Automobile workshop collaborative production scheduling method based on migrant bird optimization algorithm and production scheduling rule

By combining migratory bird optimization algorithm and heuristic rules, a multi-association workshop collaborative production scheduling method was designed, which solved the problem of random ordering capabilities of the three-dimensional library and the difference in body production sequence in the multi-association workshop collaborative production scheduling problem, and realized the production line The optimization of production shutdown time and cost has improved the speed of production scheduling decisions and the global optimization capabilities of the algorithm.

CN119047747APending Publication Date: 2024-11-29YANCHENG INST OF TECH
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411064584.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing technology is difficult to effectively solve the problem of collaborative production scheduling in multiple related vehicle workshops, especially when considering the random ordering capabilities of three-dimensional warehouses and the demand for car body production sequences in different workshops, it is difficult to minimize the production line production time and cost.

Method used

The collaborative production scheduling method of vehicle-wide vehicle multi-related workshops based on migratory bird optimization algorithm and heuristic production scheduling rules is adopted. Through statistics of customer orders, migratory bird coding strategies, heuristic decoding strategies, and the evolution strategy and cross-section of various neighborhood structures of the bird evolution strategy and the cross-section of the bird. Neighborhood search and the evolution strategy of flying birds, build mathematical models to optimize the production sequence of welding, coating and assembly workshops to meet the collaborative production needs of each workshop.

Benefits of technology

It achieves the production line suspension time and cost while meeting the production line production sequence needs of each workshop, improves the production schedule decision speed and the overall optimization ability of the algorithm, and reduces the production line suspension time and cost.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119047747A_ABST
    Figure CN119047747A_ABST
Patent Text Reader

Abstract

The invention discloses an automobile workshop collaborative production scheduling method based on a migrant bird optimization algorithm and a production scheduling rule. The method comprises the following steps: counting all customer order information in a production scheduling period to form a customer order set; processing the customer order set by adopting a migrant bird coding strategy based on a vehicle type; an initial migrant bird population is generated in a random mode, and a decoding strategy based on a heuristic rule is adopted to quickly obtain a collaborative production scheduling file corresponding to each migrant bird individual and evaluate the collaborative production scheduling file; adopting a bird flock V-shaped queue construction strategy based on non-dominated sorting to determine the construction of a leading bird group and a left and right following bird group; a leading bird evolutionary strategy based on multiple neighborhood structures and a following bird evolutionary strategy fusing crossover and neighborhood search are adopted to realize cyclic evolution of the bird flock; and outputting a collaborative production scheduling scheme corresponding to the leading bird when the tour times are reached, otherwise, re-constructing the leading bird and the left and right flying bird groups. According to the method, the decision speed is high, and a better collaborative production scheduling scheme can be obtained.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of automobile production scheduling, and in particular to a collaborative production scheduling method for automobile workshops based on a migratory bird optimization algorithm and production scheduling rules. Background Art

[0002] The automobile production line consists of four major process workshops: stamping, welding, painting and assembly. Since the stamping workshop only produces various body covering parts such as doors, side panels, hatches, and roofs, the stamping workshop has little correlation with other workshops. The body begins to take shape in the welding workshop and is given a unique body identification code. Therefore, the three major workshops of welding, painting and assembly constitute an interrelated vehicle assembly line. However, due to the differences in the optimization goals and process constraints of each workshop, each workshop has its own production scheduling needs. Studying the production scheduling model and algorithm that meets the body sequence requirements of each workshop is a hot and difficult issue in the current research field of automobile production workshop management. In recent years, many researchers at home and abroad have designed corresponding production scheduling algorithms based on the production scheduling needs of each workshop. However, most studies focus on a single workshop, establish a production scheduling model and algorithm for a single workshop, and ignore the correlation between workshops.

[0003] Considering the correlation and collaborative production relationship between the three major workshops, some researchers have studied the joint production scheduling problem of multiple workshops. However, some studies assume that each workshop produces according to the same body sequence, but in actual production, in order to better meet the production process requirements of each workshop, each workshop adopts its own production scheduling sequence. The difference in the production scheduling sequence of each workshop requires the setting of corresponding body reordering buffers between workshops. At present, the solution commonly adopted by major automobile companies' production lines is to set up a white body storage area (WBS) between the welding workshop and the painting workshop, and a painted body storage area (PBS) between the painting workshop and the general assembly workshop. Therefore, the collaborative production scheduling problem of multiple workshops not only needs to comprehensively consider the differences in production scheduling requirements of different workshops, but also needs to consider the body reordering ability of the buffer area. At present, some researchers have designed corresponding scheduling model methods for the collaborative production scheduling problem of multiple related workshops with linear buffer areas. However, these studies are mostly aimed at traditional linear WBS and PBS. Although the construction cost of traditional linear WBS and PBS is low, they also have the disadvantages of large footprint and weak reordering ability. In recent years, in order to accelerate the realization of multi-variety, small-batch mixed-flow production mode, more and more automobile companies have begun to introduce random storage buffers composed of vertical warehouses and stackers to replace traditional linear WBS and PBS. This type of random storage buffer not only occupies a small area, but also has a strong random reordering capability. However, the introduction of strong random reordering capabilities not only enhances the reordering capability, but also increases the solution space scale of the multi-related workshop collaborative production scheduling problem. It is necessary to design a corresponding collaborative production scheduling method. However, there is currently a lack of multi-related workshop collaborative production scheduling models and methods that consider the random reordering capabilities of vertical warehouses. To this end, in order to meet the different body production sequence requirements of automobile welding, painting and assembly workshops, the present invention takes minimizing the production line downtime and cost as the comprehensive optimization goal, considers the random reordering range constraints of the three-dimensional warehouse between each workshop on the white body and the color body, and establishes a mathematical model for the collaborative scheduling optimization of multiple related workshops of automobiles. It integrates the advantages of the improved migratory bird optimization algorithm and the heuristic scheduling rules, and designs a collaborative scheduling method for multiple related workshops of automobiles. Summary of the invention

[0004] To solve the above problems, the present invention provides a collaborative production scheduling method for multiple associated workshops of complete vehicles of automobiles that integrates an improved migratory bird optimization algorithm and heuristic scheduling rules, aiming at the production scheduling optimization problem of multiple associated workshops of complete vehicles of automobiles with random reordering capability of three-dimensional warehouses.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] The collaborative production scheduling method for automobile workshops based on migratory bird optimization algorithm and production scheduling rules includes the following steps:

[0007] All customer order information within the production scheduling cycle is counted to form a customer order set, where the order information includes model, configuration, color and delivery period;

[0008] A migratory bird coding strategy based on vehicle type is used to process customer order sets;

[0009] The initial migratory bird population is generated randomly, and the collaborative production files corresponding to each migratory bird individual are quickly obtained and evaluated using a decoding strategy based on heuristic rules.

[0010] The V-shaped formation construction strategy of bird flock based on non-dominated sorting is adopted to determine the construction of the leading bird and the left and right following bird flocks;

[0011] The leader-flying bird evolution strategy based on multiple neighborhood structures and the follow-flying bird evolution strategy integrating crossover and neighborhood search are used to realize the itinerant evolution of the bird flock.

[0012] When the number of rounds is reached, the collaborative production scheduling plan corresponding to the leading bird is output, otherwise the leading bird and the left and right following bird groups are rebuilt.

[0013] Furthermore, the migratory bird coding strategy based on vehicle models is as follows: the number of vehicles required to be produced for each vehicle model within the production cycle is counted, and the number of batches of each vehicle model within the production cycle is determined by rounding up the minimum batch size of the welding workshop. A complete migratory bird code is composed of the vehicle model codes, and the number of times each vehicle model code appears is the same as the batch number.

[0014] Furthermore, the decoding strategy based on heuristic rules specifically includes the following steps:

[0015] Input a customer order set, and use multiple heuristic rules to pre-sort the customer orders in sequence, where the specific pre-sorting rules are: pre-sort the customer orders for the first time according to the principle of first delivery first production; on the basis of the first pre-sorting, pre-sort the customer orders for the second time according to the principle of concentration of the same model; on the basis of the second pre-sorting, pre-sort the customer orders for the third time according to the principle of concentration of the same color within the same model; on the basis of the third pre-sorting, pre-sort the customer orders for the fourth time according to the principle of balanced low, medium and high configurations within the same color;

[0016] The production scheduling decision of the welding workshop is based on the minimum batch size of the welding workshop. As the batch upper limit, the model sequence is given according to the migratory bird individual code, and the orders of the corresponding models are taken out from the orders after the fourth pre-sorting. The obtained order sequence is directly used as the production sequence of the welding workshop

[0017] Production scheduling decision of paint shop, with the maximum adjustment range of body-in-white in stereoscopic warehouse As the upper limit of the adjustment sequence of the body in white in the three-dimensional warehouse, the production sequence of the welding workshop is collected according to the principle of concentration of the same color Reorder to obtain the production sequence set for the paint shop

[0018] The production scheduling decision of the assembly workshop is based on the maximum adjustment range of the color car body in the three-dimensional warehouse As the upper limit of the color car body adjustment sequence in the three-dimensional warehouse, the production sequence set of the paint shop is arranged according to the priority order of configuration {L, M, H} Reorder to obtain the production sequence set for the final assembly workshop Among them, H represents high configuration, M represents medium configuration, and L represents low configuration;

[0019] According to the production sequence of the three major workshops of welding, painting and assembly and Calculate the total downtime and total cost increase to evaluate migratory birds, and the decoding is completed.

[0020] Furthermore, the evolutionary strategy of the leading flyer based on multiple neighborhood structures is specifically: designing a neighborhood structure based on inversion, a neighborhood structure based on single-point mutation, and a neighborhood structure based on right-ring shift, and randomly generating neighborhood solutions of the leading flyer through the neighborhood structure to form an evolutionary solution set of the leading flyer; among them, if there are neighborhood migratory birds that dominate the leading flyer in the evolutionary solution set, then a neighborhood migratory bird that dominates the leading flyer is randomly selected to replace the leading bird; if there are no neighborhood migratory birds that dominate the leading flyer in the evolutionary solution set, the leading bird will not be updated.

[0021] Among them, the neighborhood structure based on reverse order is specifically defined as follows: randomly select two positions POS1 and POS2 in the original migratory bird code according to the code length, rearrange the vehicle model code between the original migratory bird code POS1 and POS2 in reverse order, and obtain the neighborhood migratory bird based on reverse order;

[0022] The neighborhood structure based on single-point mutation is specifically defined as follows: randomly select two positions POS1 and POS2 with different vehicle models in the original migratory bird code, swap the vehicle models corresponding to the original migratory bird codes POS1 and POS2, and obtain the neighborhood migratory bird based on single-point mutation;

[0023] The neighborhood structure based on right ring shift is specifically defined as follows: the lowest bit of the original migratory bird model code is moved to the highest bit, and the remaining position codes are shifted right by one bit in turn to obtain the neighborhood migratory bird based on the right ring shift of 1 bit; wherein, each right ring shift of 1 bit obtains a neighborhood migratory bird of the original migratory bird, and for a length of N B The migratory bird code can be obtained by right-shifting at most (N B -1) neighboring migratory birds.

[0024] Furthermore, the following bird evolution strategy integrating crossover and neighborhood search is as follows: for each following bird in the left and right following bird groups, their own evolutionary solution set is constructed, and the individuals in the evolutionary solution set include:

[0025] The shared solution provided by the preceding migratory bird: if it is the first following bird, the shared solution is the suboptimal solution in the set of solutions of the leading bird; if it is not the first following bird, the shared solution is the suboptimal solution in the set of solutions of the previous following bird;

[0026] Neighboring migratory birds of the following flying bird: According to the designed neighborhood structure based on inversion, neighborhood structure based on single point mutation and neighborhood structure based on right ring shift, six neighboring migratory birds of the following flying bird are randomly generated;

[0027] New migratory birds generated by crossing with following birds: Randomly select the following bird that is only in front of the following bird. Each selected front following bird can generate a new migratory bird with the following bird by crossing. The crossing method is as follows: first, the codes of the original following bird and a front following bird participating in the crossing are converted into codes with sequences according to the number of times each vehicle type appears in the code; second, randomly generate two positions POS1 and POS2, and rearrange the codes between the two positions of the original following bird according to their order in the front following birds to obtain a new code of the following bird with sequence; finally, remove the sequence in the code to obtain the new following bird after crossing;

[0028] Each individual in the evolutionary solution set is decoded and evaluated. If the optimal solution in the evolutionary solution set dominates the original flying bird, the optimal solution replaces the original flying bird.

[0029] Furthermore, in the collaborative scheduling scheme, the scheduling rules include: mathematical model decision variables, optimization objectives and constraints;

[0030] The decision variables of the mathematical model include: the production sequence number set of the welding workshop Paint shop production sequence number collection and the assembly workshop production sequence number collection

[0031] The optimization goals include: welding workshop optimization goal, painting workshop optimization goal, assembly workshop optimization goal, customer order delivery on-time optimization goal and overall optimization goal;

[0032] The constraints include: welding shop constraints, painting shop constraints, assembly shop constraints and three-dimensional warehouse constraints.

[0033] Furthermore, among the optimization objectives, the optimization objectives of the welding workshop include: minimizing the downtime caused by vehicle model switching and minimizing the increased costs due to vehicle model switching;

[0034] Among them, the specific formula for minimizing the downtime caused by vehicle model switching is:

[0035]

[0036] The specific formula to minimize the increased cost due to vehicle model switching is:

[0037]

[0038] Among them, θ W Indicates the downtime caused by a single vehicle model change in the welding workshop. Indicates the number of vehicle model switching times in the production sequence of the welding workshop, μ W It represents the additional production cost of a single vehicle model switch in the welding workshop;

[0039] Paint shop optimization goals include: minimizing downtime caused by spray gun and pipe cleaning and reducing the increased costs due to spray gun cleaning;

[0040] The specific formula for minimizing downtime caused by spray gun and pipeline cleaning is:

[0041]

[0042] The specific formula to reduce the cost of spray gun cleaning is:

[0043]

[0044] Among them, θ C Indicates the downtime caused by a single color switch in the paint shop, a C Indicates the cost of a single color change in a paint shop. Indicates the number of times the spray gun is cleaned when the paint shop continuously sprays the same color car body, Q i and Q j Respectively represent the order information of the i-th and j-th customers, Ω Q represents a collection of customer orders, and Represents customer orders Q i and Q j The production sequence number in the paint shop, SP i,j It is used to indicate whether the colors of adjacent car bodies in the paint shop are the same. 1 indicates different colors, otherwise 0;

[0045] The optimization goals of the general assembly shop include: reducing the amount of rest time arrears and reducing material supply costs;

[0046] Among them, the rest time arrears generated after a certain workstation in the general assembly workshop assembles the i-th vehicle is The calculation formula is:

[0047]

[0048] Among them, T as It indicates the working hours required for assembly work at a certain workstation in the general assembly workshop. represents the average rest time required by workers at each workstation in the final assembly workshop, and T represents the production rhythm;

[0049] The specific formula for optimizing the goal of reducing rest time arrears is:

[0050]

[0051] Among them, a f N represents the average cost of repairing each order in the general assembly workshop. Q Indicates the total number of customer orders, It represents the average rest time arrears of all orders. Indicates the repair rate of the general assembly workshop, which is related to the amount of rest time arrears. and Respectively expressed as:

[0052]

[0053] The specific formula for the optimization goal of reducing material supply cost is:

[0054]

[0055] in, Table 1. Average material supply cost of the final assembly workshop. Represents the standard deviation of the average consumption rate of all weighted items, specifically defined as:

[0056]

[0057] Among them, N J Indicates the number of types of key and important parts. For all customer orders in a production batch, the key parts J j The average consumption rate, The first r production sequence of the assembly workshop is the key component J j The average consumption rate is calculated as follows:

[0058]

[0059] Among them, x n,j It is used to indicate whether the vehicle with the production sequence number n in the final assembly workshop is equipped with key and important parts;

[0060] The optimization objectives for on-time delivery of customer orders include: minimizing the penalty for delay of all customer orders;

[0061] Among them, the penalty for delay The specific formula is:

[0062]

[0063] Among them, λ represents the penalty payment coefficient for order delay, Represents customer order Q i The contractual delivery time is Represents customer order Q i The actual delivery time;

[0064] The specific formula to minimize the penalty for late payment of all customer orders is:

[0065]

[0066] The overall optimization objectives include minimizing the total downtime of welding and painting shops and minimizing the total cost increase;

[0067] The specific formula for minimizing the total downtime of welding and painting workshops is:

[0068] f A =f 1,1 +f 2,1 (15)

[0069] The specific formula for minimizing the total cost increase is:

[0070] f B =f 1,2 +f 2,2 +f 3,1 +f 3,2 +f 4 (16).

[0071] Furthermore, among the constraints, the constraints for the welding workshop are:

[0072]

[0073] in, and Represents customer orders Q i and Q k The production sequence number in the welding workshop, formula (17) indicates that the production sequence number of any customer order in the welding workshop must be unique; formula (18) indicates that the production sequence numbers of any two customer orders in the welding workshop are different;

[0074] The constraints for the paint shop are:

[0075]

[0076] in, and Represents customer orders Q i and Q k The production sequence number in the paint shop, formula (19) indicates that the production sequence number of any customer order in the paint shop must be unique; formula (20) indicates that the production sequence numbers of any two customer orders in the paint shop are different;

[0077] The constraints of the general assembly workshop are:

[0078]

[0079] in, and Represents customer orders Q i and Q k The production sequence number in the assembly workshop, formula (21) indicates that the production sequence number of any customer order in the paint shop must be unique; formula (22) indicates that the production sequence numbers of any two customer orders in the paint shop should be different;

[0080] The constraints of the three-dimensional library are:

[0081]

[0082] in, Indicates the parking capacity of the stereo garage. Indicates the maximum range of the BIW reordering in the stereo library. It indicates the maximum range of color car body reordering in the three-dimensional warehouse.

[0083] The meanings of the parameters are as follows:

[0084]

[0085]

[0086] Compared with the prior art, the present invention has the following beneficial effects:

[0087] 1. The method provided by the present invention takes into account the optimization objectives and constraints in terms of automobile whole vehicle, welding, painting, final assembly, stereoscopic warehouse sequencing capability, order delivery timeliness, etc., and can meet the different body production sequence requirements of automobile whole vehicle welding, painting and final assembly workshops, and reduce production line downtime and cost.

[0088] 2. The present invention designs a decoding strategy based on heuristic scheduling rules, which can take advantage of the fast decision-making speed of heuristic rules, quickly obtain the collaborative scheduling plan of multiple associated workshops corresponding to migratory bird individuals and realize the evaluation of migratory bird individuals, thereby shortening the scheduling decision time.

[0089] 3. In order to address the defect of the traditional migratory bird optimization algorithm that the single neighborhood structure leads to slow convergence of the algorithm, the present invention designs a leading bird evolution strategy based on multiple neighborhood structures. In order to improve the global optimization ability of the algorithm, a following bird evolution strategy that integrates crossover and neighborhood search is designed, so that the improved migratory bird optimization algorithm has the characteristics of fast convergence speed and strong global optimization ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0091] Figure 1 is an overall flow chart of an embodiment of the present invention;

[0092] Figure 2 is a flowchart of a decoding strategy decision based on heuristic rules according to an embodiment of the present invention;

[0093] Figure 3 is a schematic diagram of production scheduling decision of a welding workshop according to an embodiment of the present invention;

[0094] Figure 4 is a schematic diagram of a production scheduling sequence decision for a paint shop according to an embodiment of the present invention;

[0095] Figure 5 is a schematic diagram of production scheduling decision of an assembly shop in an embodiment of the present invention;

[0096] Figure 6 It is a schematic diagram of neighborhood migratory birds based on reverse order according to an embodiment of the present invention.

[0097] Figure 7 is a schematic diagram of neighborhood migratory birds based on single-point mutation according to an embodiment of the present invention;

[0098] Figure 8 This is a schematic diagram of migratory birds in a neighborhood based on a right ring shift of 1 bit according to an embodiment of the present invention;

[0099] Fig. 9 It is a schematic diagram of generating new migratory birds by crossing with flying birds according to an embodiment of the present invention. DETAILED DESCRIPTION

[0101] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0102] Coding strategy based on vehicle model:

[0103] In order to minimize the number of vehicle model switching in the welding workshop, a coding method based on vehicle model is adopted. Taking the customer order scheduling in a certain production cycle as an example, the specific steps are as follows: First, the number of vehicles required to be produced for each model in the production cycle is counted. Secondly, the initial As the production batch of all car bodies in the welding workshop, according to the batch The number of batches of each model in the production cycle is determined by rounding up. Finally, the production code is randomly generated according to the model and batch number. In the code, the number of times each model appears is the same as the batch number. Taking the production of three models as an example, assuming that the customer order information in the production cycle is shown in Table 1, the number of bodies, batches and batches required for each model in the production cycle are shown in Table 2. A feasible code is: G i ={A,B,A,C,B,A,C}. This code indicates that the welding workshop can be put into production smoothly as follows: 3 Type A cars - 3 Type B cars - 3 Type A cars - 3 Type C cars - 1 Type B car - 2 Type A cars - 3 Type C cars.

[0104] Table 1 Customer order information list

[0105]

[0106]

[0107] Table 2 Number of bodies, batches and lot information required for each model

[0108]

[0109] Initialization of migratory bird population:

[0110] Generation of initial migratory bird population:

[0111] Assume that the size of the migratory bird population is N POP , then N is randomly generated directly according to the number of batches of each model counted in the production scheduling cycle POP different codes, which constitute the initial migratory bird population Ω={G i |i∈N and 1≤i≤N POP}, each migratory bird individual code G i The number of times each model code appears is consistent with the batch number of that model.

[0112] Decoding and evaluation of individual migratory birds:

[0113] Due to the large scale of the collaborative scheduling optimization problem of multiple related workshops in the whole vehicle, in order to improve the decoding speed, the following Figure 2 The individual heuristic decoding and evaluation process of migratory birds is shown in G i={A,B,A,C,B,A,C} as an example, combined with Table 1 and Table 2, the decoding and evaluation process is specifically described:

[0114] Step 1: Input the customer order set, and use multiple heuristic rules from the sorting rule library to pre-sort the customer orders. First, in order to improve the on-time delivery rate of orders and reduce the default costs of order delays, the customer orders are pre-sorted for the first time according to the principle of first delivery first production; secondly, in order to reduce the number of model changes in the welding workshop, reduce the total downtime and reduce costs, on the basis of the first pre-sorting, the customer orders are pre-sorted for the second time according to the principle of the same model concentration; in order to reduce the number of color changes in the painting workshop, reduce the total downtime and reduce costs, on the basis of the second pre-sorting, the customer orders are pre-sorted for the third time according to the principle of the same color concentration within the same model; finally, in order to balance the load of the general assembly workshop, on the basis of the third pre-sorting, the customer orders are pre-sorted for the fourth time according to the principle of balanced low, medium and high configurations within the same color. After completing the above four pre-sortings in sequence, enter Step 2. Taking the order shown in Table 1 as an example, the results after each pre-sorting are shown in Table 3.

[0115] Table 3 Pre-sorting results of customer order sets each time

[0116] Pre-sorting for several times The order number sequence of each pre-sorted customer order first 1,7,9,11,15,17,2,5,8,10,13,16,18,3,4,6,12,14 Second time 1,7,11,5,10,13,16,4,17,2,8,14,9,15,18,3,6,12 The third time 1,11,13,16,7,5,10,4,17,2,8,14,9,15,18,3,6,12 Fourth 13,1,16,11,7,4,5,10,2,17,8,14,15,9,18,3,6,12

[0117] Step 2: Production scheduling decision for welding workshop. As the batch upper limit, the model sequence is given according to the migratory bird individual code, and the orders of the corresponding models are taken out from the orders after the fourth pre-sorting. The obtained order sequence is directly used as the production sequence of the welding workshop After completion, go to Step 3. Taking the pre-sorting results shown in Table 3 as an example, assuming that the migratory bird individual code is G i ={A,B,A,C,B,A,C}, The schematic diagram of the production sequence decision of the welding workshop is as follows Figure 3 shown.

[0118] Step 3: Production scheduling decision for the paint shop. As the upper limit of the adjustment sequence of the body in white in the three-dimensional warehouse, the production sequence of the welding workshop is collected according to the principle of concentration of the same color Reorder to obtain the production sequence set for the paint shop After completion, go to Step 4. Figure 3 The welding shop production sequence shown and For example, the production sequence decision diagram of the paint shop is as follows Figure 4 Take customer order 8 as an example. Its production sequence number in the welding workshop is 6, while its production sequence number in the painting workshop is 4, and it needs to be adjusted forward by 2 positions.

[0119] Step 4: Production scheduling decision for the final assembly workshop. As the upper limit of the color car body adjustment sequence in the three-dimensional warehouse, the production sequence set of the paint shop is arranged according to the priority order of configuration {L, M, H} Reorder to obtain the production sequence set for the final assembly workshop After completion, go to Step 5. Figure 4 The paint shop production sequence shown and For example, the production sequence decision diagram of the assembly workshop is as follows Figure 5 Take customer order 11 as an example. Its production sequence number in the paint shop is 5, while its production sequence number in the assembly shop is 4, and it needs to be adjusted forward by 1.

[0120] Step 5: According to the production sequence of the three major workshops of welding, painting and assembly and Calculate the total downtime and total cost added value (referred to as total cost) to evaluate the migratory birds, and the decoding is completed.

[0121] Firstly, the designed decoding method is used to decode the migratory bird population Ω={G i |i∈N and 1≤i≤N POP} to decode and evaluate all migratory birds and obtain each migratory bird individual G k The corresponding total downtime f A (G k ) and the total cost f B (G k ). Then, the non-dominated sorting method is used to determine the sorting level and crowding degree of each migratory bird individual. The lower the sorting level, the better the migratory bird individual. Finally, the migratory bird with the highest crowding degree in the non-dominated front is selected from the migratory bird population as the leader, and the remaining migratory birds are evenly distributed to the left and right follower bird populations according to the sorting level. The lower the sorting level, the closer the follower bird is to the leader. The same sorting level means that the more crowded the follower bird is, the closer it is to the leader. Finally, the V-shaped queue of the migratory bird population is constructed.

[0122] In order to avoid the single evolution direction of the leader bird caused by a single neighborhood structure and speed up the convergence of the algorithm, this paper designs a leader bird neighborhood search based on multiple neighborhood structures. The three neighborhood structures are:

[0123] Neighborhood migratory birds based on reverse order:

[0124] The schematic diagram of migratory birds in the neighborhood based on reverse order is as follows Figure 6 As shown, two positions POS in the original migratory bird code are randomly selected according to the code length 1 and POS 2 , the original migratory bird code POS 1and POS 2 By rearranging the vehicle model codes in reverse order, a neighborhood migratory bird based on reverse order can be obtained.

[0125] Neighborhood structure based on single-point mutation:

[0126] The schematic diagram of the neighborhood migratory birds based on single point mutation is as follows Figure 7 As shown, two positions POS with different vehicle models in the original migratory bird code are randomly selected 1 and POS 2 , the original migratory bird code POS 1 and POS 2 By exchanging the positions of the corresponding models, a neighborhood migratory bird based on single-point mutation can be obtained.

[0127] Neighborhood structure based on right ring shift:

[0128] The schematic diagram of the neighborhood migratory birds based on the right ring shift of 1 is as follows Figure 8 As shown in the figure, the lowest vehicle model code of the original migratory bird is moved to the highest position, and the remaining position codes are shifted right by one position in turn, and the neighboring migratory birds based on the right ring shift of 1 position can be obtained. Each right ring shift of 1 position can obtain a neighboring migratory bird of the original migratory bird, so the maximum number of migratory birds that can be obtained by right ring shift is (N B -1) neighboring migratory birds, where N B Length of the code for migratory birds.

[0129] The N number of leading birds is randomly generated by the above three neighborhood structures. N The neighborhood solutions constitute the evolutionary solution set of the leading flyer, and the following strategy is used to update the neighborhood solutions: if there is a neighborhood migratory bird that dominates the leading flyer in the evolutionary solution set, a neighborhood migratory bird that dominates the leading flyer is randomly selected to replace the leading flyer; if there is no neighborhood migratory bird that dominates the leading flyer in the evolutionary solution set, the leading flyer is not updated.

[0130] Traditional following bird evolution realizes the evolution of following birds based on the neighborhood shared by the previous order, which will make the bird group tend to be consistent, causing the algorithm to easily fall into the local optimum. In order to improve this defect, this paper designs a following bird evolution strategy that combines crossover and neighborhood search. The details are as follows: For each following bird in the left and right following bird groups, it is necessary to construct their own evolutionary solution set. The individuals in the evolutionary solution set are composed of three parts:

[0131] 1. The shared solution provided by the preceding migratory bird: If it is the first following bird, the shared solution is the suboptimal solution in the set of solutions evolved by the leading bird; if it is not the first following bird, the shared solution is the suboptimal solution in the set of solutions evolved by the previous following bird.

[0132] 2. Neighboring migratory birds of the flying bird: Generate 6 random neighboring migratory birds of the flying bird according to the three neighborhood structures defined in 2.4.

[0133] 3. New migratory birds generated by crossing with flying birds: randomly select N P The following bird (front following bird) located in front of the following bird, each selected front following bird can generate a new migratory bird by crossing with the following bird. The schematic diagram of generating a new migratory bird by crossing the following bird is as follows Fig. 9 As shown, first, the original heel bird and the front heel bird codes participating in the crossover are converted into codes with sequences according to the number of times each model appears in the code. Secondly, two positions POS are randomly generated. 1 and POS 2 , rearrange the codes between the two positions of the original follow-up bird according to their order in the previous follow-up bird, obtain the new follow-up bird code with sequence after the updated position, and finally remove the sequence in the code to obtain the new follow-up bird after crossing.

[0134] Each individual in the evolutionary solution set is decoded and evaluated. If the optimal solution in the evolutionary solution set dominates the original flying bird, the optimal solution replaces the original flying bird.

[0135] The examples described in the present invention are merely descriptions of the preferred implementation modes of the present invention, and are not intended to limit the concept and scope of the present invention. Without departing from the design concept of the present invention, various modifications and improvements made to the technical solutions of the present invention by engineers and technicians in this field should all fall within the protection scope of the present invention.

Claims

1. A collaborative production scheduling method for automobile workshops based on migratory bird optimization algorithm and production scheduling rules, characterized in that: The steps include: All customer order information within the production scheduling cycle is counted to form a customer order set, where the order information includes model, configuration, color and delivery period; A migratory bird coding strategy based on vehicle type is used to process customer order sets; The initial migratory bird population is generated randomly, and the collaborative production files corresponding to each migratory bird individual are quickly obtained and evaluated using a decoding strategy based on heuristic rules. The V-shaped formation construction strategy of bird flock based on non-dominated sorting is adopted to determine the construction of the leading bird and the left and right following bird flocks; The leader-flying bird evolution strategy based on multiple neighborhood structures and the follow-flying bird evolution strategy integrating crossover and neighborhood search are used to realize the itinerant evolution of the bird flock. When the number of rounds is reached, the collaborative production scheduling plan corresponding to the leading bird is output, otherwise the leading bird and the left and right following bird groups are rebuilt.

2. The method according to claim 1, characterized in that The migratory bird coding strategy based on vehicle models is as follows: the number of vehicles required to be produced for each vehicle model within the production cycle is counted, and the number of batches of each vehicle model within the production cycle is determined by rounding up the minimum batch size of the welding workshop. A complete migratory bird code is composed of the vehicle model codes, and the number of times each vehicle model code appears is the same as the batch number.

3. The method according to claim 2, characterized in that The decoding strategy based on heuristic rules includes the following steps: Input a customer order set, and use multiple heuristic rules to pre-sort the customer orders in sequence, where the specific pre-sorting rules are: pre-sort the customer orders for the first time according to the principle of first delivery first production; on the basis of the first pre-sorting, pre-sort the customer orders for the second time according to the principle of concentration of the same model; on the basis of the second pre-sorting, pre-sort the customer orders for the third time according to the principle of concentration of the same color within the same model; on the basis of the third pre-sorting, pre-sort the customer orders for the fourth time according to the principle of balanced low, medium and high configurations within the same color; The production scheduling decision of the welding workshop is based on the minimum batch size of the welding workshop. As the batch upper limit, the model sequence is given according to the migratory bird individual code, and the orders of the corresponding models are taken out from the orders after the fourth pre-sorting. The obtained order sequence is directly used as the production sequence of the welding workshop Production scheduling decision of paint shop, with the maximum adjustment range of body-in-white in stereoscopic warehouse As the upper limit of the adjustment sequence of the body in white in the three-dimensional warehouse, the production sequence of the welding workshop is collected according to the principle of concentration of the same color Reorder to obtain the production sequence set for the paint shop The production scheduling decision of the assembly workshop is based on the maximum adjustment range of the color car body in the three-dimensional warehouse As the upper limit of the color car body adjustment sequence in the three-dimensional warehouse, the production sequence set of the paint shop is configured in the priority order of {L, M, H} Reorder to obtain the production sequence set for the final assembly workshop Among them, H represents high configuration, M represents medium configuration, and L represents low configuration; According to the production sequence of the three major workshops of welding, painting and assembly and Calculate the total downtime and total cost increase to evaluate migratory birds, and the decoding is completed.

4. The method according to claim 1, characterized in that The leader-flyer evolution strategy is based on a variety of neighborhood structures, specifically: designing a neighborhood structure based on inversion, a neighborhood structure based on single-point mutation, and a neighborhood structure based on right-ring shift. The neighborhood solutions of the leader-flyer are randomly generated through the neighborhood structure and constitute the evolutionary solution set of the leader-flyer. Among them, if there are neighborhood migratory birds that dominate the leader-flyer in the evolutionary solution set, a neighborhood migratory bird that dominates the leader-flyer is randomly selected to replace the leader-flyer. If there are no neighborhood migratory birds that dominate the leader-flyer in the evolutionary solution set, the leader-flyer will not be updated. Among them, the neighborhood structure based on reverse order is specifically defined as follows: randomly select two positions POS1 and POS2 in the original migratory bird code according to the code length, rearrange the vehicle model code between the original migratory bird code POS1 and POS2 in reverse order, and obtain the neighborhood migratory bird based on reverse order; The neighborhood structure based on single-point mutation is specifically defined as follows: randomly select two positions POS1 and POS2 with different vehicle models in the original migratory bird code, swap the vehicle models corresponding to the original migratory bird codes POS1 and POS2, and obtain the neighborhood migratory bird based on single-point mutation; The neighborhood structure based on right ring shift is specifically defined as follows: the lowest bit of the original migratory bird model code is moved to the highest bit, and the remaining position codes are shifted right by one bit in turn to obtain the neighborhood migratory bird based on the right ring shift of 1 bit; wherein, each right ring shift of 1 bit obtains a neighborhood migratory bird of the original migratory bird, and for a length of N B The migratory bird code can be obtained by right-shifting at most (N B -1) neighboring migratory birds.

5. The method according to claim 1, characterized in that The following bird evolution strategy that integrates crossover and neighborhood search is as follows: for each following bird in the left and right following bird groups, each following bird constructs its own evolutionary solution set, and the individuals in the evolutionary solution set include: The shared solution provided by the preceding migratory bird: if it is the first following bird, the shared solution is the suboptimal solution in the set of solutions of the leading bird; if it is not the first following bird, the shared solution is the suboptimal solution in the set of solutions of the previous following bird; Neighboring migratory birds of the following flying bird: According to the designed neighborhood structure based on inversion, neighborhood structure based on single point mutation and neighborhood structure based on right ring shift, six neighboring migratory birds of the following flying bird are randomly generated; New migratory birds generated by crossing with following birds: Randomly select the following bird that is only in front of the following bird. Each selected front following bird can generate a new migratory bird with the following bird by crossing. The crossing method is as follows: first, the codes of the original following bird and a front following bird participating in the crossing are converted into codes with sequences according to the number of times each vehicle type appears in the code; second, randomly generate two positions POS1 and POS2, and rearrange the codes between the two positions of the original following bird according to their order in the front following birds to obtain a new code of the following bird with sequence; finally, remove the sequence in the code to obtain the new following bird after crossing; Each individual in the evolutionary solution set is decoded and evaluated. If the optimal solution in the evolutionary solution set dominates the original flying bird, the optimal solution replaces the original flying bird.

6. The method according to claim 1, characterized in that In the collaborative scheduling scheme, the scheduling rules include: mathematical model decision variables, optimization objectives and constraints; The decision variables of the mathematical model include: the production sequence number set of the welding workshop Paint shop production sequence number collection and the assembly workshop production sequence number collection The optimization goals include: welding workshop optimization goal, painting workshop optimization goal, assembly workshop optimization goal, customer order delivery on-time optimization goal and overall optimization goal; The constraints include: welding shop constraints, painting shop constraints, assembly shop constraints and three-dimensional warehouse constraints.

7. The method according to claim 6, characterized in that Among the optimization objectives, the optimization objectives of the welding workshop include: minimizing the downtime caused by vehicle model switching and minimizing the increased costs due to vehicle model switching; Among them, the specific formula for minimizing the downtime caused by vehicle model switching is: The specific formula to minimize the increased cost due to vehicle model switching is: Among them, θ W Indicates the downtime caused by a single vehicle model change in the welding workshop. Indicates the number of vehicle model switching times in the production sequence of the welding workshop, μ W It represents the additional production cost of a single vehicle model switch in the welding workshop; Paint shop optimization goals include: minimizing downtime caused by spray gun and pipe cleaning and reducing the increased costs due to spray gun cleaning; The specific formula for minimizing downtime caused by spray gun and pipeline cleaning is: The specific formula to reduce the cost of spray gun cleaning is: Among them, θ C Indicates the downtime caused by a single color switch in the paint shop, a C Indicates the cost of a single color change in a paint shop. Indicates the number of times the spray gun is cleaned when the paint shop continuously sprays the same color car body, Q i and Q j Respectively represent the order information of the i-th and j-th customers, Ω Q represents a collection of customer orders, and Represents customer orders Q i and Q j The production sequence number in the paint shop, SP i,j It is used to indicate whether the colors of adjacent car bodies in the paint shop are the same. 1 indicates different colors, otherwise 0; The optimization goals of the general assembly shop include: reducing the amount of rest time arrears and reducing material supply costs; Among them, the rest time arrears generated after a certain workstation in the general assembly workshop assembles the i-th vehicle is The calculation formula is: Among them, T as It indicates the working hours required for assembly work at a certain workstation in the general assembly workshop. represents the average rest time required by workers at each workstation in the final assembly workshop, and T represents the production rhythm; The specific formula for optimizing the goal of reducing rest time arrears is: Among them, a f N represents the average cost of repairing each order in the general assembly workshop. Q Indicates the total number of customer orders, It represents the average rest time arrears of all orders. Indicates the repair rate of the general assembly workshop, which is related to the amount of rest time arrears. and Respectively expressed as: The specific formula for the optimization goal of reducing material supply cost is: in, Table 1. Average material supply cost of the final assembly workshop. Represents the standard deviation of the average consumption rate of all weighted items, specifically defined as: Among them, N J Indicates the number of types of key and important parts. For all customer orders in a production batch, the key parts J j The average consumption rate, The first r production sequence of the assembly workshop is the key component J j The average consumption rate is calculated as follows: Among them, x n,j It is used to indicate whether the vehicle with the production sequence number n in the final assembly workshop is equipped with key and important parts; The optimization objectives for on-time delivery of customer orders include: minimizing the penalty for delay of all customer orders; Among them, the penalty for delay The specific formula is: Among them, λ represents the penalty payment coefficient for order delay, Represents customer order Q i The contractual delivery time is Represents customer order Q i The actual delivery time; The specific formula to minimize the penalty for late payment of all customer orders is: The overall optimization objectives include minimizing the total downtime of welding and painting shops and minimizing the total cost increase; The specific formula for minimizing the total downtime of welding and painting workshops is: f A =f 1,1 +f 2,1 (15) The specific formula for minimizing the total cost increase is: f B =f 1,2 +f 2,2 +f 3,1 +f 3,2 +f4 (16)。 8. The method according to claim 6, characterized in that Among the constraints, the constraints for the welding workshop are: in, and Represents customer orders Q i and Q k The production sequence number in the welding workshop, formula (17) indicates that the production sequence number of any customer order in the welding workshop must be unique; formula (18) indicates that the production sequence numbers of any two customer orders in the welding workshop are different; The constraints for the paint shop are: in, and Represents customer orders Q i and Q k The production sequence number in the paint shop, formula (19) indicates that the production sequence number of any customer order in the paint shop must be unique; formula (20) indicates that the production sequence numbers of any two customer orders in the paint shop are different; The constraints of the general assembly workshop are: in, and Represents customer orders Q i and Q k The production sequence number in the assembly workshop, formula (21) indicates that the production sequence number of any customer order in the paint shop must be unique; formula (22) indicates that the production sequence numbers of any two customer orders in the paint shop should be different; The constraints of the three-dimensional library are: in, Indicates the parking capacity of the stereo garage. Indicates the maximum range of the BIW reordering in the stereo library. It indicates the maximum range of color car body reordering in the three-dimensional warehouse.

Citation Information

Cited By

  • Steel cutting and production scheduling method and device based on dynamic risk assessment

    CN120542877A