A flexible workshop dynamic batch scheduling method for quality inspection disturbance
Through the flexible workshop dynamic batch scheduling method and the improved Gray Wolf algorithm, the interference problem of quality inspection disturbance on workshop production is solved, and the timely response and energy consumption optimization of quality inspection disturbances in workshop production is achieved, and the production efficiency and resources are improved. Utilization.
Patent Information
- Application Number
- CN202210737349.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-27
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2042-06-27
AI Technical Summary
During the workshop production process, quality inspection disturbances and interference lead to the failure of static scheduling schemes, and the existing technology cannot respond in a timely manner and consider energy consumption indicators, resulting in production efficiency and resource waste.
A flexible workshop dynamic batch scheduling method is proposed. By establishing a mathematical model to optimize the objective function, combining the improved Gray Wolf algorithm for rescheduling and reorganizing batching, the quality inspection disturbance rescheduling strategy and reorganizing batching technology are designed to optimize the batch and batching of workpieces. Machine distribution reduces energy consumption and maintains scheduling stability.
It realizes timely response when quality inspection disturbances occur, shortens completion time, reduces machine tool energy consumption, and improves the optimization efficiency and stability of the scheduling plan. Compared with the existing technology, it has significant advantages in comprehensive cost and energy consumption.
Smart Images

Figure CN114995318B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of dynamic batch scheduling of flexible job shops, and in particular to a dynamic batch scheduling method of flexible job shops targeting quality inspection disturbances. Background Art
[0002] In the actual production process of the workshop, there are various dynamic emergencies, which will interfere with the static production scheduling plan before the start of work, making the operation plan out of touch with the actual production. Therefore, dynamic scheduling has great research significance. At the same time, since energy conservation and emission reduction have become strategic requirements for the manufacturing industry in various countries, manufacturing companies at home and abroad must consider energy consumption factors while pursuing economic benefits. Therefore, the research on workshop scheduling problems is not limited to indicators that reflect production benefits such as completion time, delivery period and machine utilization rate. More and more studies at home and abroad have considered energy consumption factors. Energy conservation and emission reduction and green manufacturing have become the focus of current and future research. Summary of the invention
[0003] In view of the problem of quality inspection disturbance interference in the actual production process of the workshop, the present invention provides a flexible workshop dynamic batch scheduling method for quality inspection disturbance.
[0004] A flexible workshop dynamic batch scheduling method for quality inspection disturbance of the present invention is specifically as follows:
[0005] Step 1: Determine the description of the flexible job shop dynamic batch scheduling problem.
[0006] Suppose there are M machine tools M={M k |k=1,2,…,m}, n types of workpieces J={J i |i=1,2,…,n}, the number of workpieces of the i-th category is R i , and each type of workpiece has O i = {O il |l=1,2,…,w i} process, and there are M machine tools that can process each process il , M il ∈M units, the processing time of each process varies with the performance of the machine tool; each type of workpiece is divided into multiple sub-batches F i ={F is |s=1,2,…,u i}Processed on different machines, each sub-batch is handled as a whole and occupies the same auxiliary time.
[0007] Step 2: Establish a mathematical model for the optimization of dynamic batch scheduling objectives in flexible job shops.
[0008] The objective functions are the workshop production cost, machine tool energy consumption cost and the difference cost of dynamic and static scheduling schemes, and the three objective function values are processed by weighted normalization.
[0009] Objective function:
[0010]
[0011]
[0012]
[0013] min{ω1·f1+ω2·f2+ω3·f3} (4)
[0014] Constraints:
[0015]
[0016] T=max(Z il ·E islk ) (6)
[0017] if S islk <T,DS islk -S islk =0 (7)
[0018]
[0019] Among them, C is represents the completion time of the sth batch of the i-th type of workpiece, t islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The single-piece working time on ilk Represents the lth process of the i-th type of workpiece on the machine tool M k Auxiliary time on S islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The start processing time on E islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The processing completion time on γ ilk represents the machine selection decision variable, when process O il Available on machine M k Processing time ilk =1, otherwise γ ilk =0, represents the auxiliary time decision variable. k If the category of the workpiece being processed on the machine tool is the same as the category of the last workpiece processed on the machine tool, on the contrary Q krepresents the maximum completion time of machine k, W k represents the load of machine k, PC k Indicates machine M k Average energy cost per unit time during load, NC k Indicates machine M k Average energy consumption cost per unit time when no load, RC k Indicates machine M k The average energy consumption cost per unit time during preparation, η represents the completion cost per unit time in the workshop, ω1 represents the completion cost weight coefficient, ω2 represents the energy consumption cost weight coefficient, ω3 represents the difference cost weight coefficient of dynamic and static scheduling schemes, DS islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the dynamic scheduling plan is on machine M k Start processing time on DW islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the dynamic scheduling plan is on machine M k The processing completion time on D islk Represents process O in the dynamic scheduling scheme il Machine M is still in the static scheduling scheme k During processing, D islk =0, otherwise D islk =1, S islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the static scheduling plan is on machine M k The start processing time on the ,ω3 represents the difference cost weight coefficient, V represents the difference unit cost of dynamic and static scheduling schemes, T represents the disturbance time node, DF is represents the sth subbatch of the i-th type of workpiece in the dynamic scheduling scheme, SF is represents the sth sub-batch of the i-th type of workpiece after the disturbance time node in the static scheduling scheme, R represents the number of new workpieces caused by the disturbance, and Z il Indicates process O il is the disturbance process, D il =1, otherwise D il =0.
[0020] Among them, formula (5) indicates that the processing completion time of a batch of workpieces in the dynamic scheduling scheme must be greater than or equal to the sum of the starting processing time, auxiliary time and working time of all workpieces in the batch; formula (6) indicates that the disturbance time node is the lth process of the sth batch of the i-th type of workpiece on machine M. k The processing completion time on the disturbance time node; Formula (7) indicates that the process before the disturbance time node and the process starting processing do not participate in dynamic scheduling; Formula (8) indicates that the total batch number of workpieces in the dynamic scheduling scheme is equal to the sum of the number of unprocessed workpieces at the disturbance time node and the number of newly added workpieces caused by the disturbance.
[0021] Step 3: Rescheduling for quality inspection disturbances.
[0022] The quality inspection disturbance processing flow is to execute the process according to the static scheduling plan. When the key process that needs to be inspected is executed, the quality inspection results of the current processing batch of workpieces are first judged: 1) For workpieces with qualified quality inspection results, the subsequent processes are continued;
[0023] 2) Workpieces that are returned for repair as a result of quality inspection need to be reprocessed in the current key process and added to the unprocessed workpiece set; 3) Workpieces that are scrapped as a result of quality inspection need to be reprocessed from the first process and also added to the unprocessed workpiece set; after obtaining the information of all workpieces, the workpieces in the processed workpiece set will not be processed, and the workpieces in the unprocessed workpiece set need to be rescheduled.
[0024] Step 4: Recombination batch for quality inspection disturbance;
[0025] When rescheduling, it is necessary to group various workpieces into batches and generate new processing tasks for rescheduling. The specific process of regrouping batches is as follows: Figure 3 As shown in the figure, the process is as follows: after the quality inspection disturbance occurs, the workpieces have different processing steps to be processed, but they can all be scheduled for processing at the rescheduling time. Therefore, it is necessary to re-batch the workpieces with the same process route, that is, to combine the workpieces with the same processing steps under various types of workpieces into a new batch of workpieces, forming a new scheduling problem.
[0026] Step 5: Optimize the solution using the improved grey wolf algorithm;
[0027] The encoding and decoding method is designed for the reorganized batch workpieces after dynamic disturbance. The encoding architecture consists of three layers: batch layer, process layer and machine layer. The batch layer encoding represents the batch information of the workpieces, the process layer encoding is the workpieces and processes, the number of times different digital codes appear represents different batches, and the machine layer represents the processing machines of each process; the decoding method is divided into two parts. First, the batch layer is decoded to obtain the specific batch division strategy to obtain the specific batch number of each batch of workpieces. Then, the process layer and machine layer are decoded and the target value of the scheduling plan is calculated. When decoding the process layer, the reorganized batch workpieces need to be calculated according to the original workpiece information;
[0028] An improved gray wolf algorithm is designed. The algorithm is mainly divided into two layers, in which the outer search mainly solves the batch solution, and the inner search solves the scheduling solution. When the gray wolf algorithm is solving, the bernoulli chaotic map is used to initialize the population, making the population distribution more uniform, improving the algorithm convergence speed and the quality of the optimal solution, and when updating the control parameters, nonlinear control parameters are used, so that the algorithm has a strong global search capability in the early stage and a fast convergence speed of local optimization in the later stage. The beneficial technical effects of the present invention are:
[0029] (1) The present invention designs a quality inspection disturbance rescheduling strategy, which can give a rescheduling plan in time when a disturbance occurs, solve the interference of workshop quality inspection disturbance on the static scheduling plan, and at the same time, the obtained rescheduling plan can effectively reduce the energy consumption of machine tools and maintain the stability of the static scheduling plan while shortening the completion time. When dealing with dynamic interference, the existing technology usually adopts a periodic rescheduling method, which cannot respond in time and does not consider energy consumption indicators.
[0030] (2) The present invention designs a batch reorganization technology, which effectively solves the problem of rescheduling the same workpiece due to different remaining process routes caused by disturbances. However, the prior art does not consider reorganizing batches of workpieces when solving the rescheduling problem, which will increase the scale of workpieces and make it difficult to find the optimal solution. At the same time, compared with the prior art method, the present method also improves the population initialization and parameter control of the gray wolf algorithm, greatly improving the algorithm's solution efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 Flowchart of the quality inspection disturbance rescheduling strategy;
[0032] Figure 2 To improve the Grey Wolf algorithm flow chart;
[0033] Figure 3 Schematic diagram of the reorganization batch strategy;
[0034] Figure 4 Schematic diagram of the encoding method;
[0035] Figure 5 Gantt chart for scheduling static solutions for quality inspection disturbances;
[0036] Figure 6 Scheduling Gantt chart for quality inspection disturbance rescheduling scheme. DETAILED DESCRIPTION
[0037] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0038] A flexible workshop dynamic batch scheduling method for quality inspection disturbance of the present invention is specifically as follows:
[0039] Step 1: Determine the description of the flexible job shop dynamic batch scheduling problem.
[0040] The dynamic batch scheduling problem of a flexible job shop studies the scheduling problem of rescheduling according to the dynamic disturbance in the workshop production process. The specific description is as follows:
[0041] Suppose there are M machine tools M={M k |k=1,2,…,m}, n types of workpieces J={J i |i=1,2,…,n}, the number of workpieces of the i-th category is Ri , and each type of workpiece has O i = {O il |l=1,2,…,w i} process, and there are M machine tools that can process each process il , M il ∈M units, the processing time of each process varies with the performance of the machine tool; each type of workpiece is divided into multiple sub-batches F i ={F is |s=1,2,…,u i} Processing on different machines, each sub-batch is treated as a whole and occupies the same auxiliary time. This problem has a static scheduling solution. During the execution of the static scheduling solution, disturbances will occur. The processes that start before the disturbance time node are divided into completed workpieces and workpieces being processed. They do not need to participate in rescheduling. The processes that start after the disturbance time node need to be dynamically scheduled. When performing dynamic scheduling, the available time of all machine tools changes from the original zero time to the disturbance time node or the original processing end time of the workpiece being processed.
[0042] Step 2: Establish a mathematical model for the optimization of dynamic batch scheduling objectives in flexible job shops.
[0043] The objective functions are the workshop production cost, machine tool energy consumption cost and the difference cost of dynamic and static scheduling schemes, and the three objective function values are processed by weighted normalization.
[0044] Objective function:
[0045]
[0046]
[0047]
[0048] min{ω1·f1+ω2·f2+ω3·f3} (4)
[0049] Constraints:
[0050]
[0051] T=max(Z il ·E islk ) (6)
[0052] if S islk <T,DS islk -S islk =0 (7)
[0053]
[0054] Among them, Cis represents the completion time of the sth batch of the i-th type of workpiece, t islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The single-piece working time on ilk Represents the lth process of the i-th type of workpiece on the machine tool M k Auxiliary time on S islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The start processing time on E islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The processing completion time on γ ilk represents the machine selection decision variable, when process O il Available on machine M k Processing time ilk =1, otherwise γ ilk =0, represents the auxiliary time decision variable. k If the category of the workpiece being processed on the machine tool is the same as the category of the last workpiece processed on the machine tool, on the contrary Q k represents the maximum completion time of machine k, W k represents the load of machine k, PC k Indicates machine M k Average energy cost per unit time during load, NC k Indicates machine M k Average energy consumption cost per unit time when no load, RC k Indicates machine M k The average energy consumption cost per unit time during preparation, η represents the completion cost per unit time in the workshop, ω1 represents the completion cost weight coefficient, ω2 represents the energy consumption cost weight coefficient, ω3 represents the difference cost weight coefficient of dynamic and static scheduling schemes, DS islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the dynamic scheduling plan is on machine M k Start processing time on DW islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the dynamic scheduling plan is on machine M k The processing completion time on D islk Represents process O in the dynamic scheduling scheme il Machine M is still in the static scheduling scheme k During processing, D islk =0, otherwise D islk =1, S islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the static scheduling plan is on machine M kThe start processing time on the ,ω3 represents the difference cost weight coefficient, V represents the difference unit cost of dynamic and static scheduling schemes, T represents the disturbance time node, DF is represents the sth subbatch of the i-th type of workpiece in the dynamic scheduling scheme, SF is represents the sth sub-batch of the i-th type of workpiece after the disturbance time node in the static scheduling scheme, R represents the number of new workpieces caused by the disturbance, and Z il Indicates process O il is the disturbance process, D il =1, otherwise D il =0.
[0055] Among them, formula (5) indicates that the processing completion time of a batch of workpieces in the dynamic scheduling scheme must be greater than or equal to the sum of the starting processing time, auxiliary time and working time of all workpieces in the batch; formula (6) indicates that the disturbance time node is the lth process of the sth batch of the i-th type of workpiece on machine M. k The processing completion time on the disturbance time node; Formula (7) indicates that the process before the disturbance time node and the process starting processing do not participate in dynamic scheduling; Formula (8) indicates that the total batch number of workpieces in the dynamic scheduling scheme is equal to the sum of the number of unprocessed workpieces at the disturbance time node and the number of newly added workpieces caused by the disturbance.
[0056] Step 3: Rescheduling for quality inspection disturbances.
[0057] In the current research on dynamic scheduling of workshops, there are few literatures that consider the impact of quality inspection disturbances on the execution of static scheduling schemes. In the research on batch dynamic scheduling of flexible workshops, there is no literature involved. However, quality inspection disturbances are important production disturbances in actual production, which have an important impact on workshop scheduling and have great research significance. At the same time, the rescheduling scheme generated according to quality inspection disturbances will also involve energy consumption issues, which need further research. In response to the quality problem of workpieces, the existing literature usually deals with the scrap rate of workpieces by estimating the scrap rate of workpieces. When performing static scheduling, the number of workpieces scheduled is increased to improve the tolerance of static scheduling schemes to workpiece quality problems and reduce the interference of quality inspection disturbances. However, this treatment method has major drawbacks. First, the estimated scrap rate may not be accurate. If the estimate is too large, it will lead to overcapacity and waste of resources. If the estimate is too small, it will lead to insufficient capacity and failure to complete on time. Second, the quality inspection process may produce reworked workpieces. If the workpiece is directly scrapped, it will lead to waste of resources. If it is arranged for rework, rescheduling is required. Therefore, this treatment method will cause the scheduling plan to be out of touch with the actual production and cannot provide continuous guidance for the actual production. Therefore, this method proposes a rescheduling strategy for quality inspection disturbance. Figure 1 .
[0058] The quality inspection disturbance processing flow executes the process according to the static scheduling plan. When executing the key process that needs to be inspected, the quality inspection results of the current processing batch of workpieces are first judged: 1) The workpieces with qualified quality inspection results continue to execute the subsequent processes; 2) The workpieces with the quality inspection result of being returned for repair need to be reprocessed in the current key process and added to the unprocessed workpiece set; 3) The workpieces with the quality inspection result of being scrapped need to be reprocessed from the first process and also added to the unprocessed workpiece set; after obtaining all the workpiece information, the workpieces in the processed workpiece set will not be processed, and the workpieces in the unprocessed workpiece set need to be rescheduled.
[0059] Step 4: Recombination batch for quality inspection disturbance;
[0060] After a quality inspection disturbance occurs, due to the progress of the processing, the remaining processing steps of different batches of the same type of workpieces are different. At the same time, the quality inspection disturbance will also result in workpieces that fail the quality inspection. Therefore, when rescheduling, it is necessary to group various types of workpieces into batches and generate new processing tasks for rescheduling. Figure 3 Specifically, after the quality inspection disturbance occurs, the workpieces to be processed are different, and the workpieces with the same process route are reorganized, that is, the workpieces with the same processing steps under various types of workpieces are combined into a new batch of workpieces, forming a new scheduling problem.
[0061] Step 5: Improve the Grey Wolf Algorithm with a two-layer search framework to optimize the solution;
[0062] The present invention designs a coding and decoding method for the reorganized batch workpieces after dynamic disturbance. The coding architecture consists of three layers: batch layer, process layer and machine layer. The process layer is encoded as workpieces and processes. The number of times different digital codes appear represents different batches. Figure 4 As shown, taking the gray number 23 in the figure as an example, 23 represents the third process of workpiece 2. This number appears for the first time in the process layer, that is, it represents the first batch.
[0063] Secondly, since the same type of workpieces are grouped into different batches for processing after reorganization, and the workpieces in each batch will be divided into batches again during rescheduling, it will cause redundancy in the encoding method. Therefore, this paper assigns new workpiece numbers to the same type of workpieces in different batches after reorganization, but still uses the original workpiece information for calculation during decoding.
[0064] The decoding method is divided into two parts. First, the batch layer is decoded to obtain the specific batch division strategy in order to obtain the specific batch number of each batch of workpieces. Then the process layer and machine layer are decoded and the target value of the scheduling plan is calculated. When decoding the process layer, the reorganized batch workpieces need to be calculated according to the original workpiece information.
[0065] In order to reschedule the reorganized batched workpieces, the improved gray wolf algorithm of the present invention is mainly divided into two layers, in which the outer search mainly solves the batching scheme, and the inner search solves the scheduling scheme, such as Figure 2 The solution process is as follows: when the grey wolf algorithm is solving, the bernoulli chaotic map is used to initialize the population, making the population distribution more uniform, improving the algorithm convergence speed and the quality of the optimal solution; when updating the control parameters, nonlinear control parameters are used, which makes the algorithm have strong global search capabilities in the early stage and fast convergence speed of local optimization in the later stage.
[0066] Example:
[0067] In order to apply theoretical research to the actual production control of the workshop, the present invention analyzes the scheduling status and scheduling needs of the flexible job workshop of an aviation structural parts factory. According to the problems existing in the existing scheduling methods and processing procedures of the workshop, the static batch scheduling model of the flexible job workshop considering energy consumption proposed in the present invention is used to design a static scheduling plan. At the same time, for dynamic disturbance events in the workshop, the dynamic batch scheduling model of the flexible job workshop considering energy consumption proposed in the present invention is used, and the insertion disturbance strategy and quality inspection disturbance strategy are adopted to obtain a satisfactory dynamic scheduling plan for the workshop.
[0068] There are a total of 18 types of workpieces that need to be processed in this workshop. Each type of workpiece needs to process 10 workpieces in one production cycle. Two types of workpieces have only two processes, and the rest have four processes. The processing time of each type of workpiece on the machine tool varies with the performance of each machine tool. The specific processing information is shown in Table 1.
[0069] Table 1 Workpiece processing information table
[0070]
[0071]
[0072]
[0073] At the disturbance time T=3325, the workshop needs to conduct quality inspection on process 1 of the 4th batch of workpiece 15. There are two workpieces in this batch. The quality inspection results are that one workpiece is qualified and one workpiece is returned for repair. The results of each algorithm under this working condition are shown in Table 2. The results of the original dynamic scheduling strategy of the workshop are also given in Table 2.
[0074] Table 2 Engineering example quality inspection disturbance experiment results
[0075]
[0076] It can be seen that the solution result of the algorithm proposed in the present invention not only reduces the comprehensive cost, completion time cost and energy consumption cost by 1.5%, 0.6% and 3.8% respectively compared with the existing scheduling strategy of the workshop, but also the scheduling scheme of the present invention is better than the algorithm compared with the present invention, which verifies the effectiveness and feasibility of the algorithm of the present invention. At the same time, the present invention applies the existing scheduling strategy of the workshop to the static scheduling scheme optimized in this article for comparison. If it is directly compared with the original static scheduling scheme of the workshop, the comprehensive cost, completion time cost and energy consumption cost are increased by 27.3%, 34.8% and 24.5% respectively. Figure 5 Gantt chart of static scheduling scheme for quality inspection disturbance, Figure 6 It is a dynamic scheduling scheme, in which workpiece 1 is actually workpiece 2, workpiece 19 is actually workpiece 10, workpiece 20 is actually workpiece 12, workpiece 21 is actually workpiece 15, workpiece 22 is actually workpiece 17, workpiece 23 is actually workpiece 18, workpiece 24 is actually workpiece 15, and workpiece 25 is actually workpiece 14.
[0077] Aiming at the influence of quality inspection disturbance on static scheduling scheme in the workshop production process, the present invention takes minimizing workshop production cost, minimizing total energy consumption cost of machine tools, and minimizing difference cost of dynamic and static scheduling schemes as optimization goals, establishes a dynamic batch scheduling problem model of flexible job shop considering energy consumption, designs specific rescheduling strategy for quality inspection disturbance, reorganizes batches for rescheduled workpieces, and adopts improved grey wolf algorithm for optimization and solution after the reorganization batches are completed. Results show that the strategy and algorithm proposed in the present invention can effectively solve the interference of quality inspection disturbance, and the obtained rescheduling scheme not only has the lowest comprehensive cost, but also has great advantages in workshop production cost and total energy consumption cost of machine tools, which can bring greater benefits to enterprises.
Claims
1. A flexible workshop dynamic batch scheduling method for quality inspection disturbance, characterized in that: Specifically: Step 1: Determine the description of the flexible job shop dynamic batch scheduling problem; Suppose there are M machine tools M={M k |k=1,2,…,m}, n types of workpieces J={J i |i=1,2,…,n}, the number of workpieces of the i-th category is R i , and each type of workpiece has O i = {O il |l=1,2,…,w i } process, and there are M machine tools that can process each process il , M il ∈M units, the processing time of each process varies with the performance of the machine tool; each type of workpiece is divided into multiple sub-batches F i ={F is |s=1,2,…,u i } Processing on different machines, each sub-batch is treated as a whole and occupies the same auxiliary time; Step 2: Establish a mathematical model for the optimization of dynamic batch scheduling objectives in flexible job shops; The objective functions are workshop production cost, machine tool energy consumption cost and dynamic and static scheduling scheme difference cost, and the three objective function values are processed by weighted normalization. Objective function: min{ω1·f1+ω2·f2+ω3·f3} (4) Constraints: T=max(Z il ·E islk )(6) if S islk <T,DS islk -S islk =0(7) Among them, C is represents the completion time of the sth batch of the i-th type of workpiece, t islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The single-piece working time on ilk Represents the lth process of the i-th type of workpiece on the machine tool M k Auxiliary time on S islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The start processing time on E islk It means that the lth process of the sth batch of the i-th type of workpiece is performed on the machine M k The processing completion time on γ ilk represents the machine selection decision variable, when process O il Available on machine M k Processing time ilk =1, otherwise γ ilk =0, represents the auxiliary time decision variable. k If the category of the workpiece being processed on the machine tool is the same as the category of the last workpiece processed on the machine tool, on the contrary Q k represents the maximum completion time of machine k, W k represents the load of machine k, PC k Indicates machine M k Average energy cost per unit time during load, NC k Indicates machine M k Average energy cost per unit time when no load, RC k Indicates machine M k The average energy consumption cost per unit time during preparation, η represents the completion cost per unit time in the workshop, ω1 represents the completion cost weight coefficient, ω2 represents the energy consumption cost weight coefficient, ω3 represents the difference cost weight coefficient of dynamic and static scheduling schemes, DS islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the dynamic scheduling plan is on machine M k Start processing time on DW islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the dynamic scheduling plan is on machine M k The processing completion time on D islk Represents process O in the dynamic scheduling scheme il Machine M is still in the static scheduling scheme k During processing, D islk =0, otherwise D islk =1, S islk Indicates that the lth process of the sth batch of the i-th type of workpiece in the static scheduling plan is on machine M k The start processing time on the , V represents the unit cost of the difference between dynamic and static scheduling schemes, T represents the disturbance time node, DF is represents the sth subbatch of the i-th type of workpiece in the dynamic scheduling scheme, SF is represents the sth sub-batch of the i-th type of workpiece after the disturbance time node in the static scheduling scheme, R represents the number of new workpieces caused by the disturbance, and Z il Indicates process O il is the disturbance process, Z il =1, otherwise Z il =0; Among them, formula (5) indicates that the processing completion time of a batch of workpieces in the dynamic scheduling scheme must be greater than or equal to the sum of the starting processing time, auxiliary time and working time of all workpieces in the batch; formula (6) indicates that the disturbance time node is the lth process of the sth batch of the i-th type of workpiece on machine M. k The processing completion time on the disturbance time node; Formula (7) indicates that the process before the disturbance time node and the process starting processing do not participate in dynamic scheduling; Formula (8) indicates that the total batch number of workpieces in the dynamic scheduling scheme is equal to the sum of the number of unprocessed workpieces at the disturbance time node and the number of newly added workpieces caused by the disturbance; Step 3: Rescheduling for quality inspection disturbances; The quality inspection disturbance processing flow is to execute the process according to the static scheduling plan. When executing the key process that needs to be inspected, first judge the quality inspection results of the current processing batch of workpieces: 1) The workpieces with qualified quality inspection results continue to execute the subsequent processes; 2) The workpieces with the quality inspection result of rework need to be reprocessed in the current key process and added to the unprocessed workpiece set; 3) The workpieces with the quality inspection result of scrap need to be reprocessed from the first process and also added to the unprocessed workpiece set; after obtaining all the workpiece information, the workpieces in the processed workpiece set will not be processed, and the workpieces in the unprocessed workpiece set need to be rescheduled; Step 4: Recombination batch for quality inspection disturbance; When rescheduling, it is necessary to group various workpieces into batches and generate new processing tasks for rescheduling. The specific process of re-batch is as follows: after the quality inspection disturbance occurs, the workpieces to be processed are different, and the workpieces with the same process route are re-batch, that is, the workpieces with the same processing steps under various workpieces are combined into a new batch of workpieces, forming a new scheduling problem. Step 5: Improve the Grey Wolf Algorithm with a two-layer search framework to optimize the solution; The encoding and decoding method is designed for the reorganized batch workpieces after dynamic disturbance. The encoding architecture consists of three layers: batch layer, process layer and machine layer. The batch layer encoding represents the batch information of the workpieces. The process layer encoding is the workpiece and process. The number of times different digital codes appear represents different batches. The machine layer represents the processing machine of each process. The decoding method is divided into two parts. First, the batch layer is decoded to obtain the specific batch division strategy to obtain the specific batch number of each batch of workpieces. Then, the process layer and machine layer are decoded and the target value of the scheduling plan is calculated. When decoding the process layer, the reorganized batch workpieces need to be calculated according to the original workpiece information. An improved Grey Wolf Algorithm is designed. The algorithm is mainly divided into two layers. The outer search is mainly used to solve the batch scheme, and the inner search is used to solve the scheduling scheme. The solution process is as follows: when the Grey Wolf Algorithm is solving, the Bernoulli chaotic map is used to initialize the population, which makes the population distribution more uniform, improves the algorithm convergence speed and the quality of the optimal solution, and when updating the control parameters, nonlinear control parameters are used, which makes the algorithm have strong global search capabilities in the early stage and fast convergence speed of local optimization in the later stage.
