A Cloud Platform Scheduling Method for Rational Allocation of Enterprise Production Resources Based on Processing and Manufacturing Granularity

Abstract

The present invention discloses a cloud platform scheduling method for the rational allocation of enterprise production resources based on the processing and manufacturing granularity. The cloud platform scheduling includes the establishment of subtasks and the initial candidate set and the final candidate set of the enterprise, and random matching is performed to form an initial population. Aiming at the problem of making full use of enterprise manufacturing resources, a strategy that an enterprise can process multiple tasks simultaneously is proposed, and a related model is established. A scheduling mathematical model is established by combining the strategy that an enterprise can process multiple tasks simultaneously with processing production constraints. A multi-objective function is established in combination with the mathematical model for scheduling in combination with an optimization algorithm. The present invention establishes a multi-objective model based on the makespan and the sum of the remaining manufacturing granularities in the fully loaded state of the enterprise. Through this scheduling method, the problem of making full use of enterprise resources can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of cloud manufacturing scheduling, and is applicable to collaborative manufacturing in enterprises. Specifically, it relates to a cloud platform scheduling method for reasonable allocation of enterprise production resources based on machining and manufacturing granularity. Background Art

[0002] With the development of Internet technology in the information field, the traditional manufacturing mode in the manufacturing industry has gradually merged with Internet technology, and the cloud manufacturing mode has emerged. Cloud manufacturing makes up for many deficiencies in traditional manufacturing, such as scattered enterprise resources, low utilization rate, and inability to quickly respond to market demands, and breaks through the constraints of factors such as geographical distance on the manufacturing mode, manufacturing demands, and manufacturing resources of enterprises. Through Internet technology, enterprises in different regions can carry out collaborative manufacturing, and reasonably optimize the allocation of enterprise resources to effectively utilize the superior resources and idle resources of enterprises.

[0003] Currently, research on resource scheduling in cloud manufacturing mainly focuses on scheduling between shop floors, and relatively few studies focus on scheduling between enterprises or at the enterprise level. In the scheduling between enterprises, the traditional scheduling method is that each enterprise can only accept one task at the same time. At this time, if the machines and manpower required for the task volume of an enterprise are less and it cannot accept another task, it will cause waste of resources. Therefore, researching a scheduling method that allows enterprises to accept multiple tasks simultaneously is of great significance for the full utilization of enterprise resources. Summary of the Invention

[0004] A cloud platform scheduling method for reasonable allocation of enterprise production resources based on machining and manufacturing granularity proposed by the present invention, on the basis of the characteristics of cloud manufacturing with enterprises as units, proposes a strategy that an enterprise can process multiple tasks simultaneously for the problem of full utilization of enterprise manufacturing resources, and establishes relevant mathematical models and scheduling models.

[0005] A cloud platform scheduling method for reasonable allocation of enterprise production resources based on machining and manufacturing granularity mainly includes the following steps:

[0006] Step 1: Randomly match subtasks with enterprises to form an initial population.

[0007] The platform gives a task set and decomposes it into a subtask set. Analyze the processing types and required manufacturing granularities of the subtasks, and classify the types of enterprises and subtasks according to the information filled in by the platform. According to the information filled in by the platform, decompose the task into multiple subtasks with processing sequences and processing types.

[0008] The processing and manufacturing granularity covers four types of processing resources: numerical control machine tools, human resources, industrial software systems, and product blanks. According to the actual production situation, the four resources are constrained (one human resource can operate 2 numerical control machine tools and 1 industrial software to process 1 cubic meter of product blanks). When the corresponding constraint relationships are achieved among the four resources, they can be converted into 1 processing and manufacturing granularity. The specific mathematical model is as follows:

[0009] g(n) = p(n) + m(n) + w(n) + r(n) (1)

[0010] All subtasks in the subtask set are matched with all enterprises in terms of processing types to form a preliminary candidate set. The subtasks are matched with enterprises to form an enterprise candidate set for the subtasks. The required processing types of the subtasks are compared with the processing types of the enterprises. If the processing types are the same, this enterprise enters the preliminary candidate set corresponding to the subtask. If the processing types are different, then the next enterprise is compared until all enterprises have been compared. The relevant constraint model is:

[0011]

[0012] Q a,s,c = F z,b ·b (3)

[0013] All subtasks are compared with the enterprises in their preliminary candidate sets in terms of processing and manufacturing granularity to form a final candidate set. The subtasks are matched with the enterprises in their preliminary candidate sets in terms of processing and manufacturing granularity. If the processing and manufacturing granularity of this enterprise is greater than the processing and manufacturing granularity required by this subtask, then this enterprise enters the final candidate enterprise set of the subtask. If the processing and manufacturing granularity of the enterprise is less than the processing and manufacturing granularity required by the subtask, then the next one is compared until all enterprises have been compared.

[0014] Z b = Z a,s , b ∈ Q a,s,c (4)

[0015]

[0016] Q a,s,f = F z,b ·F g,b ·b (6)

[0017] The subtask randomly selects enterprises from the final candidate set to form an initial sorting. Three matrices of the initial population are randomly formed, namely tasks, subtasks, and corresponding enterprises. Enterprises are matched according to the order in which the subtasks appear. After an enterprise matches a corresponding subtask, the corresponding processing and manufacturing granularity of the enterprise decreases, and the processing and manufacturing granularity of the enterprise is updated to the remaining granularity. When the subsequent subtasks perform random enterprise matching, if the enterprise is matched again, the remaining processing and manufacturing granularity of the enterprise is compared with the processing and manufacturing granularity of the subtask. If it is greater than the processing and manufacturing granularity of the subtask, processing can be carried out.

[0018] Bu b,g＝ B b,g -Bx b,g , b ∈ Q a,s,f (7)

[0019] Z b = Z a,s (8)

[0020]

[0021] Step 2: Regarding the problem of making full use of enterprise manufacturing resources, a strategy for an enterprise to process multiple tasks simultaneously is proposed, and a related model is established.

[0022] When the remaining processing and manufacturing granularity of the enterprise is not sufficient to support any subtask, if a subtask matches this enterprise again, processing will wait. When a previous subtask is completed and the released processing and manufacturing granularity is more than the granularity required by the currently waiting subtask for processing, processing will start.

[0023] g{min(TF a,b )} > Ts a+1,s,b (10)

[0024] Z b = Z(Ts a+1,s,b ) (11)

[0025] Ts a+1,s,b = min(TF a,b ) + Td a,b {min(TF a,b )}, a = 1, 2,....A (12)

[0026] Step 3: Combine a strategy for an enterprise to process multiple tasks simultaneously with processing and production constraints to establish a scheduling mathematical model.

[0027] Based on the maximum processing time of the enterprise's processing tasks, a mathematical model for the maximum completion time of tasks is established, mainly including the waiting time, processing time, transportation time of tasks, and time for various uncertain factors. The mathematical model is as follows:

[0028] Tmax = max(Ta) (13)

[0029]

[0030] T a,s = Tw a,s + T a,s,b + Tr a,s + Tu a,s (15)

[0031] Tu a,s = χ1·Tw a,s + χ2·T a,s,b + χ3·Tr a,s (16)

[0032] Establish a mathematical model of processing and manufacturing granularity in working hours, including the unused remaining manufacturing granularity during the processing tasks of all enterprises. The mathematical model is as follows:

[0033] gk min = min(gk) (17)

[0034]

[0035] Step 4: Establish a multi-objective function in combination with the mathematical model for scheduling in combination with the optimization algorithm

[0036] Combined with the diverse scheduling requirements in forging production, establish a multi-objective function with the maximum completion time and the minimum unused remaining manufacturing granularity during enterprise processing tasks as the objectives:

[0037] F(x) = min(f1, f2) (19)

[0038] f1 = T max (20)

[0039] f2 = gk min (21)

[0040] The technical details of each symbol are as follows:

[0041] a: The number of tasks (a = 1, 2, 3... A);

[0042] s: The number of subtasks (S = 1, 2, 3... S);

[0043] b: The number of enterprises (b = 1, 2, 3... B);

[0044] g: Processing and manufacturing granularity;

[0045] z: Processing type;

[0046] m(n): n numerically controlled machine tools;

[0047] w(n): n software;

[0048] p(n): n people;

[0049] r(n): n cubic meters of blanks;

[0050] g(n): n machining and manufacturing granularities;

[0051] Z a,s : The machining type required for the s-th sub-task of task a;

[0052] F j,b : The decision variable for the enterprise to enter the final candidate set of the sub-task;

[0053] TF a,b : The time for enterprise b to complete the current machining task a;

[0054] Ts a+1,s,b : The start machining time of the s-th sub-task of task a + 1 on enterprise b;

[0055] Q a,s,c : The preliminary candidate set of enterprises for the s-th sub-task of task a;

[0056] Q a,s,f : The final candidate set of enterprises for the s-th sub-task of task a;

[0057] F z,b : The decision variable for the enterprise to enter the preliminary candidate set of the sub-task;

[0058] F g,b : The decision variable for the enterprise to enter the final candidate set of the sub-task;

[0059] B b,g : Enterprise b has g machining and manufacturing granularities;

[0060] B a,s,g : The s-th sub-task of task a requires g machining and manufacturing granularities to be machined;

[0061] Pt b,z : The machining type of enterprise b is z;

[0062] Pt a,s,z : The machining type of the s-th sub-task of task a is z;

[0063] Bu b,g : The updated machining and manufacturing granularity of enterprise b is g;

[0064] Bx b,g: The granularity consumed when the remaining processing and manufacturing granularity of the current enterprise b can no longer process any subtasks;

[0065] Bx b,g,n : The granularity consumed when the remaining processing and manufacturing granularity of enterprise b for the nth time can no longer process any subtasks; TF a,b : The time for enterprise b to complete the current processing task a;

[0066] Ts a,s,b : The start processing time of task a on enterprise b;

[0067] Td a,b : The processing duration required for task a among all tasks currently processed by enterprise b;

[0068] T a : The total processing duration of task a;

[0069] T a,s : The processing duration of the s-th subtask of task a;

[0070] Tw a,s : The waiting time before task a performs the s-th subtask;

[0071] T a,s,b : The processing time of the s-th subtask of task a in enterprise b;

[0072] Tr a,s : The transportation time of the s-th subtask of task a;

[0073] Tu a,s : The delay time caused by uncertain factors of the s-th subtask of task a;

[0074] T max : The maximum completion time;

[0075] gk: The unused remaining processing and manufacturing granularity when the enterprise processes tasks;

[0076] gk min : The unused remaining minimum processing and manufacturing granularity when the enterprise processes tasks;

[0077] Z b : The processing type of enterprise b is Z;

[0078] χ n : The random coefficient;

[0079] By analyzing the problems of enterprise resource utilization existing in task allocation and scheduling in the cloud platform, the present invention proposes a strategy that an enterprise can process multiple tasks simultaneously. A scheduling mathematical model is established by combining the strategy that an enterprise can process multiple tasks simultaneously with processing production constraints, and a multi-objective function is established. Compared with the current technology, the present invention can achieve the following effects:

[0080] (1) Constraints are proposed for the mutual relationship of four types of resources in the processing and manufacturing granularity, and a related mathematical model is established.

[0081] (2) When randomly matching subtasks with enterprises, it is proposed to first match the processing type to enter the preliminary candidate set of the enterprise, then perform the processing and manufacturing granularity matching to enter the final candidate set of the enterprise, and finally the task randomly selects an enterprise in the final candidate set for processing and a related mathematical model is established.

[0082] (3) A strategy that an enterprise can process multiple tasks simultaneously is proposed. When the remaining processing and manufacturing granularity of the enterprise is not sufficient to process any task, the task that is matched to this enterprise again waits until the processing and manufacturing granularity released after a task is completed is greater than the granularity required by the task. And a related mathematical model is established

[0083] (4) Combining the strategy that an enterprise can process multiple tasks simultaneously with actual processing constraints, a multi-objective scheduling model based on the completion time and the unused processing and manufacturing granularity is established, and a multi-objective optimization function is established.

[0084] (5) The established scheduling model can solve the multi-objective scheduling problem of making full use of enterprise resources in the cloud platform by combining intelligent optimization algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] The present invention will be further described below in conjunction with the drawings and specific implementation methods.

[0086] Figure 1 is the flow chart of the scheduling method of the cloud platform;

[0087] Figure 2 is the flow chart of the method for enterprises to perform random matching; DETAILED DESCRIPTION OF THE INVENTION

[0088] As Figure 1 shown, a cloud platform scheduling method for reasonable allocation of enterprise production resources based on processing and manufacturing granularity proposed by the present invention mainly includes the following steps:

[0089] Step 1: Randomly match subtasks with enterprises to form an initial population.

[0090] As Figure 2As shown, the platform provides a task set and decomposes it into subtask sets. Analyze the processing types of the subtasks and the required manufacturing granularity, and classify the types of enterprises and subtasks according to the information filled in by the platform. Decompose the tasks into multiple subtasks with processing sequences and processing types according to the information filled in by the platform.

[0091] The manufacturing granularity covers four processing resources: numerically controlled machine tools, human resources, industrial software systems, and product blanks. According to the actual production situation, the four resources are constrained (one human resource can operate 2 numerically controlled machine tools and 1 industrial software to process 1 cubic meter of product blanks). When the corresponding constraint relationships are achieved among the four resources, they can be converted into 1 manufacturing granularity. The specific mathematical model is as follows:

[0092] g(n) = p(n) + m(n) + w(n) + r(n) (1)

[0093] As Figure 2 shown, match the processing types of all subtasks in the subtask set with all enterprises to form a preliminary candidate set. Match the subtasks with the enterprises to form an enterprise candidate set for the subtasks. Compare the required processing types of the subtasks with the processing types of the enterprises. If the processing types are the same, this enterprise enters the preliminary candidate set corresponding to the subtask. If the processing types are different, then compare the next enterprise until all enterprises have been compared. The relevant constraint model is:

[0094]

[0095] Q a,s,c = F z,b ·b (3)

[0096] As Figure 2 shown, compare the manufacturing granularity of all subtasks with the enterprises in their preliminary candidate sets to form a final candidate set. Match the manufacturing granularity of the subtasks with the enterprises in their preliminary candidate sets. If the manufacturing granularity of this enterprise is greater than the required manufacturing granularity of this subtask, then this enterprise enters the final candidate enterprise set of the subtask. If the manufacturing granularity of the enterprise is less than the required manufacturing granularity of the subtask, then compare the next one until all enterprises have been compared.

[0097] Z b = Z a,s , b ∈ Q a,s,c (4)

[0098]

[0099] Q a,s,f = F z,b ·F g,b ·b (6)

[0100] AsFigure 2 As shown in the figure, the subtasks randomly select enterprises from the final candidate set to form an initial sorting. The three matrices of the initial population are randomly formed, namely tasks, subtasks, and corresponding enterprises. The enterprises are matched according to the order in which the subtasks appear. After each enterprise matches a corresponding subtask, the processing and manufacturing granularity of the enterprise decreases accordingly, and the processing and manufacturing granularity of the enterprise is updated to the remaining granularity. When the subsequent subtasks perform random matching with enterprises, if the same enterprise is matched again, the remaining processing and manufacturing granularity of the enterprise is compared with the processing and manufacturing granularity of the subtask. If it is greater than the processing and manufacturing granularity of the subtask, processing can be carried out.

[0101] Bu b,g＝ B b,g -Bx b,g , b ∈ Q a,s,f (7)

[0102] Z b =Z a,s (8)

[0103]

[0104] Step 2: Regarding the problem of making full use of enterprise manufacturing resources, a strategy is proposed that an enterprise can process multiple tasks simultaneously, and a related model is established.

[0105] When the remaining processing and manufacturing granularity of an enterprise is not sufficient to support any subtask, if a subtask matches this enterprise again, processing will wait. When a previous subtask is completed and the released processing and manufacturing granularity is more than the granularity required by the currently waiting subtask for processing, processing will start.

[0106] g{min(TF a,b )}>Ts a+1,s,b (10)

[0107] Z b =Z(Ts a+1,s,b ) (11)

[0108] Ts a+1,s,b =min(TF a,b )+Td a,b {min(TF a,b )},(a = 1,2,....a) (12)

[0109] Step 3: Combine the strategy that an enterprise can process multiple tasks simultaneously with the processing and production constraints to establish a scheduling mathematical model.

[0110] Based on the maximum processing time of the enterprise's processing tasks, a mathematical model for the maximum completion time of tasks is established, mainly including the waiting time, processing time, transportation time of tasks, and time for various uncertain factors. The mathematical model is as follows:

[0111] Tmax = max(Ta) (13)

[0112]

[0113] T a,s = Tw a,s + T a,s,b + Tr a,s + Tu a,s (15)

[0114] Tu a,s = χ1·Tw a,s + χ2·T a,s,b + χ3·Tr a,s (16)

[0115] Establish a mathematical model of machining granularity in man - hours, including the unused remaining manufacturing granularity during the processing tasks of all enterprises. The mathematical model is as follows:

[0116] gk min = min(gk) (17)

[0117]

[0118] Step 4: Establish a multi - objective function in combination with the mathematical model for scheduling in combination with an optimization algorithm

[0119] Combined with the diverse scheduling requirements in forging production, establish a multi - objective function with the maximum completion time and the minimum unused remaining manufacturing granularity during enterprise processing tasks as the objectives:

[0120] F(x) = min(f1, f2) (19)

[0121] f1 = T max (20)

[0122] f2 = gk min (21)

[0123] The technical meanings of each symbol are as follows:

[0124] a: The number of tasks (a = 1, 2, 3... A);

[0125] s: The number of subtasks (S = 1, 2, 3... S);

[0126] b: The number of enterprises (b = 1, 2, 3... B);

[0127] g: Machining granularity;

[0128] z: Machining type;

[0129] m(n): n numerically controlled machine tools;

[0130] w(n): n software;

[0131] p(n): n people;

[0132] r(n): n cubic meters of blanks;

[0133] g(n): n processing and manufacturing granularities;

[0134] Z a,s : The processing type required for the s-th subtask of task a;

[0135] F j,b : The decision variable for the enterprise to enter the final candidate set of the subtask;

[0136] TF a,b : The time for enterprise b to complete the current processing task a;

[0137] Ts a+1,s,b : The start processing time of the s-th subtask of task a + 1 on enterprise b;

[0138] Q a,s,c : The preliminary candidate set of enterprises for the s-th subtask of task a;

[0139] Q a,s,f : The final candidate set of enterprises for the s-th subtask of task a;

[0140] F z,b : The decision variable for the enterprise to enter the preliminary candidate set of the subtask;

[0141] F g,b : The decision variable for the enterprise to enter the final candidate set of the subtask;

[0142] B b,g : Enterprise b has g processing and manufacturing granularities;

[0143] B a,s,g : The s-th subtask of task a requires g processing and manufacturing granularities to be processed;

[0144] Pt b,z : The processing type of enterprise b is z;

[0145] Pt a,s,z : The processing type of the s-th subtask of task a is z;

[0146] Bu b,g : The updated processing and manufacturing granularity of enterprise b is g;

[0147] Bx b,g: The granularity consumed when the remaining processing and manufacturing granularity of the current enterprise b can no longer process any subtasks;

[0148] Bx b,g,n : The granularity consumed when the remaining processing and manufacturing granularity of enterprise b for the nth time can no longer process any subtasks; TF a,b : The time for enterprise b to complete the current processing task a;

[0149] Ts a,s,b : The start processing time of task a on enterprise b;

[0150] Td a,b : The processing duration required for task a among all the tasks currently processed by enterprise b;

[0151] T a : The total processing duration of task a;

[0152] T a,s : The processing duration of the sth subtask of task a;

[0153] Tw a,s : The waiting time before task a performs the sth subtask;

[0154] T a,s,b : The processing time of the sth subtask of task a in enterprise b;

[0155] Tr a,s : The transportation time of the sth subtask of task a;

[0156] Tu a,s : The delay time caused by the uncertain factors of the sth subtask of task a;

[0157] T max : The maximum completion time;

[0158] gk: The unused remaining processing and manufacturing granularity when the enterprise processes tasks;

[0159] gk min : The unused remaining minimum processing and manufacturing granularity when the enterprise processes tasks;

[0160] Z b : The processing type of enterprise b is Z;

[0161] χ n : The random coefficient.

Claims

1. A cloud platform scheduling method for reasonable allocation of enterprise production resources based on processing and manufacturing granularity, characterized in that: It includes the following steps: Step 1: Sub-tasks are randomly matched with enterprises to form an initial population; The platform gives a task set and decomposes it into a sub-task set; analyze the processing types and required manufacturing granularities of the sub-tasks, and classify the types of enterprises and sub-tasks according to the information filled in by the platform; according to the information filled in by the platform, decompose the task into multiple sub-tasks with processing sequences and processing types; The manufacturing granularity covers four processing resources: numerically controlled machine tools, human resources, industrial software systems, and product blanks. According to the actual production situation, the four resources are constrained. When the corresponding constraint relationships are achieved among the four resources, it can be converted into 1 manufacturing granularity. The specific mathematical model is as follows: g(n) = p(n) + m(n) + w(n) + r(n) (1) All sub-tasks in the sub-task set are matched with all enterprises in terms of processing types to form a preliminary candidate set; the sub-tasks are matched with enterprises to form an enterprise candidate set for the sub-tasks; the required processing types of the sub-tasks are compared with the processing types of the enterprises; if the processing types are the same, this enterprise enters the preliminary candidate set corresponding to the sub-task, and if the processing types are different, then the next enterprise is compared until all enterprises are compared; the relevant constraint model is: Q a,s,c = F z,b ·b (3) All sub-tasks compare the manufacturing granularities with the enterprises in their preliminary candidate sets to form a final candidate set; the sub-tasks match the manufacturing granularities with the enterprises in their preliminary candidate sets; if the manufacturing granularity of this enterprise is greater than the required manufacturing granularity of this sub-task, then this enterprise enters the final candidate enterprise set of the sub-task; if the manufacturing granularity of the enterprise is less than the required manufacturing granularity of the sub-task, then the next one is compared until all enterprises are compared; Z b = Z a,s , b ∈ Q a,s,c (4) Q a,s,f = F z,b · F g,b · b(6) The sub-tasks randomly select enterprises from the final candidate set to form an initial sorting; three matrices of the initial population are randomly formed, namely tasks, sub-tasks, and corresponding enterprises; enterprises are matched according to the order in which the sub-tasks appear. After each sub-task is matched with an enterprise, the corresponding manufacturing granularity of the enterprise decreases accordingly, and the manufacturing granularity of the enterprise is updated to the remaining granularity; when the subsequent sub-tasks are randomly matched with enterprises again, if this enterprise is matched again, compare the remaining manufacturing granularity of the enterprise with the manufacturing granularity of the sub-task. If it is greater than the manufacturing granularity of the sub-task, then processing can be carried out; Bu b,g = B b,g - Bx b,g , b ∈ Q a,s,f (7) Z b = Z a,s (8) Step 2: Regarding the problem of making full use of enterprise manufacturing resources, propose a strategy that an enterprise can process multiple tasks simultaneously and establish a relevant model; When the remaining manufacturing granularity of the enterprise is not sufficient to support any sub-task, if a sub-task is matched with this enterprise again, processing is awaited; When the previous sub-task is completed and the released manufacturing granularity is more than the granularity required by the currently waiting sub-task for processing, then processing starts; g{min(TF a,b )} > Ts a+1,s,b (10) Z b = Z(Ts a+1,s,b ) (11) Ts a+1,s,b = min(TF a,b ) + Td a,b {min(TF a,b )}, a = 1, 2,....A (12) Step 3: Combine a strategy that an enterprise can process multiple tasks simultaneously with processing production constraints to establish a scheduling mathematical model; Based on the maximum processing time of the enterprise's processing tasks, establish a mathematical model for the maximum completion time of the tasks. Based on the waiting time, processing time, transportation time, and various uncertain factor times of the tasks, the mathematical model is as follows: Tmax = max(Ta) (13) T a,s = Tw a,s + T a,s,b + Tr a,s + Tu a,s (15) Tu a,s = χ1·Tw a,s + χ2·T a,s,b + χ3·Tr a,s (16) Establish a mathematical model of machining granularity in working hours, including the unused remaining manufacturing granularity during all enterprise processing tasks. The mathematical model is as follows: gk min = min(gk) (17) Step 4: Establish a multi-objective function in combination with the mathematical model for scheduling in combination with optimization algorithms Combined with the diverse scheduling requirements in forging production, establish a multi-objective function with the makespan and the minimum unused remaining manufacturing granularity during enterprise processing tasks as the objectives: F(x) = min(f1, f2) (19) f1 = T max (20) f2 = gk min (21) The technical meanings of each symbol are as follows: a: The task number, a = 1, 2, 3... A; s: The subtask number, S = 1, 2, 3... S; b: The enterprise number, b = 1, 2, 3... B; g: Machining granularity; z: Machining type; Z a,s : The processing type required for the s-th sub-task of task a; F j,b : Decision variable for an enterprise to enter the final candidate set of subtasks; TF a,b : The time for enterprise b to complete the current processing task a; Ts a+1,s,b : The start processing time of sub-task s of task a + 1 on enterprise b; m(n): n CNC machine tools; w(n): n industrial software systems; p(n): n human resources; r(n): n cubic meters of product blanks; g(n): n machining granularities; Q a,s,c : The initial enterprise candidate set for the s-th sub-task of task a; Q a,s,f : The final candidate set of enterprises for the s-th sub-task of task a; F z,b : Decision variable for the enterprise to enter the preliminary candidate set of subtasks; F g,b : Decision variable for the enterprise to enter the final candidate set of subtasks; B b,g : Enterprise b has g manufacturing granularities; B a,s,g : The s-th sub-task of task a requires g manufacturing granularities for processing; Pt b,z : The processing type of enterprise b is z; Pt a,s,z : The processing type of the s-th sub-task of task a is z; This b,g : The updated processing and manufacturing granularity of enterprise b is g; Bx b,g : The granularity consumed when the remaining processing and manufacturing granularity of the current enterprise b can no longer process any subtasks; Bx b,g,n : The granularity consumed when the remaining processing and manufacturing granularity of enterprise b for the nth time cannot process any subtasks anymore; TF a,b : The time for enterprise b to complete the current processing task a; Ts a,s,b : The start processing time of task a on enterprise b; Td a,b : The processing duration required for Company b to process Task a among all current tasks T a : Total processing duration of task a; T a,s : Processing duration of the s-th sub-task of task a; Tw a,s : Waiting time of task a before the s-th subtask; T a,s,b : The processing time of the s-th sub-task of task a in enterprise b; Tr a,s : Transportation time of the s-th sub-task of task a; Tu a,s : Delay time caused by uncertainties in the sth sub-task of task a; T max : Makespan; gk: The unused remaining machining granularity during enterprise processing tasks; gk min : The minimum remaining processing and manufacturing granularity that is not utilized during the enterprise's processing tasks; Z b : The processing type of Company b is Z; χ n : Random coefficient.

2. The cloud platform scheduling method for reasonable allocation of enterprise production resources based on processing and manufacturing granularity according to claim 1, characterized in that: In the correspondence between tasks and subtasks, the order in which a task appears is the subtask corresponding to this task. Random matching of corresponding enterprises is performed according to the subtasks.

3. A cloud platform scheduling method for reasonable allocation of enterprise production resources based on processing and manufacturing granularity according to claim 1, characterized in that: During the process of releasing the machining granularity of an enterprise, only when all subtasks are completed will the corresponding manufacturing resource granularity be fully released.

Images