Multi-unit integrated scheduling optimization method considering power consumption cost
Through greedy algorithms and genetic algorithms, the integrated scheduling of multi-units for cold rolling production is optimized, and combined with the power consumption cost factors, the problem of insufficient energy consumption management in traditional methods is solved, and the production efficiency and cost optimization is achieved.
Patent Information
- Application Number
- CN202510572256.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-30
- Publication Date
- 2025-08-19
AI Technical Summary
In the existing cold rolling production, the integrated scheduling technology of multi-units has shortcomings in energy consumption cost management and real-time production dynamic adjustment, resulting in low overall production efficiency and high costs.
The greedy algorithm is used to generate the initial scheduling plan, build the production line scheduling model, and optimize the genetic algorithm, combine power consumption cost factors, set the objective function and constraints, introduce random disturbance and restart mechanisms, and optimize the integrated scheduling of multiple units.
Improve production flexibility and response speed, reduce production costs, realize global optimization of production plans, and improve production efficiency.
Smart Images
Figure CN120509642A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of production line scheduling, and in particular to a multi-unit integrated scheduling optimization method considering power consumption costs. Background Art
[0002] With the development of small and medium-sized manufacturing industries, such as automobiles and home appliances, customer demands for steel products have become more differentiated and personalized. This has exacerbated the conflict between large-scale, extensive, and decentralized production methods and customers' differentiated demands for diverse and small batches, creating significant challenges for cold rolling production organization and design.
[0003] At present, the research ideas of relevant domestic companies conducting research on integrated scheduling issues are mostly concentrated on establishing rule base models based on cold rolling production processes, and optimizing and solving the models through heuristic algorithms. The scheduling strategy mainly focuses on static scheduling, and there are still many shortcomings in energy consumption cost management and combining real-time production dynamic adjustment.
[0004] The main research goal of the integrated scheduling technology for cold rolling multi-units is to formulate the optimal multi-unit continuous production plan based on order delivery requirements, product process path, process capacity, unit scheduling constraints, etc., so as to achieve the goals of reducing batch switching and work-in-process inventory, improving contract fulfillment rate, improving production efficiency and saving energy costs. Summary of the Invention
[0005] In order to solve the above problems, the present invention proposes a multi-unit integrated scheduling optimization method considering power consumption cost.
[0006] The specific plan is as follows:
[0007] A multi-unit integrated scheduling optimization method considering power consumption cost includes the following steps:
[0008] S1: Obtain information on steel coils to be arranged on each production line;
[0009] S2: Based on the information of the steel coils to be scheduled, and while satisfying the hard constraints, a greedy algorithm is used to generate the initial scheduling plan for each production line;
[0010] S3: Build a production line scheduling model and set the model's objective function and constraints;
[0011] The objective function of the model is expressed as:
[0012]
[0013] Among them, Min means taking the minimum value; U represents the set of all processes on a production line; u represents the sequence number of the process; N represents the set of all coils to be arranged; i and j both represent the sequence numbers of the coils; B represents the set of all batches; b represents the sequence number of the batch; M represents the set of all parallel equipment included in a process; m represents the sequence number of the equipment; x ij Indicates whether the j steel coil is processed and produced immediately after the i steel coil; P uij Indicates the penalty points caused by the change in width or thickness exceeding the preset threshold when switching from producing coil i to producing coil j in process u; d i Indicates whether the steel coil i meets the steel coil delivery time; P i Indicates the penalty points caused by the steel coil i not meeting the steel coil delivery time; h mi Indicates whether the i steel coil is produced on the m equipment; P mi represents the penalty score caused by the i steel coil not meeting the equipment preference of the m equipment; P uij Indicates the penalty points caused by the i coil and the j coil processed and produced immediately after it not belonging to the same batch in process u; ui h represents the time consumed by coil i in process u; ui Indicates whether the i steel coil is produced in the u process; V u represents the theoretical capacity of process u after deducting the maintenance time; f b Indicates whether the number of steel coils in batch b meets the upper and lower limits of the process preset; P b Indicates the penalty caused by the number of steel coils in batch b not meeting the upper and lower limits of the process preset; E uti represents the electricity penalty caused by processing coil i on equipment u at time t; f su Indicates whether the quantity of steel coils in the raw material storage area s corresponding to process u meets the upper and lower limits of the preset safety stock; P su represents the penalty points caused by the quantity of steel coils in the raw material storage area s corresponding to process u not meeting the upper or lower limit of the preset safety stock; α1, α2, α3, α4, α5, α6, α7, and α8 are all penalty coefficients;
[0014] S4: After using the initial schedule plan as an individual in the initial population, the production line scheduling model is solved by genetic algorithm to obtain the optimal schedule for each production line.
[0015] Furthermore, hard constraints include: waiting time constraints between two adjacent processes, steel coil delivery time constraints, the number of steel coils in a batch must meet the upper and lower limits preset by the process, and the inventory must meet the preset safety stock or maximum inventory.
[0016] Furthermore, the model's constraints include: each coil can only be processed once on one unit, each coil must be processed once on each unit throughout the entire process, the coils allocated for production at each process cannot exceed the maximum capacity of each process, if one coil is produced immediately after another in a certain process, the start time of the latter coil must be after the completion time of the previous coil, and the start time of a coil in the subsequent process must be after the completion time of the previous process.
[0017] Furthermore, during each iteration, the genetic algorithm retains the best individuals of each generation so that they can be passed on to the next generation.
[0018] Furthermore, random perturbation and random restart mechanisms are introduced into the genetic algorithm.
[0019] The present invention adopts the above technical solution, and by comprehensively considering factors such as contracts, processes, production line status and electricity consumption costs, it solves the overall non-optimal problem caused by traditional methods in optimizing a single link, significantly improving the flexibility and response speed of enterprises in complex production processes and dynamic market environments, improving production efficiency, reducing production costs, and achieving global optimization of production plans. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 Shown is a flow chart of embodiment 1 of the present invention. DETAILED DESCRIPTION
[0021] To further illustrate various embodiments, the present invention provides accompanying drawings. These drawings form part of the present disclosure and are primarily used to illustrate the embodiments and, in conjunction with the relevant description in the specification, to explain the operating principles of the embodiments. By referring to these drawings, those skilled in the art will be able to understand other possible implementations and the advantages of the present invention.
[0022] The present invention will now be further described with reference to the accompanying drawings and specific embodiments.
[0023] Example 1:
[0024] The embodiment of the present invention provides a multi-unit integrated scheduling optimization method considering power consumption cost, such as Figure 1 As shown, the method includes the following steps:
[0025] S1: Obtain the information of steel coils to be arranged on each production line.
[0026] S2: Based on the information of the steel coils to be scheduled, and while satisfying the hard constraints, a greedy algorithm is used to generate the initial scheduling plan for each production line.
[0027] The constraints for generating the initial scheduling plan in this embodiment include hard constraints and soft constraints. Hard constraints are constraints that must be met. The hard constraints set in this embodiment include: the waiting time constraint between two adjacent processes, the steel coil delivery time constraint, the number of steel coils in a batch must meet the upper and lower limits of the process preset, and the inventory must meet the preset safety stock or maximum inventory. Soft constraints are constraints that can be tolerated. The soft constraints set in this embodiment include: the processing preferences of each device (such as a preference for producing thicker steel coils, a preference for producing wider steel coils, etc.), the thickness width or thickness variation constraint of the adjacent steel coils produced, and the batch switching constraint.
[0028] In step S2, first, different penalties are set according to the importance of the above hard constraints, and then according to the solution principle of the greedy algorithm, hard constraints with larger penalties are satisfied first to generate an initial scheduling plan.
[0029] S3: Build a production line scheduling model and set the model's objective function and constraints.
[0030] The objective function of the model is expressed as:
[0031]
[0032] Among them, U represents the set of all processes on a production line, u represents the sequence number of the process; N represents the set of all coils to be arranged; i and j both represent the sequence numbers of the coils; B represents the set of all batches; b represents the sequence number of the batch; M represents the set of all parallel equipment included in a process (a process may have multiple equipment for parallel production, and only one equipment needs to be used); m represents the sequence number of the equipment; x ij Indicates whether the j steel coil is processed and produced immediately after the i steel coil. If it is, it is 1, otherwise it is 0; P uij Indicates the penalty points caused by the change in width or thickness exceeding the preset threshold when switching from producing coil i to producing coil j in process u; d i Indicates whether the i steel coil meets the steel coil delivery time, if not, it is 0, otherwise, it is 0; P i Indicates the penalty points caused by the steel coil i not meeting the steel coil delivery time; h mi Indicates whether the i steel coil is produced on the m equipment. If it is produced on the m equipment, it is 1, otherwise it is 0; P mi represents the penalty score caused by the i steel coil not meeting the equipment preference of the m equipment; P uij Indicates the penalty points caused by the i coil and the j coil processed and produced immediately after it not belonging to the same batch in process u; ui h represents the time consumed by coil i in process u; ui Indicates whether the i steel coil is produced in the u process; V urepresents the theoretical capacity of process u after deducting the maintenance time; f b Indicates whether the number of steel coils in batch b meets the upper and lower limits of the process preset. If not (i.e., greater than the upper limit or less than the lower limit), it is 1, otherwise it is 0; P b Indicates the penalty caused by the number of steel coils in batch b not meeting the upper and lower limits of the process preset; E uti represents the electricity penalty caused by the processing of steel coil i on equipment u at time t. It is necessary to find out which type of electricity consumption period the time t belongs to (including peak period, peak period, flat period, and valley period), and set different penalty points for different types of electricity consumption periods; su Indicates whether the quantity of steel coils in the raw material storage area s corresponding to process u (each process has its corresponding raw material storage area, and the products produced in the previous process are first transported to the raw material storage area of the next process for storage to be used in the production of the next process) meets the preset upper and lower limits of the safety stock; P su represents the penalty for the number of coils in raw material storage area s corresponding to process u not meeting the upper or lower limit of the preset safety stock. α1, α2, α3, α4, α5, α6, α7, and α8 represent penalty coefficients. α1 represents the penalty for width or thickness fluctuations, α2 represents the penalty for delayed coil delivery, α3 represents the penalty for equipment preference during the production process, α4 represents the penalty for batch switching between previous and next coils, α5 represents the penalty for equipment utilization, α6 represents the penalty for not meeting the upper and lower limits of batches, α7 represents the penalty for coils with different power consumption when avoiding off-peak hours and peak hours, and α8 represents the penalty for the number of coils in the warehouse not meeting the upper and lower limits of the preset safety stock. The values of these penalties are user-configurable.
[0033] The constraints of the model are expressed as:
[0034] ∑ m∈M h umi =1 (2)
[0035] ∑ u∈U h ui =n (3)
[0036] 0≤∑ u∈U ∑ i∈N t ui h ui ≤V u (4)
[0037]
[0038] Wherein, formula 2 indicates that coil i can only be processed once on unit u, h umi Indicates whether coil i is produced on equipment m in unit u, which is 1 if yes and 0 otherwise; Formula 3 indicates that coil i must be processed once on each unit in the entire process, and n represents the total number of units in the entire process; Formula 4 indicates that the coils allocated to each process cannot exceed the maximum production capacity of each process; Formula 5 indicates that if coil j is produced immediately after coil i in process u, then the start time of coil j must be after the completion time of coil i, and x uij Indicates whether the j coil is processed and produced immediately after the i coil in the u process. If it is immediately followed, it is 1, otherwise it is 0. Indicates the start time of production of coil j in process u, represents the production completion time of coil i in process u; Formula 6 indicates that the start time of production of a coil in the next process must be after the production completion time of the previous process. Indicates the start time of production of coil i in process u, Indicates the production completion time of steel coil i in process u-1.
[0039] S4: After using the initial schedule plan as an individual in the initial population, the production line scheduling model is solved by genetic algorithm to obtain the optimal schedule for each production line.
[0040] When generating the initial population, this embodiment adds the initial solution (i.e., the initial scheduling plan) generated by the greedy algorithm in step S2 as an individual to the population. That is, the initial population contains randomly generated individuals (sorted by the delivery date and time of the steel coils, and steel coils with the same batch number are placed together, and then a certain number of scheduling plans are randomly generated) and the initial solution generated by the greedy algorithm. The genetic algorithm is then used to continuously perform crossover and mutation iterations on the population to find the optimal solution. A new population is obtained by selecting using the fitness function (the inverse of the penalty value obtained by the objective function is used as the fitness p). In each iteration, the elite individuals (with the smallest total penalty) of each generation are retained to ensure that they will be selected for the next generation. Finally, the genetic algorithm is used to perform cyclic iterative adjustments until convergence, and a scheduling plan is selected until the total penalty no longer decreases.
[0041] In order to prevent the results of the above genetic algorithm from falling into local optimum during the iteration process, this embodiment not only adopts the elite retention strategy but also adds random perturbation and random restart mechanisms to ensure the diversity of the population.
[0042] The embodiments of the present invention solve the overall non-optimal problem caused by traditional methods in optimizing a single link by comprehensively considering factors such as contracts, processes, production line status and electricity consumption costs. It significantly improves the flexibility and response speed of enterprises in complex production processes and dynamic market environments, improves production efficiency, reduces production costs, and achieves global optimization of production plans.
[0043] Although the present invention has been particularly shown and described in conjunction with preferred embodiments, it will be understood by those skilled in the art that various changes in form and details may be made to the present invention without departing from the spirit and scope of the invention as defined in the appended claims, and all such changes are within the scope of protection of the present invention.
Claims
1. A multi-unit integrated scheduling optimization method considering power consumption cost, characterized in that: The following steps are involved: S1: Obtain information on steel coils to be arranged on each production line; S2: Based on the information of the steel coils to be scheduled, and while satisfying the hard constraints, a greedy algorithm is used to generate the initial scheduling plan for each production line; S3: Build a production line scheduling model and set the model's objective function and constraints; The objective function of the model is expressed as: Among them, Min means taking the minimum value; U represents the set of all processes on a production line; u represents the sequence number of the process; N represents the set of all coils to be arranged; i and j both represent the sequence numbers of the coils; B represents the set of all batches; b represents the sequence number of the batch; M represents the set of all parallel equipment included in a process; m represents the sequence number of the equipment; x ij Indicates whether the j steel coil is processed and produced immediately after the i steel coil; P uij Indicates the penalty points caused by the change in width or thickness exceeding the preset threshold when switching from producing coil i to producing coil j in process u; d i Indicates whether the steel coil i meets the steel coil delivery time; P i Indicates the penalty points caused by the steel coil i not meeting the steel coil delivery time; h mi Indicates whether the i steel coil is produced on the m equipment; P mi represents the penalty score caused by the i steel coil not meeting the equipment preference of the m equipment; P uij Indicates the penalty points caused by the i coil and the j coil processed and produced immediately after it not belonging to the same batch in process u; ui h represents the time consumed by coil i in process u; ui Indicates whether the i steel coil is produced in the u process; V u represents the theoretical capacity of process u after deducting the maintenance time; f b Indicates whether the number of steel coils in batch b meets the upper and lower limits of the process preset; P b Indicates the penalty caused by the number of steel coils in batch b not meeting the upper and lower limits of the process preset; E uti represents the electricity penalty caused by processing coil i on equipment u at time t; f su Indicates whether the quantity of steel coils in the raw material storage area s corresponding to process u meets the upper and lower limits of the preset safety stock; P su represents the penalty points caused by the quantity of steel coils in the raw material storage area s corresponding to process u not meeting the upper or lower limit of the preset safety stock; α1, α2, α3, α4, α5, α6, α7, and α8 are all penalty coefficients; S4: After using the initial schedule plan as an individual in the initial population, the production line scheduling model is solved by genetic algorithm to obtain the optimal schedule for each production line.
2. The multi-unit integrated scheduling optimization method considering power consumption cost according to claim 1 is characterized in that: Hard constraints include: waiting time constraints between two adjacent processes, steel coil delivery time constraints, the number of steel coils in a batch must meet the upper and lower limits of the process preset, and the inventory must meet the preset safety stock or maximum inventory.
3. The multi-unit integrated scheduling optimization method considering power consumption cost according to claim 1, characterized in that: The constraints of the model include: each steel coil can only be processed once on one unit, each steel coil must be processed once on each unit in the entire processing flow, the steel coils allocated for production in each process cannot exceed the maximum production capacity of each process, if one steel coil is produced immediately after another steel coil in a certain process, the start time of the production of the latter steel coil must be after the production completion time of the previous steel coil, and the start time of the production of a steel coil in the subsequent process must be after the production completion time of the previous process.
4. The multi-unit integrated scheduling optimization method considering power consumption cost according to claim 1, characterized in that: In each iteration, the genetic algorithm retains the best individuals of each generation so that they can be passed on to the next generation.
5. The multi-unit integrated scheduling optimization method considering power consumption cost according to claim 1 is characterized in that: Random perturbation and random restart mechanisms are introduced into the genetic algorithm.