Production scheduling method considering multi-cavity imbalance and machine fault

Abstract

The invention provides a production scheduling method considering multi-cavity imbalance and machine faults, which comprises the following steps: step S11, setting a multi-cavity imbalance probability and a machine fault rate for each machine, and setting a corresponding processing threshold upper limit and a corresponding processing threshold lower limit; s12, based on the multi-cavity unbalance probability, the machine failure rate, the upper limit of the machining threshold value and the lower limit of the machining threshold value, the machining time is calculated; s13, on the basis of the machining time, a dual-objective function of the maximum completion time of all the workpieces and the sum of production energy consumption of all the machines is set; and S14, solving by adopting an optimization algorithm to obtain a production scheduling scheme. When a dynamic disturbance event occurs, different tasks are reasonably arranged for processing on different machines, and theoretical and technical support is provided for enterprise processing.

Description

Technical Field

[0001] The invention relates to the technical field of production scheduling and planning, and in particular to a production scheduling method considering multi-cavity imbalance and machine failure. Background Art

[0002] With the rapid development of the manufacturing industry, sustainable development goals have gradually become the focus of attention of countries around the world. Among them, reducing electricity consumption is a key to achieving sustainable development. It can not only help companies reduce operating costs, but also effectively reduce carbon emissions and alleviate global problems such as resource scarcity. At the same time, in the context of manufacturing scheduling, optimizing complex processing workshops is a key challenge. However, in actual processing, changes such as order insertion, cancellation and modification, machine malfunction, wear, and human factors are inevitable. These interferences make the static processing scheduling plan far from the expected effect, seriously affecting the processing efficiency. In dynamic events, the machine deterioration effect is inevitable in actual processing. Unlike machine malfunction, the deterioration effect causes the machine processing performance to decline, but considering the actual production cost, the processing usually continues. For example, in the medical device processing workshop, when the mold cavity is damaged during the injection molding extrusion process, the plugging operation is often used instead of the overall mold replacement, forming a multi-cavity imbalance interference event. Multi-cavity imbalance occurs in the injection molding extrusion stage of plastic parts processing.

[0003] However, most of the current processing in enterprises is based on processing experience, which makes it difficult to plan processing strategies efficiently, and thus cannot guarantee the theory of sustainable development of low power consumption. At the same time, the existing technology lacks dynamic processing methods such as multi-cavity imbalance, ignoring the impact of multi-cavity imbalance on the actual processing process, but multi-cavity imbalance often occurs in processing workshops such as building materials, molds, and medical devices that have injection molding processing tasks. Therefore, how to carry out production scheduling planning while considering multi-cavity imbalance is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The present invention proposes a production scheduling method that takes into account multi-cavity imbalance and machine failure to solve the technical problem that the existing scheduling planning method does not take into account the situations of multi-cavity imbalance and machine failure, resulting in the generated scheduling scheme being difficult to apply.

[0005] In order to solve the above technical problems, the present invention provides a production scheduling method considering multi-cavity imbalance and machine failure, comprising the following steps: Step S11: setting the multi-cavity imbalance probability and machine failure rate for each machine, and setting the corresponding upper and lower limits of the processing threshold; Step S12: Calculating the processing time based on the multi-cavity imbalance probability, the machine failure rate, the upper processing threshold value and the lower processing threshold value; Step S13: Based on the processing time, a dual objective function of the maximum completion time of all workpieces and the total production energy consumption of all machines is set; Step S14: Use an optimization algorithm to solve and obtain a production scheduling solution.

[0006] Preferably, in step S12, the processing time is calculated The expression is: ; ; ; In the formula, T ijk Indicates the process O ij In the machine M k The time required for processing on WT k Indicates the machine M k Cumulative processing time; Indicates the machine M k The lower processing threshold of Indicates the machine M k The upper processing threshold of e k Indicates the machine M k The probability of multi-chamber imbalance; N Represents a collection of artifacts; J i Representation of workpiece i The corresponding process set; M represents a collection of machines; n Represents a random number between 0 and 1; R k Indicates the machine k The time required to repair the fault; Indicates the machine k machine failure rate; Indicates the machine M k Load time; b ijk represents the decision variable, the machine If the process is normal, it is 1; otherwise, it is 0.

[0007] Preferably, in step S13, the maximum complete processing time is expressed as: ; The total processing power consumption of all machines is expressed as: ; In the formula, Indicates the machine M k Processing power; Indicates the machine M k Standby power; T ijk Indicates the process O ij In the machine M k The time required for processing on T i Representation of workpiece i Total processing time; E ij Indicates the process O ij In the machine M k The processing completion time on E k Indicates the machine M k The processing end time; a ijk represents the decision variable, if the process O ij Select the machine M k If the process is on, it is 1; otherwise, it is 0.

[0008] Preferably, when solving the problem in step S14, one or more of the following constraints are set: 1) Each process of each workpiece can only be processed by one feasible processing machine: ; In the formula, N Represents a collection of artifacts; J i Representation of workpiece i The corresponding process set; M represents a collection of machines; a ijk represents the decision variable, if the process O ij Select the machine M k If the processing is done, it is 1, otherwise it is 0; 2) There are sequential constraints between different workpieces on the same processing machine: ; In the formula, S ij Indicates the processO ij The processing start time; T ijk Indicates the process O ij In the machine M k The time required for processing on 3) The completion time of the workpiece must not be less than the sum of the start time of the workpiece and the processing time of the workpiece: ; In the formula, C i express J i completion time; S i Representation of workpiece i The start time of R k Indicates the machine k The time required to repair the fault; 4) Each workpiece must be processed according to its specific process route, and there is a sequence between the processes of the same workpiece: ; In the formula, E ij Indicates process operation O ij In the machine M k The processing completion time on 5) Once the workpiece starts processing, it cannot be interrupted: .

[0009] Preferably, in step S14, a multi-objective genetic programming method is used to solve the problem, including the following steps: Step S21: completing population initialization according to workpiece processing information and encoding and decoding strategies; Step S22: performing grafting, mutation and elite selection operations; Step S23: When the algorithm running time reaches the maximum number of running times, the algorithm terminates and a planning solution is obtained.

[0010] Preferably, the encoding strategy in step S21 includes: setting a workpiece arrangement sequence and a machine arrangement sequence; the workpiece arrangement sequence and the machine arrangement sequence are encoded using a binary tree; the elements in the binary tree are all terminal symbol set rules; the terminal symbol set uses "+", "-", "x" and " / " as a function symbol set for connection; The terminal symbol set rules for the workpiece arrangement sequence include: 1) NAPT: average processing time of the next process; 2) NOR: number of remaining processes; 3) NAO: the average number of machines available for the next process of the workpiece; 4) NCO: the number of operations completed by the current workpiece; 5) EAT: the earliest machinable time of the workpiece; 6) TS: the time the workpiece has been processed; 7) PTC: Processing time of workpiece in current process; 8) EER: Energy efficiency ratio, which is the ratio of energy required to complete a task to processing time; 9) WT workpiece waiting time; 10) Average energy consumption of the next process of ECN; The terminal symbol set rules for machine arrangement sequences include: 1) MLU: the machine with the lowest utilization rate; 2) MSPT: the earliest available time of the machine; 3) SWT: The shortest machine processing time for the next component.

[0011] Preferably, the decoding strategy in step S21 includes: inputting workpiece terminal symbol set rules, machine terminal symbol set rules, function symbol set, population size and binary tree depth; performing individual decoding, selecting workpieces and machines one by one for allocation; recording the processing time of each machine, the workpiece processing status and the current maximum completion time during allocation, waiting for the workpiece processes to be arranged, and determining the completion time of each workpiece E ij , the maximum completion time of production is , the current individual decoding is completed; determine the operation time window and maximum completion time of all tasks arranged by all binary tree rules in the population, and the decoding is completed.

[0012] Preferably, the grafting operation in step S22 includes: replacing branches by pruning the binary tree to compare the quality of the result of the grafted binary tree; according to the greedy rule, taking the better new binary tree as the winning tree; if the quality of the binary tree is not improved, keeping the original binary tree unchanged.

[0013] Preferably, the mutation operation in step S22 includes: selecting a binary tree, pruning the terminal symbol rules or function symbols in the branches, and mutating them into other rules in the set; according to the greedy rule, if the mutated binary tree is excellent, it is used as the winning tree, and if the quality of the binary tree is not improved, the original binary tree is kept unchanged; wherein the mutated category is subjected to high-quality mutation according to the mutation probability of the terminal symbol of the high-quality binary tree.

[0014] Preferably, the elite selection operation in step S22 includes: evaluating the multi-objective processing binary tree scheduling result by using the hypervolume index, assuming that the reference point is , then the HV value is the binary tree scheduling result x To reference point The hypervolume between is expressed as: ; Calculate the selection probability of a binary tree P ( x i )for: ; In the formula, represents the maximum completion time; It represents the total production energy consumption of all machines; N Represents a collection of parts; Then, non-dominated sorting is used for elite selection, and the ones with high sorting levels are used as new binary tree groups. The frequency of terminal symbols in the new binary tree groups is counted as the mutation probability. The non-dominated sorting includes: 1) Calculate the crowding degree of the new population; 2) Determine the terminator mutation probability based on the combined scheduling rule with a high dominance level; 3) According to the set population number, select individuals with high dominance level to form a new population.

[0015] The beneficial effects of the present invention include at least: the present invention provides a mathematical model for multi-cavity imbalance and machine malfunction processing that occurs in the actual processing of injection molding companies such as medical devices, plates and various types of plastic products. For processing tasks in the two situations of machine failure and multi-cavity imbalance, the present invention selects a processable machine for processing under the premise of satisfying the processing sequence. When a dynamic disturbance event occurs, different tasks are reasonably arranged to be processed on different machines, providing theoretical and technical support for enterprise processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of a method flow chart of an embodiment of the present invention; Figure 2 A schematic diagram of a multi-cavity imbalance in an enterprise according to an embodiment of the present invention; Figure 3 A schematic diagram of a production scheduling machine failure and multi-cavity imbalance according to an embodiment of the present invention; Figure 4 Scatter plots of processing time and energy consumption for all cases considered in performance evaluation in five independent tests comparing embodiments of the present invention; Figure 5 The Pareto surface and Gantt chart of the embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the protection scope of the present invention.

[0018] Example 1 like Figure 1 As shown, an embodiment of the present invention provides a production scheduling method taking multi-cavity imbalance and machine failure into consideration.

[0019] Before describing the embodiment, the present embodiment first describes the production task taking into account multi-cavity imbalance and machine failure as follows.

[0020] Multi-cavity imbalance often occurs when the mold cavity is damaged. Because workers will seal the damaged muzzle, muzzle imbalance will occur during the processing. In the production of medical devices, this deterioration effect is particularly obvious, such as Figure 2 As shown. Multi-cavity imbalance also often occurs in general plastics, complete vehicles and mold production. Multi-cavity imbalance does not mean that the machine has failed. More importantly, in order to ensure profits during the processing, multi-cavity imbalance is allowed to exist without affecting the integrity of the processing. At the same time, it cannot be ignored that machine failure is still one of the most common disturbance events in actual production. In the fault stage, if the workpiece fails during the processing stage, you can still choose to postpone the processing or reschedule it according to the processing status and recovery time. Figure 3 The processing tasks of the two cases of machine failure and multi-cavity imbalance shown in the figure are processed by selecting a processable machine under the premise of meeting the processing sequence. The difficulty of this problem lies in how to reasonably arrange different tasks to be processed on different machines and optimize the evaluation indicators when a dynamic disturbance event occurs.

[0021] The difference between multi-cavity imbalance and machine failure is that: ① When multi-cavity imbalance occurs, the machine can still continue processing; ② As the processing progresses, the cavity will be randomly blocked, and the degree of blockage will directly affect the processing time.

[0022] When the number of processing machines is fixed, the operation time of different workpieces in different machines is different, and the optimization goal is to minimize the processing time and total processing energy consumption. The processing time takes the maximum complete processing time of all machines. The complete processing time of each machine includes processing operation time and standby time. Standby time is mainly the waiting time caused by priority constraints. The machine will generate corresponding energy consumption in different processing time periods, that is, the machine production energy consumption includes processing operation energy consumption and standby energy consumption.

[0023] The method of this embodiment includes the following steps: Step S11: setting the multi-cavity imbalance probability and machine failure rate for each machine, and setting the corresponding upper and lower limits of the processing threshold.

[0024] Specifically, first, according to the processing equipment conditions, determine the interference factors that occur during the processing and the machines that have multi-cavity imbalance and failures during the processing; and according to the processing information of the production order, determine the process of each workpiece, the number of processable machines and the processable machine information corresponding to each process.

[0025] For the convenience of description, the symbols and decision variables used in the embodiments of the present invention are as follows: i : Workpiece number, j : Workpiece The total number of processes, k : Machine number, workpiece set is N , the machine set is M ; J i : Workpiece i Corresponding process; M k : No. k Machines; P k :machine M k Processing power; P´k :machine M k Standby power; O ij : Workpiece i No. j process; E k :machine M k The processing end time; e k :machine M k The probability of multi-chamber imbalance; WT k :machine M k Cumulative processing time; :machine M k The lower processing threshold of :machine M k The upper processing threshold of C i : J i completion time; T ijk :Process O ij In the machine M k The processing time required E ij :Process O ij In the machine M k The processing completion time on S ij :Process O ij The processing start time; n : A random number between 0 and 1; R k :machine k The time required to repair the fault; T i : Workpiece i Total processing time; S i : Workpiece i Start time a ijk :Decision variables, if the process O ij Select the machine M k If the processing is done, it is 1, otherwise it is 0; b ijk : Decision variables, machine If the processing is normal, it is 1, otherwise it is 0; Step S12: Calculate the processing time based on the multi-cavity imbalance probability, the machine failure rate, the upper processing threshold value and the lower processing threshold value.

[0026] Specifically, the processing time Calculate using the following expression.

[0027] ; ; ; In the formula, T ijk Indicates the process O ij In the machine M k The time required for processing on WT k Indicates the machine M k Cumulative processing time; Indicates the machine M k The lower processing threshold of Indicates the machine M k The upper processing threshold of e k Indicates the machine M k The probability of multi-chamber imbalance; N Represents a collection of artifacts; J i Representation of workpiece i The corresponding process set; M represents a collection of machines; n Represents a random number between 0 and 1; R k Indicates the machine k The time required to repair the fault; Indicates the machine k machine failure rate; Indicates the machine M k Load time; b ijk represents the decision variable, the machine If the process is normal, it is 1; otherwise, it is 0.

[0028] Step S13: Based on the processing time, a dual objective function of the maximum completion time of all workpieces and the total production energy consumption of all machines is set.

[0029] Specifically, the completion time and processing energy consumption are minimized, and the expression is: ; Sub-goal 1: The maximum completion time of all workpieces, expressed as: .

[0030] Sub-goal 2: The total production energy consumption of all machines is expressed as: .

[0031] Step S14: Use an optimization algorithm to solve and obtain a production scheduling solution.

[0032] The optimization algorithms used in this embodiment include, but are not limited to, genetic algorithm, particle swarm optimization, non-dominated sorting genetic algorithm, differential evolution algorithm and simulated annealing algorithm.

[0033] In order to better meet the actual processing conditions, the following constraints are set in this embodiment.

[0034] 1) Each process of each workpiece can only be processed by one feasible processing machine: ; In the formula, N Represents a collection of artifacts; J i Representation of workpiece i The corresponding process set; M represents a collection of machines; a ijk represents the decision variable, if the process O ij Select the machine M k If the processing is done, it is 1, otherwise it is 0; 2) There are sequential constraints between different workpieces on the same processing machine: ; In the formula, S ij Indicates the process O ij The processing start time; T ijk Indicates the process O ij In the machine M k The time required for processing on 3) The completion time of the workpiece must not be less than the sum of the start time of the workpiece and the processing time of the workpiece: ; In the formula, C i express J i completion time; S i Representation of workpiece i The start time of 4) Each workpiece must be processed according to its specific process route, and there is a sequence between the processes of the same workpiece: ; In the formula, E ij Indicates process operation O ij In the machine M k The processing completion time on 5) Once the workpiece starts processing, it cannot be interrupted: .

[0035] Example 2 This embodiment provides an optimization algorithm based on Embodiment 1. By designing an adaptive terminator mutation mechanism for the genetic programming algorithm, an improved multi-objective genetic programming algorithm is obtained to replace the traditional optimization algorithm in step S14 of Embodiment 1. The solution method is as follows.

[0036] Step S21: initializing algorithm parameters; Step S22: completing population initialization according to workpiece processing information and encoding and decoding strategies; Step S23: executing the grafting stage, mutation stage, and elite selection stage respectively; Step S24: When the algorithm running time reaches the maximum number of running times, the algorithm terminates and outputs the optimization result.

[0037] Specifically, in the coding strategy, the production and processing arrangement includes the workpiece arrangement sequence and the machine arrangement sequence. The workpiece and machine arrangement sequence are encoded using a binary tree. The elements of the binary tree are all terminal symbol set rules. The terminal symbol set uses "+", "-", "x", and " / " as function symbol sets for connection. The terminal symbol set rules of the workpiece arrangement are: ①NAPT: the average processing time of the next process; ②NOR: the number of remaining processes; ③NAO: the average number of machines available for the next process of the workpiece; ④NCO: the number of operations completed by the current workpiece; ⑤EAT: the earliest processable time of the workpiece; ⑥TS: the time the workpiece has been processed; ⑦PTC: the processing time of the workpiece in the current process; ⑧EER: energy efficiency ratio (the ratio of energy required to complete the task to processing time); ⑨WT workpiece waiting time; ⑩ECN the average energy consumption of the next process. The terminal set rules of machine arrangement are: ①MLU: the machine with the lowest utilization rate; ②MSPT: the earliest available time of the machine; ③SWT: the shortest processing time of the machine for the next workpiece.

[0038] The specific encoding process is as follows: ① Input the workpiece terminal symbol set rule, machine terminal symbol set rule, function symbol set, population size, and binary tree depth; ② According to the determined binary tree depth, the terminal set is used to fill in the complete production binary tree rule, in which some nodes are not filled in as incomplete binary tree rules; When decoding, it is necessary to determine the time window for workpiece processing. When arranging workpieces and corresponding machines according to the generation rules, it is necessary to consider the process sequence of the workpieces. When adding time windows, it is necessary to make sure that there is no idle time between people, so as to ensure the shortest complete processing time and the minimum energy consumption. The specific decoding process is as follows: Step S221: inputting workpiece terminal symbol set rules, machine terminal symbol set rules, function symbol set, population size P, and binary tree depth D; Step S222: Execute individual decoding, select workpieces and machines one by one for allocation; Step S223: When allocating, record the processing time of each machine, the workpiece processing status, the current maximum completion time, wait for the workpiece process to be arranged, and determine the completion time of each workpiece E ij , the maximum completion time of production is , the current individual decoding is completed.

[0039] Step S224: Determine the operation time windows and maximum completion times of all tasks arranged by all binary tree rules in the population, and the decoding is completed.

[0040] In the grafting stage, the quality of the grafted binary tree is compared by replacing branches through pruning. According to the greedy rule, the better new binary tree is selected as the winning tree. If the quality of the binary tree does not improve, the original binary tree remains unchanged.

[0041] The specific steps of the grafting stage are as follows: ① Input binary tree individuals ; ② Randomly select different Binary tree , randomly select different branches, perform grafting, and get a new binary tree and ; ③If , then use the non-dominated solution Replace binary tree ;like , then use the non-dominated solution Replace binary tree ; Otherwise, no non-dominated solution is found, and the binary tree is maintained No change; ④ Output binary tree group .

[0042] In the mutation phase, a binary tree is selected and mutated to other rules in the set by pruning the terminal symbol rules or function symbols in the branches. According to the greedy rule, if the mutated binary tree is excellent, it will be the winning tree. If the quality of the binary tree is not improved, the original binary tree will be kept unchanged. The mutated category is mutated according to the mutation probability of the terminal symbol of the high-quality binary tree.

[0043] The specific steps of the mutation phase are as follows: ① Input binary tree individuals ; ② Randomly select terminal symbols or function symbols on the branches, perform mutation, and obtain a new binary tree ; ③If , then use the non-dominated solution Replace binary tree , otherwise no non-dominated solution is found, and the binary tree is maintained No change; ④ Output binary tree group .

[0044] In the elite selection stage, the multi-objective production binary tree scheduling results are evaluated by the super volume index, and the reference point is set as , then the HV value is the binary tree scheduling result x To reference point The hypervolume between is expressed as ; The selection probability of a binary tree can be calculated P ( x i )for ; Use non-dominated sorting for elite selection, take the high sorting level as the new binary tree group, and count the frequency of terminal symbols in the new binary tree group as the mutation probability. The non-dominated sorting includes 1) Calculate the crowding degree of the new population; 2) Determine the terminator mutation probability based on the combined scheduling rule with a high dominance level; 3) According to the set population number, select individuals with high dominance level to form a new population.

[0045] Example 3 This embodiment takes the production workshop of a medical infusion device manufacturing enterprise as an example, performs performance analysis on the planning method of embodiment 1 and the optimization algorithm of embodiment 2, constructs production problems of multi-cavity imbalance and machine failure, and analyzes the application performance of the method of the present invention in actual engineering cases. There are 10 types of workpieces to be produced, with a quantity of 100,000 pieces each. The number of processes for each workpiece is 3 and has strict sequence constraints. The processing equipment and its related energy consumption information are shown in Table 1, and the detailed information of the processing tasks is shown in Table 2. Table 1 gives the types of available machines, and gives the operating energy consumption and standby energy consumption. Table 2 gives the processes of 10 types of workpieces and the machines available for each process, and gives the processing time on each machine.

[0046] Table 1

[0047] Table 2

[0048] The dispatching rules of 10 terminal symbol sets (NAPT, NOR, NAO, NCO, EAT, TS, PTC, EER, WT, ECN), genetic programming (GP) and the designed multi-objective genetic programming algorithm (MOGPA) were compared. The population size of all algorithms was set to 200, and the algorithm termination condition was to run 500 times. The crossover and mutation probabilities of MOGPA were set to 0.8 and 0.2 respectively. The dispatching rules of the optimal terminal symbol set, GP and MOGPA were compared under 5 standard data sets. Figure 4 shown.

[0049] from Figure 4 It can be seen that compared with the optimal scheduling rules of 10 terminal symbol sets, MOGPA achieved relatively excellent results, with a maximum optimization rate of 30% in terms of total construction period and total energy consumption. The deviation box plot in the lower right corner is compared with the standard of the benchmark rule. Compared with the ordinary scheduling rules, the vast majority of cases achieved optimality under the same objectives, further proving the superiority of MOGPA in handling production under multi-cavity imbalance and machine failure.

[0050] In the above example, each machine will fail in the whole example, and the maximum number of failures is set to 5. Multiple cavity imbalances will occur on injection molding and extrusion machines. Combined with the actual production situation, the maximum number of cavity imbalances on the two types of machines is 50%. The solution results are as follows: Figure 5 shown.

[0051] Figure 5 (a) is the Pareto surface of the solution case. The processing Gantt chart is drawn with the solution result of maximum completion time of 65 and total energy consumption of 25.3409. Figure 5 (b), where the red frame line represents machine failure and the dashed line following the processing represents multi-cavity imbalance. Figure 5 (c) is a Gantt chart with the goal of minimizing the maximum completion time. The optimal completion time is 55. Figure 5 (d) is a Gantt chart with the goal of minimizing total energy consumption, and the minimum energy consumption is 24.985. (c) and (d) are both optimal solutions obtained without disturbance factors. Combining (a) and (b), it can be found that in the actual case, the algorithm proposed in this embodiment has achieved the optimal solution in considering the maximum completion time and energy consumption.

[0052] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. Only the preferred embodiments of the present invention are expressed. The description is more specific and detailed, but it cannot be understood as limiting the scope of the present invention. As long as there is no contradiction in the combination of these technical features, they should be considered as within the scope of this specification.

[0053] It should be noted that, for those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these modifications and improvements all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A production scheduling method considering multi-cavity imbalance and machine failure, characterized by: The following steps are involved: Step S11: setting the multi-cavity imbalance probability and machine failure rate for each machine, and setting the corresponding upper and lower limits of the processing threshold; Step S12: Calculating the processing time based on the multi-cavity imbalance probability, the machine failure rate, the upper processing threshold value and the lower processing threshold value; Step S13: Based on the processing time, a dual objective function of the maximum completion time of all workpieces and the total production energy consumption of all machines is set; Step S14: Use an optimization algorithm to solve and obtain a production scheduling solution.

2. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 1, characterized in that: In step S12, the processing time is calculated The expression is: ； ； ； In the formula, T ijk Indicates the process O ij In the machine M k The time required for processing on WT k Indicates the machine M k Cumulative processing time; Indicates the machine M k The lower processing threshold of Indicates the machine M k The upper processing threshold of e k Indicates the machine M k The probability of multi-chamber imbalance; N Represents a collection of artifacts; J i Representation of workpiece i The corresponding process set; M represents a collection of machines; n Represents a random number between 0 and 1; R k Indicates the machine k The time required to repair the fault; Indicates the machine k machine failure rate; Indicates the machine M k Load time; b ijk represents the decision variable, the machine If the process is normal, it is 1; otherwise, it is 0.

3. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 1, characterized in that: In step S13, the maximum complete processing time is expressed as: ； The total processing power consumption of all machines is expressed as: ； In the formula, Indicates the machine M k Processing power; Indicates the machine M k Standby power; T ijk Indicates the process O ij In the machine M k The time required for processing on T i Representation of workpiece i Total processing time; E ij Indicates the process O ij In the machine M k The processing completion time on E k Indicates the machine M k The processing end time; a ijk represents the decision variable, if the process O ij Select the machine M k If the process is on, it is 1; otherwise, it is 0.

4. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 1, characterized in that: In step S14, one or more of the following constraints are set when solving the problem: 1) Each process of each workpiece can only be processed by one feasible processing machine: ； In the formula, N Represents a collection of artifacts; J i Representation of workpiece i The corresponding process set; M represents a collection of machines; a ijk represents the decision variable, if the process O ij Select the machine M k If the processing is done, it is 1, otherwise it is 0; 2) There are sequential constraints between different workpieces on the same processing machine: ； In the formula, S ij Indicates the process O ij The processing start time; T ijk Indicates the process O ij In the machine M k The time required for processing on 3) The completion time of the workpiece must not be less than the sum of the start time of the workpiece and the processing time of the workpiece: ； In the formula, C i express J i completion time; S i Indicates the start time of the workpiece; 4) Each workpiece must be processed according to its specific process route, and there is a sequence between the processes of the same workpiece: ； In the formula, E ij Indicates process operation O ij In the machine M k The processing completion time on 5) Once the workpiece starts processing, it cannot be interrupted: 。 5. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 1, characterized in that: In step S14, a multi-objective genetic programming method is used to solve the problem, including the following steps: Step S21: completing population initialization according to workpiece processing information and encoding and decoding strategies; Step S22: performing grafting, mutation and elite selection operations; Step S23: When the algorithm running time reaches the maximum number of running times, the algorithm terminates and a planning solution is obtained.

6. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 5, characterized in that: The encoding strategy in step S21 includes: setting a workpiece arrangement sequence and a machine arrangement sequence; the workpiece arrangement sequence and the machine arrangement sequence are encoded using a binary tree; the elements in the binary tree are all terminal symbol set rules; the terminal symbol set uses "+", "-", "x" and " / " as a function symbol set for connection; The terminal symbol set rules for the workpiece arrangement sequence include: 1) NAPT: average processing time of the next process; 2) NOR: number of remaining processes; 3) NAO: the average number of machines available for the next process of the workpiece; 4) NCO: the number of operations completed by the current workpiece; 5) EAT: the earliest machinable time of the workpiece; 6) TS: the time the workpiece has been processed; 7) PTC: Processing time of workpiece in current process; 8) EER: Energy efficiency ratio, which is the ratio of energy required to complete a task to processing time; 9) WT workpiece waiting time; 10) Average energy consumption of the next process of ECN; The terminal symbol set rules for machine arrangement sequences include: 1) MLU: the machine with the lowest utilization rate; 2) MSPT: the earliest available time of the machine; 3) SWT: The shortest machine processing time for the next component.

7. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 6, characterized in that: The decoding strategy in step S21 includes: inputting workpiece terminal symbol set rules, machine terminal symbol set rules, function symbol set, population size and binary tree depth; performing individual decoding, selecting workpieces and machines one by one for allocation; recording the processing time of each machine, the workpiece processing status and the current maximum completion time during allocation, waiting for the workpiece processes to be arranged, and determining the completion time E of each workpiece ij , the maximum completion time of production is , the current individual decoding is completed; determine the operation time window and maximum completion time of all tasks arranged by all binary tree rules in the population, and the decoding is completed.

8. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 7, characterized in that: The grafting operation in step S22 includes: replacing branches by pruning the binary tree to compare the quality of the result of the grafted binary tree; according to the greedy rule, taking the better new binary tree as the winning tree; if the quality of the binary tree is not improved, keeping the original binary tree unchanged.

9. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 8, characterized in that: The mutation operation in step S22 includes: selecting a binary tree, pruning the terminal symbol rules or function symbols in the branches, and mutating them into other rules in the set; according to the greedy rule, if the mutated binary tree is excellent, it is used as the winning tree; if the quality of the binary tree is not improved, the original binary tree is kept unchanged; wherein the mutated category is subjected to high-quality mutation according to the mutation probability of the terminal symbol of the high-quality binary tree.

10. A production scheduling method considering multi-cavity imbalance and machine failure according to claim 9, characterized in that: The elite selection operation in step S22 includes: evaluating the multi-objective processing binary tree scheduling result by using the hypervolume index, assuming that the reference point is , then the HV value is the binary tree scheduling result x To reference point The hypervolume between is expressed as: ； Calculate the probability of selection of a binary tree P ( x i )for: ； In the formula, represents the maximum completion time; It represents the total production energy consumption of all machines; N Represents a collection of parts; Then, non-dominated sorting is used for elite selection, and the ones with high sorting levels are used as new binary tree groups. The frequency of terminal symbols in the new binary tree groups is counted as the mutation probability. The non-dominated sorting includes: 1) Calculate the crowding degree of the new population; 2) Determine the terminator mutation probability based on the combined scheduling rule with a high dominance level; 3) According to the set population number, select individuals with high dominance level to form a new population.

Images