Intelligent scheduling decision method and system for high time sensitivity specific steel grade

Abstract

The application provides a kind of intelligent scheduling decision method and system for high time sensitivity specific steel grade, comprising: step S1: obtaining steelmaking production line production data, including: steelmaking production line equipment real-time state, molten steel parameters, each link processing and waiting time, equipment instantaneous capacity and current overall production progress;Step S2: according to the process constraints of target product and production target, a scheduling decision model is constructed;Step S3: according to the constructed scheduling decision model, based on the obtained steelmaking production line production data, an initial scheduling scheme is generated using the greedy algorithm;Step S4: and under the condition of meeting the preset process constraints, the initial scheduling scheme is fine-tuned according to the production target, to obtain the fine-tuned scheduling scheme, and the fine-tuned scheduling scheme is converted into the execution instruction of each station equipment;Step S5: based on the execution instruction of each station equipment, drive the operation of each station equipment on the production line.

Description

Technical Field

[0001] This invention relates to the field of automation control technology, specifically to an intelligent scheduling decision-making method and system for special steel grades with high time sensitivity, and more specifically to a multi-station hybrid scheduling decision-making method for steelmaking special steel grades with high time sensitivity. Background Technology

[0002] Existing technologies primarily optimize multiple objectives such as minimizing waiting time, minimizing casting interruptions, and minimizing production time. They can also improve equipment utilization through "flexible workstations / flexible paths." However, with the surge in demand for special steel grades in high-end equipment manufacturing, their "high time sensitivity" characteristics are becoming increasingly prominent. For example, specific casting start times within a specific time window are required, and workstation delays must be reduced to ensure steel quality. Current technologies are insufficient to meet the stringent requirements of highly time-sensitive steel grades on production time variations.

[0003] Patent document CN108303958B (application number: 201810030207.1) discloses a multi-objective flexible scheduling method for the continuous casting process of steelmaking, which optimizes the expected values ​​of multiple objectives such as average waiting time, interruption time penalty and waiting time overrun penalty. However, this method can only solve the multi-objective optimization process in steelmaking and is difficult to effectively cope with the strict requirements of high time-sensitive steel grades on the changes in production time. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this invention is to provide an intelligent scheduling decision-making method and system for special steel grades with high time sensitivity.

[0005] The present invention provides an intelligent scheduling decision-making method for highly time-sensitive special steel grades, comprising:

[0006] Step S1: Obtain production data of the steelmaking production line, including: real-time status of steelmaking production line equipment, molten steel parameters, processing and waiting time of each stage, instantaneous capacity of equipment and current overall production progress;

[0007] Step S2: Construct a scheduling decision model based on the process constraints and production targets of the target product;

[0008] Step S3: Based on the constructed scheduling decision model and the acquired steelmaking production line production data, generate an initial scheduling scheme using a greedy algorithm.

[0009] Step S4: Under the condition of satisfying the preset process constraints, the initial scheduling plan is fine-tuned according to the production target to obtain the fine-tuned scheduling plan, and the fine-tuned scheduling plan is converted into the execution instructions of each workstation equipment;

[0010] Step S5: Drive the operation of each workstation on the production line based on the execution instructions of each workstation.

[0011] Preferably, the scheduling decision model includes:

[0012] Optimization objectives include: ;

[0013] in, , , These represent the weights of multiple objectives, which are dynamically adjusted according to the production status. Indicates time deviation; Indicates production costs; Indicates temperature loss;

[0014] The time deviation Represented as:

[0015]

[0016] in, This indicates the number of minutes that the process completion time exceeds the upper limit of the time window; This indicates the number of minutes that the process started before the lower limit of the time window; They represent the weights, and ;

[0017] The temperature loss Represented as:

[0018]

[0019] in, This indicates the actual temperature of furnace i in process j; This indicates the minimum temperature requirement for process j;

[0020] The constraints include process time window constraints, process sequence constraints, and maximum cycle constraints of adjacent units;

[0021] The process time window constraints include:

[0022]

[0023] in, Indicates the start time of furnace i in process j; , [] indicates the allowable time window for furnace i in process j;

[0024] The process sequence constraints include:

[0025]

[0026] in, This indicates the standard processing time for process j; This represents the transportation time from process j to process j+1; This indicates the start time of furnace i in process j+1;

[0027] The maximum period constraint of adjacent units includes:

[0028]

[0029] in, This represents the maximum turnaround time from process j to j+1; This indicates the start time of furnace i+1 in process j;

[0030] Equipment maintenance avoidance:

[0031]

[0032] in, Indicates the equipment involved in the scheduling; Indicates the maintenance period.

[0033] Preferably, step S3 includes:

[0034] Step S3.1: Calculate the time window urgency for each heat based on the obtained steelmaking production line production data;

[0035] Time window urgency = (current time - earliest start time of this batch) / (latest start time of this batch - earliest start time of this batch); the closer the urgency is to 1, the more urgent the time window is, and the more priority it needs to be scheduled.

[0036] Step S3.2: Arrange the furnace batches in descending order of time window urgency to prioritize resource allocation for furnaces with high time sensitivity;

[0037] Step S3.3: Obtain the furnace with the highest urgency and arrange the time windows of each process in sequence according to the production process order;

[0038] Step S3.4: For the same process, obtain available equipment that meets the process and equipment requirements; calculate the violation score of each available equipment based on the objective function in the scheduling decision model, and select the available equipment with the smallest score;

[0039] Step S3.5: After all furnaces have been equipped, time window violations are corrected to obtain the initial scheduling plan.

[0040] Preferably, step S4 includes:

[0041] The initial scheduling scheme is divided into a freeze segment and a fine-tuning segment according to the preset requirements;

[0042] The decision variables are changed to all furnace runs in the fine-tuning segment, the objective function remains unchanged, and the constraints also include a maximum offset constraint; wherein, the maximum offset constraint includes:

[0043]

[0044] in, Indicates the adjusted time. This represents the initial scheduling scheme provided by the greedy algorithm. This indicates the maximum offset that is set.

[0045] The current optimal scheduling scheme is obtained by using the Gurobi solver.

[0046] Preferably, the method further includes: real-time acquisition of the task's real-time execution progress, sudden changes in equipment status, and real-time changes in molten steel parameters, and real-time transmission of the real-time acquisition of the task's real-time execution progress, sudden changes in equipment status, and real-time changes in molten steel parameters to the scheduling decision model, thereby realizing real-time adjustment of the scheduling scheme.

[0047] An intelligent scheduling decision-making system for highly time-sensitive special steel grades provided by the present invention includes:

[0048] Module M1: Acquires production data from the steelmaking production line, including: real-time status of steelmaking production line equipment, molten steel parameters, processing and waiting time at each stage, instantaneous equipment capacity, and current overall production progress;

[0049] Module M2: Constructs a scheduling decision model based on the process constraints and production targets of the target product;

[0050] Module M3: Based on the constructed scheduling decision model and the acquired steelmaking production line production data, an initial scheduling scheme is generated using a greedy algorithm.

[0051] Module M4: Under the premise of meeting the preset process constraints, the initial scheduling plan is fine-tuned according to the production target to obtain the fine-tuned scheduling plan, and the fine-tuned scheduling plan is converted into the execution instructions of each workstation equipment;

[0052] Module M5: Drives the operation of each workstation on the production line based on the execution instructions of each workstation.

[0053] Preferably, the scheduling decision model includes:

[0054] Optimization objectives include: ;

[0055] in, , , These represent the weights of multiple objectives, which are dynamically adjusted according to the production status. Indicates time deviation; Indicates production costs; Indicates temperature loss;

[0056] The time deviation Represented as:

[0057]

[0058] in, This indicates the number of minutes that the process completion time exceeds the upper limit of the time window; This indicates the number of minutes that the process started before the lower limit of the time window; They represent the weights, and ;

[0059] The temperature loss Represented as:

[0060]

[0061] in, This indicates the actual temperature of furnace i in process j; This indicates the minimum temperature requirement for process j;

[0062] The constraints include process time window constraints, process sequence constraints, and maximum cycle constraints of adjacent units;

[0063] The process time window constraints include:

[0064]

[0065] in, Indicates the start time of furnace i in process j; , [] indicates the allowable time window for furnace i in process j;

[0066] The process sequence constraints include:

[0067]

[0068] in, This indicates the standard processing time for process j; This represents the transportation time from process j to process j+1; This indicates the start time of furnace i in process j+1;

[0069] The maximum period constraint of adjacent units includes:

[0070]

[0071] in, This represents the maximum turnaround time from process j to j+1; This indicates the start time of furnace i+1 in process j;

[0072] Equipment maintenance avoidance:

[0073]

[0074] in, Indicates the equipment involved in the scheduling; Indicates the maintenance period.

[0075] Preferably, the module M3 includes:

[0076] Module M3.1: Calculates the time window urgency for each heat based on the acquired steelmaking production line production data;

[0077] Time window urgency = (current time - earliest start time of this batch) / (latest start time of this batch - earliest start time of this batch); the closer the urgency is to 1, the more urgent the time window is, and the more priority it needs to be scheduled.

[0078] Module M3.2: Arrange furnace batches in descending order of time window urgency to prioritize resource allocation for furnaces with high time sensitivity;

[0079] Module M3.3: Obtain the furnace with the highest urgency and arrange the time windows of each process in sequence according to the production process order;

[0080] Module M3.4: For the same process, obtain available equipment that meets the process and equipment requirements; calculate the violation score of each available equipment based on the objective function in the scheduling decision model, and select the available equipment with the lowest score;

[0081] Module M3.5: After all furnaces have been equipped, time window violations are corrected to obtain the initial scheduling scheme.

[0082] Preferably, the module M4 includes:

[0083] The initial scheduling scheme is divided into a freeze segment and a fine-tuning segment according to the preset requirements;

[0084] The decision variables are changed to all furnace runs in the fine-tuning segment, the objective function remains unchanged, and the constraints also include a maximum offset constraint; wherein, the maximum offset constraint includes:

[0085]

[0086] in, Indicates the adjusted time. This represents the initial scheduling scheme provided by the greedy algorithm. This indicates the maximum offset that is set.

[0087] The current optimal scheduling scheme is obtained by using the Gurobi solver.

[0088] Preferably, the system further includes: real-time acquisition of task execution progress, sudden changes in equipment status, and real-time changes in molten steel parameters, and real-time transmission of the real-time acquisition of task execution progress, sudden changes in equipment status, and real-time changes in molten steel parameters to the scheduling decision model, thereby realizing real-time adjustment of the scheduling scheme.

[0089] Compared with the prior art, the present invention has the following beneficial effects:

[0090] 1. This invention, through a complete system architecture of scheduling decision model, intelligent optimization algorithm, and closed-loop execution and feedback, and more specifically, by combining the intelligent optimization algorithm to generate high-quality initial solutions with the MIP solver for precise optimization, successfully solves the core problems commonly found in existing technologies for multi-station scheduling of steelmaking for highly time-sensitive special steel grades, such as uneven distribution of production tasks, low efficiency of equipment resource utilization, production rhythm easily disturbed and causing delays, and difficulty in balancing efficiency, cost, and quality requirements in multi-objective optimization. This invention can significantly improve the overall efficiency of steelmaking production and the stability of product quality, and accurately meet the extremely strict requirements of special steel grades for production time windows.

[0091] 2. This invention deeply integrates multi-objective optimization, dynamic response mechanism and advanced intelligent algorithm. Through accurate time prediction and intelligent scheduling, it realizes refined time sequence control and resource coordination of each work station in steelmaking. Its purpose is to ensure seamless connection of production process, minimize delays, meet the stringent requirements of special steel grades for production time, and ultimately improve overall production efficiency and product quality stability. Attached Figure Description

[0092] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0093] Figure 1 This is a structural diagram of an intelligent scheduling and decision-making system for special steel grades with high time sensitivity.

[0094] Figure 2 This is a flowchart of an intelligent scheduling decision-making method for special steel grades with high time sensitivity.

[0095] Figure 3 This is a data flow diagram for intelligent scheduling decision-making methods for special steel grades with high time sensitivity. Detailed Implementation

[0096] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0097] Example 1

[0098] The present invention provides an intelligent scheduling decision-making method for highly time-sensitive special steel grades, comprising:

[0099] Step S1: Obtain production data of the steelmaking production line, including: real-time status of steelmaking production line equipment, molten steel parameters, processing and waiting time of each stage, instantaneous capacity of equipment and current overall production progress;

[0100] Step S2: Construct a scheduling decision model based on the process constraints and production targets of the target product;

[0101] Step S3: Based on the constructed scheduling decision model and the acquired steelmaking production line production data, generate an initial scheduling scheme using a greedy algorithm.

[0102] Step S4: Under the condition of satisfying the preset process constraints, the initial scheduling plan is fine-tuned according to the production target to obtain the fine-tuned scheduling plan, and the fine-tuned scheduling plan is converted into the execution instructions of each workstation equipment;

[0103] Step S5: Drive the operation of each workstation on the production line based on the execution instructions of each workstation.

[0104] Specifically, the scheduling decision model includes:

[0105] Optimization objectives include: ;

[0106] in, , , These represent the weights of multiple objectives, which are dynamically adjusted according to the production status. Indicates time deviation; Indicates production costs; Indicates temperature loss;

[0107] The time deviation Represented as:

[0108]

[0109] in, This indicates the number of minutes that the process completion time exceeds the upper limit of the time window; This indicates the number of minutes that the process started before the lower limit of the time window; They represent the weights, and ;

[0110] The temperature loss Represented as:

[0111]

[0112] in, This indicates the actual temperature of furnace i in process j; This indicates the minimum temperature requirement for process j;

[0113] The constraints include process time window constraints, process sequence constraints, and maximum cycle constraints of adjacent units;

[0114] The process time window constraints include:

[0115]

[0116] in, Indicates the start time of furnace i in process j; , [] indicates the allowable time window for furnace i in process j;

[0117] The process sequence constraints include:

[0118]

[0119] in, This indicates the standard processing time for process j; This represents the transportation time from process j to process j+1; This indicates the start time of furnace i in process j+1;

[0120] The maximum period constraint of adjacent units includes:

[0121]

[0122] in, This represents the maximum turnaround time from process j to j+1; This indicates the start time of furnace i+1 in process j;

[0123] Equipment maintenance avoidance:

[0124]

[0125] in, Indicates the equipment involved in the scheduling; Indicates the maintenance period.

[0126] Specifically, step S3 includes:

[0127] Step S3.1: Calculate the time window urgency for each heat based on the obtained steelmaking production line production data;

[0128] Time window urgency = (current time - earliest start time of this batch) / (latest start time of this batch - earliest start time of this batch); the closer the urgency is to 1, the more urgent the time window is, and the more priority it needs to be scheduled.

[0129] Step S3.2: Arrange the furnace batches in descending order of time window urgency to prioritize resource allocation for furnaces with high time sensitivity;

[0130] Step S3.3: Obtain the furnace with the highest urgency and arrange the time windows of each process in sequence according to the production process order;

[0131] Step S3.4: For the same process, obtain available equipment that meets the process and equipment requirements; calculate the violation score of each available equipment based on the objective function in the scheduling decision model, and select the available equipment with the smallest score;

[0132] Step S3.5: After all furnaces have been arranged with equipment, time window violations are repaired. This mainly involves finding out if the same equipment can be swapped with furnaces after the current furnace to resolve time window conflicts and obtain the initial scheduling plan.

[0133] Specifically, step S4 includes:

[0134] According to preset requirements, the initial scheduling scheme is divided into a freezing segment and a fine-tuning segment. In this embodiment, the freezing segment is the first 30 minutes after the pouring or processing has started and cannot be moved. The fine-tuning segment is from the end of the freezing segment to the next 4 hours, and only the schedule is rearranged in this segment. This can reduce the amount of data to ensure the speed of solving the problem, and at the same time help to generate feasible solutions.

[0135] The decision variables are changed to all furnace runs in the fine-tuning segment, the objective function remains unchanged, and the constraints also include a maximum offset constraint; wherein, the maximum offset constraint includes:

[0136]

[0137] in, Indicates the adjusted time. This represents the initial scheduling scheme provided by the greedy algorithm. This indicates the maximum offset that is set.

[0138] Prevent excessive drift;

[0139] The solution is obtained using the Gurobi solver, starting with the initial solution generated by the greedy algorithm. If the solver fails to converge within the set execution time, it returns the current optimal feasible solution.

[0140] Specifically, the method further includes: real-time acquisition of the task's real-time execution progress, sudden changes in equipment status, and real-time changes in molten steel parameters, and real-time transmission of the real-time acquisition of the task's real-time execution progress, sudden changes in equipment status, and real-time changes in molten steel parameters to the scheduling decision model, thereby realizing real-time adjustment of the scheduling scheme.

[0141] The present invention also provides an intelligent scheduling decision system for high time-sensitive special steel grades. The intelligent scheduling decision system for high time-sensitive special steel grades can be implemented by executing the process steps of the intelligent scheduling decision method for high time-sensitive special steel grades. That is, those skilled in the art can understand the intelligent scheduling decision method for high time-sensitive special steel grades as a preferred embodiment of the intelligent scheduling decision system for high time-sensitive special steel grades.

[0142] Example 2

[0143] Example 2 is a preferred example of Example 1.

[0144] According to the present invention, an intelligent scheduling decision-making system for highly time-sensitive special steel grades is provided, such as... Figures 1 to 3 As shown, it includes:

[0145] The data acquisition module is responsible for collecting production data in real time and comprehensively from key workstations in the steelmaking production line, such as converters, refining furnaces, and continuous casting machines. The data covers real-time equipment status, key steel parameters, processing and waiting times at each stage, instantaneous equipment capacity, and the current overall production progress. The real-time equipment status includes running, stopped, and fault states; the key steel parameters include temperature and composition (C, Si, Mn, P, S, etc.).

[0146] All data serves as foundational information and is transmitted to the subsequent production product process constraint library module and production target setting and weight adjustment module for processing.

[0147] The production process constraint library module is responsible for storing and managing detailed, customized process specifications for various time-sensitive special steel grades. These specifications include hard constraints such as allowable temperature ranges, composition ranges, processing time windows, and transfer time limits for each workstation. This constraint information is dynamically invoked by the hybrid scheduling decision engine module as the constraint conditions for scheduling decisions.

[0148] The production target setting and weight adjustment module provides a human-machine interface, allowing operators or the system to set or adjust scheduling targets and their priority weights based on current production conditions, such as order urgency and equipment load. Common targets include minimizing total completion time, maximizing equipment utilization, minimizing waiting time, and ensuring steel quality. Target setting is based on partial data from the data acquisition module, such as order information, inventory status, and manual input. The final target function and weights are then invoked into the hybrid scheduling decision engine module.

[0149] The hybrid scheduling decision engine module, designed for the production needs of highly time-sensitive specialty steel grades, comprehensively utilizes process constraints from the production product process constraint library module and production targets from the production target setting and weight adjustment module. It combines a two-layer optimization structure integrating multi-objective mixed-integer programming and intelligent optimization algorithms to accurately and efficiently generate and optimize multi-station production scheduling plans. This engine can precisely handle complex constraints, especially excelling in time constraints, and makes fine trade-offs among multiple objectives to ensure that the high time sensitivity requirements of the production process are met.

[0150] Specifically, the hybrid scheduling decision engine module includes:

[0151] The multi-objective modeling submodule focuses on the production characteristics of highly time-sensitive special steel grades. It is responsible for accurately abstracting the actual production scheduling problem and constructing a complex mathematical model that includes strict time window constraints, equipment resource conflicts, process path dependencies, and multiple objective functions (such as time, cost, and quality). It particularly emphasizes the important role of the time factor in the model to ensure that the model can accurately reflect various constraints and objectives in a highly time-sensitive production environment.

[0152] The complex mathematical model includes:

[0153] Optimization goal:

[0154] Time deviation:

[0155]

[0156] in, This indicates the number of minutes that the process completion time exceeds the upper limit of the time window; This indicates the number of minutes that the process started before the lower limit of the time window; The penalty for arriving early is lower than for arriving late, because equipment can wait, but molten steel cannot be reworked;

[0157] Temperature loss:

[0158]

[0159] in, This indicates the actual temperature of furnace i in process j; This indicates the minimum temperature requirement for process j; Indicates production costs;

[0160] The constraints include:

[0161] Process time window constraints:

[0162]

[0163] Indicates the start time of furnace i in process j; , [] indicates the allowable time window for furnace i in process j;

[0164] Process sequence constraints:

[0165]

[0166] This indicates the standard processing time for process j; This represents the transportation time from process j to process j+1;

[0167] Maximum period constraint of adjacent cells:

[0168]

[0169] This represents the maximum turnaround time from process j to j+1;

[0170] Equipment maintenance avoidance:

[0171] .

[0172] The hybrid solution submodule employs a phased optimization strategy of "intelligent heuristics + precise solution," specifically optimized for the production needs of special steel grades with high time sensitivity.

[0173] Specifically, the hybrid solution submodule includes:

[0174] Module M4.2.1: This module uses a greedy algorithm to generate an initial solution through scoring, paying particular attention to the time factor. It ensures that the initial solution satisfies all constraints in the production process constraint library module while maximally meeting the stringent production time requirements of highly time-sensitive special steel grades. Through efficient processing by intelligent optimization algorithms, it provides a high-quality initial solution for subsequent accurate solutions.

[0175] Module M4.2.2: Under the condition of satisfying all constraints, the optimization objective of the module is set and weighted according to the production target, especially for time-sensitive optimization objectives, and the initial solution is fine-tuned. The MIP solver can accurately handle time-related constraints and objectives, ensuring that the final scheduling scheme is highly feasible and accurate in terms of time, meeting the production requirements of highly time-sensitive special steel grades;

[0176] The decision output submodule transforms the optimal or suboptimal scheduling scheme output by the hybrid solution submodule into an instruction format that each workstation device can directly understand and execute, such as the work sequence, start / end time, parameter settings, etc., and clearly conveys it to the execution feedback module for execution.

[0177] The execution feedback module is responsible for receiving scheduling instructions from the hybrid scheduling decision engine module and driving each workstation to execute tasks according to plan. Simultaneously, it monitors and collects feedback information in real time, including actual task progress, sudden equipment status changes (such as new malfunctions), and real-time changes in molten steel parameters, and feeds this dynamic data back to the hybrid scheduling decision engine module. This enables the scheduling system to make agile adjustments based on real-time operating conditions, effectively addressing uncertainties in production and ensuring the effective implementation of scheduling plans and the stable operation of the production process.

[0178] Those skilled in the art will understand that, in addition to implementing the system, apparatus, and their modules provided by this invention in purely computer-readable program code, the same program can be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system, apparatus, and their modules provided by this invention can be considered a hardware component, and the modules included therein for implementing various programs can also be considered structures within the hardware component; alternatively, modules for implementing various functions can be considered both software programs implementing the method and structures within the hardware component.

[0179] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A method for intelligent scheduling decision of specific steel grades for high time sensitivity, characterized in that, The method comprises the following steps: Step S1: acquiring steelmaking production line production data, including: real-time state of steelmaking production line equipment, molten steel parameters, processing and waiting time of each link, instantaneous production capacity of equipment and current overall production progress; Step S2: constructing a scheduling decision model according to process constraints and production targets of target products; Step S3: generating an initial scheduling scheme by using a greedy algorithm based on the acquired steelmaking production line production data according to the constructed scheduling decision model; Step S4: fine-tuning the initial scheduling scheme according to the production target under the condition of meeting the preset process constraints, obtaining a fine-tuned scheduling scheme, and converting the fine-tuned scheduling scheme into execution instructions of each station equipment; Step S5: driving the operation of each station equipment on the production line based on the execution instructions of each station equipment; The scheduling decision model comprises: Optimization objectives include: ; wherein, , , respectively represent multi-objective weights, which are dynamically adjusted according to production states; represents a time deviation; represents a production cost; represents a temperature loss; The time offset is represented as: wherein, represents the number of minutes that the end time of the process exceeds the upper limit of the time window; represents the number of minutes that the start time of the process is earlier than the lower limit of the time window; respectively represent weights, and ; The temperature loss is represented as: wherein, represents the actual temperature of the heat treatment furnace for the batch i at the process j; represents the minimum temperature requirement for the process j; The constraint conditions comprise process time window constraints, process sequence constraints and adjacent unit maximum cycle constraints; The process time window constraints comprise: wherein, denotes the start time of the batch i at the process j; , denotes the allowed time window of the batch i at the process j; The process sequence constraints comprise: wherein, represents the standard processing duration of step j; represents the transport time from step j to j+1; represents the start time of heat i at step j+1; The adjacent unit maximum cycle constraints comprise: wherein, represents the maximum flow-through time of the process j to j+1; represents the start time of the heat i+1 at the process j; Equipment maintenance avoidance: wherein, represents a device participating in scheduling; represents a maintenance time period; The step S3 comprises: Step S3.1: calculating a time window urgency for each heat according to the acquired steelmaking production line production data; The time window urgency=(current time-earliest start time of the heat) / (latest start time of the heat-earliest start time of the heat); when the urgency is closer to 1, it indicates that the time window is more urgent and needs to be scheduled in priority; Step S3.2: arranging the heats in descending order of time window urgency, so as to preferentially allocate resources to high time sensitivity heats; Step S3.3: acquiring a heat with the highest urgency, and arranging the time windows of each process in sequence according to the production process sequence; Step S3.4: obtaining available equipment that meets the process and equipment requirements for the same process; calculating the violation scores of the available equipment based on the objective function in the scheduling decision model, and selecting the available equipment with the smallest score; Step S3.5: repairing the time window violation after the arrangement of the equipment for all heats, and obtaining an initial scheduling scheme.

2. The intelligent scheduling decision method for specific steel grades with high time sensitivity according to claim 1, characterized in that, The step S4 comprises: Dividing the initial scheduling scheme into a frozen section and a fine-tuning section according to preset requirements; Changing the decision variables into all heats in the fine-tuning section, and the objective function remains unchanged, and the constraint conditions further comprise a maximum offset constraint; wherein the maximum offset constraint comprises: wherein, denotes the adjusted time, denotes the initial scheduling scheme provided by the greedy algorithm, denotes the set maximum offset; Solving by using a Gurobi solver to obtain a current optimal scheduling scheme.

3. The intelligent scheduling decision method for specific steel grades with high time sensitivity according to claim 1, characterized in that, The method further comprises: collecting real-time execution progress of tasks, equipment state mutation and real-time changes of molten steel parameters in real time, and transmitting the collected real-time execution progress of tasks, equipment state mutation and real-time changes of molten steel parameters to the scheduling decision model in real time, so as to realize real-time adjustment of the scheduling scheme.

4. An intelligent scheduling decision system for high time-sensitive specific steel grades, characterized by, The method comprises the following steps: Module M1: acquiring steelmaking production line production data, including: real-time state of steelmaking production line equipment, molten steel parameters, processing and waiting time of each link, instantaneous production capacity of equipment and current overall production progress; Module M2: constructing a scheduling decision model according to process constraints and production targets of target products; Module M3: generating an initial scheduling scheme by using a greedy algorithm based on the acquired steelmaking production line production data according to the constructed scheduling decision model; Module M4: fine-tune the initial scheduling scheme according to the production target under the condition of meeting the preset process constraints, obtain a fine-tuned scheduling scheme, and convert the fine-tuned scheduling scheme into execution instructions of each workstation device; Module M5: drive the operation of each workstation device on the production line based on the execution instructions of each workstation device; The scheduling decision model comprises: Optimization objectives include: ; wherein, , , respectively represent multi-objective weights, which are dynamically adjusted according to production states; represents a time deviation; represents a production cost; represents a temperature loss; The time offset is represented as: wherein, represents the number of minutes that the end time of the process exceeds the upper limit of the time window; represents the number of minutes that the start time of the process is earlier than the lower limit of the time window; respectively represent weights, and ; The temperature loss is represented as: wherein, represents the actual temperature of the heat treatment step j for the furnace campaign i; represents the minimum temperature requirement for the heat treatment step j. The constraint conditions comprise a process time window constraint, a process sequence constraint, and a maximum cycle constraint of adjacent units; The process time window constraint comprises: wherein, represents the start time of the batch i at the process j; , represents the allowed time window of the batch i at the process j; The process sequence constraint comprises: wherein, represents the standard processing time for process j; represents the transport time from process j to j+1; represents the start time of heat i at process j+1; The maximum cycle constraint of adjacent units comprises: wherein, represents the maximum flow-through time of the process j to j+1; represents the start time of the heat i+1 at the process j; Device maintenance avoidance: wherein, represents a device participating in scheduling; represents a maintenance time period; The module M3 comprises: Module M3.1: calculate the time window urgency for each heat according to the obtained steelmaking production line production data; The time window urgency=(current time-earliest start time of the heat) / (latest start time of the heat-earliest start time of the heat); when the urgency is closer to 1, the time window is more urgent and needs to be scheduled first; Module M3.2: arrange the heats in descending order of time window urgency, so as to preferentially allocate resources to high time sensitivity heats; Module M3.3: obtain the heat with the highest urgency, and arrange the time windows of each process in sequence according to the production process sequence; Module M3.4: obtain available devices that meet the process and device requirements for the same process; calculate the violation scores of the available devices based on the objective function in the scheduling decision model, and select the available device with the smallest score; Module M3.5: repair the time window violation after all heats are arranged with devices to obtain an initial scheduling scheme.

5. The intelligent dispatch decision system for specific steel grades with high time sensitivity according to claim 4, characterized in that, The module M4 comprises: Divide the initial scheduling scheme into a frozen section and a fine-tuning section according to preset requirements; Change the decision variables to all heats in the fine-tuning section, the objective function remains unchanged, and the constraint conditions further comprise a maximum offset constraint; wherein the maximum offset constraint comprises: wherein, denotes the adjusted time, denotes the initial scheduling scheme provided by the greedy algorithm, denotes the set maximum offset; Solve using a Gurobi solver to obtain the current optimal scheduling scheme.

6. The intelligent dispatch decision system for specific steel grades with high time sensitivity according to claim 4, characterized in that, The system further comprises: real-time collection of real-time execution progress of tasks, device state mutations, and real-time changes of molten steel parameters, and real-time transmission of the real-time collected real-time execution progress of tasks, device state mutations, and real-time changes of molten steel parameters to the scheduling decision model, thereby realizing real-time adjustment of the scheduling scheme.

Images