An integrated feeding plan optimization method and system for wide and thick plates based on NSGA-Ⅲ

Abstract

The present invention discloses a method and system for optimizing the integrated feeding plan of wide and thick plates based on NSGA-III, the method comprising: obtaining order and process related parameters, and defining decision variables; designing two-stage objective functions and constraints, and constructing a two-stage multi-objective optimization model based on the obtained order and process related parameters, and the defined decision variables; wherein the first-stage objective function is used to solve the feeding balance problem and optimize the order execution timing and relative position; the second-stage objective function is used to continue to optimize the furnace order and slab sequence corresponding to the order based on the optimization result of the first-stage multi-objective optimization model; based on the NSGA-III algorithm, the two-stage multi-objective optimization model is solved to obtain an optimized feeding plan. The scheme of the present invention can significantly optimize the feeding sequence, improve the coordination and balance of materials between processes, reduce material relocation and waiting, and improve production efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of wide and thick plate multi-process production plan optimization, and in particular to a wide and thick plate integrated feeding plan optimization method and system based on NSGA-III. Background Art

[0002] In the field of wide and thick plate production, the integrated feeding stage consisting of steelmaking and continuous casting processes is the key link in feeding subsequent hot rolling, heat treatment, etc. The main processes include converters, refining and continuous casting. Each process has multiple production equipment and has a strong selectivity for equipment. It belongs to a mixed flow scheduling problem of multiple processes and multiple parallel machines. It is necessary to schedule the furnace plan and casting plan reasonably to achieve the goal of continuous and efficient batch production.

[0003] At present, the optimization research of feeding plan mainly adopts the multi-objective optimization idea, establishes the mathematical model of steelmaking-continuous casting production planning problem, and uses search algorithm, genetic algorithm and ant colony algorithm to solve it. In the past 6 or 7 years, NSGA algorithm has been introduced to solve the scheduling problem of steelmaking-continuous casting, mainly using NSGA-II algorithm. Related research is as follows:

[0004] Reference 1 (Multi-objective optimization of steelmaking-continuous casting scheduling based on NSGA-II, Applied Research, 1007-9416 (2016) 08-0061-02) proposed a multi-objective optimization method for steelmaking-continuous casting scheduling based on NSGA-II; Reference 2 (Application research of improved NSGA2 in steelmaking-continuous casting scheduling, Control Engineering, Vol. 27, No. 3, March 2020) used the improved NSGA2 algorithm based on priority strategy to solve the steelmaking-continuous casting scheduling problem with multiple refining methods; Reference 3 (A steelmaking-continuous casting scheduling optimization method considering energy constraints, Journal of Huazhong University of Science and Technology (Natural Science Edition), Vol. 50, No. 6, June 2022) On the basis of considering energy constraints, the fitness hierarchical evolution and prophet population strategy were used to improve the non-dominated sorting multi-objective genetic evolution algorithm (NSGA-Ⅱ) to solve the steelmaking-continuous casting production scheduling problem.

[0005] The main problem with the above research is that it does not take into account the connection between subsequent production processes such as hot rolling, finishing, and heat treatment, resulting in uneven distribution of subsequent materials in time and space and reduced capacity utilization of equipment in each process. Summary of the invention

[0006] The present invention provides an integrated feeding plan optimization method and system for wide and thick plates based on NSGA-Ⅲ, so as to solve the technical problems existing in the prior art that the connection of subsequent hot rolling, finishing, heat treatment and other processes is not considered, resulting in uneven distribution of subsequent materials in time and space and reduced capacity utilization of equipment in each process.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] On the one hand, the present invention provides an optimization method for integrated feeding plan of wide and thick plates based on NSGA-III, and the optimization method for integrated feeding plan of wide and thick plates based on NSGA-III comprises:

[0009] Obtain order and process related parameters and define decision variables;

[0010] Design two-stage objective functions and constraints, and build a two-stage multi-objective optimization model based on the acquired order and process-related parameters and the defined decision variables; the first-stage objective function is used to solve the problem of material balance and optimize the timing and relative position of order execution; the second-stage objective function is used to continue to optimize the furnace order and slab sequence corresponding to the order based on the optimization results of the first-stage multi-objective optimization model;

[0011] Based on the NSGA-III algorithm, the two-stage multi-objective optimization model is solved to obtain the optimized feeding plan.

[0012] Furthermore, the order and process related parameters include: order number o, slab number s, the latest delivery date T′ required for the slab number s in the order number o OS and production end date T os , the weight W of the slab numbered s S , Heat number f, thickness H of the same heat f 、Width B f And steel grade G f , the weight W of the heat numbered f f 、Upper limit of single furnace capacity W U and the lower limit W l , the target rolling width B of the slab numbered a in the heat numbered f fa The target rolling width of the slab numbered b and the target rolling thickness H of the slab numbered a in the heat numbered f are fa and the target rolling thickness H of the slab numbered b fb , steel type G of slab numbered a in heat numbered f fa and steel grade G of slab numbered b fb , casting number j, delivery date T of heat number a in casting number j ja and the delivery date T of the batch number b jb , the minimum error time allowed for delivery time T c , steelmaking-continuous casting process number G, equipment number m, average target rolling width B of the previous heat f and the average target rolling width B′ of the next heat f, the start processing time S of the heat numbered f on the equipment numbered m in the process numbered g gmf , the start processing time S of the heat numbered f on the equipment numbered m' in the next process numbered g' g′m′f , the processing time T of the heat numbered f on the equipment numbered m in the process numbered g gmf The time T for transporting the heat numbered f from the equipment numbered m in the process numbered g to the equipment numbered m' in the next process numbered g' (g,g′)(m,m′)f , the start processing time S of the heat numbered f on the equipment numbered m in the pouring numbered i jmf , the start processing time S of the next heat numbered f′ on the equipment numbered m in the pouring numbered j jmf ′, the processing time T of the heat numbered f on the equipment numbered m in the pouring numbered j jmf , the start time T of the furnace numbered f f , the start time T' of the heat numbered f' that is adjacent to the heat numbered f f , the number of furnaces that can be accommodated in the same casting batch q, the maximum number of furnaces J u , steel grade content G in heat numbered f in pouring numbered j jf , the thickness H of the heat numbered a in the pouring numbered j ja , the thickness H of the casting numbered b in the casting numbered j jb , the width B of the heat numbered f in the pouring numbered j jf 、Width of the first furnace B jfs 、The width of the last furnace B jfe , Single width decrease maximum value ΔB r , the total range limit of the same casting batch width decreasing ΔB t , the optimal proportion of online cutting volume R, the number of slabs with the same heat treatment temperature point in the neighborhood of slab numbered s N s , and the minimum batch production volume Q of heat treatment min .

[0013] Furthermore, the decision variables include:

[0014]

[0015]

[0016]

[0017]

[0018]

[0019]

[0020] Furthermore, the objective function of the first stage aims at minimizing the delay, minimizing the remaining capacity of the furnace, maximizing the number of furnaces in the same casting, optimizing the online cutting ratio of the feeding steel plate, and batch production of heat treatment at the same temperature; the expression of the objective function of the first stage is:

[0021] Minimize delays: Among them, n represents the maximum value of the slab number;

[0022] Minimize the remaining capacity of the furnace: Where k represents the total number of heats;

[0023] Maximize the number of heats in the same pour: Among them, q represents the maximum value of the device number;

[0024] The online cutting ratio of steel feed steel plates is optimal, that is, a batch plan is evenly divided into r time periods, so that the absolute value of the difference between the ratio of the number of steel plates that meet the thickness specifications to the total number of steel plates and the specified optimal ratio is minimized. The expression is:

[0025] Heat treatment batch production at the same temperature:

[0026] Furthermore, the objective function of the second stage aims to minimize the steelmaking waiting time and minimize the average width reduction of the product in each furnace; the expression of the objective function of the second stage is:

[0027] Minimize steelmaking waiting time: min(f1+f2), where:

[0028]

[0029]

[0030] Among them, p represents the maximum value of the process number; l represents the maximum value of the pouring number;

[0031] Minimize the average product width reduction in each furnace:

[0032] Furthermore, the constraints include:

[0033] Each slab can only be placed in one heat;

[0034] The total weight of slabs accommodated in each heat is limited by the maximum furnace capacity;

[0035] The steel grade, thickness and width of slabs in the same heat are the same;

[0036] In steelmaking production, each heat can only be processed by one piece of equipment.

[0037] Each steelmaking equipment can process the next heat only after processing one heat;

[0038] The same batch of heat can enter the next process only after the previous process is completed;

[0039] All heats are scheduled and only scheduled to one pouring;

[0040] Limitation on the number of heats included in the same pouring;

[0041] The thickness of slabs in each heat in the same casting is the same;

[0042] The steel grades of adjacent heats in the same pour are similar, measured by the difference in composition: C jf -C j(f+1) ≤R c ; R c is a constant;

[0043] The width of the billet in each heat within the same casting is not increased, the single casting range is limited, and the continuous casting total range is limited: 0≤B jf -B j(f+1) ≤ΔB r ; where ΔB r It is the maximum value of single width decrease;

[0044] Where, ΔB t The total range limit of decreasing width of the furnaces in the same casting batch;

[0045] The target rolling width of each slab in the same heat increases;

[0046] The delivery time within the same pouring is similar: T ja- T jb ≤T c ,a,b∈{1,2,…,k} and a≠b.

[0047] Furthermore, the two-stage multi-objective optimization model is solved based on the NSGA-III algorithm to obtain an optimized feeding plan, including:

[0048] Adopting adaptive mechanism to improve NSGA-III algorithm, we get improved NSGA-III algorithm;

[0049] The improved NSGA-III algorithm is used to solve the optimization model according to the objective function of the first stage, and a preliminary solution set of the feeding plan is obtained;

[0050] Selecting some solutions from the preliminary feeding plan solution set as the initial population of the second-stage NSGA-III algorithm, using the improved NSGA-III algorithm, solving the optimization model according to the second-stage objective function, and obtaining an optimized feeding plan solution set;

[0051] The optimal decision method of fuzzy membership is adopted to select the only optimal solution from the optimized feeding plan solution set solved by the second stage NSGA-III algorithm as the optimized feeding plan.

[0052] Furthermore, the improved NSGA-III algorithm is used to solve the optimization model according to the first-stage objective to obtain a preliminary feed plan solution set, including:

[0053] First, chromosome encoding is performed. Chromosome design follows the basic principle that any chromosome can be decoded to form a feasible solution;

[0054] Population initialization, defining a set of reference points, using a double-layer reference point generation method to reduce the number of reference points while ensuring their wide distribution;

[0055] Crossover and mutation operation: the parent population generates a temporary population through random selection, simulation of two-point crossover and polynomial mutation, introduces an adaptive feedback mechanism to generate the child population, and merges the parent population and the child population to obtain a new population;

[0056] Target normalization processing, calculate the minimum value of the current generation population in each dimension of the target, generate an ideal point set, transform the target function, subtract the value of the ideal point from the target value of all individuals, that is, translate the population to meet the function requirements, calculate the extreme points of each target, construct a linear hyperplane and calculate the intercept, and finally calculate the target normalization value of each individual;

[0057] The selection operation uses the niche method to select dominant individuals to enter the next generation population and determine whether the number of iterations has been reached. If not, select again according to the rules. If it has been reached, the population is output as the algorithm solution result.

[0058] Furthermore, the optimal decision method of fuzzy membership is adopted to select the only optimal solution from the optimized feeding plan solution set solved by the second-stage NSGA-III algorithm as the optimized feeding plan, including:

[0059] The fuzzy membership function is used to dedimensionalize each objective function, and the membership of each non-dominated solution is standardized and then sorted. The smallest non-dominated solution is the only optimal solution for the model.

[0060] On the other hand, the present invention also provides an integrated feeding plan optimization system for wide and thick plates based on NSGA-III, and the integrated feeding plan optimization system for wide and thick plates based on NSGA-III comprises:

[0061] Data acquisition module, used to obtain order and process related parameters and define decision variables;

[0062] The optimization problem modeling module is used to design two-stage objective functions and constraints, and to construct a two-stage multi-objective optimization model based on the order and process-related parameters acquired by the data acquisition module and the defined decision variables; wherein the first-stage objective function is used to solve the problem of material supply balance and optimize the timing and relative position of order execution; the second-stage objective function is used to continue to optimize the furnace order and slab sequence corresponding to the order based on the optimization results of the first-stage multi-objective optimization model;

[0063] The optimization model solving module is used to solve the two-stage multi-objective optimization model constructed by the optimization problem modeling module based on the NSGA-III algorithm to obtain an optimized feeding plan.

[0064] On the other hand, the present invention further provides an electronic device, comprising a processor and a memory; wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the above method.

[0065] In yet another aspect, the present invention further provides a computer-readable storage medium, wherein the storage medium stores at least one instruction, and the instruction is loaded and executed by a processor to implement the above method.

[0066] The beneficial effects brought about by the technical solution provided by the present invention include at least:

[0067] The present invention starts from the perspective of connecting the entire process of steel production and maximizing the utilization of materials. It constructs a two-stage multi-objective optimization model according to the goals of first improving the balance of feeding, optimizing the timing and relative position of order execution, and then optimizing the furnace and slab sequence corresponding to the order. The NSGA-III algorithm improved by the adaptive mechanism is used to solve the two-stage multi-objective optimization model to obtain the optimal scheduling solution. On the one hand, the two-stage solution effectively improves the effectiveness and accuracy of the solution; on the other hand, it can significantly optimize the feeding sequence, improve the rolling production efficiency, reduce material transshipment and waiting, and improve production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0069] Figure 1 It is a schematic diagram of the execution flow of the wide and thick plate integrated feeding plan optimization method based on NSGA-III provided in an embodiment of the present invention;

[0070] Figure 2 is a schematic diagram of chromosome encoding provided by an embodiment of the present invention;

[0071] Figure 3 It is a schematic diagram of a plan for 30 furnaces and 2 castings provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0072] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0073] First embodiment

[0074] In order to solve the technical problems in the prior art that the connection of subsequent hot rolling, finishing, heat treatment and other processes is not considered, resulting in uneven distribution of subsequent materials in time and space and reduced capacity utilization of equipment in each process, this embodiment combines the characteristics of the wide and thick plate production process, comprehensively considers production efficiency, benefit optimization, process constraints and other conditions, and optimizes the design of the integrated feeding plan from the perspective of the whole process connection from steelmaking to heat treatment, and provides an integrated feeding plan optimization method for wide and thick plates based on NSGA-Ⅲ for integrated feeding plan scheduling optimization. The optimization method can be implemented by an electronic device, which can be a terminal or a server. The execution flow of the method is as follows: Figure 1 As shown, the following steps are included:

[0075] S1, obtain order and process related parameters and define decision variables;

[0076] Specifically, the order and process-related parameters obtained in this embodiment include order, slab, furnace, casting and other related parameters, specifically: the order number is o, o∈{1,2,…,m}; the latest delivery date and production end date required by the slab s in order o are T′ respectively OS , T os , T os ≤T′ OS +D (D is the maximum time that the order can be delayed); the slab number is s, s∈{1,2,…,n}; the weight of slab s is W S; The heat number is f, f∈{1,2,…,k}, with a total of k heats; the thickness, width and steel type of the same heat are H f， B f , G f ;;The weight of heat f is W f ; The upper and lower limits of the single furnace capacity are W U and W l ; The target rolling widths of slabs a and b in heat f are B fa , B fb ; The target rolling thickness of slabs a and b in heat f is H fa , H fb ; The steel grades of slabs a and b in heat f are G fa , G fb ; The casting number is j, j∈{1,2,…,l}; The delivery dates of heats a and b in casting j are T ja and T jb ,a,b∈{1,2,…,k} and a≠b; the minimum error time allowed for delivery is T c ; The process number of steelmaking-continuous casting is G, g∈{1,2,…,p}; The equipment number is m, m∈{1,2,…,q}; The average target rolling width of the adjacent previous heat and the next heat is B f and B′ f ; The starting processing time of heat f on process g equipment m is S gmf ; The starting processing time of furnace f on the next process g′ equipment m′ is S g′m′f ; The processing time of heat f on equipment m in process g is T gmf ; The time for transporting furnace f from equipment m in process g to equipment m' in the next process g' is T (g,g′)(m,m′)f ; The start processing time of furnace f on equipment m of casting j is S jmf ; The start time of the next heat f′ on the casting j equipment m is S jmf′ ; The processing time of furnace f on equipment m for casting j is T jmf ; The start time of furnace f is T f ; The start time of the next furnace f' is T' f ; The number of furnaces that can be accommodated in the same casting is q; the maximum number of furnaces is J u ; The steel grade content of heat f in casting j is G jf ; The thickness of heat a in pouring j is H ja ; The thickness of heat b in casting j is H jb ,a,b∈{1,2,…,k}; the width of the furnace f in the casting j is B jf ; The width of the first furnace is The width of the last furnace is The maximum value of a single width decrease is ΔB r ; The total range of decreasing width of the same casting batch is limited to ΔB t ; The optimal proportion of online cutting volume is R; The number of slabs with the same heat treatment temperature point in the neighborhood of slab s is N s ; The minimum batch production volume of heat treatment is Q min .

[0077] Specifically, this embodiment randomly selects 6428 order items from the monthly order set according to the date range, corresponding to a total order volume of 74297 tons. The orders are merged and grouped according to delivery date, steel type, thickness group, heat treatment requirements, corresponding billet thickness, billet width, etc., to generate 258 aggregated order data items, some of which are shown in Table 1.

[0078] Table 1 Part of the order data

[0079]

[0080]

[0081] Furthermore, the decision variables include:

[0082]

[0083]

[0084]

[0085] Among them, t represents the time period specified by a batch plan, and r represents the number of batch plans that can be evenly divided into t time periods in sequence;

[0086]

[0087]

[0088]

[0089] S2, design two-stage objective functions and constraints, and build a two-stage multi-objective optimization model based on the acquired order and process-related parameters and decision variables; the first-stage objective function is used to solve the problem of material balance and optimize the timing and relative position of order execution; the second-stage objective function is used to continue to optimize the furnace order and slab sequence corresponding to the order based on the optimization results of the first-stage multi-objective optimization model;

[0090] It should be noted that, in this embodiment, a two-stage objective function is designed to greatly reduce the coupling degree of the problem and reduce the complexity of the solution. The first stage solution mainly solves the problem of material balance and the execution timing and relative position of the order; wherein, the first stage objective function aims to minimize the delay, minimize the remaining capacity of the furnace, maximize the number of furnaces in the same casting, optimize the online cutting ratio of the feeding steel plate, and batch production of heat treatment at the same temperature; the expression of the first stage objective function is:

[0091] Minimize delays and ensure order delivery time:

[0092] Minimize the remaining capacity of the furnace:

[0093] Maximize the number of heats in the same pour:

[0094] The online cutting ratio of steel feeding steel plates is optimal, that is, a batch plan is evenly divided into r time periods, each time period is t hours, so that the absolute value of the difference between the ratio of the number of steel plates that meet the thickness specifications to the total number of steel plates and the specified optimal ratio is minimized, promoting the balance of the processing capacity of the finishing process equipment. The expression is:

[0095] Heat treatment batch production at the same temperature:

[0096] Among them, the two optimization goals of "optimal online cutting ratio of steel feeding and steel plates and batch production of heat treatment at the same temperature" will help optimize the utilization rate of online cutting and offline fire cutting equipment in the subsequent hot rolling process, and ensure the optimized feeding of subsequent heat treatment processes.

[0097] Furthermore, the objective function of the second stage aims to minimize the steelmaking waiting time and the average width reduction of the product in each furnace; the furnaces and slabs corresponding to the order are restored, and the execution sequence is accurately optimized and solved; the expression of the objective function of the second stage is:

[0098] Minimize the steelmaking waiting time. The steelmaking waiting time is mainly composed of two times: the interval time caused by the interruption of molten steel pouring in each furnace and the waiting time of processing equipment in each process such as LF and RH. The minimum steelmaking waiting time is expressed as: min(f1+f2), where:

[0099]

[0100]

[0101] Minimize the average product width reduction in each furnace, that is, the average target rolling width of each furnace in the same casting is reduced:

[0102] Among them, the optimization goal of "minimizing the average product width reduction per furnace" will improve the disorder of the rolling width corresponding to the feeding plan, promote the consistency of the feeding of the upper process with the demand of the lower process, and improve production efficiency.

[0103] The model constraints include:

[0104] 1) Each slab can only be placed in one heat:

[0105] 2) The total weight of slabs accommodated in each furnace is limited by the maximum furnace capacity:

[0106] 3) Slabs of the same batch with the same steel grade, thickness and width: G fa＝ G fb ,H fa＝ H, B fa＝ B fb a,b∈{1,2,…,n};

[0107] 4) In each heat of steelmaking production, the same process can only be processed by one of the LF, RH, continuous casting and other equipment:

[0108] 5) Each steelmaking equipment can process the next heat only after processing one heat (the transportation time between heats is negligible): S gmf +T gmf ≤S gm(f+1) ;

[0109] 6) The same batch can only enter the next process after the previous process is completed: S gmf +T gmf +T (g，g+1)mf +≤S (g+1)mf ;

[0110] 7) All heats are arranged and only arranged in one pouring:

[0111] 8) Limitation on the number of heats included in the same pouring:

[0112] 9) The thickness of slabs in each heat of the same casting is the same: H ja =H jb ,a,b∈{1,2,…,k};

[0113] 10) The steel grades of adjacent heats in the same pour are similar, measured by the difference in composition: C jf- C j(f+1) ≤R c ;

[0114] In the formula, R c ——given constant;

[0115] 11) The width of the billet in each heat within the same casting is not increased, the single casting range is limited, and the continuous casting total range is limited: 0≤B jf -B j(f+1) ≤ΔB r ; In the formula, ΔB r ——The maximum value of single width decrease. Where, ΔB t The total range limit of decreasing width of the furnaces in the same casting batch;

[0116] 12) The target rolling width of each slab in the same heat increases: B fa -B f(a+i) ≤0,i∈{1,2,…,k};

[0117] 13) The delivery time within the same pouring is similar: T ja- T jb ≤T c ,a,b∈{1,2,…,k} and a≠b.

[0118] S3, solve the two-stage multi-objective optimization model based on the NSGA-III algorithm to obtain the optimized feeding plan.

[0119] Specifically, in this example, the above S3 includes the following steps:

[0120] S31, adopting the adaptive mechanism to improve the NSGA-III algorithm, and obtaining the improved NSGA-III algorithm;

[0121] S32, using the improved NSGA-III algorithm, solves the optimization model according to the first stage objective function to obtain a preliminary feed plan solution set; specifically, it includes:

[0122] 1) First, chromosome encoding is performed. Chromosome design follows the basic principle that any chromosome can be decoded to form a feasible solution.

[0123] 2) Population initialization: define a set of reference points and adopt a double-layer reference point generation method to reduce the number of reference points while ensuring their wide distribution.

[0124] 3) Crossover and mutation operation. The parent population generates a temporary population through random selection, simulated binary crossover (SBX) and polynomial mutation, and introduces an adaptive feedback mechanism to generate a child population Qt with a population size of N. The parent population Pt and the child population Qt are merged to obtain a new population Rt, and the population size becomes 2N.

[0125] 4) Target normalization. Calculate the minimum value of the current generation population in each dimension of the target, generate an ideal point set, transform the target function, subtract the value of the ideal point from the target value of all individuals, that is, translate the population to meet the function requirements, calculate the extreme points of each target, construct a linear hyperplane and calculate the intercept, and finally use the formula to calculate the target normalization value of each individual.

[0126] 5) Selection operation. Using the niche method, select dominant individuals to enter the next generation population, and determine whether the number of iterations has been reached. If not, select again according to the rules. If it has been reached, the population is output as the first stage solution result.

[0127] The parameters of the first-stage NSGA-III algorithm are set as follows: population size is 100, crossover probability is 0.86, and mutation probability is 0.15. Figure 2 According to the NSGA-III algorithm, the double-layer reference point method is then used to initialize the population and define a set of reference points. The target dimension of the first stage model in this embodiment is 5. On the basis of ensuring the uniform distribution of reference points, the number of boundary equal segments is set to 5, generating 126 reference points. Then, crossover mutation operation, target normalization processing and selection operation are performed to obtain the 3-day casting machine production specification scheduling plan, as shown in Table 2.

[0128] Table 2 Casting machine production specification scheduling plan

[0129]

[0130] S33, select some solutions from the preliminary solution set of the feeding plan as the initial population of the second-stage NSGA-III algorithm, adopt the improved NSGA-III algorithm, solve the optimization model according to the second-stage objective function, and obtain the optimized feeding plan solution set; specifically: in combination with the needs of scheduling, according to the number of pours and the number of pours in the overall scheduling plan, select a solution set of a certain number of pours and furnaces as the initial population, as the initial population of the second-stage NSGA-III algorithm according to the second-stage objectives, perform cross-mutation operations again, target normalization and selection operations, and generate a new solution set.

[0131] Specifically, this embodiment selects a preliminary scheduling plan of 2 pourings and 30 furnaces from the first stage scheduling plan as the initial population for the second stage. The smelting time of each process equipment and the NSGA-III algorithm parameter settings are shown in Table 3. Then, the NSGA-III crossover mutation, target normalization and selection operations are performed to obtain a Pareto solution set containing multiple optimal solutions.

[0132] Table 3 Smelting time of each process equipment in the second stage and NSGA-III algorithm parameter settings

[0133]

[0134] S34, using the optimal decision-making method of fuzzy membership, selects the only optimal solution from the optimized feeding plan solution set solved by the second-stage NSGA-III algorithm as the optimized feeding plan; specifically: use the fuzzy membership function to dedimensionalize each objective function, standardize the membership of each non-dominated solution and then sort them, and the smallest non-dominated solution is the only optimal solution for the model.

[0135] Based on the above, this embodiment obtains a plan for 30 heats and 2 pours, such as Figure 3 shown.

[0136] Furthermore, this embodiment also conducted a comparative test on the operating efficiency of the two-stage NSGA-III algorithm solution. The test showed that from the perspective of algorithm operation efficiency, the program operation efficiency and effect of the two-stage NSGA-III model solution are much better than the one-time high-dimensional solution model. The running time of the two-stage solution model is only about 8% of the one-time solution model, and the feed optimization effect of the former is significantly better than the latter. The comparison indicators are shown in Table 4.

[0137] Table 4 Comparison of the running efficiency of the two-stage NSGA-III model solution and the one-stage NSGA-III model solution

[0138]

[0139] Furthermore, this embodiment also compares the order of rolling width of adjacent slabs under different feeding modes, including comparative analysis of the rolling width trends obtained by the feeding plan model considering the hot rolling plan factor and not considering the hot rolling plan factor. As can be seen from Table 5, the rolling width corresponding to the feeding plan considering the hot rolling plan factor changes smoothly, while the rolling width without considering this factor changes greatly. This shows that the data model design in the present invention takes "minimizing the average product width reduction of the furnace" as the optimization goal, which is effective and necessary for improving rolling efficiency.

[0140] In addition, this embodiment further compares the orderliness of the rolling widths of adjacent slabs under two different feeding modes, one considering the rolling width target and the other not considering the rolling width target. The test comparison results in Table 5 show the impact of the integrated ordered feeding model on billet supply. The results of 10 calculations show that when the rolling width target is not considered, the average disorder of the rolling width corresponding to the feeding plan is as high as 17.97, while after considering this target, the average disorder is reduced to 2.31. This shows that the feeding model can significantly improve the contradiction between the feeding of the upper process and the demand of the lower process, which plays a very important role in integrated production scheduling.

[0141] Table 5 Comparison of the order of rolling width of adjacent slabs under two different feeding modes

[0142]

[0143] In summary, this embodiment shows that: from the perspective of the connection between casting and rolling, the casting plan formed takes into account the hot rolling loading sequence and the factors of billet warehouse transfer and stacking to the maximum extent. The feeding optimization model solves the problems that cannot be solved unilaterally by billet warehouse management and hot rolling plan optimization; from the perspective of the whole process feeding optimization, the system solves the balance problem between multiple optimization objectives better through the two-stage NSGA-III solution. The steelmaking plan obtained by system optimization makes it possible to optimize the utilization rate of online cutting and offline fire cutting equipment in the later process of hot rolling from the perspective of resource allocation, and also ensures the optimized feeding of subsequent heat treatment processes; the hot rolling plan roll scheduling is generally required to obey the trend of rolling width from wide to narrow. The steelmaking production plan takes the rolling process requirements into consideration in the integrated plan, which can effectively improve the consistency of material processing sequence between processes, reduce material transfer and waiting, and improve production efficiency.

[0144] Second embodiment

[0145] This embodiment provides a wide and thick plate integrated feeding plan optimization system based on NSGA-III, and the wide and thick plate integrated feeding plan optimization system based on NSGA-III includes the following modules:

[0146] Data acquisition module, used to obtain order and process related parameters and define decision variables;

[0147] The optimization problem modeling module is used to design two-stage objective functions and constraints, and to construct a two-stage multi-objective optimization model based on the order and process-related parameters acquired by the data acquisition module and the defined decision variables; wherein the first-stage objective function is used to solve the problem of material supply balance and optimize the timing and relative position of order execution; the second-stage objective function is used to continue to optimize the furnace order and slab sequence corresponding to the order based on the optimization results of the first-stage multi-objective optimization model;

[0148] The optimization model solving module is used to solve the two-stage multi-objective optimization model constructed by the optimization problem modeling module based on the NSGA-III algorithm to obtain an optimized feeding plan.

[0149] The integrated material feeding plan optimization system for wide and thick plates based on NSGA-Ⅲ of the present embodiment corresponds to the integrated material feeding plan optimization method for wide and thick plates based on NSGA-Ⅲ of the above-mentioned first embodiment; wherein, the functions implemented by each functional module in the integrated material feeding plan optimization system for wide and thick plates based on NSGA-Ⅲ of the present embodiment correspond one by one to each process step in the integrated material feeding plan optimization method for wide and thick plates based on NSGA-Ⅲ of the above-mentioned first embodiment; therefore, they will not be repeated here.

[0150] Third embodiment

[0151] This embodiment provides an electronic device, which includes a processor and a memory; wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the method of the first embodiment.

[0152] The electronic device may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) and one or more memories, wherein the memory stores at least one instruction, and the instruction is loaded by the processor to execute the above method.

[0153] Fourth embodiment

[0154] This embodiment provides a computer-readable storage medium, which stores at least one instruction, and the instruction is loaded and executed by a processor to implement the method of the first embodiment. The computer-readable storage medium may be a ROM, a random access memory, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc. The instructions stored therein may be loaded by a processor in a terminal to execute the method.

[0155] In addition, it should be noted that the present invention can be provided as a method, an apparatus or a computer program product. Therefore, the embodiments of the present invention can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code.

[0156] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0157] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device, so that a series of operation steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable terminal device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0158] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0159] Finally, it should be noted that the above is a preferred embodiment of the present invention. It should be pointed out that although the preferred embodiment of the present invention has been described, for those skilled in the art, once the basic creative concept of the present invention is known, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications should also be regarded as the protection scope of the present invention. Therefore, the attached claims are intended to be interpreted as including the preferred embodiment and all changes and modifications that fall within the scope of the embodiments of the present invention.

Claims

1. A method for optimizing the integrated feeding plan for wide and thick plates based on NSGA-Ⅲ, characterized in that: include: Obtain order and process related parameters and define decision variables; Design two-stage objective functions and constraints, and build a two-stage multi-objective optimization model based on the acquired order and process-related parameters and the defined decision variables; the first-stage objective function is used to solve the problem of material balance and optimize the timing and relative position of order execution; the second-stage objective function is used to continue to optimize the furnace order and slab sequence corresponding to the order based on the optimization results of the first-stage multi-objective optimization model; Based on the NSGA-III algorithm, the two-stage multi-objective optimization model is solved to obtain the optimized feeding plan; The order and process related parameters include: order number o, slab number s, and the latest delivery date T' required for the slab number s in the order number o. OS and production end date T os , the weight W of the slab numbered s S , heat number f, thickness H of the same heat f 、Width B f And steel grade G f , the weight W of the heat numbered f f 、Upper limit of single furnace capacity W U and the lower limit W l , the target rolling width B of the slab numbered a in the heat numbered f fa The target rolling width of the slab numbered b and the target rolling thickness H of the slab numbered a in the heat numbered f are fa and the target rolling thickness H of the slab numbered b fb , steel type G of slab numbered a in heat numbered f fa and steel grade G of slab numbered b fb , casting number j, delivery date T of heat number a in casting number j ja and the delivery date T of the batch number b jb , the minimum error time allowed for delivery T c , steelmaking-continuous casting process number G, equipment number m, average target rolling width B of the previous heat f and the average target rolling width B′ of the next heat f , the start processing time S of the heat numbered f on the equipment numbered m in the process numbered g gmf , the start processing time S of the heat numbered f on the equipment numbered m' in the next process numbered g' g′m′f , the processing time T of the heat numbered f on the equipment numbered m in the process numbered g gmf The time T for transporting the heat numbered f from the equipment numbered m in the process numbered g to the equipment numbered m' in the next process numbered g' (g,g′)(m,m′)f , the start processing time S of the heat numbered f on the equipment numbered m in the pouring numbered i jmf , the start processing time S of the next heat numbered f′ on the equipment numbered m in the pouring numbered j jmf′ , the processing time T of the heat numbered f on the equipment numbered m in the pouring numbered j jmf , the start time T of the furnace numbered f f , the start time T' of the heat numbered f' that is adjacent to the heat numbered f f , the number of furnaces that can be accommodated in the same casting batch q, the maximum number of furnaces J u , steel grade content G in heat numbered f in pouring numbered j jf , the thickness H of the heat numbered a in the pouring numbered j ja , the thickness H of the casting numbered b in the casting numbered j jb , the width B of the heat numbered f in the pouring numbered j jf , Width of the first furnace Width of the last furnace Maximum value of single width decrease ΔB r , the total range limit of the same casting batch width decreasing ΔB t , the optimal proportion of online cutting volume R, the number of slabs with the same heat treatment temperature point in the neighborhood of slab numbered s N s , and the minimum batch production volume Q of heat treatment min ; The decision variables include: The objective function of the first stage aims to minimize the delay, minimize the remaining capacity of the furnace, maximize the number of furnaces in the same casting, optimize the online cutting ratio of the feeding steel plate, and batch production of heat treatment at the same temperature; the expression of the objective function of the first stage is: Minimize delays: Among them, n represents the maximum value of the slab number; Minimize the remaining capacity of the furnace: Where k represents the total number of heats; Maximize the number of heats in the same pour: Among them, q represents the maximum value of the device number; The online cutting ratio of steel feed steel plates is optimal, that is, a batch plan is evenly divided into r time periods, so that the absolute value of the difference between the ratio of the number of steel plates that meet the thickness specifications to the total number of steel plates and the specified optimal ratio is minimized. The expression is: Heat treatment batch production at the same temperature: The objective function of the second stage aims to minimize the steelmaking waiting time and minimize the average width reduction of the product in each furnace; the expression of the objective function of the second stage is: Minimize steelmaking waiting time: min(f1+f2), where: Among them, p represents the maximum value of the process number; l represents the maximum value of the pouring number; Minimize the average product width reduction in each furnace:

2. The method for optimizing the integrated feeding plan for wide and thick plates based on NSGA-III as claimed in claim 1, characterized in that: The constraints include: Each slab can only be placed in one heat; The total weight of slabs accommodated in each heat is limited by the maximum furnace capacity; The steel grade, thickness and width of slabs in the same heat are the same; In steelmaking production, each heat can only be processed by one piece of equipment. Each steelmaking equipment can process the next heat only after processing one heat; The same batch of heat can enter the next process only after the previous process is completed; All heats are scheduled and only scheduled to one pouring; Limitation on the number of heats included in the same pouring; The thickness of slabs in each heat in the same casting is the same; The steel grades of adjacent heats in the same pour are similar, measured by the difference in composition: C jf- C j(f+1) ≤R c ; R c is a constant; The width of the billet in each heat within the same casting is not increased, the single casting range is limited, and the continuous casting total range is limited: 0≤B jf -B j(f+1) ≤ΔB r ; where ΔB r It is the maximum value of single width decrease; Where, ΔB t The total range limit of decreasing width of the furnaces in the same casting batch; The target rolling width of each slab in the same heat increases; The delivery time within the same pouring is similar: T ja- T jb ≤T c ,a,b∈{1,2,…,k} and a≠b.

3. The method for optimizing the integrated feeding plan for wide and thick plates based on NSGA-III as claimed in claim 1, characterized in that: The NSGA-III algorithm is used to solve the two-stage multi-objective optimization model to obtain an optimized feeding plan, including: Adopting adaptive mechanism to improve NSGA-III algorithm, we get improved NSGA-III algorithm; The improved NSGA-III algorithm is used to solve the optimization model according to the objective function of the first stage, and a preliminary solution set of the feeding plan is obtained; Selecting some solutions from the preliminary feeding plan solution set as the initial population of the second-stage NSGA-III algorithm, using the improved NSGA-III algorithm, solving the optimization model according to the second-stage objective function, and obtaining an optimized feeding plan solution set; The optimal decision method of fuzzy membership is adopted to select the only optimal solution from the optimized feeding plan solution set solved by the second stage NSGA-III algorithm as the optimized feeding plan.

4. The method for optimizing the integrated feeding plan for wide and thick plates based on NSGA-III as claimed in claim 3, characterized in that: The improved NSGA-III algorithm is used to solve the optimization model according to the first stage goal to obtain a preliminary feed plan solution set, including: First, chromosome encoding is performed. Chromosome design follows the basic principle that any chromosome can be decoded to form a feasible solution; Population initialization, defining a set of reference points, using a double-layer reference point generation method to reduce the number of reference points while ensuring their wide distribution; Crossover and mutation operation: the parent population generates a temporary population through random selection, simulation of two-point crossover and polynomial mutation, introduces an adaptive feedback mechanism to generate the child population, and merges the parent population and the child population to obtain a new population; Target normalization processing, calculate the minimum value of the current generation population in each dimension of the target, generate an ideal point set, transform the target function, subtract the value of the ideal point from the target value of all individuals, that is, translate the population to meet the function requirements, calculate the extreme points of each target, construct a linear hyperplane and calculate the intercept, and finally calculate the target normalization value of each individual; The selection operation uses the niche method to select dominant individuals to enter the next generation population and determine whether the number of iterations has been reached. If not, select again according to the rules. If it has been reached, the population is output as the algorithm solution result.

5. The method for optimizing the integrated feeding plan for wide and thick plates based on NSGA-III as claimed in claim 3, characterized in that: The optimal decision method using fuzzy membership selects a unique optimal solution from the optimized feeding plan solution set obtained by the second-stage NSGA-III algorithm as the optimized feeding plan, including: The fuzzy membership function is used to dedimensionalize each objective function, and the membership of each non-dominated solution is standardized and then sorted. The smallest non-dominated solution is the only optimal solution for the model.

6. An integrated feeding plan optimization system for wide and thick plates based on NSGA-Ⅲ, characterized in that: include: Data acquisition module, used to obtain order and process related parameters and define decision variables; The optimization problem modeling module is used to design two-stage objective functions and constraints, and to construct a two-stage multi-objective optimization model based on the order and process-related parameters acquired by the data acquisition module and the defined decision variables; wherein the first-stage objective function is used to solve the problem of material supply balance and optimize the timing and relative position of order execution; the second-stage objective function is used to continue to optimize the furnace order and slab sequence corresponding to the order based on the optimization results of the first-stage multi-objective optimization model; An optimization model solving module is used to solve the two-stage multi-objective optimization model constructed by the optimization problem modeling module based on the NSGA-III algorithm to obtain an optimized feeding plan; The order and process related parameters include: order number o, slab number s, the latest delivery date T required for the slab number s in the order number o O ' S and production end date T os , the weight W of the slab numbered s S , heat number f, thickness H of the same heat f 、Width B f And steel grade G f , the weight W of the heat numbered f f 、Upper limit of single furnace capacity W U and the lower limit W l , the target rolling width B of the slab numbered a in the heat numbered f fa The target rolling width of the slab numbered b and the target rolling thickness H of the slab numbered a in the heat numbered f are fa and the target rolling thickness H of the slab numbered b fb , steel type G of slab numbered a in heat numbered f fa and steel grade G of slab numbered b fb , casting number j, delivery date T of heat number a in casting number j ja and the delivery date T of the batch number b jb , the minimum error time allowed for delivery T c , steelmaking-continuous casting process number G, equipment number m, average target rolling width B of the previous heat f and the average target rolling width B of the next heat f ', the start processing time S of the heat numbered f on the equipment numbered m in the process numbered g gmf , the start processing time S of the heat numbered f on the equipment numbered m' in the next process numbered g' g′m′f , the processing time T of the heat numbered f on the equipment numbered m in the process numbered g gmf The time T for transporting the heat numbered f from the equipment numbered m in the process numbered g to the equipment numbered m' in the next process numbered g' (g,g′)(m,m′)f , the start processing time S of the heat numbered f on the equipment numbered m in the pouring numbered i jmf , the start processing time S of the next heat numbered f′ on the equipment numbered m in the pouring numbered j jmf′ , the processing time T of the heat numbered f on the equipment numbered m in the pouring numbered j jmf , the start time T of the furnace numbered f f , the start time T of the heat numbered f' that is adjacent to the heat numbered f f ', the number of furnaces that can be accommodated in the same casting, q, the maximum number of furnaces, J u , steel grade content G in heat numbered f in pouring numbered j jf , the thickness H of the heat numbered a in the pouring numbered j ja , the thickness H of the casting numbered b in the casting numbered j jb , the width B of the heat numbered f in the pouring numbered j jf 、Width of the first furnace B jfs 、The width of the last furnace B jfe , Single width decrease maximum value ΔB r , the total range limit of the same casting batch width decreasing ΔB t , the optimal proportion of online cutting volume R, the number of slabs with the same heat treatment temperature point in the neighborhood of slab numbered s N s , and the minimum batch production volume Q of heat treatment min ; The decision variables include: The objective function of the first stage aims to minimize the delay, minimize the remaining capacity of the furnace, maximize the number of furnaces in the same casting, optimize the online cutting ratio of the feeding steel plate, and batch production of heat treatment at the same temperature; the expression of the objective function of the first stage is: Minimize delays: Among them, n represents the maximum value of the slab number; Minimize the remaining capacity of the furnace: Where k represents the total number of heats; Maximize the number of heats in the same pour: Among them, q represents the maximum value of the device number; The online cutting ratio of steel feed steel plates is optimal, that is, a batch plan is evenly divided into r time periods, so that the absolute value of the difference between the ratio of the number of steel plates that meet the thickness specifications to the total number of steel plates and the specified optimal ratio is minimized. The expression is: Heat treatment batch production at the same temperature: The objective function of the second stage aims to minimize the steelmaking waiting time and minimize the average width reduction of the product in each furnace; the expression of the objective function of the second stage is: Minimize steelmaking waiting time: min(f1+f2), where: Among them, p represents the maximum value of the process number; l represents the maximum value of the pouring number; Minimize the average product width reduction in each furnace:

Images