Electric furnace production line production plan automatic optimization scheduling method and system

Abstract

The invention discloses an electric furnace production line production plan automatic optimization scheduling method and system, and belongs to the field of automatic control. Actual production data and manufacturing standard data are collected from a manufacturing management and production process control system, and different steel grade characteristics and time standards are determined; extracting and importing a to-be-discharged heat plan from a manufacturing management system; according to a process rule condition and a shutdown maintenance plan, setting a production plan automatic scheduling constraint condition; determining an optimal production plan sorting result on the basis of a genetic algorithm model according to the plan of the heat needing to be arranged and the constraint condition; and according to the steel grade plan sequence and the corresponding time standard, with the process with the longest processing period as the center, the production time table of each process is determined in sequence, and the predicted yield, time consumption information and risk warning information in the plan period are displayed. The scheduling optimization model is used for calculation, the optimized scheduling of the production plan is completed, the planned yield, the planned time consumption and the planned risk point warning are calculated, and the scheduling efficiency, precision and intelligent level of the production plan of the electric furnace production line are improved.

Description

Technical Field

[0001] This invention belongs to the field of steelmaking continuous casting process control, and particularly relates to an automatic optimization scheduling method and system for production planning of electric furnace production lines. Background Technology

[0002] Electric arc furnace steelmaking production lines mainly use scrap steel and molten iron as raw materials and electricity as the main energy source. They obtain qualified molten steel through processes such as electric arc furnace smelting (EAF), ladle refining (LF), and vacuum treatment (VD). Finally, the molten steel is cast into continuously cast billets of various cross-sectional shapes and sizes through the continuous casting process (CC).

[0003] Electric arc furnace (EAF) production lines are characterized by a wide variety of products, small batches, complex furnace conditions, and numerous interfering factors. Different products exhibit variations in steel composition, processing time, production rhythm, and process paths. Furthermore, shutdown and maintenance plans must be arranged based on furnace conditions and equipment status. Therefore, many factors need to be considered in actual production scheduling. The steelmaking plans originally arranged based on contracts and orders often fail to meet the needs of stable and smooth production on-site, frequently requiring manual scheduling for secondary adjustments and organization of the production plan. This places high demands on the experience and technical skills of production schedulers, sometimes making it difficult to ensure the rationality of all furnace scheduling plans. This results in key performance indicators such as output, quality, and energy consumption failing to achieve optimal results, thus affecting the product quality pass rate and the overall production line efficiency.

[0004] The invention patent with authorization announcement date of May 29, 2018, and authorization announcement number CN105550751B, discloses a method for scheduling steelmaking continuous casting production using a priority strategy hybrid genetic algorithm. The method includes: establishing a production scheduling plan objective function, establishing a constraint set, and using a priority strategy hybrid genetic algorithm to iteratively calculate the objective function and obtain decision variables. Specifically, this involves: initializing the model; calculating feasible solutions: designing a segmented combination of real-number codes composed of continuous casting machine start-up time information and furnace processing equipment information, randomly generating operation times based on distribution patterns, and obtaining a time-conflict-free scheduling plan through backward calculation and conflict elimination methods; and population genetic optimization: using the processing weight assignment of executable equipment to quantify the matching relationship between processing equipment in reality, and introducing it into the genetic operation in the form of equipment selection priority strategy for population evolution. This technical solution utilizes a priority measurement hybrid strategy genetic algorithm to solve the uncertainty problem of equipment selection and operation time in production, obtaining an optimized executable steelmaking continuous casting production scheduling plan. It involves the selection of production equipment and operating time for multiple steelmaking and continuous casting production lines, but does not address the issue of simultaneously considering process rule constraints and process maintenance plan constraints when scheduling multiple steel types.

[0005] The invention patent CN108303958B, with an authorization announcement date of March 19, 2019, discloses a multi-objective flexible scheduling method for steelmaking continuous casting processes. The method includes: constructing an uncertain steelmaking-continuous casting scheduling model; obtaining input parameter values ​​from the steel plant's MES database to form an offline scheduling optimization problem; determining an initial population; defining a multi-objective reference point based on decision-maker preferences to transform the multi-objective optimization problem into a single-objective optimization problem; iteratively optimizing the initial population using a chemical reaction optimization algorithm based on the transformed objective function to obtain a set of candidate scheduling schemes; during the iterative optimization process, using a T-test method based on stochastic simulation to coarsely sort and select the current population; and using a knowledge-based neighborhood search strategy to improve the convergence performance of the chemical reaction optimization algorithm. Finally, a precise sorting and selection strategy based on indifference regions is used to obtain the optimal scheduling scheme. This technical solution uses a chemical reaction optimization algorithm to achieve multi-objective optimization of average waiting time, interruption time penalty, and waiting time out-of-bounds penalty. However, this technical solution mainly considers time constraints and does not take into account the constraints of multi-steel grade sorting in the production plan.

[0006] The practical technical problem that this invention aims to solve is how to use a scheduling optimization model to calculate and assist in the automatic completion of production plan optimization scheduling under the constraints of process experience rules and downtime maintenance plans, and automatically calculate planned output, planned time consumption, and planned risk point warnings, so as to achieve the goal of minimizing the conflict between quality and efficiency for two electric furnace production lines under multiple constraint rules. Summary of the Invention

[0007] The technical problem this invention aims to solve is to provide an automatic optimization scheduling method and system for electric arc furnace (EAF) production lines. By employing a hybrid model algorithm that considers both process rule constraints and process time constraints, it can eliminate product quality risks and process time conflicts in EAF production lines. Under the constraints of process experience rules and downtime maintenance plans, it utilizes a scheduling optimization model to assist in automatically optimizing the production schedule. It also automatically calculates planned output, planned time consumption, and warnings of planned risk points. This better meets the scheduling requirements for multiple steel types and small batches in EAF steelmaking, helping production scheduling automatically complete multi-steel type optimization and multi-objective optimization scheduling. It provides a convenient and efficient auxiliary tool for scheduling work, improving the efficiency, accuracy, and intelligence level of EAF production line scheduling.

[0008] The technical solution of this invention is: to provide an automatic optimization scheduling method for production planning in electric furnace production lines, characterized by:

[0009] 1) Collect production performance data and manufacturing standard data from the manufacturing management system and production process control system to determine the characteristics and time standards of different steel grades;

[0010] 2) Extract and import the required furnace schedule from the manufacturing management system;

[0011] 3) Set automatic scheduling constraints for production plans based on process rules and downtime maintenance plans;

[0012] 4) Based on the required furnace schedule and constraints, determine the optimal production schedule ranking result using a genetic algorithm model;

[0013] 5) Based on the steel grade plan and corresponding time standards, take the process with the longest processing cycle as the center, determine the production timetable for each process in sequence, and display the expected output, time consumption information and risk warning information within the planning period.

[0014] The automatic optimization scheduling method for electric furnace production lines described in this invention, based on the production plan issued by the manufacturing management system, and under the constraints of various process rules and downtime maintenance plans, utilizes a scheduling optimization model to calculate and assist in the automatic completion of optimized production scheduling. It also automatically calculates planned output, planned time consumption, and warnings of planned risk points, providing a convenient and efficient auxiliary tool for scheduling work and improving the efficiency, accuracy, and intelligence level of production planning scheduling for electric furnace production lines.

[0015] Specifically, in step 1), production performance data and manufacturing standard data are stored in a relational database with the heat number as the primary index field. Production performance data includes the steel grade designation for each heat and the start and end times of each processing step. Time standards include reference values ​​for the processing cycle of each steel grade in each process, inter-process transfer time, and preparation time for each process. The processing cycle of each process is recorded in the database as the difference between the end and start times of each process for each heat. The inter-process transfer time is recorded in the database as the difference between the start time of the next process and the end time of the previous process for each heat. Preparation time is recorded in the database and is the difference between the start time of each process and the end time of the previous process for each heat of steel. The reference values ​​for the processing cycle of each process, the inter-process transfer time, and the preparation time of each process are determined by the moving average of the most recent n valid samples stored in the database. Valid samples are heat samples after data cleaning. The data cleaning process marks heat records with processing cycles, inter-process transfer times, and preparation times within a reasonable range as valid samples. Manufacturing standard data includes the characteristics of each steel grade, including the standard composition, process characteristics, cross-sectional specifications, and standard casting speed corresponding to each steel grade mark.

[0016] Specifically, in step 2), the required furnace schedule consists of the number of furnaces to be added to the automatic scheduling, the schedule number, the steel grade mark, and the planned continuous casting section specifications.

[0017] The extraction of the required furnace run plan includes the following steps:

[0018] 2.1) After selecting plan import, the plan for N furnaces to be produced will be automatically extracted from the production plan pool;

[0019] 2.2) Automatically determine whether the N+1 heat is a different steel type from the Nth heat. If it is a different steel type, the extraction of the heat schedule plan for the Nth heat will end. If it is the same steel type, the heat will be added to the heat schedule plan and the system will continue to determine whether the steel type of the next heat is a different steel type until a different steel type is encountered and the extraction will end.

[0020] 2.3) Import and save the extracted furnace schedule plans to the database.

[0021] Specifically, in step 3), the automatic scheduling constraints of the production plan include at least the constraint rules for the round billet production line and the constraint rules for the large billet production line.

[0022] The constraint rules for the round and square billet production line include at least the following:

[0023] Based on the difficulty of the steel grades, priority should be given to arranging low-speed steel grades for furnace grouping and scheduling;

[0024] The furnace grate should be prioritized according to the decreasing carbon content of the steel grades;

[0025] Furnace scheduling is based on prioritizing single and double furnace steel grades according to the first and last furnaces.

[0026] The constraint rules for the large billet production line include at least the following:

[0027] The grate should be prioritized according to the decreasing order of alloy element content in high alloy steel grades;

[0028] The furnace grate should be prioritized according to the decreasing boron and titanium content of the steel grades.

[0029] Specifically, in step 4), determining the optimal production plan ranking result based on the genetic algorithm model includes the following steps:

[0030] 4.1) Randomly sort the required furnace schedules to generate a first-generation chromosome population;

[0031] 4.2) Calculate the penalty value based on the scheduling constraints and fitness function;

[0032] 4.3) Select, crossover, and mutate chromosomes to form a new generation of chromosome population;

[0033] 4.4) If the number of iterations has not reached the set value, continue iterating to step 4.1); if the number of iterations reaches the set value, stop iterating and select the plan sort corresponding to the chromosome with the smallest penalty value as the final plan sort.

[0034] Furthermore, in step 4.2), the fitness function F = C max -Z, where C max Z is a sufficiently large constant, and Z is the objective function established based on the scheduling constraints.

[0035] The objective function is to set a smaller penalty value for sorting plans that meet the scheduling constraints and a larger penalty value for sorting plans that do not meet the constraints.

[0036] The objective function model is as follows:

[0037]

[0038] Where Z is the objective function for scheduling optimization; minZ represents the function calculation result that takes the minimum penalty value as the final selected sorting result; N is the number of process rule constraints; M is the number of shutdown maintenance plan constraints; t is the start time of the shutdown maintenance plan event obtained from the model scheduling calculation; T j The start time of the shutdown and maintenance plan event pre-set for production scheduling; F j D is the constraint factor for the shutdown and maintenance plan. i P represents the constraint coefficient for process rules; i This is a penalty value function established based on process rules and conditions.

[0039] Specifically, in step 5), the risk warning in the production plan, used to assess the risks affecting continuous casting, specifically includes:

[0040] Determine if the start time of the next batch in each process is earlier than the end time of the previous batch. If so, provide an alarm message: "Cannot keep up with continuous casting."

[0041] Determine whether the start time of the next continuous casting heat is delayed compared to the end time of the previous heat. If the delay is too long, issue a continuous casting ladle change timeout alarm.

[0042] Production staff can make timely interventions and adjustments to the process rhythm based on alarm prompts and working conditions, thereby reducing the risk of production losses caused by process time conflicts.

[0043] The technical solution of the present invention also provides an automatic optimization scheduling system for electric furnace production lines, characterized by:

[0044] The scheduling data acquisition module, scheduling analysis and optimization module, and scheduling information display module are connected in sequence.

[0045] The scheduling data acquisition module is connected to the manufacturing management system and is used to collect data related to the production plan scheduling of the electric furnace production line and send it to the scheduling analysis and optimization module.

[0046] The scheduling data acquisition module is connected to the production control system and is used to collect process standard data related to the scheduling of the electric furnace production line.

[0047] The scheduling data acquisition module is connected to the process control system and is used to collect production performance data related to production scheduling.

[0048] The scheduling analysis and optimization module is connected to the scheduling data acquisition module;

[0049] The scheduling analysis and optimization module includes: a scheduling data storage module, a scheduling data governance module, a scheduling constraint configuration module, and a scheduling calculation and analysis module;

[0050] The scheduling information display module is used to display the optimized scheduling results on the operator terminal, so that production schedulers can organize production and make timely adjustments, making the production rhythm more reasonable and smooth.

[0051] The aforementioned electric arc furnace production line automatic optimization scheduling system, based on the daily production plan pools of the two production lines of the electric arc furnace plant issued by the manufacturing system, combined with the actual performance data of different steel grades and production scheduling related processes and expert knowledge, uses a genetic algorithm to assist production planning and scheduling personnel in intelligently optimizing the production scheduling of steel grades in the original plan pool, and providing the optimal furnace assembly sequence and scheduling time arrangement.

[0052] Specifically, the scheduling data acquisition module is a process control server connected to the production information network.

[0053] Specifically, the scheduling analysis and optimization module uses a workshop edge server, which has a central processing unit, memory, hard disk storage and network interface, and is connected to the basic automation and production information network via Ethernet.

[0054] Specifically, the scheduling data storage module includes a relational database for storing production plan data indexed by time, production performance data indexed by furnace number, and manufacturing standard data indexed by steel grade mark sent by the data acquisition module.

[0055] The scheduling data storage module stores the optimized scheduling analysis results returned after calculation and analysis by the scheduling calculation and analysis module;

[0056] The scheduling data governance module is used to clean, transform, and process the production scheduling data for each heat of steel collected by the data acquisition module to ensure the quality and consistency of the data and provide complete and accurate data for the scheduling calculation and analysis module.

[0057] The scheduling constraint configuration module includes a process rule condition constraint configuration module and a shutdown and maintenance plan constraint configuration module, which are used to provide calculation constraints for the scheduling calculation and analysis module.

[0058] The scheduling calculation and analysis module is used to perform calculations using a scheduling optimization model based on the production plan data and manufacturing standard data in the scheduling data storage module, the production performance data after being processed by the scheduling data management module, and the constraints determined by the scheduling constraint configuration module. The scheduling results are then returned to the scheduling data storage module and output to the scheduling information display module for display.

[0059] Furthermore, the process rule condition constraint configuration module consists of various rule items formulated by process experts and production schedulers based on their experience and knowledge, including rule number, rule name, and rule priority, which are used to determine the sequential connection constraint relationship between different steel grades;

[0060] The shutdown and maintenance plan constraint configuration module is used to set the time constraints for electric furnace shutdown and startup, including the exhaust plan, startup plan, and pit cleaning plan within the planning period.

[0061] Furthermore, the scheduling results include the optimized production plan arrangement results for different steel grades within the planning period, production plan risk point warnings, and information on the number of furnaces produced and the time consumed within the planning period.

[0062] The automatic production planning optimization scheduling system for electric furnace production lines described in this invention, under the constraints of process experience rules and downtime maintenance plans, uses a scheduling optimization model to calculate and assist in the automatic completion of production planning optimization. It also automatically calculates planned output, planned time consumption, and warnings of planned risk points, providing a convenient and quick auxiliary tool for scheduling work and improving the efficiency, accuracy, and intelligence level of production planning scheduling for electric furnace production lines.

[0063] Compared with the prior art, the advantages of the present invention are:

[0064] 1. The technical solution of the present invention addresses the problem that existing production plans rely entirely on manual scheduling and adjustment, which is labor-intensive and prone to errors. It provides an automatic optimization scheduling method and system for electric furnace production lines.

[0065] 2. The technical solution of this invention adopts a hybrid model algorithm that takes into account both process rule constraints and process time constraints. Therefore, it can better meet the planning and scheduling requirements of electric arc furnace steelmaking for multiple steel types and small batches while eliminating product quality risks and process time conflicts in electric arc furnace production lines. It helps production scheduling to automatically complete the multi-steel type optimization sorting and multi-objective optimization scheduling of production plans, and automatically provides planned output, production time and risk warning information, thus providing a convenient and fast intelligent auxiliary tool for production scheduling.

[0066] 3. The technical solution of the present invention, under the constraints of process experience rules and shutdown maintenance plans, uses a scheduling optimization model to calculate and assist in the automatic completion of production plan optimization scheduling, and automatically calculates the planned output, planned time consumption and planned risk point warnings, thereby improving the scheduling efficiency, accuracy and intelligence level of electric furnace production line production plan. Attached Figure Description

[0067] Figure 1 This is a flowchart of the automatic optimization scheduling method for production planning in electric furnace production lines;

[0068] Figure 2 It is a flowchart of the required furnace schedule extracted from the manufacturing management system;

[0069] Figure 3 It is a flowchart for determining the optimal production plan based on a genetic algorithm;

[0070] Figure 4 This is a block diagram of the automatic production planning and scheduling system for this electric furnace production line. Detailed Implementation

[0071] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0072] This invention provides an automatic optimization scheduling method and system for electric furnace production lines. Based on the production plan issued by the manufacturing management system, and under the constraints of various process rules and downtime maintenance plans, it uses a scheduling optimization model to calculate and assist in the automatic completion of production plan optimization scheduling. It also automatically calculates the planned output, planned time consumption, and planned risk point warnings, thus providing a convenient and quick auxiliary tool for scheduling work and improving the efficiency, accuracy, and intelligence level of production planning scheduling for electric furnace production lines.

[0073] The technical solution of this invention is applied to the automatic scheduling of two production lines: electric furnace round billet (old line) and large billet (new line).

[0074] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0075] Firstly, such as Figure 1As shown, the technical solution of the present invention provides a method for automatically optimizing and scheduling production plans for electric furnace production lines, comprising the following steps:

[0076] 1) Collect production performance data and manufacturing standard data from the manufacturing management system and production process control system to determine the characteristics and time standards of different steel grades;

[0077] The production performance data and manufacturing standard data are stored in a relational database with furnace number as the primary index field.

[0078] The production performance data includes the steel grade designation for each heat of steel and the start and end times of each process.

[0079] The time standard includes reference values ​​for the processing cycle of each steel grade in each process, the transfer time between processes, and the preparation time of each process;

[0080] The processing cycle of each process is recorded in the database and is the difference between the end time and the start time of each process for each heat of steel.

[0081] The inter-process transfer time is recorded in the database and is the difference between the start time of the next process and the end time of the previous process for each heat of steel.

[0082] The preparation time for each process is recorded in the database and is the difference between the start time of each process and the end time of the previous process for each heat of steel.

[0083] The reference values ​​for the processing cycle of each process, the inter-process transfer time, and the preparation time of each process are determined by the moving average of the most recent n valid samples stored in the database.

[0084] The valid samples are furnace samples after data cleaning; the data cleaning process involves marking furnace records with processing cycles, inter-process transfer times, and preparation times for each process within a reasonable range as valid samples.

[0085] The manufacturing standard data includes the characteristics of each steel grade, including the standard composition, process characteristics, cross-sectional specifications and standard drawing speed corresponding to each steel grade mark;

[0086] 2) Extract and import the required furnace schedule from the manufacturing management system;

[0087] The required furnace schedule consists of the number of furnaces to be included in the automatic scheduling, the schedule number, the steel grade mark, and the planned continuous casting section specifications.

[0088] The number of furnace cycles can be entered and determined on the system settings screen;

[0089] like Figure 2As shown, the extraction of the required furnace schedule includes the following steps:

[0090] 2.1) After selecting plan import, the plan for N furnaces to be produced will be automatically extracted from the production plan pool;

[0091] 2.2) Automatically determine whether the (n+1)th heat is a different type of steel than the Nth heat. If it is a different type of steel, the extraction of the heat schedule plan for the nth heat will end. If it is the same type of steel, the heat will be added to the heat schedule plan and the system will continue to determine whether the steel type of the next heat is a different type of steel until a different type of steel is encountered and the extraction ends.

[0092] 2.3) Import and save the extracted furnace schedule plans to the database;

[0093] 3) Set automatic scheduling constraints for production plans based on process rules and downtime maintenance plans;

[0094] The process rules and conditions are the basis for determining the order of connection between different steel grades, including rule number, rule name and rule priority, which are defined and selected by process experts and production schedulers based on electric furnace production process requirements and production organization experience.

[0095] The shutdown and maintenance plan is the basis for scheduling the shutdown and startup times of the electric furnace, including: exhaustion plan, startup plan, and pit cleaning plan;

[0096] The exhaustion plan includes the exhaustion plan time and the selection of exhaustion steel grades for the electric furnace process; the exhaustion plan time is the end time of the electric furnace plan processing for the last furnace before the electric furnace production is shut down, and is used to arrange the production schedule before the exhaustion shutdown.

[0097] The furnace start-up plan includes the start-up time of the electric furnace process and the selection of the steel grade to be started; the start-up time is the start time of the electric furnace plan processing for the first furnace after the electric furnace production is interrupted, and is used to generate a new continuous scheduling plan from this point.

[0098] The pit cleaning plan includes the pit cleaning plan time for the VD process and the selection of the pit cleaning steel grade; the pit cleaning plan time is the period during which the VD tank pit is cleaned, during which VD process steel grades are not scheduled for production, and the production schedule for non-VD process steel grades is arranged accordingly.

[0099] 4) Based on the required furnace schedule and constraints, determine the optimal production schedule ranking result using a genetic algorithm model;

[0100] like Figure 3 As shown, the process of determining the optimal production plan ranking result based on the genetic algorithm model includes the following steps:

[0101] 4.1) Randomly sort the required furnace schedules to generate a first-generation chromosome population;

[0102] 4.2) Calculate the penalty value based on the scheduling constraints and fitness function;

[0103] The fitness function F = C max -Z, where C max Z is a sufficiently large constant, and Z is the objective function established based on the scheduling constraints.

[0104] The objective function is to set a smaller penalty value for sorting plans that meet the scheduling constraints and a larger penalty value for sorting plans that do not meet the constraints.

[0105] The objective function model is as follows:

[0106]

[0107] Where Z is the objective function for scheduling optimization;

[0108] minZ represents the function calculation result that takes the smallest penalty value as the final selected sorting result;

[0109] N represents the number of process rule constraints;

[0110] M represents the number of constraints in the boiler shutdown and maintenance plan;

[0111] t represents the start time of the shutdown maintenance plan event calculated by the model scheduling; for the pit cleaning plan, t is the pit cleaning start time, which is the process completion time of the last furnace of the VD process; for the exhaustion plan, t is the exhaustion start time, which is the process completion time of the last furnace of the EAF process; for the furnace start plan, t is the furnace start time, which is the process completion time of the first furnace of the EAF process.

[0112] T j The start time of the pre-set shutdown and maintenance plan event for production scheduling;

[0113] F j This is the constraint coefficient for the shutdown and maintenance plan. Its purpose is to minimize the difference between the set value and the calculated value of the shutdown and maintenance plan time during the iteration process. The larger the value, the more important the constraint is, that is, the higher the priority level.

[0114] D i D represents the constraint coefficient for process rules. The higher the priority of the process rule conditions, the higher the coefficient. i The larger the coefficient, the more likely the unselected process rule condition D is to be affected. i The coefficient is 0;

[0115] Where F j and D i The coefficients are set using an exponential function with a base of 10, and the function range is 10.2 Up to 10 2*(M+N) Rules with higher priority are assigned larger coefficients;

[0116] P i The penalty value function is established based on the process rule conditions. The more the process rule conditions are satisfied, the smaller the penalty value. Different penalty value functions can be established according to different process rule conditions.

[0117] 4.3) Select, crossover, and mutate chromosomes to form a new generation of chromosome population;

[0118] 4.4) If the number of iterations has not reached the set value, continue iterating to step 4.1); if the number of iterations reaches the set value, stop iterating and select the plan sort corresponding to the chromosome with the smallest penalty value as the final plan sort.

[0119] 5) Based on the steel grade plan and corresponding time standards, take the process with the longest processing cycle as the center, determine the production timetable for each process in sequence, and display the expected output, time consumption information and risk warning information within the planning period.

[0120] The risk warnings in the production plan are used to assess the risks affecting continuous casting and pouring, including:

[0121] Determine if the start time of the next batch in each process is earlier than the end time of the previous batch. If so, provide an alarm message: "Cannot keep up with continuous casting."

[0122] Determine whether the start time of the next continuous casting heat is delayed compared to the end time of the previous heat. If the delay is too long, issue a continuous casting ladle change timeout alarm.

[0123] Production staff can make timely interventions and adjustments to the process rhythm based on alarm prompts and operating conditions, thereby reducing the risk of production losses caused by process time conflicts.

[0124] This embodiment provides the penalty value function calculation model for the typical furnace scheduling constraint rules of electric furnace production lines as follows:

[0125] A. Constraint rule model for round / square billet production line:

[0126] Rule 1: Prioritize the scheduling of furnace batches for steel grades with lower casting speeds based on their difficulty level.

[0127] This rule prioritizes the production of low-speed steel grades before difficult steel grades, and also prioritizes the centralized production of difficult steel grades. This is beneficial for difficult steel grades with longer refining cycles to keep up with the continuous casting process, reducing the risk of casting interruption. The penalty value algorithm for this rule is as follows:

[0128]

[0129] in,

[0130] N----Number of furnace runs required for scheduling;

[0131] P1----Penalty value for constraint rule 1;

[0132] x i ----Is the steel grade (tap mark) of the i-th furnace a difficult-to-finish steel grade? If so, then x i =1, if not, then x i =0;

[0133] y i-1 ----Is the steel grade (tap mark) of the i-1th heat a low-speed steel grade? If so, then y i-1 =1, if not, then y i-1 =0;

[0134] w1----is the coefficient for the continuous arrangement of difficult steel grades; in this embodiment, w1 is taken as -2.

[0135] w2----is the coefficient for the continuous arrangement of difficult steel grades and low-speed steel grades. In this embodiment, w2 is taken as -1.

[0136] Rule 2: Prioritize furnace scheduling based on the decreasing carbon (C) content of the steel grades:

[0137] This means that furnace scheduling is prioritized based on the carbon (C) content of the steel grades, from high carbon to low carbon. This is mainly used to gradually transition from producing high carbon steel to medium and low carbon steel, reducing the risk of subsequent steel grades exceeding the carbon content limit. The penalty value algorithm for this rule is as follows:

[0138] P2=1-cos(x,y) (3)

[0139]

[0140] in:

[0141] P2----is the penalty value for constraint rule 2;

[0142] N----Number of furnace runs required for scheduling;

[0143] x---- is the n-dimensional coordinate formed by the current furnace group sequence;

[0144] x i ---- represents the i-th dimension coordinate of the current furnace group sequence;

[0145] y----an n-dimensional coordinate system formed by the furnace grouping sequence with decreasing carbon (C) content;

[0146] y i----The i-th dimension of the furnace grouping sequence with decreasing carbon (C) content

[0147] The above formula treats the furnace grouping sequence as a spatial vector in N-dimensional space. By calculating the cosine value of the two spatial vectors, the closer the cosine value is to 1, the more similar the two furnace grouping sequences are. That is, the current furnace grouping sequence in the iteration process is more in line with the requirement of decreasing carbon content.

[0148] After numerical transformation, the penalty value is smaller when the furnace grouping sequence conforms to the decreasing order of carbon content.

[0149] Rule 3: The furnace scheduling should prioritize single and double furnace steel grades based on the first and last furnaces.

[0150] This rule prioritizes steel grades suitable for single-furnace or double-furnace production for the first or last furnace in the furnace grouping plan. This reduces the probability of single or double-furnace steel grades appearing in the middle of the scheduling, and lowers the risk of difficulty in continuous casting. The penalty value algorithm for this rule is as follows:

[0151] P3 = -X1 - X N (5)

[0152] in:

[0153] P3----is the penalty value for constraint rule 3;

[0154] N----Number of furnace runs required for scheduling;

[0155] X1----Whether the first heat belongs to a single or double heat steel grade. If yes, then X1 = 1; otherwise, then X1 = 0.

[0156] X N ----Is the Nth heat a single or double-heat steel grade? If so, then X N =1, if not, then X N =0;

[0157] The above formula indicates that if the first or Nth heat of steel is a single or double heat cast steel (the same steel is only cast 1 or 2 times in the plan), the penalty value is reduced.

[0158] B. Constraint rule model for large billet production line:

[0159] Rule 1: Prioritize furnace scheduling based on the decreasing alloy element content of high-alloy steel grades.

[0160] The penalty value algorithm of this rule is similar to the algorithm of Rule 2 (decreasing carbon content) for round and square billet production lines. It represents the alloy element content of steel grades according to the total content of chromium, nickel and molybdenum. It is mainly used for the gradual transition after the production of high alloy steel to reduce the risk of excessive alloy elements in subsequent steel production. The more the furnace grouping sequence meets the decreasing order of high alloy element content, the smaller the penalty value.

[0161] Rule 2: Prioritize furnace grouping and scheduling based on the decreasing boron (B) and titanium (Ti) content of the steel grades.

[0162] The penalty value algorithm for this rule is similar to the carbon content decreasing algorithm. The more the furnace grouping sequence satisfies the decreasing order of boron (B) and titanium (Ti) content, the smaller the penalty value.

[0163] The above typical scheduling constraint rule mathematical models are the core of the model calculation module in this embodiment.

[0164] Secondly, such as Figure 4 As shown, this application provides an automatic optimization scheduling system for electric furnace production lines, including: a scheduling data acquisition module 1, a scheduling analysis and optimization module 2, and a scheduling information display module 3.

[0165] The scheduling data acquisition module 1 is used to collect data information related to the production plan scheduling of the electric furnace production line and send it to the scheduling analysis and optimization module 2.

[0166] The scheduling data acquisition module 1 is connected to the manufacturing management system 4 and is used to collect the daily production plans of the two production lines issued by the manufacturing management system 4, that is, the daily plan pool to be scheduled, including information such as PO number, steel tapping mark, cross section, and status.

[0167] The scheduling data acquisition module 1 is connected to the production control system 5 and is used to collect process standard data related to the scheduling of the electric furnace production line, including electric furnace standard feeding settings, VOD calming time, and casting speed settings.

[0168] The scheduling data acquisition module 1 is connected to the process control system 6 and is used to collect production performance data related to production scheduling, including the start and end times of each heat in the EAF, LF, VD (VOD), and CC processes, as well as the tapping mark, standard casting speed corresponding to the cross-sectional specifications, standard composition, etc., and to calculate the processing cycle and transfer time of each heat; historical data related to the production plan, including time series data of indicators such as electric furnace production heats, process processing time, production efficiency, equipment failure rate, and equipment maintenance time; and steel grade process parameters, namely, production information such as the composition, width, casting speed, whether it is a difficult steel grade, and whether it is a lightly pressed steel grade in the two production lines of large square billet and round square billet.

[0169] In this example, the scheduling data acquisition module 1 is a process control server connected to the production information network.

[0170] like Figure 4As shown, the scheduling analysis and optimization module 2 is connected to the scheduling data acquisition module 1 and includes: a scheduling data storage module 21, a scheduling data management module 22, a scheduling constraint configuration module 23, and a scheduling calculation and analysis module 24.

[0171] In this embodiment, the scheduling analysis and optimization module 2 adopts a workshop edge server, which has a central processing unit, memory, hard disk storage and network interface, and is connected to the basic automation and production information network via Ethernet;

[0172] The scheduling data storage module 21 includes a relational database for storing production plan data indexed by time, production performance data indexed by furnace number, and manufacturing standard data indexed by steel grade mark sent by the data acquisition module 1.

[0173] The scheduling data storage module 21 stores the optimized scheduling analysis results returned after calculation and analysis by the scheduling calculation and analysis module 24, including: the production plan sequence of the two production lines of the electric furnace on the same day, displaying the start and end times of each furnace and each process in the form of a train timetable, and providing risk warnings for the time consumption and risk points of each process in the plan.

[0174] The scheduling data management module 22 is used to clean, transform and process the production scheduling data of each heat of steel collected by the data acquisition module to ensure the quality and consistency of the data and provide complete and accurate data for the scheduling calculation and analysis module 24.

[0175] The scheduling constraint configuration module 23 includes a process rule condition constraint configuration module 231 and a shutdown and maintenance plan constraint configuration module 232, which are used to provide calculation constraints for the scheduling calculation and analysis module 24.

[0176] The process rule condition constraint configuration module 231 consists of various rule items formulated by process experts and production schedulers based on their experience and knowledge, including rule number, rule name, and rule priority, which are used to determine the sequential connection constraint relationship between different steel grades;

[0177] The shutdown and maintenance plan constraint configuration module 232 is used to set the time constraints for the shutdown and startup of the electric furnace, including the exhaust plan, startup plan and pit cleaning plan within the planning period.

[0178] The scheduling calculation and analysis module 24 is used to perform calculations using a scheduling optimization model based on the production plan data and manufacturing standard data in the scheduling data storage module 21, the production performance data after processing by the scheduling data management module 22, and the constraints determined by the scheduling constraint configuration module 23. The scheduling results are then returned to the scheduling data storage module 21 and output to the scheduling information display module 3 for display.

[0179] The scheduling results include the optimized production plan arrangement for different steel grades within the planning period, warnings of production plan risk points, and information on the number of furnaces produced and the time consumed within the planning period.

[0180] The scheduling information display module 3 is used to display the optimized scheduling results on the operator terminal, so that production schedulers can organize production and make timely adjustments, making the production rhythm more reasonable and smooth.

[0181] This embodiment uses a genetic algorithm to assist production planning and scheduling personnel in intelligently optimizing the production schedule of steel grades in the original plan pool based on the daily production plan pool of the two production lines of the electric furnace plant issued by the manufacturing system, combined with the actual performance data of different steel grades and production scheduling related processes and expert knowledge. The algorithm provides the optimal furnace grouping sequence and scheduling time arrangement.

[0182] The technical solution of this invention, by adopting a hybrid model algorithm that takes into account both process rule constraints and process time constraints, can eliminate product quality risks and process time conflicts in electric arc furnace production lines. It utilizes a scheduling optimization model to assist in the automatic completion of optimized production planning and automatically calculates planned output, planned time consumption, and planned risk point warnings. This better meets the planning and scheduling requirements of electric arc furnace steelmaking for multiple steel types and small batches, and helps production scheduling automatically complete multi-steel type optimization sorting and multi-objective optimization scheduling of production plans. It provides a convenient and quick auxiliary tool for scheduling work, improving the efficiency, accuracy, and intelligence level of production planning and scheduling in electric arc furnace production lines.

[0183] This invention can be widely used in the formulation and management of production plans for electric furnace production lines.

Claims

1. A method for automatically optimizing and scheduling production plans for electric furnace production lines, characterized by: 1) Collect production performance data and manufacturing standard data from the manufacturing management system and production process control system to determine the characteristics and time standards of different steel grades; 2) Extract and import the required furnace schedule from the manufacturing management system; 3) Set automatic scheduling constraints for production plans based on process rules and downtime maintenance plans; 4) Based on the required furnace schedule and constraints, determine the optimal production schedule ranking result using a genetic algorithm model; 5) Based on the steel grade plan and corresponding time standards, take the process with the longest processing cycle as the center, determine the production timetable for each process in sequence, and display the expected output, time consumption information and risk warning information within the planning period.

2. The automatic optimization scheduling method for electric furnace production line production planning according to claim 1, characterized in that: The aforementioned automatic production planning optimization scheduling method for electric furnace production lines, based on the production plan issued by the manufacturing management system, and under the constraints of various process rules and downtime maintenance plans, utilizes a scheduling optimization model to calculate and assist in the automatic completion of production plan optimization scheduling. It also automatically calculates planned output, planned time consumption, and planned risk point warnings, providing a convenient and efficient auxiliary tool for scheduling work and improving the efficiency, accuracy, and intelligence level of production planning scheduling for electric furnace production lines.

3. The automatic optimization scheduling method for electric furnace production line production planning according to claim 1, characterized in that... In step 1), production performance data and manufacturing standard data are stored in a relational database with furnace number as the primary index field. Production performance data includes the steel grade designation for each heat of steel and the start and end times of each processing step; The time standard includes reference values ​​for the processing cycle of each steel grade in each process, the transfer time between processes, and the preparation time of each process; The processing cycle of each process is recorded in the database, which is the difference between the end time and the start time of each process for each heat of steel. The inter-process transfer time is recorded in the database and is the difference between the start time of the next process and the end time of the previous process for each heat of steel. The preparation time for each process is recorded in the database, which is the difference between the start time of each process and the end time of the previous process for each heat of steel. The reference values ​​for the processing cycle of each process, the inter-process transfer time, and the preparation time of each process are determined by the moving average of the most recent n valid samples stored in the database. Valid samples are furnace samples after data cleaning; the data cleaning process is to mark furnace records with processing cycles, inter-process transfer times and preparation times of each process within a reasonable range as valid samples. Manufacturing standard data includes the characteristics of each steel grade, including the standard composition, process characteristics, cross-sectional specifications, and standard drawing speed corresponding to each steel grade designation.

4. The automatic optimization scheduling method for electric furnace production line production planning according to claim 1, characterized in that... In step 2), the required furnace schedule consists of the number of furnaces to be added to the automatic scheduling, the schedule number, the steel grade mark, and the planned continuous casting section specifications. The extraction of the required furnace run plan includes the following steps: 2.1) After selecting plan import, the plan for N furnaces to be produced will be automatically extracted from the production plan pool; 2.2) Automatically determine whether the (n+1)th heat is a different type of steel than the nth heat. If it is a different type of steel, the extraction of the heat schedule plan for the nth heat will end. If it is the same type of steel, the heat will be added to the heat schedule plan and the system will continue to determine whether the steel type of the next heat is a different type of steel until a different type of steel is encountered and the extraction ends. 2.3) Import and save the extracted furnace schedule plans to the database.

5. The automatic optimization scheduling method for electric furnace production line production planning according to claim 1, characterized in that... In step 3), the automatic scheduling constraints of the production plan include at least the constraint rules for the round billet production line and the constraint rules for the large billet production line. The constraint rules for the round and square billet production line include at least the following: Based on the difficulty of the steel grades, priority should be given to arranging low-speed steel grades for furnace grouping and scheduling; The furnace grate should be prioritized according to the decreasing carbon content of the steel grades; Furnace scheduling is based on prioritizing single and double furnace steel grades according to the first and last furnaces. The constraint rules for the large billet production line include at least the following: The grate should be prioritized according to the decreasing order of alloy element content in high alloy steel grades; The furnace grate should be prioritized according to the decreasing boron and titanium content of the steel grades.

6. The automatic optimization scheduling method for electric furnace production line production planning according to claim 1, characterized in that... Step 4) involves determining the optimal production plan ranking result based on the genetic algorithm model, which includes the following steps: 4.1) Randomly sort the required furnace schedules to generate a first-generation chromosome population; 4.2) Calculate the penalty value based on the scheduling constraints and fitness function; 4.3) Select, crossover, and mutate chromosomes to form a new generation of chromosome population; 4.4) If the number of iterations has not reached the set value, continue iterating to step 4.1); if the number of iterations reaches the set value, stop iterating and select the plan sort corresponding to the chromosome with the smallest penalty value as the final plan sort.

7. The automatic optimization scheduling method for electric furnace production line production planning according to claim 6, characterized in that... In step 4.2), the fitness function F = C max -Z, where C max Z is a sufficiently large constant, and Z is the objective function established based on the scheduling constraints. The objective function is to set a smaller penalty value for sorting plans that meet the scheduling constraints and a larger penalty value for sorting plans that do not meet the constraints. The objective function model is as follows: Where Z is the objective function for scheduling optimization; minZ represents the function calculation result that takes the minimum penalty value as the final selected sorting result; N is the number of process rule constraints; M is the number of shutdown maintenance plan constraints; t is the start time of the shutdown maintenance plan event obtained from the model scheduling calculation; T j The start time of the shutdown and maintenance plan event pre-set for production scheduling; F j D is the constraint factor for the shutdown and maintenance plan. i P represents the constraint coefficient for process rules; i This is a penalty value function established based on process rules and conditions.

8. The automatic optimization scheduling method for electric furnace production line production planning according to claim 1, characterized in that... In step 5), the risk warning in the production plan is used to determine the risks affecting continuous casting and pouring, specifically including: Determine if the start time of the next batch in each process is earlier than the end time of the previous batch. If so, provide an alarm message: "Cannot keep up with continuous casting." Determine whether the start time of the next continuous casting heat is delayed compared to the end time of the previous heat. If the delay is too long, issue a continuous casting ladle change timeout alarm. Production staff can make timely interventions and adjustments to the process rhythm based on alarm prompts and working conditions, thereby reducing the risk of production losses caused by process time conflicts.

9. An automatic production planning optimization scheduling system for an electric furnace production line, characterized by: The scheduling data acquisition module, scheduling analysis and optimization module, and scheduling information display module are connected in sequence. The scheduling data acquisition module is connected to the manufacturing management system and is used to collect data related to the production plan scheduling of the electric furnace production line and send it to the scheduling analysis and optimization module. The scheduling data acquisition module is connected to the production control system and is used to collect process standard data related to the scheduling of the electric furnace production line. The scheduling data acquisition module is connected to the process control system and is used to collect production performance data related to production scheduling. The scheduling analysis and optimization module is connected to the scheduling data acquisition module; The scheduling analysis and optimization module includes: a scheduling data storage module, a scheduling data governance module, a scheduling constraint configuration module, and a scheduling calculation and analysis module; The scheduling information display module is used to display the optimized scheduling results on the operator terminal, so that production schedulers can organize production and make timely adjustments, making the production rhythm more reasonable and smooth. The aforementioned electric arc furnace production line automatic optimization scheduling system, based on the daily production plan pools of the two production lines of the electric arc furnace plant issued by the manufacturing system, combined with the actual performance data of different steel grades and production scheduling related processes and expert knowledge, uses a genetic algorithm to assist production planning and scheduling personnel in intelligently optimizing the production scheduling of steel grades in the original plan pool, and providing the optimal furnace assembly sequence and scheduling time arrangement.

10. The automatic optimization scheduling system for electric furnace production lines according to claim 9, characterized in that: The scheduling data acquisition module is a process control server connected to the production information network.

11. The automatic optimization scheduling system for electric furnace production lines according to claim 9, characterized in that: The scheduling analysis and optimization module uses a workshop edge server, which has a central processing unit, memory, hard disk storage and network interface, and is connected to the basic automation and production information network via Ethernet.

12. The automatic optimization scheduling system for electric furnace production lines according to claim 9, characterized in that: The scheduling data storage module includes a relational database for storing production plan data indexed by time, production performance data indexed by furnace number, and manufacturing standard data indexed by steel grade mark sent by the data acquisition module. The scheduling data storage module stores the optimized scheduling analysis results returned after calculation and analysis by the scheduling calculation and analysis module; The scheduling data governance module is used to clean, transform, and process the production scheduling data for each heat of steel collected by the data acquisition module to ensure the quality and consistency of the data and provide complete and accurate data for the scheduling calculation and analysis module. The scheduling constraint configuration module includes a process rule condition constraint configuration module and a shutdown and maintenance plan constraint configuration module, which are used to provide calculation constraints for the scheduling calculation and analysis module. The scheduling calculation and analysis module is used to perform calculations using a scheduling optimization model based on the production plan data and manufacturing standard data in the scheduling data storage module, the production performance data after being processed by the scheduling data management module, and the constraints determined by the scheduling constraint configuration module. The scheduling results are then returned to the scheduling data storage module and output to the scheduling information display module for display.

13. The automatic optimization scheduling system for electric furnace production lines according to claim 12, characterized in that: The process rule condition constraint configuration module consists of various rule items formulated by process experts and production schedulers based on their experience and knowledge, including rule number, rule name, and rule priority, which are used to determine the sequential connection constraint relationship between different steel grades; The shutdown and maintenance plan constraint configuration module is used to set the time constraints for electric furnace shutdown and startup, including the exhaust plan, startup plan, and pit cleaning plan within the planning period.

14. The automatic optimization scheduling system for electric furnace production lines according to claim 12, characterized in that: The scheduling results include the optimized production plan arrangement for different steel grades within the planning period, warnings of production plan risk points, and information on the number of furnaces produced and the time consumed within the planning period.

15. The automatic optimization scheduling system for electric furnace production lines according to claim 9, characterized in that: The aforementioned automatic production planning optimization scheduling system for electric furnace production lines, under the constraints of process experience rules and downtime maintenance plans, uses a scheduling optimization model to calculate and assist in the automatic completion of production planning optimization. It also automatically calculates planned output, planned time consumption, and warnings of planned risk points, providing a convenient and quick auxiliary tool for scheduling work and improving the efficiency, accuracy, and intelligence level of production planning scheduling for electric furnace production lines.

Images