Hybrid flow shop scheduling method based on historical information ant colony algorithm

Abstract

The invention discloses a hybrid flow shop scheduling method based on a historical information ant colony algorithm. The method comprises the following steps: 1, constructing a total time and energy consumption model of glass workpiece processing; 2, corresponding constraint conditions are set, and a double-target hybrid flow shop scheduling model is constructed; and 3, solving the double-target hybrid flow shop scheduling model by using a historical information ant colony algorithm according to related information of the glass workpiece and a machine to obtain a production and processing scheme, so that the glass workpiece is produced and processed. According to the method, an optimal production and processing scheme combining time consumption and energy consumption can be obtained, so that the production and processing efficiency of the glass workpiece can be improved.

Description

Technical Field

[0001] The invention belongs to the field of hybrid flow shop scheduling optimization control and intelligent decision-making, and specifically, is a hybrid flow shop scheduling method based on historical information ant colony algorithm. Background Art

[0002] The hybrid flow shop scheduling problem is an important research topic in the field of production scheduling and is widely used in industries such as semiconductors, glass production, and steelmaking. Unlike traditional flow shops, each process in a hybrid flow shop can have multiple machines, and these machines may be of different types. The number of machines and processing time of different processes can also be different. The goal of hybrid flow shop scheduling is usually to optimize production efficiency and reduce production cycle time by reasonably arranging the processing order of each task.

[0003] In recent years, the global energy consumption problem has become increasingly prominent, and many countries and regions have successively formulated strict environmental protection laws and regulations. Among them, the energy consumption of the manufacturing industry accounts for about 50% of the total. For energy-intensive factories, if they cannot effectively reduce energy consumption, they will face elimination. Therefore, managers urgently need to consider production methods that balance energy saving and efficiency. In view of this, how to balance production efficiency and energy consumption has become an issue worth studying.

[0004] Tempered glass is a glass product that has been specially treated to enhance its strength and impact resistance. It is widely used in the fields of construction, automobiles, electronic equipment, and home appliances. As a typical manufacturing scenario, the production of tempered glass has long been widely studied by academia and industry with a single goal of reducing its production time. However, the issue of considering both completion time and energy consumption has not been fully explored. Summary of the invention

[0005] The present invention aims to solve the deficiencies of the above-mentioned prior art and proposes a hybrid flow shop intelligent scheduling method and application based on historical information ant colony algorithm, in order to obtain the optimal production and processing plan combining time consumption and energy consumption, thereby improving the efficiency and accuracy of workshop scheduling.

[0006] In order to achieve the above-mentioned object of the invention, the present invention adopts the following technical scheme:

[0007] The hybrid assembly line intelligent scheduling optimization method based on historical information ant colony algorithm of the present invention is characterized in that it is applied to each glass workpiece entering the first buffer zone for waiting after being processed by the cutting parallel machine, and then transporting each glass workpiece from the first buffer zone to the spraying parallel machine for processing, and then entering the second buffer zone for waiting, and then transporting each glass workpiece from the second buffer zone to the batch processing machine to form each batch, and then each batch is tempered on the batch processing machine until all glass workpieces are processed in the assembly line, and any parallel machine of the cutting parallel machine and the spraying parallel machine is recorded as the first parallel machine. The hybrid flow shop intelligent scheduling optimization method is carried out according to the following steps:

[0008] Step 1: Establish the first objective function with maximum completion time as the goal And the second objective function with the minimum energy consumption as the goal As the two objective functions of the hybrid flow shop scheduling model, let any oth objective function be , ;

[0009] Step 2: Establish the constraints of the hybrid flow shop scheduling model;

[0010] Step 3: Use the historical information ant colony algorithm to solve the hybrid flow shop scheduling model to obtain a production and processing plan.

[0011] The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm described in the present invention is also characterized in that in step 1, two objective functions of the hybrid flow shop scheduling model are established using equations (1) and (2):

[0012] (1)

[0013] (2)

[0014] In formula (1), is the maximum completion time, Indicates that the bth batch is in Completion time on batch processing machines;

[0015] In formula (2), TEC is energy consumption, is the total number of time periods, represents the total number of parallel machines, represents the total number of batch processors, f(t) represents the power consumption in the tth time period, Indicates Taiwan Is the parallel machine in working state in the tth time period? If so, let =1; otherwise, let =0; Indicates Taiwan Is the parallel machine in the tth time period in the standby state? If so, let =1; otherwise, let =0; Indicates Taiwan Is the parallel machine in the tth time period in the shutdown state? If so, let =1; otherwise, let =0; Indicates Is the batch processing machine in working state in the tth time period? If so, let =1; otherwise, let =0; Indicates Is the batch processor in the tth time period in the standby state? If so, let =1; otherwise, let =0; Indicates Is the batch processor in the tth time period in a shutdown state? If so, let =1; otherwise, let =0; , and Respectively represent Energy consumption of a quasi-parallel machine in working, standby and off states; , and Respectively represent the energy consumption of the batch machine in working state, standby state and shutdown state.

[0016] Furthermore, in step 2, the constraints of the hybrid flow shop scheduling model are established using equations (3) to (45):

[0017] (3)

[0018] (4)

[0019] (5)

[0020] (6)

[0021] (7)

[0022] (8)

[0023] (9)

[0024] (10)

[0025] (11)

[0026] (12)

[0027] (13) (14)

[0028] (15)

[0029] (16)

[0030] (17)

[0031] (18)

[0032] (19)

[0033] (20)

[0034] (21)

[0035] (22)

[0036] (23)

[0037] (24)

[0038] (25)

[0039] (26)

[0040] (27)

[0041] (28)

[0042] (29)

[0043] (30)

[0044] (31)

[0045] (32)

[0046] (33)

[0047] (34)

[0048] (35)

[0049] (36)

[0050] (37)

[0051] (38)

[0052] (39)

[0053] (40)

[0054] (41)

[0055] (42)

[0056] (43)

[0057] (44)

[0058] (45)

[0059] In formula (3) to formula (45), Indicates The glass workpiece is Processing time on a quasi-parallel machine, Indicates The arrival time of the glass workpieces, Indicates The glass workpiece is Completion time on a parallel machine, Indicates The size of the glass workpiece, Indicates The capacity of batch processing machines, Indicates The batch is in Processing time on batch processing machines, Indicates The batch is in The start time of processing on the batch processing machine, Indicates The glass workpiece is Processing time on batch processing machines; Indicates Is the glass workpiece in Taiwan If the type of parallel processing is =1; otherwise, let =0; Indicates Is the glass workpiece in Taiwan If the type of parallel processing is Otherwise, let ; Indicates whether the jth glass workpiece is waiting in the first buffer in the tth time period. If so, let Otherwise, let ; Indicates The qualification constraint of the glass workpiece is whether it can be Taiwan If possible, then Otherwise, let ; Indicates Is the glass workpiece in The first batch processor In a batch, if so, then let Otherwise, let ; Indicates the tth time period Is the batch in If it is processed on a batch processing machine, then Otherwise, let ; Indicates Does the batch contain at least one glass workpiece? If so, Otherwise, let ; is the number of stations i-th in the t-th time period Auxiliary binary variables for parallel machines, Indicates the tth time period Auxiliary binary variables for batch processing; and are the capacities of the first buffer and the second buffer respectively; Indicates high battery; Indicates low battery; Indicates the end time period of high power; Indicates the end time period of low battery; Indicates The batch processor Is it in working state during the time period? If so, =1; otherwise, let =0; Indicates Taiwan Class parallel machine Is it in working state during the time period? If so, =1; otherwise, let =0.

[0060] Furthermore, the step 3 comprises:

[0061] Step 3.1, define the maximum number of iterations as and the total number of ants is , randomly generate the first preference vector With the second preference vector ; Initialize the heuristic information matrix; define the current generation as k; initialize k=1; initialize the kth generation pheromone matrix ;

[0062] Step 3.2, set the ant's index a=1;

[0063] Step 3.3, initialize the parameters of the glass workpiece and the machine, set the first taboo table DE1 of n dimensions, and assign an initial value to DE1 according to the serial number of the glass workpiece. , set the n-dimensional second taboo table DE2 and the n-dimensional third taboo table DE3, and assign initial values ​​to both ; n represents the total number of glass workpieces;

[0064] Step 3.4: When l=1 and DE1 is not (0,0,...,0), execute The scheduling of the lth parallel machine, thus Taiwan A glass workpiece is processed on a parallel machine, and the processed glass workpiece is stored in buffer 1, and the value of the index position of the processed glass workpiece in DE1 is set to 0, and the value of the index position of the processed glass workpiece in DE2 is modified to the index number of the processed glass workpiece itself; thereby DE1 and DE2 are updated; and then Middle Daidi Taiwan Class parallel machine The oth target of the glass workpiece The pheromone is updated to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece ;

[0065] Step 3.5: When l=2 and DE2 is not (0,0,...,0), execute Taiwan The scheduling of parallel machines is Taiwan A glass workpiece is processed on a parallel machine, and the processed glass workpiece is stored in buffer 2, and the value of the index position of the processed glass workpiece in DE2 is set to 0, and the value of the index position of the processed glass workpiece in DE3 is modified to the index number of the processed glass workpiece itself; thereby DE2 and DE3 are updated; and then Middle Daidi Taiwan Class parallel machine The oth target of the glass workpiece The pheromone is updated to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece ;

[0066] Step 3.6, when DE3 is not (0,0,...,0), execute the scheduling of the mth batch processing machine, so that the bth batch is tempered on the mth batch processing machine, and set the value of the index position of the glass workpiece after the tempering process in DE3 to 0, thereby updating DE3; then Middle Daidi Taiwan Class parallel machine The oth target of the glass workpiece The pheromone is updated to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece ;

[0067] Step 3.7: If and and , then go to step 3.4; otherwise, it means that a workshop scheduling plan corresponding to the a-th ant of the k-th generation is obtained;

[0068] Step 3.8: When Assign to , go to step 3.3 and execute sequentially, otherwise, it means that the kth generation is obtained The workshop scheduling plan of each ant is compared with the historical workshop scheduling plan set Merge to get the kth generation merged population, and select the one with the largest fitness The workshop scheduling plan is used as the new historical workshop scheduling plan set ; When k=1, the historical workshop scheduling plan set That is the kth generation The workshop scheduling plan for ants;

[0069] Step 3.9, if , then output the latest historical workshop scheduling plan for production, otherwise, execute step 3.10;

[0070] Step 3.10: Use formula (61) to get Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact :

[0071] (61)

[0072] In formula (61), represents the volatility coefficient of the global pheromone, express The workshop scheduling plan of any ant in the Indicates Daidi Class On parallel machines The pheromone change of the oth target of the glass workpiece; and:

[0073] (62)

[0074] In formula (62), Indicates Daidi Class On parallel machines The oth target value corresponding to the glass workpiece;

[0075] Step 3.11: Use formula (63) to get +1 generation The glass workpiece and The pheromone of the oth target when glass workpieces are simultaneously assigned to the bth batch:

[0076] (63)

[0077] In formula (63), Indicates Daidi The glass workpiece and The oth target value when glass workpieces are simultaneously assigned to the bth batch, Indicates Daidi The glass workpiece and The pheromone change of the oth target when the glass workpieces are simultaneously assigned to the bth batch is:

[0078] (64)

[0079] In formula (64), Indicates Daidi The glass workpiece and The oth target value when glass workpieces are simultaneously assigned to the bth batch;

[0080] Step 3.12: Assign to Then, go to step 3.2 and execute it sequentially.

[0081] Further, the step 3.4 includes:

[0082] Step 3.4.1: The ath ant randomly selects an unscheduled glass workpiece from DE1, and then selects the ath ant according to the qualification constraints of the glass workpiece using formula (46). Taiwan Class parallel machines, using equations (47) and (48) we can get Taiwan First candidate list for class parallel machines With the second candidate list :

[0083] (46)

[0084] (47)

[0085] (48)

[0086] In formula (46), represents the oth objective function The preference weight of Indicates Taiwan The oth target value of the parallel machine; in equations (47) and (48), Indicates that the jth glass workpiece arrives at The time of a parallel machine, Indicates Taiwan Completion time of a quasi-parallel machine;

[0087] Step 3.4.2: Use equation (49) to get the jth glass workpiece reaching the Taiwan First candidate list for class parallel machines Heuristic information , using formula (50) to obtain the jth glass workpiece arriving at the Taiwan Second candidate list for parallel machines Heuristic information ;

[0088] (49)

[0089] (50)

[0090] Step 3.4.3: The ath ant uses formula (51) to calculate the Taiwan The probability of selecting the jth glass workpiece in a parallel machine , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for processing, and storing the processed glass workpiece in buffer 1, and setting the value of the index position of the processed glass workpiece in DE1 to 0, and modifying the value of the index position of the processed glass workpiece in DE2 to the index number of the processed glass workpiece itself; thereby updating DE1 and DE2;

[0091] (51)

[0092] In formula (51), and There are 2 weights respectively; Indicates Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact, Indicates The glass workpiece arrives at Taiwan First candidate list for class parallel machines Heuristic information, Indicates The glass workpiece arrives at Taiwan Second candidate list for parallel machines Heuristic information, Indicates Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact;

[0093] Step 3.4.4: The ath ant uses formula (52) to Update to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece :

[0094] (52)

[0095] In formula (52), represents the volatility coefficient of local pheromone, It is the initial pheromone.

[0096] 6. The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm according to claim 5, characterized in that step 3.5 comprises:

[0097] Step 3.5.1: The ath ant randomly selects an unscheduled glass workpiece from DE2, and then selects the ath ant according to the qualification constraints of the glass workpiece using formula (46). Taiwan Class parallel machines, using equations (47) and (48) we can get Taiwan First candidate list for class parallel machines With the second candidate list ;

[0098] Step 3.5.2: Use equation (49) to get the jth glass workpiece reaching the Taiwan First candidate list for class parallel machines Heuristic information , using formula (50) to obtain the jth glass workpiece arriving at the Taiwan Second candidate list for parallel machines Heuristic information ;

[0099] Step 3.5.3: The ath ant uses formula (51) to calculate the Taiwan The probability of selecting the jth glass workpiece in a parallel machine , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for processing, and storing the processed glass workpiece in buffer 2, and setting the value of the index position of the processed glass workpiece in DE2 to 0, and modifying the value of the index position of the processed glass workpiece in DE3 to the index number of the processed glass workpiece itself; thereby updating DE2 and DE3;

[0100] Step 3.5.4: The ath ant uses formula (52) to Update to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece .

[0101] Further, the step 3.6 includes:

[0102] Step 3.6.1, the ath ant uses formula (53) to select the mth batch processor, which is used to add the glass workpieces that meet the qualification constraints and arrive at the mth batch processor earliest to the bth batch, and uses formula (54) and formula (55) to construct the first batch processor that arrives at the mth batch processor earliest. glass workpieces are assigned to the first candidate list of the bth batch With glass workpieces are assigned to the second candidate list of the bth batch :

[0103] (53)

[0104] (54)

[0105] (55)

[0106] In formulas (56) and (57), The mth batch processing machine The arrival time of the glass workpieces, Indicates the start processing time of the bth batch; Indicates The size of the glass workpiece, represents the capacity of the mth batch processor, Indicates the size of the b-th batch;

[0107] Step 3.6.2: Use formula (56) to get glass workpieces are assigned to the first candidate list of the bth batch Heuristic information , using formula (57) to obtain glass workpieces are assigned to the second candidate list of the bth batch Heuristic information :

[0108] (56)

[0109] (57)

[0110] Step 3.6.3: The ath ant uses formula (58) to calculate the list of the bth batch. The probability of selecting the jth glass workpiece in , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for tempering treatment, and setting the index position value of the glass workpiece after the tempering treatment in DE3 to 0, thereby updating DE3;

[0111] (58)

[0112] In formula (58), and There are 2 weights respectively; Indicates Daidi When the glass workpieces are assigned to the bth batch, the pheromone of the oth target is: Indicates glass workpieces are assigned to the first candidate list of the bth batch Heuristic information, Indicates glass workpieces are assigned to the second candidate list of the bth batch Heuristic information, Indicates Daidi When the glass workpiece is assigned to the bth batch, the pheromone of the oth target;

[0113] Step 3.6.4: Use formula (59) to calculate the Daidi When glass workpieces are assigned to the bth batch, the pheromone of the oth target :

[0114] (59)

[0115] In formula (61), , Indicates The first The glass workpiece and The pheromone of the oth target when the glass workpieces are simultaneously assigned to the bth batch; represents the number of glass workpieces contained in the bth batch;

[0116] Step 3.6.5: Use formula (60) to get The first The glass workpiece and The updated pheromone of the oth target when glass workpieces are simultaneously assigned to the bth batch :

[0117] (60).

[0118] An electronic device of the present invention includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the hybrid flow shop scheduling method, and the processor is configured to execute the program stored in the memory.

[0119] The present invention provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and the computer program executes the steps of the hybrid flow shop scheduling method when the computer program is executed by a processor.

[0120] Compared with the prior art, the present invention has the following beneficial effects:

[0121] 1. According to the actual production situation of the mixed flow workshop, the present invention sets up multiple constraints, establishes a dual-objective optimization model for the completion time and energy consumption required for the production of the mixed flow workshop, and uses the historical information ant colony algorithm to solve the problem model to obtain the optimal dual-objective solution. The production plan uses an intelligent method to replace or assist the dispatchers in making decisions, thereby greatly improving the production efficiency of the workshop.

[0122] 2. The historical information ant colony algorithm used in the present invention has good global search capabilities and fast convergence speed, and is very suitable for solving large-scale hybrid flow shop scheduling problems. The hybrid flow shop scheduling problem of glass workpiece production has multiple complex processing steps. The standard ant colony algorithm is easily trapped in the local optimum and cannot search for the optimal production plan due to its high blindness and high randomness. Therefore, the ant colony algorithm is targetedly designed and improved using the historical workshop scheduling plan set, so that the effect of solving the tempered glass production scheduling problem will be better. BRIEF DESCRIPTION OF THE DRAWINGS

[0123] Figure 1 This is a flow chart of the scheduling method of the present invention;

[0124] Figure 2 This is the production process flow chart of tempered glass;

[0125] Figure 3 This is the flow chart of the historical information ant colony algorithm of the present invention. DETAILED DESCRIPTION

[0126] In this embodiment, a hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm is described. Figure 2 , is to obtain the process information of tempered glass production. When the customer places a production order, the manufacturer needs to prepare the required glass sheets within a certain period of time, and then form the final tempered glass after three stages of cutting, spraying and tempering. In the cutting stage, multiple parallel machines are processed at the same time, and each parallel machine can only process a specific type of glass sheet. Therefore, it is necessary to select the corresponding parallel machine according to customer needs to complete the glass (glass workpiece) cutting operation. Subsequently, in the spraying stage, a suitable one will be selected from multiple parallel machines with different spraying functions according to the needs, and a special coating, such as anti-reflective coating and anti-scratch coating, will be applied to the surface of the cut glass. Finally, in the tempering stage, multiple furnaces with different capacities are used to temper the coated glass. As long as the total size of the glass processed at one time does not exceed the capacity of the furnace, multiple glasses can be tempered at the same time. This process can be abstracted as a parallel batch machine scheduling problem with different machine capacities. It is worth noting that between the two stages, due to the limited inventory space (i.e., the buffer capacity is limited), the glass in the current stage can only be operated in the next stage or placed in the storage space if and only if the equipment in the next stage is idle or the inventory space is idle; otherwise, the current equipment will not be able to process the next (batch) of glass. Specifically, if Figure 1 As shown in the figure, the hybrid flow shop intelligent scheduling optimization method is carried out in the following steps:

[0127] Step 1: Use equations (1) and (2) to establish two objective functions of the hybrid flow shop scheduling model, including: the first objective function with the maximum completion time as the goal And the second objective function with the minimum energy consumption as the goal , and let any oth objective function be , :

[0128] (1)

[0129] (2)

[0130] In formula (1), is the maximum completion time, Indicates that the bth batch is in Completion time on batch processing machines;

[0131] In formula (2), TEC is energy consumption, is the total number of time periods, represents the total number of parallel machines, represents the total number of batch processors, f(t) represents the power consumption in the tth time period, Indicates Taiwan Is the parallel machine in working state in the tth time period? If so, let =1; otherwise, let =0; Indicates Taiwan Is the parallel machine in the tth time period in the standby state? If so, let =1; otherwise, let =0; Indicates Taiwan Is the parallel machine in the tth time period in the shutdown state? If so, let =1; otherwise, let =0; Indicates Is the batch processing machine in working state in the tth time period? If so, let =1; otherwise, let =0; Indicates Is the batch processor in the tth time period in the standby state? If so, let =1; otherwise, let =0; Indicates Is the batch processor in the tth time period in a shutdown state? If so, let =1; otherwise, let =0; , and Respectively represent Energy consumption of a quasi-parallel machine in working, standby and off states; , and Respectively represent the energy consumption of the batch machine in working state, standby state and shutdown state.

[0132] Step 2: Use equations (3) to (45) to establish the constraints of the hybrid flow shop scheduling model:

[0133] (3)

[0134] (4)

[0135] (5)

[0136] (6)

[0137] (7)

[0138] (8)

[0139] (9)

[0140] (10)

[0141] (11)

[0142] (12)

[0143] (13) (14)

[0144] (15)

[0145] (16)

[0146] (17)

[0147] (18)

[0148] (19)

[0149] (20)

[0150] (21)

[0151] (22)

[0152] (23)

[0153] (24)

[0154] (25)

[0155] (26)

[0156] (27)

[0157] (28)

[0158] (29)

[0159] (30)

[0160] (31)

[0161] (32)

[0162] (33)

[0163] (34)

[0164] (35)

[0165] (36)

[0166] (37)

[0167] (38)

[0168] (39)

[0169] (40)

[0170] (41)

[0171] (42)

[0172] (43)

[0173] (44)

[0174] (45)

[0175] Constraint (3) means that each glass workpiece is processed on only one parallel machine at each stage;

[0176] Constraint (4) indicates that the glass workpiece can only be processed on a parallel machine that can process the glass workpiece;

[0177] Constraint (5) indicates that The time period Taiwan The working status of the parallel machine;

[0178] Constraint (6) represents the calculation of the glass workpiece Processing time in each type of parallel machine;

[0179] Constraint (7) represents the glass workpiece Assign to It can be processed after the parallel machine is installed;

[0180] Constraint (8) means that if The glass workpiece is Taiwan For parallel on-machine processing, There is at least one time period for each parallel machine Processing Glass workpieces;

[0181] Constraint (9) indicates that The preparation time of each glass workpiece must be before the start time of the first type of parallel machine.

[0182] Constraint (10) means that Class parallel machine, if The glass workpiece is In the time period If the number of parallel machines is Greater than or equal to The start time of glass workpiece processing.

[0183] Constraint (11) means that when The glass workpiece is In the time period Class is processed on parallel machines, then The completion time of glass workpiece processing is greater than or equal to .

[0184] Constraint (12) represents the restriction on A parallel machine can only be in one state in one time period;

[0185] Constraints (13) and (14) indicate that Taiwan If a parallel machine wants to time period is idle, then it has worked for at least one time period before , and in the following time period, there is still a need to be Glass workpieces machined in parallel.

[0186] Constraint (15) represents the restriction on Taiwan Class parallel machine time period and The time period cannot be in standby mode;

[0187] Constraints (16) and (17) indicate that once a parallel machine is shut down, it will not be restarted. The parallel machine If the time period is closed, arrive No glass workpieces will be processed during the time period;

[0188] Constraint (18) indicates that After a glass workpiece has been processed by the first type of parallel machine, the second type of parallel machine can be started;

[0189] Constraint (19) indicates that the calculation The waiting time of a glass workpiece in the first buffer is The difference between the start time of the second parallel machine and the completion time of the first parallel machine for each glass workpiece;

[0190] Constraints (20) and (21) indicate that the total size of the glass workpieces retained in the buffer must be less than or equal to the capacity of the buffer;

[0191] Constraint (22) indicates that the calculation The waiting time of the glass workpiece in the buffer zone, i.e. The start time of the glass workpiece in batch machining minus the completion time in type 2 parallel machining.

[0192] Constraint (23) indicates that each glass workpiece can only be processed in one batch of one batch processing machine;

[0193] Constraint (24) indicates that the size of the batch does not exceed the capacity of the batch processor that processes the batch;

[0194] Constraint (25) indicates that the processing time of the calculated batch is determined by the maximum processing time of the glass workpieces in the batch;

[0195] Constraint (26) indicates that The batch processor Whether the time period is in working state;

[0196] Constraint (27) indicates that The batch is in Processing time on batch processing machines;

[0197] Constraint (28) means that when The first batch processor When a batch is not formed, its processing time is 0;

[0198] Constraint (29) indicates that the batch On a batch processor, if and only if After the batch is generated, the first +1 batch;

[0199] Constraint (30) means that when The first batch processor When a batch contains at least one glass workpiece, is equal to 1;

[0200] Constraint (31) indicates coordination and The value of

[0201] Constraint (32) indicates that After the glass workpiece has been processed by the second type of parallel machine, the batch processing machine will start processing;

[0202] Constraint (33) indicates that Batch processing machine Batch and +1 order between batches;

[0203] Constraints (34) and (35) indicate that batch processing takes at least one time period and cannot be preempted;

[0204] Constraint (36) represents the calculation of the maximum completion time, that is, the maximum completion time of all non-empty batches of all batch processors;

[0205] Constraint (37) states that all batches must be Complete processing before the time period;

[0206] Constraint (38) indicates that The batch processor A time period can only be in one state;

[0207] Constraints (39) and (40) indicate that The first batch processing machine can only be processed when the first batch processing machine has worked for at least one time cycle and there are still glass workpieces to be processed. A time period is in standby mode;

[0208] Constraint (41) indicates that The batch processor time period and The time period cannot be in standby mode;

[0209] Constraints (42) and (43) represent the setting of The batch processing machine is in shutdown state, and once the batch processing machine is shut down, it will not be restarted;

[0210] Constraint (44) indicates that the calculation Batch processing machine Completion time of each batch.

[0211] Constraint (45) represents the amount of electricity for different time periods.

[0212] In formula (3) to formula (45), Indicates The glass workpiece is Processing time on a quasi-parallel machine, Indicates The arrival time of the glass workpieces, Indicates The glass workpiece is Completion time on a parallel machine, Indicates The size of the glass workpiece, Indicates The capacity of batch processing machines, Indicates The batch is in Processing time on batch processing machines, Indicates The batch is in The start time of processing on the batch processing machine, Indicates The glass workpiece is Processing time on batch processing machines; Indicates Is the glass workpiece in Taiwan If the type of parallel processing is =1; otherwise, let =0; Indicates Is the glass workpiece in Taiwan If the type of parallel processing is Otherwise, let ; Indicates whether the jth glass workpiece is waiting in the first buffer in the tth time period. If so, let Otherwise, let ; Indicates The qualification constraint of the glass workpiece is whether it can be Taiwan If possible, then Otherwise, let ; Indicates Is the glass workpiece in The first batch processor In a batch, if so, then let Otherwise, let ; Indicates the tth time period Is the batch in If it is processed on a batch processing machine, then Otherwise, let ; Indicates Does the batch contain at least one glass workpiece? If so, Otherwise, let ; is the number of stations i-th in the t-th time period Auxiliary binary variables for parallel machines, Indicates the tth time period Auxiliary binary variables for batch processing; and are the capacities of the first buffer and the second buffer respectively; Indicates high battery; Indicates low battery; Indicates the end time period of high power; Indicates the end time period of low battery; Indicates The batch processor Is it in working state during the time period? If so, =1; otherwise, let =0; Indicates Taiwan Class parallel machine Is it in working state during the time period? If so, =1; otherwise, let =0.

[0213] Step 3: Use the historical information ant colony algorithm to solve the hybrid flow shop scheduling model and obtain a production and processing plan; specifically, Figure 3 shown.

[0214] Step 3.1, define the maximum number of iterations as and the total number of ants is , randomly generate the first preference vector With the second preference vector ; Initialize the heuristic information matrix; define the current generation as k; initialize k=1; initialize the kth generation pheromone matrix ;

[0215] Step 3.2, set the ant's index a=1;

[0216] Step 3.3, initialize the parameters of the glass workpiece and the machine, set the first taboo table DE1 of n dimensions, and assign an initial value to DE1 according to the serial number of the glass workpiece. , set the n-dimensional second taboo table DE2 and the n-dimensional third taboo table DE3, and assign initial values ​​to both ; n represents the total number of glass workpieces;

[0217] Step 3.4: When l=1 and DE1 is not (0,0,...,0), execute the scheduling strategy of the lth parallel machine:

[0218] Step 3.4.1: The ath ant randomly selects an unscheduled glass workpiece from DE1, and then selects the ath ant according to the qualification constraints of the glass workpiece using formula (46). Taiwan Class parallel machines, using equations (47) and (48) we can get Taiwan First candidate list for class parallel machines With the second candidate list :

[0219] (46)

[0220] (47)

[0221] (48)

[0222] In formula (46), represents the oth objective function The preference weight of Indicates Taiwan The oth target value of the parallel machine; in equations (47) and (48), Indicates that the jth glass workpiece arrives at The time of a parallel machine, Indicates Taiwan Completion time of a parallel machine.

[0223] Step 3.4.2: Use equation (49) to get the jth glass workpiece reaching the Taiwan First candidate list for class parallel machines Heuristic information , using formula (50) to obtain the jth glass workpiece arriving at the Taiwan Second candidate list for parallel machines Heuristic information ;

[0224] (49)

[0225] (50)

[0226] The present invention adopts whether the glass workpiece to be processed delays the start processing time of the parallel machine to construct and Its function is to delay the parallel machine from idle waiting time. Glass workpieces are treated as first choice, thus reducing time delay.

[0227] Step 3.4.3: The ath ant uses formula (51) to calculate the Taiwan The probability of selecting the jth glass workpiece in a parallel machine , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for processing, and storing the processed glass workpiece in buffer 1, and setting the value of the index position of the processed glass workpiece in DE1 to 0, and modifying the value of the index position of the processed glass workpiece in DE2 to the index number of the processed glass workpiece itself; thereby updating DE1 and DE2;

[0228] (51)

[0229] In formula (51), and There are 2 weights respectively; Indicates Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact, Indicates The glass workpiece arrives at Taiwan First candidate list for class parallel machines Heuristic information, Indicates The glass workpiece arrives at Taiwan Second candidate list for parallel machines Heuristic information, Indicates Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact.

[0230] Step 3.4.4: The ath ant uses formula (52) to Update to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece :

[0231] (52)

[0232] In formula (52), represents the volatility coefficient of local pheromone, It is the initial pheromone.

[0233] Step 3.5: When l=2 and DE2 is not (0,0,...,0), execute the scheduling of the parallel machine in the second stage:

[0234] Step 3.5.1: The ath ant randomly selects an unscheduled glass workpiece from DE2, and then selects the ath ant according to the qualification constraints of the glass workpiece using formula (46). Taiwan Class parallel machines, using equations (47) and (48) we can get Taiwan First candidate list for class parallel machines With the second candidate list ;

[0235] Step 3.5.2: Use equation (49) to get the jth glass workpiece reaching the Taiwan First candidate list for class parallel machines Heuristic information , using formula (50) to obtain the jth glass workpiece arriving at the Taiwan Second candidate list for parallel machines Heuristic information .

[0236] Step 3.5.3: The ath ant uses formula (51) to calculate the Taiwan The probability of selecting the jth glass workpiece in a parallel machine , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for processing, and storing the processed glass workpiece in buffer 2, and setting the value of the index position of the processed glass workpiece in DE2 to 0, and modifying the value of the index position of the processed glass workpiece in DE3 to the index number of the processed glass workpiece itself; thereby updating DE2 and DE3;

[0237] Step 3.5.4: The ath ant uses formula (52) to Update to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece .

[0238] Step 3.6: When DE3 is not (0,0,...,0), execute the batch processing machine scheduling:

[0239] Step 3.6.1, the ath ant uses formula (53) to select the mth batch processor, which is used to add the glass workpieces that meet the qualification constraints and arrive at the mth batch processor earliest to the bth batch, and uses formula (54) and formula (55) to construct the first batch processor that arrives at the mth batch processor earliest. glass workpieces are assigned to the first candidate list of the bth batch With glass workpieces are assigned to the second candidate list of the bth batch :

[0240] (53)

[0241] (54)

[0242] (55)

[0243] In formulas (56) and (57), The mth batch processing machine The arrival time of the glass workpieces, Indicates the start processing time of the bth batch; Indicates The size of the glass workpiece, represents the capacity of the mth batch processor, Represents the size of the b-th batch.

[0244] Step 3.6.2: Use formula (56) to get glass workpieces are assigned to the first candidate list of the bth batch Heuristic information , using formula (57) to obtain glass workpieces are assigned to the second candidate list of the bth batch Heuristic information :

[0245] (56)

[0246] (57)

[0247] The present invention uses whether the glass workpiece to be processed will delay the start processing time of the current batch, while considering the capacity constraint of the batch processor, to construct and Its function is to give priority to glass workpieces that will not delay the start of the current batch, thus reducing the idle time of the batch processing machine.

[0248] Step 3.6.3: The ath ant uses formula (58) to calculate the list of the bth batch. The probability of selecting the jth glass workpiece in , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for processing, and setting the index position value of the processed glass workpiece in DE3 to 0, thereby updating DE3;

[0249] (58)

[0250] In formula (58), and There are 2 weights respectively; Indicates Daidi When the glass workpieces are assigned to the bth batch, the pheromone of the oth target is: Indicates glass workpieces are assigned to the first candidate list of the bth batch Heuristic information, Indicates glass workpieces are assigned to the second candidate list of the bth batch Heuristic information, Indicates Daidi The pheromone of the oth target when the glass workpiece is assigned to the bth batch.

[0251] Step 3.6.4: Use formula (59) to calculate the Daidi When glass workpieces are assigned to the bth batch, the pheromone of the oth target :

[0252] (59)

[0253] In formula (61), , Indicates The first The glass workpiece and The pheromone of the oth target when the glass workpieces are simultaneously assigned to the bth batch; Indicates the number of glass workpieces contained in the bth batch.

[0254] Step 3.6.5: Use formula (60) to get The first The glass workpiece and The updated pheromone of the oth target when glass workpieces are simultaneously assigned to the bth batch :

[0255] (60)

[0256] Step 3.7: If and and , then go to step 3.4; otherwise, it means that a workshop scheduling plan corresponding to the a-th ant of the k-th generation is obtained.

[0257] Step 3.8: When Assign to , go to step 3.3 and execute sequentially, otherwise, it means that the kth generation is obtained The workshop scheduling plan of each ant is compared with the historical workshop scheduling plan set Merge to get the kth generation merged population, and select the one with the largest fitness The workshop scheduling plan is used as the new historical workshop scheduling plan set ; When k=1, the historical workshop scheduling plan set That is the kth generation A workshop scheduling scheme for ants.

[0258] The present invention uses the historical workshop scheduling plan set for iterative search. Its function is that every time all the ants of the kth generation have constructed a scheduling plan, all the scheduling plans of this generation and the scheduling plans in the historical workshop scheduling plan set are hierarchically sorted by non-dominated scheduling plans, thereby dividing the Pareto scheduling plans of different levels with different advantages and disadvantages. Starting from the first layer of Pareto scheduling plans, the Pareto scheduling plans of each layer are selected according to the congestion distance, and the selection is completed layer by layer until The historical workshop scheduling plan set is reconstructed by using these scheduling plans, which ensures that the best quality previous plan is selected. A scheduling plan.

[0259] Step 3.9, if , then output the latest historical workshop scheduling plan for production, otherwise, execute step 3.10;

[0260] Step 3.10: Use formula (61) to get Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact :

[0261] (61)

[0262] In formula (61), represents the volatility coefficient of the global pheromone, express The workshop scheduling plan of any ant in the Indicates Daidi Class On parallel machines The pheromone change of the oth target of the glass workpiece; and:

[0263] (62)

[0264] In formula (62), Indicates Daidi Class On parallel machines The oth target value corresponding to the oth glass workpiece.

[0265] Step 3.11: Use formula (63) to get +1 generation The glass workpiece and The pheromone of the oth target when the glass workpieces are simultaneously assigned to the bth batch.

[0266] (63)

[0267] In formula (63), Indicates Daidi The glass workpiece and The oth target value when glass workpieces are simultaneously assigned to the bth batch, Indicates Daidi The glass workpiece and The pheromone change of the oth target when the glass workpieces are simultaneously assigned to the bth batch is:

[0268] (64)

[0269] In formula (64), Indicates Daidi The glass workpiece and The oth target value when glass workpieces are simultaneously assigned to the bth batch.

[0270] Step 3.12: Assign to Then, go to step 3.2 and execute it sequentially.

[0271] In this embodiment, an electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the above method, and the processor is configured to execute the program stored in the memory.

[0272] In this embodiment, a computer-readable storage medium stores a computer program on the computer-readable storage medium, and the computer program executes the steps of the above method when executed by a processor.

Claims

1. A hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm, characterized in that: It is used to process each glass workpiece into the first buffer zone after being processed by the cutting parallel machine, and then transport each glass workpiece from the first buffer zone to the spraying parallel machine for processing, and then enter the second buffer zone to wait, and then transport each glass workpiece from the second buffer zone to the batch processing machine to form each batch, and then each batch is tempered on the batch processing machine until all glass workpieces are processed in the production line, and any parallel machine of the cutting parallel machine and the spraying parallel machine is recorded as the first parallel machine. The hybrid flow shop intelligent scheduling optimization method is carried out according to the following steps: Step 1: Establish the first objective function with maximum completion time as the goal And the second objective function with the minimum energy consumption as the goal As the two objective functions of the hybrid flow shop scheduling model, let any oth objective function be , ; Step 2: Establish the constraints of the hybrid flow shop scheduling model; Step 3: Use the historical information ant colony algorithm to solve the hybrid flow shop scheduling model to obtain a production and processing plan.

2. The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm according to claim 1 is characterized in that: In step 1, two objective functions of the hybrid flow shop scheduling model are established using equations (1) and (2): (1) (2) In formula (1), is the maximum completion time, Indicates that the bth batch is in Completion time on batch processing machines; In formula (2), TEC is energy consumption, is the total number of time periods, represents the total number of parallel machines, represents the total number of batch processors, f(t) represents the power consumption in the tth time period, Indicates Taiwan Is the parallel machine in working state in the tth time period? If so, let =1; otherwise, let =0; Indicates Taiwan Is the parallel machine in the tth time period in the standby state? If so, let =1; otherwise, let =0; Indicates Taiwan Is the parallel machine in the tth time period in the shutdown state? If so, let =1; otherwise, let =0; Indicates Is the batch processing machine in working state in the tth time period? If so, let =1; otherwise, let =0; Indicates Is the batch processor in the tth time period in the standby state? If so, let =1; otherwise, let =0; Indicates Is the batch processor in the tth time period in a shutdown state? If so, let =1; otherwise, let =0; , and Respectively represent Energy consumption of a quasi-parallel machine in working, standby and off states; , and Respectively represent the energy consumption of the batch machine in working state, standby state and shutdown state.

3. The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm according to claim 2 is characterized in that: In step 2, the constraints of the hybrid flow shop scheduling model are established using equations (3) to (45): (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) (41) (42) (43) (44) (45) In formula (3) to formula (45), Indicates The glass workpiece is Processing time on a quasi-parallel machine, Indicates The arrival time of the glass workpieces, Indicates The glass workpiece is Completion time on a parallel machine, Indicates The size of the glass workpiece, Indicates The capacity of batch processing machines, Indicates The batch is in Processing time on batch processing machines, Indicates The batch is in The start time of processing on the batch processing machine, Indicates The glass workpiece is Processing time on batch processing machines; Indicates Is the glass workpiece in Taiwan If the type of parallel processing is =1; otherwise, let =0; Indicates Is the glass workpiece in Taiwan If the type of parallel processing is Otherwise, let ; Indicates whether the jth glass workpiece is waiting in the first buffer in the tth time period. If so, let Otherwise, let ; Indicates The qualification constraint of the glass workpiece is whether it can be Taiwan If possible, then Otherwise, let ; Indicates Is the glass workpiece in The first batch processor In a batch, if so, then let Otherwise, let ; Indicates the tth time period Is the batch in If it is processed on a batch processing machine, then Otherwise, let ; Indicates Does the batch contain at least one glass workpiece? If so, Otherwise, let ; is the number of stations i-th in the t-th time period Auxiliary binary variables for parallel machines, Indicates the tth time period Auxiliary binary variables for batch processing; and are the capacities of the first buffer and the second buffer respectively; Indicates high battery; Indicates low battery; Indicates the end time period of high power; Indicates the end time period of low battery; Indicates The batch processor Is it in working state during the time period? If so, =1; otherwise, let =0; Indicates Taiwan Class parallel machine Is it in working state during the time period? If so, =1; otherwise, let =0.

4. The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm according to claim 3 is characterized in that: The step 3 comprises: Step 3.1, define the maximum number of iterations as and the total number of ants is , randomly generate the first preference vector With the second preference vector ; Initialize the heuristic information matrix; define the current generation as k; initialize k=1; initialize the kth generation pheromone matrix ; Step 3.2, set the ant's index a=1; Step 3.3, initialize the parameters of the glass workpiece and the machine, set the first taboo table DE1 of n dimensions, and assign an initial value to DE1 according to the serial number of the glass workpiece. , set the n-dimensional second taboo table DE2 and the n-dimensional third taboo table DE3, and assign initial values ​​to both ; n represents the total number of glass workpieces; Step 3.4: When l=1 and DE1 is not (0,0,...,0), execute The scheduling of the lth parallel machine, thus Taiwan A glass workpiece is processed on a parallel machine, and the processed glass workpiece is stored in buffer 1, and the value of the index position of the processed glass workpiece in DE1 is set to 0, and the value of the index position of the processed glass workpiece in DE2 is modified to the index number of the processed glass workpiece itself; thereby DE1 and DE2 are updated; and then Middle Daidi Taiwan Class parallel machine The oth target of the glass workpiece The pheromone is updated to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece ; Step 3.5: When l=2 and DE2 is not (0,0,...,0), execute Taiwan The scheduling of parallel machines is Taiwan A glass workpiece is processed on a parallel machine, and the processed glass workpiece is stored in buffer 2, and the value of the index position of the processed glass workpiece in DE2 is set to 0, and the value of the index position of the processed glass workpiece in DE3 is modified to the index number of the processed glass workpiece itself; thereby DE2 and DE3 are updated; and then Middle Daidi Taiwan Class parallel machine The oth target of the glass workpiece The pheromone is updated to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece ; Step 3.6, when DE3 is not (0,0,...,0), execute the scheduling of the mth batch processing machine, so that the bth batch is tempered on the mth batch processing machine, and set the value of the index position of the glass workpiece after the tempering process in DE3 to 0, thereby updating DE3; then Middle Daidi Taiwan Class parallel machine The oth target of the glass workpiece The pheromone is updated to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece ; Step 3.7: If and and , then go to step 3.4; otherwise, it means that a workshop scheduling plan corresponding to the a-th ant of the k-th generation is obtained; Step 3.8: When Assign to , go to step 3.3 and execute sequentially, otherwise, it means that the kth generation is obtained The workshop scheduling plan of each ant is compared with the historical workshop scheduling plan set Merge to get the kth generation merged population, and select the one with the largest fitness The workshop scheduling plan is used as the new historical workshop scheduling plan set ; When k=1, the historical workshop scheduling plan set That is the kth generation The workshop scheduling plan for ants; Step 3.9, if , then output the latest historical workshop scheduling plan for production, otherwise, execute step 3.10; Step 3.10: Use formula (61) to get Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact : (61) In formula (61), represents the volatility coefficient of the global pheromone, express The workshop scheduling plan of any ant in the Indicates Daidi Class On parallel machines The pheromone change of the oth target of the glass workpiece; and: (62) In formula (62), Indicates Daidi Class On parallel machines The oth target value corresponding to the glass workpiece; Step 3.11: Use formula (63) to get +1 generation The glass workpiece and The pheromone of the oth target when glass workpieces are simultaneously assigned to the bth batch: (63) In formula (63), Indicates Daidi The glass workpiece and The oth target value when glass workpieces are simultaneously assigned to the bth batch, Indicates Daidi The glass workpiece and The pheromone change of the oth target when the glass workpieces are simultaneously assigned to the bth batch is: (64) In formula (64), Indicates Daidi The glass workpiece and The oth target value when glass workpieces are simultaneously assigned to the bth batch; Step 3.12: Assign to Then, go to step 3.2 and execute it sequentially.

5. The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm according to claim 4 is characterized in that: The step 3.4 comprises: Step 3.4.1: The ath ant randomly selects an unscheduled glass workpiece from DE1, and then selects the ath ant according to the qualification constraints of the glass workpiece using formula (46). Taiwan Class parallel machines, using equations (47) and (48) we can get Taiwan First candidate list for class parallel machines With the second candidate list : (46) (47) (48) In formula (46), represents the oth objective function The preference weight of Indicates Taiwan The oth target value of the parallel machine; in equations (47) and (48), Indicates that the jth glass workpiece arrives at The time of a parallel machine, Indicates Taiwan Completion time of a quasi-parallel machine; Step 3.4.2: Use equation (49) to get the jth glass workpiece reaching the Taiwan First candidate list for class parallel machines Heuristic information , using formula (50) to obtain the jth glass workpiece arriving at the Taiwan Second candidate list for parallel machines Heuristic information ; (49) (50) Step 3.4.3: The ath ant uses formula (51) to calculate the Taiwan The probability of selecting the jth glass workpiece in a parallel machine , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for processing, and storing the processed glass workpiece in buffer 1, and setting the value of the index position of the processed glass workpiece in DE1 to 0, and modifying the value of the index position of the processed glass workpiece in DE2 to the index number of the processed glass workpiece itself; thereby updating DE1 and DE2; (51) In formula (51), and There are 2 weights respectively; Indicates Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact, Indicates The glass workpiece arrives at Taiwan First candidate list for class parallel machines Heuristic information, Indicates The glass workpiece arrives at Taiwan Second candidate list for parallel machines Heuristic information, Indicates Daidi Taiwan Class parallel machine The pheromone of the oth target of the glass artifact; Step 3.4.4: The ath ant uses formula (52) to Update to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece : (52) In formula (52), represents the volatility coefficient of local pheromone, It is the initial pheromone.

6. The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm according to claim 5 is characterized in that: The step 3.5 comprises: Step 3.5.1: The ath ant randomly selects an unscheduled glass workpiece from DE2, and then selects the ath ant according to the qualification constraints of the glass workpiece using formula (46). Taiwan Class parallel machines, using equations (47) and (48) we can get Taiwan First candidate list for class parallel machines With the second candidate list ; Step 3.5.2: Use equation (49) to get the jth glass workpiece reaching the Taiwan First candidate list for class parallel machines Heuristic information , using formula (50) to obtain the jth glass workpiece arriving at the Taiwan Second candidate list for parallel machines Heuristic information ; Step 3.5.3: The ath ant uses formula (51) to calculate the Taiwan The probability of selecting the jth glass workpiece in a parallel machine , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for processing, and storing the processed glass workpiece in buffer 2, and setting the value of the index position of the processed glass workpiece in DE2 to 0, and modifying the value of the index position of the processed glass workpiece in DE3 to the index number of the processed glass workpiece itself; thereby updating DE2 and DE3; Step 3.5.4: The ath ant uses formula (52) to Update to get Daidi Taiwan Class parallel machine The updated pheromone of the oth target of the glass workpiece .

7. The hybrid flow shop intelligent scheduling optimization method based on historical information ant colony algorithm according to claim 6 is characterized in that: The step 3.6 includes: Step 3.6.1, the ath ant uses formula (53) to select the mth batch processor, which is used to add the glass workpieces that meet the qualification constraints and arrive at the mth batch processor earliest to the bth batch, and uses formula (54) and formula (55) to construct the first batch processor that arrives at the mth batch processor earliest. glass workpieces are assigned to the first candidate list of the bth batch With glass workpieces are assigned to the second candidate list of the bth batch : (53) (54) (55) In formulas (56) and (57), The mth batch processing machine The arrival time of the glass workpieces, Indicates the start processing time of the bth batch; Indicates The size of the glass workpiece, represents the capacity of the mth batch processor, Indicates the size of the b-th batch; Step 3.6.2: Use formula (56) to get glass workpieces are assigned to the first candidate list of the bth batch Heuristic information , using formula (57) to obtain glass workpieces are assigned to the second candidate list of the bth batch Heuristic information : (56) (57) Step 3.6.3: The ath ant uses formula (58) to calculate the list of the bth batch. The probability of selecting the jth glass workpiece in , thereby obtaining the selection probabilities of all glass workpieces and selecting the glass workpiece corresponding to the maximum selection probability for tempering treatment, and setting the index position value of the glass workpiece after the tempering treatment in DE3 to 0, thereby updating DE3; (58) In formula (58), and There are 2 weights respectively; Indicates Daidi When the glass workpiece is assigned to the bth batch, the pheromone of the oth target is: Indicates glass workpieces are assigned to the first candidate list of the bth batch Heuristic information, Indicates glass workpieces are assigned to the second candidate list of the bth batch Heuristic information, Indicates Daidi When the glass workpiece is assigned to the bth batch, the pheromone of the oth target; Step 3.6.4: Use formula (59) to calculate Daidi When glass workpieces are assigned to the bth batch, the pheromone of the oth target : (59) In formula (61), , Indicates The first The glass workpiece and The pheromone of the oth target when the glass workpieces are simultaneously assigned to the bth batch; represents the number of glass workpieces contained in the bth batch; Step 3.6.5: Use formula (60) to get The first The glass workpiece and The updated pheromone of the oth target when glass workpieces are simultaneously assigned to the bth batch : (60)。 8. An electronic device, comprising a memory and a processor, characterized in that: The memory is used to store a program that supports the processor to execute the hybrid flow shop scheduling method described in any one of claims 1 to 7, and the processor is configured to execute the program stored in the memory.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the hybrid flow shop scheduling method described in any one of claims 1 to 7 are executed.

Images