Automobile order production scheduling method considering cache region

Abstract

The invention belongs to the technical field of automobile manufacturing, and particularly relates to an automobile order production scheduling method considering a cache region, and the method comprises the following steps: carrying out the priority ranking of order production scheduling constraint conditions; constructing an iterative algorithm model, and generating an initial population by adopting an integer coding mode; a fitness value function is calculated, an iterative algorithm model is introduced into a traditional genetic algorithm, and multi-target genetic algorithm iterative optimization is carried out through a non-dominated sorting method; judging whether the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained; if the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained, terminating; if the maximum number of iterations is not reached or the optimal solution of the target fitness function is not obtained, repeating until the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained; compared with an existing method, the automobile order production scheduling method considering the cache region has the advantages that the solving speed is higher, and meanwhile the constraint satisfaction rate is higher.

Description

Technical Field

[0001] The invention belongs to the technical field of automobile manufacturing, and in particular relates to an automobile order scheduling method considering a buffer area. Background Art

[0002] The current production mode of the automobile manufacturing industry is often scheduled in stages, from the initial steel processing stamping workshop production, to the steel welding workshop production into the white body, then the white body painting workshop production, and finally the assembly workshop production. Since the welding workshop, painting workshop, and assembly workshop have completely different requirements for the order queue sequence, many automobile manufacturers will set up WBS buffers and PBS buffers between the welding workshop, painting workshop, and assembly workshop. By reordering the car bodies produced by the previous workshop in the buffer area, the car bodies that have completed the staged processing in the previous workshop are stored in it, and then the car bodies stored in it are released to the next workshop in the specified order, thereby completing the reordering of the order queue. Whether it is a PBS buffer area or a WBS buffer area, its essence is an automated transportation system. At present, the heuristic method and intelligent optimization algorithm are mainly used to solve the order scheduling problem with a buffer area. Although the heuristic method is superior to the intelligent optimization algorithm in terms of method complexity, the convergence of the method is difficult to guarantee. The intelligent optimization method mainly includes the method (Genetic Intelligent optimization methods are more suitable for complex problems under large-scale and multi-constrained conditions. They find the optimal production scheduling plan by simulating natural selection and mechanisms. At the same time, intelligent optimization methods do not rely on the gradient of the fitness function, making the solution more applicable.

[0003] Reference [1] takes minimizing color switching during painting and equalizing the consumption rate of vehicles in the adjustment lane as dual objective functions, constructs a WBS buffer vehicle scheduling model, effectively improves the production sequence rate and production efficiency of the WBS buffer, and ensures the uniformity of the consumption rate of vehicles in the adjustment lane; Reference [2] considers the use of the buffer area and provides a scheduling method for urgent orders in the automobile production process. In the scheduling control, the idle time of the production line workstations is fully considered, and based on the joint action of the two strategies, the lag index is used as the optimization indicator, so that production time can be effectively saved and production resources can be fully utilized; Reference [3] proposes a flexible job shop scheduling evolution based on digital twins The method decoding method solves the problem that the evolutionary method decoding method in the existing flexible job shop scheduling has low computational efficiency and is difficult to handle the complex constraints of the actual production process; Reference [4] solves the problem that the material quality inspection task in the existing technology cannot be automatically scheduled by performing real-time inspection on the material information of the target production order, realizes the automatic scheduling of the material quality inspection task, and improves production efficiency; Reference [5] not only responds to major emergencies in a fast and positive state, but also optimizes the allocation of scarce medical resources, which can improve the utilization rate of human resources, reduce the health risks of nurses, and effectively improve the efficiency of major emergency management in hospitals; Reference [6] selects different constraints to meet different user needs. It can meet the needs of nurses' scheduling to the greatest extent and meet the personal scheduling intentions of each nurse; Reference [7] significantly improves the efficiency and accuracy of drug management by integrating advanced intelligent identification, automated inventory management, drug distribution and data analysis technologies, ensures drug safety and compliance, and supports hospitals to make data-driven decisions, thereby optimizing the overall drug supply chain; Reference [8] selects the corresponding plate material residues from the residue archive through the residue intelligent picking method for production scheduling, rationally utilizes plate resources, improves the utilization rate and selection efficiency of plate material residues, and reduces production costs; Reference [9] derives the shortage and procurement plan of raw and auxiliary materials through the procurement module; the production management module converts the report feedback from MES into The work results are backfilled into the corresponding production plan, achieving the effect of reducing costs and increasing efficiency; Reference

[10] traverses all unflown routes, analyzes three types of sorting points, and sorts them according to the sorting rules. This method has a small amount of computation, reduces the demand for system computing power, and saves the hardware cost required to build the system; Reference

[11] comprehensively analyzes the optimized driving routes and information interaction, and makes mobile sewage treatment vehicle scheduling and task allocation decisions, solving the technical problems of inaccurate data processing and poor adaptability in the management process of mobile sewage treatment vehicles in existing technologies; Reference

[12] gives priority to bottleneck processes, which not only ensures the flexibility of plan adjustment, but also improves timeliness and referenceability.

[0004] References are as follows:

[0005] 1.[Tang Qian, 2022]WBS buffer zone vehicle scheduling method;

[0006] 2.[Cai Hongwei, 2022] Emergency order scheduling method for automobile production process;

[0007] 3. [Yan Jihong, 2024] Decoding method for flexible job shop scheduling evolutionary method based on digital twin;

[0008] 4.[Jiang Chenyang, 2024] A production task scheduling method, device, electronic device and storage medium;

[0009] 5. [Ma Ruiguo, 2023] An optimization method for nurse scheduling in emergency situations;

[0010] 6. [Zhang Zhichao, 2023] Nurse scheduling method based on demand selection;

[0011] 7.[Zhang Jiancheng, 2024] An intelligent hospital drug management system;

[0012] 8. [Zhang Jijuan, 2024] A method and system for optimizing the utilization of surplus board materials of panel-type custom furniture;

[0013] 9.[Xu Huaping, 2024] APS automatic scheduling system, method and related equipment;

[0014] 10. [Wang Xiaohu, 2024] Log generation method, device and electronic equipment for APS system;

[0015] 11.[Zeng Yuanjun, 2024] A port entry sorting method and device for solving multiple sorting problems;

[0016] 12.[Zang Yixu, 2024] An order scheduling method, device, electronic device and medium.

[0017] In order to overcome the problems of long solution time and low constraint satisfaction rate in existing methods, the present invention provides a car order scheduling method considering the cache area, which has a faster solution speed and a higher constraint satisfaction rate compared with the existing method. Summary of the invention

[0018] The purpose of the present invention is to provide a method for scheduling automobile orders taking into account a buffer zone, which has a faster solution speed and a higher constraint satisfaction rate.

[0019] In order to achieve the above technical objectives, the technical solution adopted by the present invention is as follows:

[0020] A method for scheduling automobile orders taking into account a buffer zone comprises the following steps:

[0021] Step S100: Initialize parameters and prioritize order scheduling constraints according to order scheduling constraint information;

[0022] Step S200: constructing an iterative algorithm model according to the priority order of the order scheduling constraints, and generating an initial population using integer coding;

[0023] Step S300: Calculate the fitness value function, introduce an iterative algorithm model into the traditional genetic algorithm, perform multi-objective genetic algorithm iterative optimization through a non-dominated sorting method, and update the order sequence;

[0024] Step S400: Determine whether the maximum number of iterations has been reached or the optimal solution of the target fitness function has been obtained:

[0025] The process terminates if the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained;

[0026] If the maximum number of iterations is not reached or the optimal solution of the target fitness function is not obtained, step S300 is repeated until the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained.

[0027] The constraint information of step S100 includes but is not limited to the following six constraint conditions:

[0028] ① Order location constraint: an order needs to be scheduled to a specified location;

[0029] ② Assembly pure electric constraints, which means that all vehicles with "power type" of "pure electric" cannot be produced continuously;

[0030] ③Continuous clustering constraint, which means that the clusters of different "car type classifications" must contain a specified number of cars;

[0031] ④ Mixed scheduling constraint, which means that vehicles with different “configurations” must be scheduled in order;

[0032] ⑤ Assembly balance constraint, which means that a certain model in the "model classification" needs to be evenly scheduled for production;

[0033] ⑥ Buffer area constraint means that after each order passes through the buffer area from the welding workshop to the final assembly workshop, the sequence difference must not exceed the specified value.

[0034] In step S200, the following iterative algorithm model (fitness functions F1 and F2) is constructed based on the six constraints for non-dominated sorting in a multi-objective genetic algorithm:

[0035] Formula 1: min F1 = f1 + f6;

[0036] Formula 2: min F2=f2+f3+f4+f5;

[0037] In the algorithm model, the mathematical model of each parameter is as follows:

[0038] Formula 3: Represents the sum of squared differences between the actual position and the specified position of each order;

[0039] Formula 4: It represents the sum of the squared differences between the length of each agglomerate whose "power type" is "pure electric" and 1;

[0040] Formula 5: The sum of the squared differences between the length of each block and the specified value for each "car type classification";

[0041] Formula 6: It represents the sum of the difference between the index of the last car in the first "configuration" and the index of the first car in the second "configuration";

[0042] Formula 7: It indicates the sum of the squared differences between the number of vehicles of each "vehicle type classification" and the specified value in each vehicle type block of length 40;

[0043] Formula 8: It represents the sum of the square differences between the absolute value of the sequence difference of the same order in two workshops and the specified value;

[0044] in,

[0045] n represents the order quantity; X i Indicates the position of the i-th order in the production scheduling result; L i represents the production location specified by the i-th order; Y k Indicates the length of the block of the k-th vehicle type with "power type" as "pure electric"; Z i,k N represents the length of the kth cluster of vehicles in the i-th “vehicle type classification”; i It represents the length of the vehicle block specified by the i-th “vehicle type classification”; W i,last Indicates the index of the last vehicle in the i-th “configuration” in the production scheduling result; W i+1,first represents the index of the first vehicle in the i+1th “configuration”; M i,m A represents the number of vehicles of the mth “model classification” in the i-th model block of length 40; M represents the number of vehicles of the specified mth “model classification” in each model block of length 40; A i represents the index of the i-th order in the production scheduling result of the general assembly workshop; B i Represents the index of the i-th order in the production scheduling result of the welding workshop; index represents the specified sequence difference.

[0046] In the iterative algorithm model, the constraints of each parameter are as follows:

[0047] Formula 9: Z i,k <N i,max , N i,max Indicates the maximum vehicle type cluster length of the i-th “vehicle type classification”;

[0048] Formula 10: Z i,k >N i,min , N i,min Indicates the minimum vehicle type block length of the i-th “vehicle type classification”;

[0049] Equations 9 and 10 indicate that the length of each “vehicle type classification” must be within the specified interval;

[0050] Formula 11: This means that in the final production scheduling result, vehicles with different "configurations" cannot be scheduled in an intersecting manner, and there is a clear scheduling order;

[0051] Formula 12: M-1≤M i,m ≤M+1, indicating that in each car type block of length 40, the number of cars of a specified "car type classification" must be in the specified range;

[0052] Formula 13: |A i -B i |≤index, indicating that the sequence difference of each order in the production scheduling results of the general assembly workshop and the welding workshop shall not exceed the specified value.

[0053] In order to overcome the problems of long solution time and low constraint satisfaction rate in existing methods, the present invention provides a car order scheduling method considering the cache area, which has a faster solution speed and a higher constraint satisfaction rate compared with the existing method.

[0054] Through a deep understanding of actual automobile production scenarios, the present invention establishes a unique mathematical model that better reflects the constraints on order sequences in actual automobile production scenarios; and through the individual insertion of each order, that is, according to the sequence difference constraints of the general assembly workshop and the welding workshop, each iteration inserts a certain order in the general assembly workshop into the specified index interval, which greatly improves the solution speed and solution quality. When applied, it can achieve the goal of reducing costs and increasing efficiency for automobile manufacturers. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The present invention is further illustrated by means of the non-limiting examples given in the accompanying drawings.

[0056] Figure 1 It is a schematic diagram of the method of the present invention;

[0057] Figure 2 It is a schematic diagram of algorithm iteration of the present invention;

[0058] Figure 3 When the order quantity is 400 in the embodiment of the present invention, the algorithm iteration result chart;

[0059] Figure 4 When the order quantity is 800 in the embodiment of the present invention, the algorithm iteration result chart is shown;

[0060] Figure 5 When the order quantity is 1200 in the embodiment of the present invention, the algorithm iteration result chart;

[0061] Figure 6 The graph of algorithm iteration results is shown when the order quantity is 1600 according to the embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the present invention, the technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0063] A method for scheduling automobile orders taking into account a buffer zone comprises the following steps:

[0064] Step S100: Initialize parameters and prioritize order scheduling constraints according to order scheduling constraint information;

[0065] The constraint information includes but is not limited to the following six constraints:

[0066] ① Order location constraint: an order needs to be scheduled to a specified location;

[0067] ② Assembly pure electric constraints, which means that all vehicles with "power type" of "pure electric" cannot be produced continuously;

[0068] ③Continuous clustering constraint, which means that the clusters of different "car type classifications" must contain a specified number of cars;

[0069] ④ Mixed scheduling constraint, which means that vehicles with different “configurations” must be scheduled in order;

[0070] ⑤ Assembly balance constraint, which means that a certain model in the "model classification" needs to be evenly scheduled for production;

[0071] ⑥ Buffer zone constraint: after each order passes through the buffer zone from the welding workshop to the final assembly workshop, the sequence difference must not exceed the specified value;

[0072] Step S200: constructing an iterative algorithm model according to the priority order of the order scheduling constraints, and generating an initial population using integer coding;

[0073] Among them, the following iterative algorithm model (fitness functions F1 and F2) is constructed based on the six constraints for non-dominated sorting in a multi-objective genetic algorithm:

[0074] Formula 1: min F1 = f1 + f6;

[0075] Formula 2: min F2=f2+f3+f4+f5;

[0076] Formula 3: Represents the sum of squared differences between the actual position and the specified position of each order;

[0077] Formula 4: It represents the sum of the squared differences between the length of each agglomerate whose "power type" is "pure electric" and 1;

[0078] Formula 5: The sum of the squared differences between the length of each block and the specified value for each "car type classification";

[0079] Formula 6: It represents the sum of the difference between the index of the last car in the first "configuration" and the index of the first car in the second "configuration";

[0080] Formula 7: It indicates the sum of the squared differences between the number of vehicles of each "vehicle type classification" and the specified value in each vehicle type block of length 40;

[0081] Formula 8: It represents the sum of the square differences between the absolute value of the sequence difference of the same order in two workshops and the specified value;

[0082] Formula 9: Z i,k <N i,max , N i,max Indicates the maximum vehicle type cluster length of the i-th “vehicle type classification”;

[0083] Formula 10: Z i,k >N i,min , N i,min Indicates the minimum vehicle type block length of the i-th “vehicle type classification”;

[0084] Equations 9 and 10 indicate that the length of each “vehicle type classification” must be within the specified interval;

[0085] Formula 11: This means that in the final production scheduling result, vehicles with different "configurations" cannot be scheduled in an intersecting manner, and there is a clear scheduling order;

[0086] Formula 12: M-1≤M i,m ≤M+1, indicating that in each car type block of length 40, the number of cars of a specified "car type classification" must be in the specified range;

[0087] Formula 13: |A i -Bi |≤index, indicating that the sequence difference of each order in the production scheduling results of the general assembly workshop and the welding workshop shall not exceed the specified value.

[0088] The parameters have the following meanings:

[0089] n represents the order quantity; X i Indicates the position of the i-th order in the production scheduling result; L i represents the production location specified by the i-th order; Y k Indicates the length of the block of the k-th vehicle type with "power type" as "pure electric"; Z i,k N represents the length of the kth cluster of vehicles in the i-th “vehicle type classification”; i It represents the length of the vehicle block specified by the i-th “vehicle type classification”; W i,last Indicates the index of the last vehicle in the i-th “configuration” in the production scheduling result; W i+1,first represents the index of the first vehicle in the i+1th “configuration”; M i,m A represents the number of vehicles of the mth “model classification” in the i-th model block of length 40; M represents the number of vehicles of the specified mth “model classification” in each model block of length 40; A i represents the index of the i-th order in the production scheduling result of the general assembly workshop; B i Represents the index of the i-th order in the production scheduling result of the welding workshop; index represents the specified sequence difference;

[0090] Step S300: Calculate the fitness value function, introduce an iterative algorithm model into the traditional genetic algorithm, perform multi-objective genetic algorithm iterative optimization through a non-dominated sorting method, and update the order sequence;

[0091] Step S400: Determine whether the maximum number of iterations has been reached or the optimal solution of the target fitness function has been obtained:

[0092] The process terminates if the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained;

[0093] If the maximum number of iterations is not reached or the optimal solution of the target fitness function is not obtained, step S300 is repeated until the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained.

[0094] The iteration of the iterative algorithm of step S300 and step S400 is as follows: Figure 2 shown.

[0095] Example:

[0096] Taking the actual order data provided by an automobile manufacturer as an example, 400 orders, 800 orders, 1200 orders, and 1600 orders were used as test data respectively. The method was tested in the PTYHON 3.8 environment on a workstation equipped with an Intel(R) Core(TM) i7-14700k processor, 32G memory, and Intel(R) UHD Graphics 770.

[0097] Under the four order sizes, the distribution of solutions at the beginning of the iteration and after the iteration is completed and the Pareto front. The red dot represents the Pareto front. Figures 3 to 6 As shown, we can know that:

[0098] ① For order data of different scales, this method can obtain good results, that is, this method achieves optimization in solution quality;

[0099] ② In addition to the quality of the solution, the method invention also has good stability and can obtain a better solution more quickly for order data of different sizes.

[0100] For actual automobile production and manufacturing scenarios, due to the constraints of block length and order sequence difference, most of the existing methods cannot guarantee a high constraint satisfaction rate. When solving the same mathematical model, the constraint satisfaction rate can often only reach 73%-85%. It has been verified that the method of the present invention can reach 83% to 93%.

[0101] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the technology may modify or change the above embodiments without violating the spirit and scope of the present invention. Therefore, all equivalent modifications or changes made by those with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed by the present invention should still be covered by the claims of the present invention.

Claims

1. A method for scheduling automobile orders taking into account a buffer zone, characterized in that: The following steps are involved: Step S100: Initialize parameters and prioritize order scheduling constraints according to order scheduling constraint information; Step S200: constructing an iterative algorithm model according to the priority order of the order scheduling constraints, and generating an initial population using integer coding; Step S300: Calculate the fitness value function, introduce an iterative algorithm model into the traditional genetic algorithm, perform multi-objective genetic algorithm iterative optimization through a non-dominated sorting method, and update the order sequence; Step S400: Determine whether the maximum number of iterations has been reached or the optimal solution of the target fitness function has been obtained: The process terminates if the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained; If the maximum number of iterations is not reached or the optimal solution of the target fitness function is not obtained, step S300 is repeated until the maximum number of iterations is reached or the optimal solution of the target fitness function is obtained.

2. The method for scheduling automobile orders taking buffer areas into consideration according to claim 1, characterized in that: The constraint information of step S100 includes but is not limited to the following six constraint conditions: ① Order location constraint: an order needs to be scheduled to a specified location; ② Assembly pure electric constraints, which means that all vehicles with "power type" as "pure electric" cannot be produced continuously; ③Continuous clustering constraint, which means that the clusters of different "car type classifications" must contain a specified number of cars; ④ Mixed scheduling constraint, which means that vehicles with different "configurations" must be scheduled in order; ⑤ Assembly balance constraint, which means that a certain model in "model classification" needs to be evenly scheduled for production; ⑥ Buffer area constraint means that after each order passes through the buffer area from the welding workshop to the final assembly workshop, the sequence difference must not exceed the specified value.

3. The method for scheduling automobile orders taking buffer areas into consideration according to claim 2, characterized in that: In step S200, the following iterative algorithm model (fitness functions F1 and F2) is constructed based on the six constraints for non-dominated sorting in a multi-objective genetic algorithm: Formula 1: min F1 = f1 + f6; Formula 2: min F2=f2+f3+f4+f5.

4. The method for scheduling automobile orders taking buffer areas into consideration according to claim 3, characterized in that: In the algorithm model, the mathematical model of each parameter is as follows: Formula 3: Represents the sum of squared differences between the actual position and the specified position of each order; Formula 4: Indicates the sum of the squared differences between the length of each agglomerate whose "power type" is "pure electric" and 1; Formula 5: Represents the sum of the squared differences between the length of each block of each "car type classification" and the specified value; Formula 6: It represents the sum of the difference between the index of the last car in the first "configuration" and the index of the first car in the second "configuration"; Formula 7: Indicates the sum of the squared differences between the number of vehicles of each "vehicle type classification" and the specified value in each vehicle type block of length 40; Formula 8: It represents the sum of the square differences between the absolute value of the sequence difference of the same order in two workshops and the specified value; in, n represents the order quantity; X i Indicates the position of the i-th order in the production scheduling result; L i represents the production location specified by the i-th order; Y k Indicates the length of the block of the k-th vehicle type with "power type" as "pure electric"; Z i,k N represents the length of the kth cluster of vehicles in the i-th "vehicle type classification"; i It represents the length of the vehicle type block specified by the i-th "vehicle type classification"; W i,last Indicates the index of the last vehicle in the i-th "configuration" in the production scheduling result; W i+1,first represents the index of the first vehicle in the i+1th "configuration"; M i,m A represents the number of vehicles of the mth "model classification" in the i-th model block of length 40; M represents the number of vehicles of the specified mth "model classification" in each model block of length 40; A i represents the index of the i-th order in the production scheduling result of the general assembly workshop; B i Represents the index of the i-th order in the production scheduling result of the welding workshop; index represents the specified sequence difference.

5. The method for scheduling automobile orders taking buffer areas into consideration according to claim 4, characterized in that: In the iterative algorithm model, the constraints of each parameter are as follows: Formula 9: Z i,k <N i,max , N i,max Indicates the maximum vehicle type cluster length of the i-th "vehicle type classification"; Formula 10: Z i,k >N i,min , N i,min Indicates the minimum vehicle type block length of the i-th "vehicle type classification"; Equations 9 and 10 indicate that the length of each "vehicle type classification" must be within the specified range; Formula 11: W i+1,first >W i,last , which means that in the final production scheduling result, vehicles with different "configurations" cannot be scheduled crosswise, and there is a clear scheduling order; Formula 12: M-1≤M i,m ≤M+1, indicating that in each car type block of length 40, the number of cars of a specified "car type classification" must be within the specified range; Formula 13: |A i -B i |≤index, indicating that the sequence difference of each order in the production scheduling results of the general assembly workshop and the welding workshop shall not exceed the specified value.

Images