品是否可以在对应工序的具体设备上生产是通过产品的物理化学属性和具体设备工艺参数来确定的，其中，连铸工序的工艺参数包括结晶器厚度、结晶器允许调整的最大浇铸宽度和最小铸宽度、连铸机允许浇铸的最大宽度和最小宽度、连铸机的拉速度和中间包使用寿命；热轧工序的工艺参数包括最大轧制厚度和最小轧制厚度，最大轧制宽度和最小轧制宽度，允许轧制的硬度组；对于冷轧工序，设备工艺参数包括：轧机工作辊(及支撑辊)允许轧制的最大公里数，轧机允许不同规格属性的变化量(包括正向宽度最大跳跃、反向宽度最大跳跃)，轧机允许不同规格属性材料单位时间生产量(轧机规格属性一般包括：宽度、厚度、粗糙度)；
本发明实施例中，步骤2所述的按照钢种计算不同产品在连铸机上的浇铸宽度范围，计算公式如下：
Figure PCTCN2016105581-appb-000060
Figure PCTCN2016105581-appb-000061
其中，
Figure PCTCN2016105581-appb-000062
表示客户合同i订购的最终产品(热卷或冷卷)的订货宽度，g(i)表示客户合同i订购的最终产品的钢种，pg(i)表示钢种g(i)在热轧工序上允许的最大宽展测压量，σ为连铸机运行过程中允许的最小在线调宽幅度，Dmax表示连铸机的最大工艺设定宽度，(Dmin)表示连铸机的最小工艺设定宽度；
本发明实施例中，步骤2所述的确定钢种间的优充替代关系是通过计算不同钢种中的碳、磷、硫、锰、硅元素含量的差异来实现的；具体而言是通过公式(30)计算出任意钢种g与任意钢种h包含的任意化学元素的交叉度来确定的：
Figure PCTCN2016105581-appb-000063
其中，
Figure PCTCN2016105581-appb-000064
表示钢种h中所包含的化学元素c(包括碳、磷、硫、锰、硅)的下限要求；
Figure PCTCN2016105581-appb-000065
表示钢种g中所包含的化学元素c(包括碳、磷、硫、锰、硅)的下限要求；
Figure PCTCN2016105581-appb-000066
表示钢种h中所包含的化学元素c(包括碳、磷、硫、锰、硅)的上限要求；
Figure PCTCN2016105581-appb-000067
表示钢种g中所包含的化学元素c(包括碳、磷、硫、锰、硅)的上限要求；degh表示钢种h与钢种g在化学元素c上的交叠度系数；
通过公式(31)计算钢种g与钢种h的替代关系：
Figure PCTCN2016105581-appb-000068
如果sgh＝0，则钢种g与钢种h不可替代，否则钢种g与钢种h可替代；
本发明实施例中，步骤2所述的确定不同钢种在中间包中的混浇关系及成本是通过计算不同钢种对应混交代码和索引代码的一致值来实现的；具体而言通过公式(32)比较任意钢种g与任意钢种h对应的混浇代码和索引代码之间关系来确定混浇类型的；
qgh＝Ψ(mg，mh)+Ψ(ng，nh)           (32)
其中，Ψ(a，b)为自定义比较函数，当a＝b时Ψ(a，b)＝1，否则为0；mg和ng分别表示钢种g对应的混浇代码和索引代码；
如果qgh＝0，则钢种g与钢种h严禁混浇，单位重量混浇费用f(g，h)＝∞；如果qgh＝1，则钢种g与钢种h允许混浇并定义为混浇类型一；如果qgh＝2，则钢种g与钢种h允许混浇并定义为混浇类型二；
步骤3、根据客户合同需求产品的钢种、品种属性、可选制造流程和宽度范围，判断产品订单所属分组，若客户需求总欠量大于或等于中间包允许的最大工艺连浇炉数，则属于大合同组，执行步骤6；若客户需求的产品总欠量小于中间包允许的最大工艺连浇炉数，则属于小合同组，执行步骤4至步骤5；
步骤4、采用构建数学模型的方式描述多产品在炼钢工序上的组批生产决策；
方法步骤如图4所示，具体包括以下步骤：
步骤4-1、将炼钢生产过程中多产品组批方案映射为数学模型决策变量；
具体如下：
设定连续决策变量xigwk，表示合同i在钢种为g宽度为w的第k个批次的生产板坯重量；设定整数决策变量zigwk，表示合同i在钢种为g宽度为w的第k个批次的生产板坯块数；设定整数决策变量z0gwk，表示钢种为g宽度为w的第k个批次内无委托板坯块数；设定0-1决策变量ygwk，当钢种为g宽度为w的第k个批次决定生产时，ygwk取值为1；否则ygwk取值为0；
步骤4-2、将炼钢生产过程的工艺限制映射为数学模型约束条件，具体如下：
步骤4-2-1、建立产品钢种替代关系的工艺约束；
即对任意钢种g，确定由该钢种生产的产品集合Ng；
Figure PCTCN2016105581-appb-000069
其中，N表示给定小合同组的全部产品集，gi表示产品i的钢种，G＝∪i∈Ngi表示产品集N中包含的全部钢种集，sgig表示产品i的钢种gi与任意钢种g的替代关系；
步骤4-2-2、建立产品在连铸设备上的浇铸宽度范围的工艺约束；
即对连铸机结晶器设定的任意宽度w，确定板坯浇铸为该宽度的产品集合Pw；
Figure PCTCN2016105581-appb-000070
其中，
Figure PCTCN2016105581-appb-000071
表示浇铸产品集N时连铸机的结晶器所需设定的全部宽度集合；
Figure PCTCN2016105581-appb-000072
分别表示产品i允许的最大浇铸宽度以及最小浇铸宽度；
步骤4-2-3、建立转炉每一批次冶炼容量的限制的工艺约束，即限定在同一批次内冶炼的客户合同需求的板坯和无委托板坯的总重量应接近转炉标准冶炼容量，超出转炉标准冶炼容量部分的重量和不足转炉标准冶炼容量部分的重量都要小于一块板坯的重量；所述的无委托板坯是指为满足转炉冶炼过程中要求满批生产工艺而产出的没有和客户合同管理的剩余材料
具体公式如下：
Figure PCTCN2016105581-appb-000073
Figure PCTCN2016105581-appb-000074
Figure PCTCN2016105581-appb-000075
其中，C表示转炉标准冶炼容量，Qi表示产品i的生产欠量，
Figure PCTCN2016105581-appb-000076
表示钢种为g且宽度为w的批次数上限，
Figure PCTCN2016105581-appb-000077
表示一个批次中不足转炉标准冶炼容量的部分，
Figure PCTCN2016105581-appb-000078
表示一个批次中超出转炉标准冶炼容量的部分，lstd表示连铸生产中无委托板坯的标准长度，hstd表示连铸生产中无委托板坯的标准厚度，ρ表示钢水的密度；
步骤4-2-4、建立每一炉钢水在连铸机上浇铸过程中，两流钢水消耗量的炉流平衡工艺约束，即要求同一炉钢水在连铸机两流的浇铸时间相等，在模型上映射为两流铸出的板坯块数相等；
具体公式如下：
Figure PCTCN2016105581-appb-000079
其中，ngwk表示辅助整数变量，表示钢种为g宽度为w的第k个批次奇流生产的板坯块数；
步骤4-2-5、建立板坯在连铸设备上切割长度范围的工艺约束，即受连铸机切割工艺和客户订货长度的限制，要求一炉钢水中铸造出的任意板坯的长度在规定范围内；
具体公式如下：
Figure PCTCN2016105581-appb-000080
其中，hi表示产品i所需的板坯厚度，
Figure PCTCN2016105581-appb-000081
和
Figure PCTCN2016105581-appb-000082
表示产品i所需的板坯最大和最小长度；
步骤4-2-6、构建客户订货量的柔性管理约束，即不足或超出客户订货量的部分要小于一块板坯的重量；
Figure PCTCN2016105581-appb-000083
Figure PCTCN2016105581-appb-000084
Figure PCTCN2016105581-appb-000085
其中，
Figure PCTCN2016105581-appb-000086
表示欠量不足部分，
Figure PCTCN2016105581-appb-000087
表示欠量超出部分；
步骤4-3、将炼钢生产过程中优化的工艺指标映射为数学模型目标函数，实现最小化所有 批次产出的无委托板坯总重量、最小化钢种之间优充替代总量、最小化所有批次中产出板坯重量同转炉标准冶炼容量偏差总量、最小化所有客户合同的订货量偏差总量；
本发明实施例中，将无委托板坯的库存成本指标映射为式(33)的目标函数，即最小化所有批次产出的无委托板坯总重量；
Figure PCTCN2016105581-appb-000088
本发明实施例中，将钢种之间优充替代损失成本指标映射为式(34)的目标函数，即最小化钢种之间优充替代总量；
Figure PCTCN2016105581-appb-000089
本发明实施例中，将转炉批生产的作业效率指标映射为式(35)的目标函数，即最小化所有批次中产出板坯重量同转炉标准冶炼容量偏差；
Figure PCTCN2016105581-appb-000090
本发明实施例中，将客户对订货重量满意度的管理指标映射为式(36)的目标函数，即最小化所有客户合同的订货量偏差；
Figure PCTCN2016105581-appb-000091
综上所述，目标函数，具体如下：
Figure PCTCN2016105581-appb-000092
其中，F0表示小合同组生产组批总成本，所述的小合同组生产组批总成本即为所有批次无委托板坯总量、钢种之间优充替代总量、所有客户合同的订货量偏差和转炉标准冶炼容量偏差的线性加权总和，λ1，λ2，λ3，λ4∈[0，1]，表示不同目标的权重系数，且λ1+λ2+λ3+λ4＝1。
步骤5、构建一种实数矩阵与组批方案之间的相互映射关系，并以所建立的实数矩阵作为被控对象，实现基于多对象并行迭代改进策略获取最终优化的组批方案，进而获得小合同组在连铸工序的预组批方案，方法流程如图5所示，具体如下：
步骤5-1、构建一种实数矩阵与组批方案之间的相互映射关系，具体如下：
步骤5-1-1、构建一个实数矩阵，该矩阵的维数为全部产品数、钢种和宽度的乘积，矩阵中的元素为某一合同分配到某一钢种且某一宽度的所有批次内的板坯重量占上述合同生产欠量的比率；
本发明实施例中，设计如公式(12)所表达的|N|×(|G|×|W|)维实数矩阵A：
Figure PCTCN2016105581-appb-000093
其中，A表示一个|N|×(|G|×|W|)维实数矩阵，N表示给定的小合同组的全部产品集合，W表示浇铸产品集，N时连铸机的结晶器所需设定的全部可能宽度集合，a|N|，|G|，|W|表示对应合同N分配到钢种为G且宽度为W的所有批次内的板坯重量占合同N生产欠量的比率关系；
步骤5-1-2、获得某一合同分配到目标钢种且目标宽度的所有批次内的板坯重量和分配到目标钢种且目标宽度的所有批次内的所有合同板坯的重量，并对所有钢种和宽度组合，按所有批次内的所有合同板坯的重量值从大到小进行排序，并按该顺序重复执行步骤5-1-3至步骤5-1-9，得到所有钢种和宽度组合内合同的组批方案：
给定矩阵A，按公式(13)计算合同i分配到钢种为g且宽度为w的所有批次内的板坯重量bigw，按公式(14)计算分配到钢种为g且宽度为w的所有批次内的所有合同板坯的重量Bgw：
Figure PCTCN2016105581-appb-000094
Figure PCTCN2016105581-appb-000095
其中，bigw表示合同i分配到钢种为g且宽度为w的所有批次内的板坯重量，Bgw表示分配到钢种为g且宽度为w的所有批次内的所有合同板坯的重量，aigw表示对应合同i分配到钢种为g且宽度为w的所有批次内的板坯重量占合同i生产欠量的比率关系，Qi表示产品i的生产欠量。
步骤5-1-3、对任意钢种和宽度组合(g，w)，统计所有合同分配到该钢种和宽度组合批次内的板坯重量，记为(b1gw b2gw … b|N|gw)T，创建一个不包含任何合同的空批次k，置该批次内已经包含的板坯重量Ek＝0；
步骤5-1-4、在板坯重量向量(b1gw b2gw … b|N|gw)T中选取第一个bigw＞0的合同i，并比较空批次k的剩余容量C-Ek与上述第一个板坯重量bigw的大小，若剩余容量C-Ek大于等于第一个板坯重量bigw，则执行步骤5-1-5，否则，执行步骤5-1-6；
步骤5-1-5、将客户订货量的柔性管理约束条件中对应产品的生产欠量Qi替换为该产品板坯重量bigw，并按照步骤4-2-5至步骤4-2-6限定的工艺条件获取zigwk块的板坯，将上述板坯放到空批次k中，更新该批次的板坯重量Ek并设置bigw＝0；
步骤5-1-6、将客户订货量的柔性管理约束条件中对应产品的生产欠量Qi替换为剩余容 量C-Ek，并按照步骤4-2-5至步骤4-2-6限定的工艺条件获取zigwk块的板坯，将其放到空批次k中，更新Ek并置bigw＝0；
步骤5-1-7、在不添加无委托板坯的情况下，判断空批次k内的包含板坯是否满足转炉每一批次冶炼容量的限制的工艺约束条件，若是，则执行步骤5-1-8，否则，执行步骤5-1-9；
步骤5-1-8、判断空批次k内的包含板坯是否满足两流钢水消耗量的炉流平衡工艺约束条件，若是，则直接创建下一个不包含任何合同的空批次，并设置该批次内已经包含的板坯重量为0，否则，通过从该空批次k中增加或者移除一块板坯来修复该批次使其满足炉流平衡约束，再创建下一个不包含任何合同的空批次，并设置该批次内已经包含的板坯重量Ek＝0；
步骤5-1-9、判断板坯重量向量(b1gw b2gw … b|N|gw)T是否等于0，若是，则在最后一个不为空的批次内，按转炉每一批次冶炼容量的限制的工艺约束条件和两流钢水消耗量的炉流平衡工艺约束条件添加无委托板坯，否则，返回执行步骤5-1-4；
步骤5-1-10、所有批次内的所有合同板坯的重量均执行完步骤5-1-3至步骤5-1-9，获得所有钢种和宽度组合内合同的组批方案；
步骤5-2、以所建立的实数矩阵作为被控对象，实现基于多对象并行迭代改进策略获取最终优化的组批方案，具体包括：
步骤5-2-1、随机生成NP个与步骤5-1-1所述实数矩阵结构相同的实数矩阵，并将所有构建的实数矩阵放入集合{A1，A2，…，ANP}中，每个实数矩阵Aj(j＝1，2，…，NP)中满足条件i∈Ng∩Pw的三元组(i，g，w)，aigw设置为1，不满足的元素设置为0；
其中，aigw是从均匀分布实数区间[L，U]内产生的随机数，L和U分别是区间下界和上界；NP为预先设定基于多对象并行迭代改进策略算法的种群规模参数；
步骤5-2-2、将生成的所有实数矩阵Aj(j＝1，2，…，NP)返回执行步骤5-1-1至步骤5-1-10，建立每个实数矩阵Aj与组批方案的对应关系，根据组批方案获得决策变量(x，z，y)的取值，将其代入目标函数中(即公式(11)中)，获得每个实数矩阵对应的目标函数f(Aj)；
步骤5-2-3、将所获得的目标函数f(Aj)从小到大进行排序，并将排名前二分之一的实数矩阵分为一组，记为S，将排名后二分之一的实数矩阵分为一组，记为I，即S和I满足max{f(Aj)|Aj∈S}≤min{f(Aj)|Aj∈I}；
步骤5-2-4、根据每个实数矩阵对应的目标函数所在分组，对每个实数矩阵Aj(j＝1，2，…，NP)进行变异操作和交叉操作，获得操作后实数矩阵，再将操作后所有实数矩阵返回执行步骤5-1-1至步骤5-1-10，建立每个操作后实数矩阵与组批方案的对应关系，根据组批方案获得决策变量的取值，将其代入目标函数中，获得每个操作后实数矩阵对应的目标函数；具体步骤 如下：
步骤5-2-4-1、在由实数矩阵构成的集合{A1，A2，…，ANP}中选取三个不同于目标实数矩阵Aj且互不相同的实数矩阵Ar1、Ar2、Ar3，即j≠r1≠r2≠r3；
步骤5-2-4-2、从均匀分布的实数区间[j/NP，1]随机生成一个变异步长因子Fj；
步骤5-2-4-3、将实数矩阵Aj、Ar1、Ar2、Ar3进行差分运算，获得变异操作后的实数矩阵Vj；
Figure PCTCN2016105581-appb-000096
其中，Ar*当前目标函数值最小的实数矩阵，满足f(Ar*)＝min{f(Aj)|j＝1，2，…，NP}；S′表示将所获得的目标函数从小到大进行排序，排名前二分之一的实数矩阵分组，I表示将所获得的目标函数从小到大进行排序，排名后二分之一的实数矩阵分组；
步骤5-2-4-4、从均匀分布的实数区间[i/NP，1]为矩阵的每个元素随机生成一个交叉概率因子CRigw；
步骤5-2-4-5、将每对实数矩阵Aj和Vj执行交叉操作，生成实数矩阵Uj；
Figure PCTCN2016105581-appb-000097
其中，
Figure PCTCN2016105581-appb-000098
表示实数矩阵Uj内部的元素，
Figure PCTCN2016105581-appb-000099
表示实数矩阵Vj内部的元素，
Figure PCTCN2016105581-appb-000100
表示实数矩阵Aj内部的元素，j＝1，2，…，NP，i∈N，w∈W，k∈{1，2，…，Kgw}；W表示浇铸产品集N时连铸机的结晶器所需设定的全部可能宽度集合；Kgw表示钢种为g且宽度为w的批次数上限；
Figure PCTCN2016105581-appb-000101
表示服从正态分布的(0，1)之间随机数；
步骤5-2-4-6、判断是否元素
Figure PCTCN2016105581-appb-000102
或者
Figure PCTCN2016105581-appb-000103
若是，则对实数矩阵Uj进行边界条件处理，否则，完成交叉操作；
具体公式如下：
Figure PCTCN2016105581-appb-000104
其中，L表示元素
Figure PCTCN2016105581-appb-000105
的取值范围下限，U表示元素
Figure PCTCN2016105581-appb-000106
的取值范围上限。
步骤5-2-5、对每个实数矩阵Uj(j＝1，2，…，NP)执行步骤(5.1.1-5.1.6)，建立Uj与组批方案的对应关系，将组批方案对应决策变量(x，z，y)的取值代入公式(19)，得到该组批方案下的加权目标函数，记为f(Uj)；比较f(Uj)和f(Aj)，按公式(26)对矩阵进行更新，对更新后的矩阵 集合{A1，A2，…，ANP}重复执行步骤5-2-2至步骤5-2-4直到集合{A1，A2，…，ANP}不再被更新；
步骤5-2-6、在最终的矩阵集合{A1，A2，…，ANP}中选择目标函数值f(Aj)最小的实数矩阵Aj*，，将该矩阵返回执行步骤5-1-1至步骤5-1-10，获得最终的优化组批方案；
步骤5-3、将获得的炼钢批次按照钢种和宽度进行合并，即具有相同钢种和宽度的炼钢批次合并为一个连浇批组，完成小合同组在连铸工序的预组批方案的指定；
步骤6、制定大合同组在炼钢工序的分批方案和连铸工序的预分批方案；
本发明实施例中，对于给定的大合同分组所需求产品，计算该组内所有产品的生产欠量
Figure PCTCN2016105581-appb-000107
按转炉冶炼满批生产工艺要求，计算该大合同组需要生产的炼钢批次数
Figure PCTCN2016105581-appb-000108
按照中间包最大连浇炉数Tmax，将K个炼钢批次在连铸工序上分解为
Figure PCTCN2016105581-appb-000109
个连浇批组，其中前
Figure PCTCN2016105581-appb-000110
个连浇批组包含Tmax炉，
Figure PCTCN2016105581-appb-000111
个连浇批组包含
Figure PCTCN2016105581-appb-000112
炉；
本发明实施例中，对机组产能进行了计算，对库存结构进行了统计，具体为：通过计算炼钢、精炼、连铸和热轧设备的标准产能与设备检修计划导致的停机时间之差得到炼钢、精炼、连铸和热轧设备在计划期内各天的生产能力信息；对给定的全部客户合同需求产品，按合同不同维度，对合同多个属性进行统计分析，包括获取不同流向合同欠量分布信息、不同制造流程合同欠量分布信息、不同品种属性合同欠量分布信息、不同精炼方式欠量比例信息，确定重点合同交货期信息、板坯合同交货期信息，获取热轧工序特殊品种(烫辊材、难轧材、IF钢、箱板)集批需求信息、冷轧工序不同材料组别需求信息；统计全流程网络中各库存设备上的在制品库存信息；根据机组产能和库存结构统计结果，确定计划时间段内RH精炼每天需求炉次数范围、热轧工序烫辊材和难轧材每天需求重量范围、冷轧工序各个材料组别每天需求重量范围；
步骤7、采用构建定量化的数学模型的方式确定连浇批组在连铸设备上排产决策；具体包括：选取连浇批组排产的决策变量；定量化描述连浇批组排产所追求的目标；定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求；方法流程如图6所示，具体步骤如下：
步骤7-1、选取连浇批组排产的决策变量；具体如下：
设定0-1决策变量urls表示连浇批组r是否被分配在连铸设备s的第l个位置上；设定辅助变量Qls表示连铸设备s的第l个位置的连浇批组浇铸完成时间；设定辅助变量Tr表示连浇批组r的浇铸完成时间；设定辅助变量ti表示合同i的浇铸完成时间；设定辅助变量
Figure PCTCN2016105581-appb-000113
表示热轧厂h前库在第d天的烫棍材计划库存量；设定辅助变量
Figure PCTCN2016105581-appb-000114
表示热轧厂h前库在第d 天的难轧材计划库存量；设定辅助变量Δfd表示炼钢厂在第d天为f流向计划生产的板坯量；
步骤7-2、定量化描述连浇批组排产所追求的目标；
具体包括：最大化中间包利用率、最小化异钢种连浇板坯数、最小化调宽板坯数、最小化烫辊材库存量偏差、最小化难轧材库存量偏差、最小化热轧和冷轧各流向需求量的偏差和最小化客户合同拖期时间；
所述的最大化中间包利用率；
即要求浇铸完所有连浇批组所使用的中间包个数最小，具体公式为：
Figure PCTCN2016105581-appb-000115
其中，FI表示所有连浇批组排产所更新中间包总成本；S表示连铸设备集合；R表示所有连浇批组集合；ar表示连浇批组r中包含的炉数；Tunmax表示一个中间包最大可浇铸的炉数；bs表示分配到连铸设备s上的不可连浇的批组个数；lτ，
Figure PCTCN2016105581-appb-000116
表示第τ个不可混浇的批组位置，即连浇批组
Figure PCTCN2016105581-appb-000117
和
Figure PCTCN2016105581-appb-000118
的钢种严禁混浇或者连浇批组
Figure PCTCN2016105581-appb-000119
和
Figure PCTCN2016105581-appb-000120
宽度差异超过连铸机最大允许的在线调宽幅度。
所述的最小化异钢种连浇板坯数；
具体公式为：
Figure PCTCN2016105581-appb-000121
其中，F2表示连浇批组排产所有异钢种连浇总成本；
Figure PCTCN2016105581-appb-000122
表示连铸设备s上第l个位置上分配的连铸批组，
Figure PCTCN2016105581-appb-000123
表示连铸批组
Figure PCTCN2016105581-appb-000124
的钢种；
所述的最小化调宽板坯数；
具体公式为：
Figure PCTCN2016105581-appb-000125
其中，F3表示连浇批组排产所有调宽总成本；
Figure PCTCN2016105581-appb-000126
表示连铸批组
Figure PCTCN2016105581-appb-000127
的宽度，函数h(.)在前后两个参数相同时取0，否则取1；
所述的最小化烫辊材库存量偏差；
即要求热轧前库的烫棍材计划库存量同目标库存量的偏差最小以保证热轧生产的顺畅性；
具体公式如下：
Figure PCTCN2016105581-appb-000128
其中，F4表示所有烫辊材库存风险成本；H表示热轧厂集合，D表示计划期内的天数集合，
Figure PCTCN2016105581-appb-000129
表示热轧厂h在第d天需求的烫棍材目标库存量；
所述的最小化难轧材库存量偏差；
即要求热轧前库的难轧材计划库存量同最大和最小允许的难轧材库存量的偏差最小以减少难轧材过多导致物流堵塞；
具体公式如下：
Figure PCTCN2016105581-appb-000130
其中，F5表示所有难轧材库存风险成本；
Figure PCTCN2016105581-appb-000131
表示热轧厂h在第d天最大的难轧材库存量；
Figure PCTCN2016105581-appb-000132
表示热轧厂h在第d天最小允许的难轧材库存量；
所述的最小化客户合同拖期时间；
具体公式如下：
Figure PCTCN2016105581-appb-000133
其中，F6表示客户合同满意度收益；NR表示有严格交货期要求的合同集合，Eari表示合同i的最早交货期，Duei表示合同i的最晚交货期；
步骤7-3、定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求；具体如下：
制定连浇批组在连铸设备上的分配关系约束，即要求每个连浇批组只能分配到一个连铸设备上的一个位置，每个连铸设备上的每个位置最多只能分配一个连浇批组，每个连铸设备上未分配连浇批组的位置一定是在已分配连浇批组的位置之后，具体公式如下：
Figure PCTCN2016105581-appb-000134
Figure PCTCN2016105581-appb-000135
Figure PCTCN2016105581-appb-000136
制定可行分配规则约束，即只有连浇批组内所包含的合同生产制程与给定的连铸设备兼 容时才允许将连浇批组分配到该连铸设备上，具体公式如下；
urls≤vrs        (27)
式中，vrs表示连浇批组r内所包含的合同生产制程与连铸设备s兼容性参数，vrs∈{0，1}。
具体包括：连浇批组在连铸设备上的分配关系约束和可行分配规则约束；
步骤8、将步骤7所建立的数学模型为定量计算依据，通过建立一种实数向量与连浇批组在连铸设备上排产方案之间的相互映射关系，并以所建立的实数向量作为被控对象获得基于多对象并行迭代改进策略来获得连浇批组在连铸设备上排产方案；即获得连浇批组对于连铸设备的分配和顺序；
具体如下：
步骤8-1、设定一个2|R|维实数向量PP＝[a1，a2…aR|b1，b2…b|R|]，并确定实数向量PP内部数值；
其中，ar和br为无量纲实参数，取值范围为[0，1]；1≤r≤|R|，R表示所有连浇批组集合；
步骤8-2、按br值确定分配到任意连铸设备s上的连浇批组集合Rs；
即
Figure PCTCN2016105581-appb-000137
步骤8-3、按ar值确定分配到任意连铸设备s上所有连浇批组的排序；
本发明实施例中，对于任意连铸设备s，按ar值确定分配到该设备上所有连浇批组的排序，即集合Rs中ar值最小的连浇批组arg min{ar|r∈Rs}排在连铸设备s的第一个位置，ar值第二小的连浇批组arg min{ar|r∈Rs}排在连铸设备s的第二个位置，以此类推。
步骤9、组批计划与排产集成方案调整，下发和执行。
采用本发明所述的对组批计划和生产调度方案进行集成方法得到实施例中为期一周的生产合同数据的组批及两岸调度方案如下表所示：
Figure PCTCN2016105581-appb-000138
Figure PCTCN2016105581-appb-000139
本发明实施例中，对组批计划和生产调度方案进行集成获得最终的生产组织方案，根据月初及月末计划特殊性、铁水实际供给量、重点合同交货期信息、不过RH精炼的清冷钢合同预留及全流程物流衔接等情况，再结合实际现场生产波动情况对生产组织方案进一步微调后，下达给炼钢阶段各生产制造单元，各生产制造单元按该方案进行备料和执行生产，以达到钢铁物质流在全流程工序设备和时间维度上的均衡与准时分布。

Claims (8)

  1. 一种面向全流程生产的炼钢组批与排产方法，其特征在于，包括以下步骤：
    步骤1、采用构建有向网络拓扑图的方式描述生产环境；
    其中，有向网络拓扑图上的每个节点表示一个具体的生产机组或库存设备，包括：转炉、精炼炉、连铸机、板坯库、热轧机组、平整机组、热卷库、酸洗机组和酸轧机组；有向网络拓扑图上的每条弧表示从一个机组或库存设备到另一个机组或库存设备之间存在具体的物料转移过程，包括：钢水、板坯、热卷和冷卷；
    步骤2、根据不同客户合同对最终产品的质量要求，设置产品的工艺参数，包括：确定产品的制造流程在有向网络拓扑图上的映射、按照钢种计算不同产品在连铸机上的浇铸宽度范围、确定钢种间的优充替代关系、确定不同钢种在中间包中的混浇关系及成本；
    步骤3、根据客户合同需求产品的钢种、品种属性、可选制造流程和宽度范围，判断产品订单所属分组，若客户需求总欠量大于或等于中间包允许的最大工艺连浇炉数，则属于大合同组，执行步骤6；若客户需求的产品总欠量小于中间包允许的最大工艺连浇炉数，则属于小合同组，执行步骤4至步骤5；
    步骤4、采用构建数学模型的方式描述多产品在炼钢工序上的组批生产决策；
    具体包括以下步骤：
    步骤4-1、将炼钢生产过程中多产品组批方案映射为数学模型决策变量；
    步骤4-2、将炼钢生产过程的工艺限制映射为数学模型约束条件，具体如下：
    步骤4-2-1、建立产品钢种替代关系的工艺约束；
    步骤4-2-2、建立产品在连铸设备上的浇铸宽度范围的工艺约束；
    步骤4-2-3、建立转炉每一批次冶炼容量的限制的工艺约束，即限定在同一批次内冶炼的客户合同需求的板坯和无委托板坯的总重量应接近转炉标准冶炼容量，超出转炉标准冶炼容量部分的重量和不足转炉标准冶炼容量部分的重量都要小于一块板坯的重量；
    所述的无委托板坯是指为满足转炉冶炼过程中要求满批生产工艺而产出的没有和客户合同管理的剩余材料；
    步骤4-2-4、建立每一炉钢水在连铸机上浇铸过程中，两流钢水消耗量的炉流平衡工艺约束，即要求同一炉钢水在连铸机两流的浇铸时间相等，在模型上映射为两流铸出的板坯块数相等；
    步骤4-2-5、建立板坯在连铸设备上切割长度范围的工艺约束，即受连铸机切割工艺和客户订货长度的限制，要求一炉钢水中铸造出的任意板坯的长度在规定范围内；
    步骤4-2-6、构建客户订货量的柔性管理约束，即不足或超出客户订货量的部分要小于 一块板坯的重量；
    步骤4-3、将炼钢生产过程中优化的工艺指标映射为数学模型目标函数，实现最小化所有批次产出的无委托板坯总重量、最小化钢种之间优充替代总量、最小化所有批次中产出板坯重量同转炉标准冶炼容量偏差总量、最小化所有客户合同的订货量偏差总量；
    步骤5、构建一种实数矩阵与组批方案之间的相互映射关系，并以所建立的实数矩阵作为被控对象，实现基于多对象并行迭代改进策略获取最终优化的组批方案，进而获得小合同组在连铸工序的预组批方案，具体如下：
    步骤5-1、构建一种实数矩阵与组批方案之间的相互映射关系，具体如下：
    步骤5-1-1、构建一个实数矩阵，该矩阵的维数为全部产品数、钢种和宽度的乘积，矩阵中的元素为某一合同分配到某一钢种且某一宽度的所有批次内的板坯重量占上述合同生产欠量的比率；
    步骤5-1-2、获得某一合同分配到目标钢种且目标宽度的所有批次内的板坯重量和分配到目标钢种且目标宽度的所有批次内的所有合同板坯的重量，并对所有钢种和宽度组合，按所有批次内的所有合同板坯的重量值从大到小进行排序，并按该顺序重复执行步骤5-1-3至步骤5-1-9；
    步骤5-1-3、确定所有合同分配到任意钢种和宽度组合批次内的板坯重量向量，并构建空批次，设置该批次内已经包含的板坯重量为0；
    步骤5-1-4、在板坯重量向量选取第一个板坯重量大于0的合同，并比较空批次的剩余容量与上述第一个板坯重量的大小，若剩余容量大于等于第一个板坯重量，则执行步骤5-1-5，否则，执行步骤5-1-6；
    步骤5-1-5、将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为该产品板坯重量，并按照步骤4-2-5至步骤4-2-6限定的工艺条件获取整数块的板坯，将上述板坯放到空批次中，更新该批次的板坯重量并设置在板坯重量向量中该产品板坯重量为0；
    步骤5-1-6、将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为剩余容量，并按照步骤4-2-5至步骤4-2-6限定的工艺条件获取整数块的板坯，将其放到空批次中，更新该批次的板坯重量并设置在板坯重量向量中该产品板坯重量为0；
    步骤5-1-7、在不添加无委托板坯的情况下，判断空批次内的包含板坯是否满足转炉每一批次冶炼容量的限制的工艺约束条件，若是，则执行步骤5-1-8，否则，执行步骤5-1-9；
    步骤5-1-8、判断空批次内的包含板坯是否满足两流钢水消耗量的炉流平衡工艺约束条件，若是，则直接创建下一个不包含任何合同的空批次，并设置该批次内已经包含的板坯重 量为0，否则，通过从该空批次中增加或者移除一块板坯来修复该批次使其满足炉流平衡约束，再创建下一个不包含任何合同的空批次，并设置该批次内已经包含的板坯重量为0；
    步骤5-1-9、判断板坯重量向量是否等于0，若是，则在最后一个不为空的批次内，按转炉每一批次冶炼容量的限制的工艺约束条件和两流钢水消耗量的炉流平衡工艺约束条件添加无委托板坯，否则，返回执行步骤5-1-4；
    步骤5-1-10、所有批次内的所有合同板坯的重量均执行完步骤5-1-3至步骤5-1-9，获得所有钢种和宽度组合内合同的组批方案；
    步骤5-2、以所建立的实数矩阵作为被控对象，实现基于多对象并行迭代改进策略获取最终优化的组批方案，具体包括：
    步骤5-2-1、随机生成NP个与步骤5-1-1所述实数矩阵结构相同的实数矩阵，并将所有构建的实数矩阵放入集合中，每个矩阵中满足目标钢种且目标宽度的元素设置为1，不满足的元素设置为0；
    其中，NP为预先设定基于多对象并行迭代改进策略算法的种群规模参数；
    步骤5-2-2、将生成的所有实数矩阵返回执行步骤5-1-1至步骤5-1-10，建立每个实数矩阵与组批方案的对应关系，根据组批方案获得决策变量的取值，将其代入目标函数中，获得每个实数矩阵对应的目标函数；
    步骤5-2-3、将所获得的目标函数从小到大进行排序，并将排名前二分之一的实数矩阵分为一组，将排名后二分之一的实数矩阵分为一组；
    步骤5-2-4、根据每个实数矩阵对应的目标函数所在分组，对每个实数矩阵进行变异操作和交叉操作，获得操作后实数矩阵，再将操作后所有实数矩阵返回执行步骤5-1-1至步骤5-1-10，建立每个操作后实数矩阵与组批方案的对应关系，根据组批方案获得决策变量的取值，将其代入目标函数中，获得每个操作后实数矩阵对应的目标函数；
    步骤5-2-5、判断操作前后实数矩阵对应的目标函数的大小，选择对应目标函数较小的实数矩阵为更新后的实数矩阵，获得更新后的矩阵集合，并返回执行步骤5-2-2至步骤5-2-4，直至矩阵集合不再更新，获得最终的矩阵集合；
    步骤5-2-6、在最终的矩阵集合中选择目标函数值最小的实数矩阵，将该矩阵返回执行步骤5-1-1至步骤5-1-10，获得最终的优化组批方案；
    步骤5-3、将获得的炼钢批次按照钢种和宽度进行合并，即具有相同钢种和宽度的炼钢批次合并为一个连浇批组，完成小合同组在连铸工序的预组批方案的指定；
    步骤6、制定大合同组在炼钢工序的分批方案和连铸工序的预分批方案；
    步骤7、采用构建定量化的数学模型的方式确定连浇批组在连铸设备上排产决策；具体包括：选取连浇批组排产的决策变量；定量化描述连浇批组排产所追求的目标；定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求；具体步骤如下：
    步骤7-1、选取连浇批组排产的决策变量；
    步骤7-2、定量化描述连浇批组排产所追求的目标；
    具体包括：最大化中间包利用率、最小化异钢种连浇板坯数、最小化调宽板坯数、最小化烫辊材库存量偏差、最小化难轧材库存量偏差、最小化热轧和冷轧各流向需求量的偏差和最小化客户合同拖期时间；
    步骤7-3、定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求；
    具体包括：连浇批组在连铸设备上的分配关系约束和可行分配规则约束；
    步骤8、将步骤7所建立的数学模型为定量计算依据，通过建立一种实数向量与连浇批组在连铸设备上排产方案之间的相互映射关系，并以所建立的实数向量作为被控对象获得基于多对象并行迭代改进策略来获得连浇批组在连铸设备上排产方案；
    即获得连浇批组对于连铸设备的分配和顺序；
    步骤9、组批计划与排产集成方案调整，下发和执行。
  2. 根据权利要求1所述的面向全流程生产的炼钢组批与排产方法，其特征在于，
    步骤4-1所述的将炼钢生产过程中多产品组批方案映射为数学模型决策变量；具体如下：
    设定连续决策变量xigwk，表示合同i在钢种为g宽度为w的第k个批次的生产板坯重量；设定整数决策变量zigwk，表示合同i在钢种为g宽度为w的第k个批次的生产板坯块数；设定整数决策变量z0gwk，表示钢种为g宽度为w的第k个批次内无委托板坯块数；设定0-1决策变量ygwk，当钢种为g宽度为w的第k个批次决定生产时，ygwk取值为1；否则ygwk取值为0；
    步骤4-2-1所述的建立产品钢种替代关系的工艺约束；
    即对任意钢种g，确定由该钢种生产的产品集合Ng；
    Figure PCTCN2016105581-appb-100001
    其中，N表示给定小合同组的全部产品集，gi表示产品i的钢种，G＝∪i∈Ngi表示产品集N中包含的全部钢种集，sgig表示产品i的钢种gi与任意钢种g的替代关系；
    步骤4-2-2所述的建立产品在连铸设备上的浇铸宽度范围的工艺约束；
    即对连铸机结晶器设定的任意宽度w，确定板坯浇铸为该宽度的产品集合Pw；
    Figure PCTCN2016105581-appb-100002
    其中，
    Figure PCTCN2016105581-appb-100003
    表示浇铸产品集N时连铸机的结晶器所需设定的全部宽度集合；
    Figure PCTCN2016105581-appb-100004
    分别表示产品i允许的最大浇铸宽度以及最小浇铸宽度；
    步骤4-2-3所述的建立转炉每一批次冶炼容量的限制的工艺约束，即限定在同一批次内冶炼的客户合同需求的板坯和无委托板坯的总重量应接近转炉标准冶炼容量，超出转炉标准冶炼容量部分的重量和不足转炉标准冶炼容量部分的重量都要小于一块板坯的重量；
    具体公式如下：
    Figure PCTCN2016105581-appb-100005
    Figure PCTCN2016105581-appb-100006
    Figure PCTCN2016105581-appb-100007
    其中，C表示转炉标准冶炼容量，Qi表示产品i的生产欠量，
    Figure PCTCN2016105581-appb-100008
    表示钢种为g且宽度为w的批次数上限，
    Figure PCTCN2016105581-appb-100009
    表示一个批次中不足转炉标准冶炼容量的部分，
    Figure PCTCN2016105581-appb-100010
    表示一个批次中超出转炉标准冶炼容量的部分，lstd表示连铸生产中无委托板坯的标准长度，hstd表示连铸生产中无委托板坯的标准厚度，ρ表示钢水的密度；
    步骤4-2-4所述的建立每一炉钢水在连铸机上浇铸过程中，两流钢水消耗量的炉流平衡工艺约束，即要求同一炉钢水在连铸机两流的浇铸时间相等，在模型上映射为两流铸出的板坯块数相等；
    具体公式如下：
    Figure PCTCN2016105581-appb-100011
    其中，ngwk表示辅助整数变量，表示钢种为g宽度为w的第k个批次奇流生产的板坯块数；
    步骤4-2-5所述的建立板坯在连铸设备上切割长度范围的工艺约束，即受连铸机切割工艺和客户订货长度的限制，要求一炉钢水中铸造出的任意板坯的长度在规定范围内；
    具体公式如下：
    Figure PCTCN2016105581-appb-100012
    其中，hi表示产品i所需的板坯厚度，
    Figure PCTCN2016105581-appb-100013
    和
    Figure PCTCN2016105581-appb-100014
    表示产品i所需的板坯最大和最小长度；
    步骤4-2-6所述的构建客户订货量的柔性管理约束，即不足或超出客户订货量的部分要 小于一块板坯的重量；
    Figure PCTCN2016105581-appb-100015
    Figure PCTCN2016105581-appb-100016
    Figure PCTCN2016105581-appb-100017
    其中，
    Figure PCTCN2016105581-appb-100018
    表示欠量不足部分，
    Figure PCTCN2016105581-appb-100019
    表示欠量超出部分；
    步骤4-3所述的目标函数，具体如下：
    Figure PCTCN2016105581-appb-100020
    其中，F0表示小合同组生产组批总成本，所述的小合同组生产组批总成本即为所有批次无委托板坯总量、钢种之间优充替代总量、所有客户合同的订货量偏差和转炉标准冶炼容量偏差的线性加权总和，λ1，λ2，λ3，λ4∈[0，1]，表示不同目标的权重系数，且λ1+λ2+λ3+λ4＝1。
  3. 根据权利要求1所述的面向全流程生产的炼钢组批与排产方法，其特征在于，步骤5-1-1所述的实数矩阵，具体公式如下：
    Figure PCTCN2016105581-appb-100021
    其中，A表示一个|N|(|G|×|W|)维实数矩阵，N表示给定的小合同组的全部产品集合，W表示浇铸产品集，N时连铸机的结晶器所需设定的全部可能宽度集合，a|N|，|G|，|W|表示对应合同N分配到钢种为G且宽度为W的所有批次内的板坯重量占合同N生产欠量的比率关系。
  4. 根据权利要求1所述的面向全流程生产的炼钢组批与排产方法，其特征在于，步骤5-1-2所述的获得某一合同分配到目标钢种且目标宽度的所有批次内的板坯重量和分配到目标钢种且目标宽度的所有批次内的所有合同板坯的重量，具体计算公式如下：
    Figure PCTCN2016105581-appb-100022
    Figure PCTCN2016105581-appb-100023
    其中，bigw表示合同i分配到钢种为g且宽度为w的所有批次内的板坯重量，Bgw表示分配到钢种为g且宽度为w的所有批次内的所有合同板坯的重量，aigw表示对应合同 i分配到钢种为g且宽度为w的所有批次内的板坯重量占合同i生产欠量的比率关系，Qi表示产品i的生产欠量。
  5. 根据权利要求1所述的面向全流程生产的炼钢组批与排产方法，其特征在于，
    步骤5-1-5所述的将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为该产品板坯重量，即将公式
    Figure PCTCN2016105581-appb-100024
    中的产品i的生产欠量替换为产品板坯重量bigw；
    步骤5-1-6所述的将客户订货量的柔性管理约束条件中对应产品的生产欠量替换为剩余容量，即将公式
    Figure PCTCN2016105581-appb-100025
    中的产品i的生产欠量替换为剩余容量C-Ek，C表示转炉标准冶炼容量，Ek表示批次内已经包含的板坯重量。
  6. 根据权利要求1所述的面向全流程生产的炼钢组批与排产方法，其特征在于，步骤5-2-4所述的根据每个实数矩阵对应的目标函数所在分组，对每个实数矩阵进行变异操作和交叉操作，具体步骤如下：
    步骤5-2-4-1、在由实数矩阵构成的集合{A1，A2，...，ANP}中选取三个不同于目标实数矩阵Aj且互不相同的实数矩阵Ar1、Ar2、Ar3，即j≠r1≠r2≠r3；
    步骤5-2-4-2、从均匀分布的实数区间[j/NP，1]随机生成一个变异步长因子Fj；
    步骤5-2-4-3、将实数矩阵Aj、Ar1、Ar2、Ar3进行差分运算，获得变异操作后的实数矩阵Vj；
    Figure PCTCN2016105581-appb-100026
    其中，
    Figure PCTCN2016105581-appb-100027
    当前目标函数值最小的实数矩阵，满足
    Figure PCTCN2016105581-appb-100028
    S′表示将所获得的目标函数从小到大进行排序，排名前二分之一的实数矩阵分组，I表示将所获得的目标函数从小到大进行排序，排名后二分之一的实数矩阵分组；
    步骤5-2-4-4、从均匀分布的实数区间[j/NP，1]为矩阵的每个元素随机生成一个交叉概率因子CRigw；
    步骤5-2-4-5、将每对实数矩阵Aj和Vj执行交叉操作，生成实数矩阵Uj；
    Figure PCTCN2016105581-appb-100029
    其中，
    Figure PCTCN2016105581-appb-100030
    表示实数矩阵Uj内部的元素，
    Figure PCTCN2016105581-appb-100031
    表示实数矩阵Vj内部的元素，
    Figure PCTCN2016105581-appb-100032
    表 示实数矩阵Aj内部的元素，j＝1，2，...，NP，i∈N，w∈W，k∈{1，2，...，Kgw}；W表示浇铸产品集N时连铸机的结晶器所需设定的全部可能宽度集合；Kgw表示钢种为g且宽度为w的批次数上限；
    Figure PCTCN2016105581-appb-100033
    表示服从正态分布的(0，1)之间随机数；
    步骤5-2-4-6、判断是否元素
    Figure PCTCN2016105581-appb-100034
    或者
    Figure PCTCN2016105581-appb-100035
    若是，则对实数矩阵Uj进行边界条件处理，否则，完成交叉操作；
    具体公式如下：
    Figure PCTCN2016105581-appb-100036
    其中，L表示元素
    Figure PCTCN2016105581-appb-100037
    的取值范围下限，U表示元素
    Figure PCTCN2016105581-appb-100038
    的取值范围上限。
  7. 根据权利要求1所述的面向全流程生产的炼钢组批与排产方法，其特征在于，
    步骤7-1所述的选取连浇批组排产的决策变量，具体如下：
    设定0-1决策变量urls表示连浇批组r是否被分配在连铸设备s的第l个位置上；设定辅助变量Qls表示连铸设备s的第l个位置的连浇批组浇铸完成时间；设定辅助变量Tr表示连浇批组r的浇铸完成时间；设定辅助变量ti表示合同i的浇铸完成时间；设定辅助变量
    Figure PCTCN2016105581-appb-100039
    表示热轧厂h前库在第d天的烫棍材计划库存量；设定辅助变量
    Figure PCTCN2016105581-appb-100040
    表示热轧厂h前库在第d天的难轧材计划库存量；设定辅助变量Δfd表示炼钢厂在第d天为f流向计划生产的板坯量；
    步骤7-2所述的最大化中间包利用率；
    即要求浇铸完所有连浇批组所使用的中间包个数最小，具体公式为：
    Figure PCTCN2016105581-appb-100041
    其中，F1表示所有连浇批组排产所更新中间包总成本；S表示连铸设备集合；R表示所有连浇批组集合；ar表示连浇批组r中包含的炉数；Tunmax表示一个中间包最大可浇铸的炉数；bs表示分配到连铸设备s上的不可连浇的批组个数；Iτ，
    Figure PCTCN2016105581-appb-100042
    表示第τ个不可混浇的批组位置；
    步骤7-2所述的最小化异钢种连浇板坯数；
    具体公式为：
    Figure PCTCN2016105581-appb-100043
    其中，F2表示连浇批组排产所有异钢种连浇总成本；
    Figure PCTCN2016105581-appb-100044
    表示连铸设备s上第l个位置上分配的连铸批组，
    Figure PCTCN2016105581-appb-100045
    表示连铸批组
    Figure PCTCN2016105581-appb-100046
    的钢种；
    步骤7-2所述的最小化调宽板坯数；
    具体公式为：
    Figure PCTCN2016105581-appb-100047
    其中，F3表示连浇批组排产所有调宽总成本；
    Figure PCTCN2016105581-appb-100048
    表示连铸批组
    Figure PCTCN2016105581-appb-100049
    的宽度，函数h(.)在前后两个参数相同时取0，否则取1；
    步骤7-2所述的最小化烫辊材库存量偏差；
    即要求热轧前库的烫棍材计划库存量同目标库存量的偏差最小以保证热轧生产的顺畅性；
    具体公式如下：
    Figure PCTCN2016105581-appb-100050
    其中，F4表示所有烫辊材库存风险成本；H表示热轧厂集合，D表示计划期内的天数集合，
    Figure PCTCN2016105581-appb-100051
    表示热轧厂h在第d天需求的烫棍材目标库存量；
    步骤7-2所述的最小化难轧材库存量偏差；
    即要求热轧前库的难轧材计划库存量同最大和最小允许的难轧材库存量的偏差最小以减少难轧材过多导致物流堵塞；
    具体公式如下：
    Figure PCTCN2016105581-appb-100052
    其中，F5表示所有难轧材库存风险成本；
    Figure PCTCN2016105581-appb-100053
    表示热轧厂h在第d天最大的难轧材库存量；
    Figure PCTCN2016105581-appb-100054
    表示热轧厂h在第d天最小允许的难轧材库存量；
    步骤7-2所述的最小化客户合同拖期时间；
    具体公式如下：
    Figure PCTCN2016105581-appb-100055
    其中，F6表示客户合同满意度收益；NR表示有严格交货期要求的合同集合，Eari表示合同i的最早交货期，Duei表示合同i的最晚交货期；
    步骤7-3所述的定量化描述制定连浇批组排产所需要遵循的工艺约束和管理要求，具体如下：
    制定连浇批组在连铸设备上的分配关系约束，即要求每个连浇批组只能分配到一个连铸设备上的一个位置，每个连铸设备上的每个位置最多只能分配一个连浇批组，每个连铸设备上未分配连浇批组的位置一定是在已分配连浇批组的位置之后，具体公式如下：
    Figure PCTCN2016105581-appb-100056
    Figure PCTCN2016105581-appb-100057
    Figure PCTCN2016105581-appb-100058
    制定可行分配规则约束，即只有连浇批组内所包含的合同生产制程与给定的连铸设备兼容时才允许将连浇批组分配到该连铸设备上，具体公式如下；
    urls≤vrs   (27)
    式中，vrs表示连浇批组r内所包含的合同生产制程与连铸设备s兼容性参数，vrs∈{0，1}。
  8. 根据权利要求1所述的面向全流程生产的炼钢组批与排产方法，其特征在于，步骤8所述的建立一种实数向量与连浇批组在连铸设备上排产方案之间的相互映射关系，具体如下：
    步骤8-1、设定一个2|R|维实数向量PP＝[a1，a2…a|R||b1，b2…b|R|]，并确定实数向量PP内部数值；
    其中，ar和br为无量纲实参数，取值范围为[0，1]；，1≤r≤|R|，R表示所有连浇批组集合；
    步骤8-2、按br值确定分配到任意连铸设备s上的连浇批组集合Rs；
    即
    Figure PCTCN2016105581-appb-100059
    步骤8-3、按ar值确定分配到任意连铸设备s上所有连浇批组的排序；
    即对连浇批组集合Rs按照ar值由小到大进行排序，确定分配到连铸设备s上所有连浇批组的排序。
PCT/CN2016/105581 2015-11-23 2016-11-14 一种面向全流程生产的炼钢组批与排产方法 WO2017088674A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/529,745 US10713399B2 (en) 2015-11-23 2016-11-14 Method of batching and scheduling for steelmaking production with plant-wide process consideration

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510824047.4A CN105483310B (zh) 2015-11-23 2015-11-23 一种面向全流程生产的炼钢组批与排产方法
CN201510824047.4 2015-11-23

Publications (1)

Publication Number Publication Date
WO2017088674A1 true WO2017088674A1 (zh) 2017-06-01

Family

ID=55670581

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/105581 WO2017088674A1 (zh) 2015-11-23 2016-11-14 一种面向全流程生产的炼钢组批与排产方法

Country Status (3)

Country Link
US (1) US10713399B2 (zh)
CN (1) CN105483310B (zh)
WO (1) WO2017088674A1 (zh)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109858780A (zh) * 2019-01-14 2019-06-07 沈阳建筑大学 一种炼钢-连铸生产调度优化方法
CN110059886A (zh) * 2019-04-25 2019-07-26 哈尔滨理工大学 考虑设备批处理的单组工序同时结束的综合调度方法
CN110555563A (zh) * 2019-09-05 2019-12-10 湖北文理学院 大规模钢筋分组及下料优化方法及装置
CN111931975A (zh) * 2020-06-19 2020-11-13 北京化工大学 一种考虑下游扰乱约束下的裂解炉炉群调度建模与方法
CN112204588A (zh) * 2018-05-31 2021-01-08 利乐拉瓦尔集团及财务有限公司 通过模拟在现有生产工厂中对包装乳制品的生产的调度
CN112232584A (zh) * 2020-10-28 2021-01-15 中冶南方工程技术有限公司 一种钢铁企业大宗原燃料汽车取样组批方法和系统
CN112232540A (zh) * 2019-08-02 2021-01-15 中冶南方工程技术有限公司 一种基于多目标优化的钢铁企业汽车取样组批方法
CN114913932A (zh) * 2022-07-15 2022-08-16 北矿智云科技(北京)有限公司 一种配料方法、装置、电子设备和存储介质
CN115129017A (zh) * 2022-08-30 2022-09-30 武汉理工大学 基于混合阿基米德算法的热冷加工离散制造生产调度方法
CN116880427A (zh) * 2023-09-06 2023-10-13 中南大学 一种基于入料条件估计和分工况综合分析的智能控制方法及系统
CN117331358A (zh) * 2023-10-27 2024-01-02 沧州信安保销毁设备有限公司 仓库物料上料控制系统

Families Citing this family (50)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2937747A1 (de) * 2014-04-24 2015-10-28 Siemens Aktiengesellschaft Auf Modellierung einer Beizlinie beruhende Optimierung einer Sequenz von zu beizenden Bändern
CN105483310B (zh) * 2015-11-23 2017-05-10 东北大学 一种面向全流程生产的炼钢组批与排产方法
CN106670416B (zh) * 2017-01-11 2018-07-17 青岛特殊钢铁有限公司 一种适用于碳、硅、锰元素含量均不交叉的异钢种连浇方法
TWI637248B (zh) * 2017-04-24 2018-10-01 中國鋼鐵股份有限公司 煉鋼製程設計方法及系統
CN107993011A (zh) * 2017-12-04 2018-05-04 安徽工业大学 一种硅钢生产计划系统
CN108038605B (zh) * 2017-12-05 2021-12-17 深圳市智物联网络有限公司 一种智慧工厂的plard方法及plard系统
CN108588323B (zh) * 2018-01-09 2020-05-01 东北大学 一种炼钢连铸工业过程优化控制方法
DE102018106393B4 (de) * 2018-03-19 2023-09-28 Thyssenkrupp Ag Verfahren zur Steuerung einer Produktionsanlage eines Walzwerkes
CN108876050B (zh) * 2018-06-27 2021-08-10 东北大学 一种钢铁企业合同主制程的设定与自动转换方法
JP7099111B2 (ja) * 2018-07-18 2022-07-12 日本製鉄株式会社 鋼片の製造プロセスの制御装置、方法及びプログラム
CN110766253B (zh) * 2018-07-27 2023-04-18 宝山钢铁股份有限公司 炼钢连铸计划物流流向设定的炼钢量分配控制方法
CN109447428B (zh) * 2018-10-15 2021-07-20 宁波吉利汽车研究开发有限公司 整车生产线序排产的确定方法、装置及智能终端
CN111445050B (zh) * 2019-01-17 2023-07-28 中国科学院沈阳自动化研究所 一种特种合金铝锭生产的组炉方法
CN110245808B (zh) * 2019-06-25 2023-05-12 湘潭大学 一种基于需量控制的钢包炉优化调度方法
CN110378799B (zh) 2019-07-16 2022-07-12 东北大学 基于多尺度深度卷积网络的氧化铝综合生产指标决策方法
CN110458417A (zh) * 2019-07-17 2019-11-15 首钢京唐钢铁联合有限责任公司 一种资源投入分配的方法及装置
CN110516936B (zh) * 2019-08-13 2022-04-08 武汉攀升鼎承科技有限公司 一种智能生产流水线调配系统
CN110516402B (zh) * 2019-09-05 2023-05-30 中冶南方工程技术有限公司 一种优化电弧炉废钢配料的方法
CN110647124B (zh) * 2019-10-11 2020-07-24 重庆大学 考虑铸轧协调的炼钢连铸热轧一体化生产计划方法及系统
CN110750079A (zh) * 2019-10-25 2020-02-04 福州大学 一种允许工序跳跃的混合流水车间调度优化方法
CN110825045B (zh) * 2019-11-05 2022-03-22 青岛大学 一种提高多装置多产品加工方案切换效率的方法
CN110825056B (zh) * 2019-11-26 2020-12-04 北京工业大学 一种具有可变参数连续加工与间歇加工的混合流水车间调度方法
CN111933223B (zh) * 2020-07-03 2024-01-19 大冶特殊钢有限公司 炼钢合金化过程自动配料方法
CN111899796B (zh) * 2020-07-03 2024-01-19 大冶特殊钢有限公司 使用返回料配料的炼钢合金化过程自动配料方法
CN112182797B (zh) * 2020-09-11 2022-07-05 新兴铸管股份有限公司 一种连铸冷床钢坯跟踪方法
CN112365125A (zh) * 2020-10-16 2021-02-12 重庆恢恢信息技术有限公司 通过大数据分析建筑物料快速使用分配工作方法
CN112232583B (zh) * 2020-10-28 2022-08-05 中冶南方工程技术有限公司 基于预报的大宗原燃料汽车的取样组批方法和系统
CN112699524A (zh) * 2020-11-26 2021-04-23 冶金自动化研究设计院 应用于特钢厂高速工具钢炼钢流程的仿真优化调度模型
CN112508375A (zh) * 2020-11-26 2021-03-16 冶金自动化研究设计院 一种高速工具钢热连轧工序智能排产与报料系统
CN112784398A (zh) * 2020-12-08 2021-05-11 阳春新钢铁有限责任公司 一种基于TFe元素平衡模型的含铁物料资源化利用方法
DE102020133761A1 (de) * 2020-12-16 2022-06-23 RESOUREX GmbH Verfahren zur Optimierung einer Materialausnutzung
CN112836353B (zh) * 2021-01-12 2024-02-27 中国航空工业集团公司北京长城航空测控技术研究所 一种面向批次生产的锻造调度方法
CN112926828B (zh) * 2021-01-22 2023-09-01 东北大学 一种面向中厚板产线的炼钢生产数据解析与优化排程方法
EP4288849A1 (en) * 2021-02-04 2023-12-13 C3.ai, Inc. Constrained optimization and post-processing heuristics for optimal production scheduling for process manufacturing
CN113240238A (zh) * 2021-04-09 2021-08-10 东南大学 一种前向反向法的生产智能调度方法
CN113139720B (zh) * 2021-04-19 2023-10-13 武汉理工大学 一种具有学习效应的铸造车间生产调度优化方法
CN113343567B (zh) * 2021-05-31 2022-03-22 江西理工大学 一种真空熔铸生产工艺参数优化方法及系统
JP2023006476A (ja) * 2021-06-30 2023-01-18 新東工業株式会社 情報表示システム、情報表示方法、及び、情報表示プログラム
CN113449526B (zh) * 2021-08-27 2022-02-08 杭萧钢构股份有限公司 一种钢结构生产调度策略的适用性分析方法及系统
CN113850491B (zh) * 2021-09-17 2023-01-31 北京科技大学 一种连退同品规带钢排产优化方法
CN113988617A (zh) * 2021-10-27 2022-01-28 中国航空综合技术研究所 基于综合效能评估的系统测试诊断需求分解计算方法
CN114037362B (zh) * 2021-12-27 2024-05-24 重庆大学 转炉故障时考虑氧气供需平衡的钢厂重调度方法及调度系统
CN114298567A (zh) * 2021-12-30 2022-04-08 重庆大学 连铸机浇次计划排程及开浇时间动态决策方法及系统
CN114386719B (zh) * 2022-03-22 2022-08-05 宁波钢铁有限公司 一种炉次批量计划优化方法、装置及存储介质
CN115037616B (zh) * 2022-04-22 2023-09-29 东北大学 一种基于tssdn网络的确定性流调度及其更新方法
CN115074479B (zh) * 2022-05-09 2023-12-12 包头钢铁（集团）有限责任公司 一种钢水罐加废钢的温度重量计算方法
CN114662802B (zh) * 2022-05-23 2022-09-09 宁波钢铁有限公司 一种基于扰动因素的热轧板带排程方法及系统
CN115896397B (zh) * 2023-01-09 2023-12-26 湖南华联云创信息科技有限公司 一种lf精炼智能渣系控制方法及系统
CN117289670B (zh) * 2023-11-27 2024-02-13 山东瑞芝生物科技股份有限公司 一种基于物联网的灵芝加工用调控系统及方法
CN117807424B (zh) * 2024-02-29 2024-04-30 山东钢铁股份有限公司 工业大数据驱动的宽厚钢板质量动态在线识别方法及装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003088952A (ja) * 2001-09-18 2003-03-25 Kawasaki Heavy Ind Ltd 鋳造計画作成方法および鋳造計画作成装置
CN1792501A (zh) * 2005-12-27 2006-06-28 上海宝信软件股份有限公司 炼钢连铸生产工艺中的组连浇批次方法
CN101303588A (zh) * 2008-06-04 2008-11-12 东北大学 一种炼钢—连铸炉次批量计划自动编制方法及系统
CN105483310A (zh) * 2015-11-23 2016-04-13 东北大学 一种面向全流程生产的炼钢组批与排产方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1885328B (zh) * 2006-06-20 2012-09-26 东北大学 一种炼钢-连铸中间包批量计划方法及系统
CN102364510A (zh) * 2011-09-16 2012-02-29 中冶南方工程技术有限公司 流程制造业工序关系的表达方法
JP6299155B2 (ja) * 2013-11-05 2018-03-28 新日鐵住金株式会社 キャスト計画立案装置、方法及びプログラム

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2003088952A (ja) * 2001-09-18 2003-03-25 Kawasaki Heavy Ind Ltd 鋳造計画作成方法および鋳造計画作成装置
CN1792501A (zh) * 2005-12-27 2006-06-28 上海宝信软件股份有限公司 炼钢连铸生产工艺中的组连浇批次方法
CN101303588A (zh) * 2008-06-04 2008-11-12 东北大学 一种炼钢—连铸炉次批量计划自动编制方法及系统
CN105483310A (zh) * 2015-11-23 2016-04-13 东北大学 一种面向全流程生产的炼钢组批与排产方法

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112204588A (zh) * 2018-05-31 2021-01-08 利乐拉瓦尔集团及财务有限公司 通过模拟在现有生产工厂中对包装乳制品的生产的调度
CN109858780B (zh) * 2019-01-14 2023-01-31 沈阳建筑大学 一种炼钢-连铸生产调度优化方法
CN109858780A (zh) * 2019-01-14 2019-06-07 沈阳建筑大学 一种炼钢-连铸生产调度优化方法
CN110059886A (zh) * 2019-04-25 2019-07-26 哈尔滨理工大学 考虑设备批处理的单组工序同时结束的综合调度方法
CN112232540B (zh) * 2019-08-02 2022-08-05 中冶南方工程技术有限公司 一种基于多目标优化的钢铁企业汽车取样组批方法
CN112232540A (zh) * 2019-08-02 2021-01-15 中冶南方工程技术有限公司 一种基于多目标优化的钢铁企业汽车取样组批方法
CN110555563B (zh) * 2019-09-05 2022-02-08 湖北文理学院 大规模钢筋分组及下料优化方法及装置
CN110555563A (zh) * 2019-09-05 2019-12-10 湖北文理学院 大规模钢筋分组及下料优化方法及装置
CN111931975A (zh) * 2020-06-19 2020-11-13 北京化工大学 一种考虑下游扰乱约束下的裂解炉炉群调度建模与方法
CN112232584A (zh) * 2020-10-28 2021-01-15 中冶南方工程技术有限公司 一种钢铁企业大宗原燃料汽车取样组批方法和系统
CN112232584B (zh) * 2020-10-28 2023-11-24 中冶南方工程技术有限公司 一种钢铁企业大宗原燃料汽车取样组批方法和系统
CN114913932B (zh) * 2022-07-15 2022-10-14 北矿智云科技(北京)有限公司 一种配料方法、装置、电子设备和存储介质
CN114913932A (zh) * 2022-07-15 2022-08-16 北矿智云科技(北京)有限公司 一种配料方法、装置、电子设备和存储介质
CN115129017B (zh) * 2022-08-30 2022-12-23 武汉理工大学 基于混合阿基米德算法的热冷加工离散制造生产调度方法
CN115129017A (zh) * 2022-08-30 2022-09-30 武汉理工大学 基于混合阿基米德算法的热冷加工离散制造生产调度方法
CN116880427A (zh) * 2023-09-06 2023-10-13 中南大学 一种基于入料条件估计和分工况综合分析的智能控制方法及系统
CN116880427B (zh) * 2023-09-06 2023-11-14 中南大学 基于入料条件估计和分工况分析的智能控制方法及系统
CN117331358A (zh) * 2023-10-27 2024-01-02 沧州信安保销毁设备有限公司 仓库物料上料控制系统

Also Published As

Publication number Publication date
US10713399B2 (en) 2020-07-14
CN105483310B (zh) 2017-05-10
US20170316131A1 (en) 2017-11-02
CN105483310A (zh) 2016-04-13

Similar Documents

Publication Publication Date Title
WO2017088674A1 (zh) 一种面向全流程生产的炼钢组批与排产方法
CN106779220B (zh) 一种炼钢-连铸-热轧集成调度方法与系统
Lee et al. Primary production scheduling at steelmaking industries
CN1885328B (zh) 一种炼钢-连铸中间包批量计划方法及系统
CN104376424B (zh) 一种钢铁企业冷轧区多产线钢卷协调调度方法
CN104635683B (zh) 一种冶金企业全流程机组产能分配控制方法
CN108876050B (zh) 一种钢铁企业合同主制程的设定与自动转换方法
CN114386719B (zh) 一种炉次批量计划优化方法、装置及存储介质
JP2015090532A (ja) キャスト計画立案装置、方法及びプログラム
US20110258087A1 (en) Analytics for setting up strategic inventory systems to handle small lot orders in the steel industry
JP6593080B2 (ja) 製鋼圧延計画立案装置、製鋼圧延計画立案方法、およびプログラム
US7130712B1 (en) Method and system for small-lot orders to optimize production runs in the steel industry
Moengin et al. Model of integrated production-inventory-distribution system: The case of billet steel manufacturing
CN111062571B (zh) 一种面向铝行业的锭型选择与合同组批的集成优化方法
Zhang et al. Solving the order planning problem at the steelmaking shops by considering logistics balance on the plant-wide process
Lü et al. Multiproduct and multistage integrated production planning model and algorithm based on an available production capacity network
JP5652069B2 (ja) 最適チャ−ジ編成装置及び最適チャージ編成方法
CN112926828B (zh) 一种面向中厚板产线的炼钢生产数据解析与优化排程方法
CN106707994A (zh) 钢铁轧制动态调整系统及其方法
JP2000176634A (ja) 鉄鋼製品の生産計画策定装置
JP6477309B2 (ja) 製鋼生産スケジュール作成装置、製鋼生産スケジュール作成方法、操業方法、及び製鋼製品の製造方法
CN114298567A (zh) 连铸机浇次计划排程及开浇时间动态决策方法及系统
Pan et al. Application of extremal optimization approach to the integrated scheduling problem of continuous casting and hot rolling process
Zhang et al. A mixed integer programming model and improved genetic algorithm for order planning of iron-steel plants
RU43384U1 (ru) Автоматизированная система управления в реальном времени производством продукции по заказам клиентов на предприятии (варианты)

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 15529745

Country of ref document: US

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16867898

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16867898

Country of ref document: EP

Kind code of ref document: A1