A method and system for scheduling production of automobile parts based on big data analysis

Abstract

The present invention relates to the technical field of automobile parts production scheduling, specifically a method and system for automobile parts production scheduling based on big data analysis, including: constructing an automobile knowledge graph based on the relationship between automobile parts, constructing a relationship triple based on the automobile knowledge graph; obtaining target automobile parts to be produced, matching the target automobile parts to be produced with the automobile knowledge graph to obtain multiple production scheduling combination paths, and selecting the multiple production scheduling combination paths based on the inventory of the automobile parts to obtain the optimal production scheduling combination path. The present invention presents various types of automobile parts and their relationships in a graphical manner by constructing an automobile knowledge graph, which can more clearly reflect the interdependence between parts. This structured information helps the production scheduling system to quickly identify the optimal production scheduling combination path when facing a complex production environment, and significantly reduces possible resource waste and production bottlenecks.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile parts production scheduling, and in particular to an automobile parts production scheduling method and system based on big data analysis. Background Art

[0002] Auto parts production scheduling refers to the activities of rationally arranging and managing production resources, time and processes during the manufacturing process of auto parts. Its main goal is to optimize production efficiency, reduce costs and ensure on-time delivery.

[0003] Traditional methods often rely on information managed independently by various departments and lack effective integration, resulting in poor information flow in the production scheduling process. This information island phenomenon will lead to delayed decision-making and inability to respond to changes in the production environment in real time. In addition, many traditional methods use fixed scheduling strategies and cannot be flexibly adjusted to adapt to changes in market demand. This leads to excess or shortage of inventory on the production line when facing demand fluctuations, resulting in economic losses. In addition, in a complex production environment, traditional methods are difficult to effectively identify resource waste and bottleneck problems in the production process, which may lead to low production efficiency and irrational use of resources. In addition, the method fails to fully consider environmental factors and fails to quantify the carbon emissions and environmental impact of equipment, which may lead to the inability of enterprises to effectively control environmental costs during operations and lack awareness of sustainable development. Summary of the invention

[0004] The technical problem to be solved by the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a method and system for automobile parts production scheduling based on big data analysis.

[0005] The technical solution adopted to solve the above technical problems is: a method for scheduling the production of automobile parts based on big data analysis, including:

[0006] Building an automobile knowledge graph based on the relationships between automobile parts, and building a relationship triple based on the automobile knowledge graph;

[0007] Acquire target auto parts to be produced, match the target auto parts to be produced with the auto knowledge graph to obtain multiple production scheduling combination paths, and select the multiple production scheduling combination paths based on the inventory of the auto parts to obtain the optimal production scheduling combination path;

[0008] Determine the sub-parts required for the target automobile part to be produced based on the optimal production scheduling combination path, determine the processes required for the sub-parts based on the total number of sub-part types, and determine the processing equipment required for the sub-parts based on the processes required for the sub-parts;

[0009] Constructing a first objective function and a first constraint condition based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, and constructing a second objective function and a second constraint condition based on the types of processing equipment required for the sub-assemblies;

[0010] A production scheduling optimization model is constructed based on the first objective function, the first constraint condition, the second objective function and the second constraint condition, and the production scheduling optimization model is solved based on the improved ant algorithm to obtain a production scheduling solution.

[0011] Preferably, the automobile knowledge graph includes multiple entity nodes and multiple relationship edges, wherein the entity nodes correspond to the automobile parts, and the relationship edges correspond to the relationships between the automobile parts, wherein the relationships between the automobile parts include inclusion relationships and category relationships, the relationship triples include a head entity node, a relationship edge and a tail entity node, and the production scheduling combination path is composed of multiple relationship triplets.

[0012] Preferably, the target automobile parts to be produced are matched with the automobile knowledge graph to obtain multiple production scheduling combination paths, including:

[0013] Matching the target automobile parts to be produced with multiple entity nodes in the automobile knowledge graph to obtain target entity nodes that match the multiple entity nodes in the automobile knowledge graph;

[0014] Taking the target entity node as the starting node, traversing all entity nodes connected to the starting node in the automobile knowledge graph based on a depth-first search algorithm, and recording the traversal paths to obtain multiple recorded paths;

[0015] The multiple record paths are linked based on the relationship triples to obtain multiple relationship triple sequences, wherein the multiple relationship triple sequences correspond to multiple production scheduling combination paths.

[0016] Preferably, the multiple production scheduling combination paths are selected based on the inventory of the auto parts to obtain the optimal production scheduling combination path, including:

[0017] Taking the first entity node of the plurality of production scheduling combination paths as a central entity node, and obtaining the next entity node of the central entity node;

[0018] The weight coefficient from the central entity node to the next entity node is calculated to obtain a weight coefficient set, and the next entity node corresponding to the maximum weight coefficient in the weight coefficient set is taken as the optimal entity node, wherein the calculation formula of the weight coefficient is as follows:

[0019] ;

[0020] in, Represents the weight coefficient from the central entity node to the next entity node, represents the rectified linear unit activation function, A function representing the correlation between two entity nodes. The weight matrix represents the feature transformation of the entity node, and the weight matrix is ​​related to the inventory of auto parts. Represents the central entity node The embedding vector of Indicates the next entity node The embedding vector of

[0021] The optimal entity node is used as a new central entity node, and the above operation is repeated until all entity nodes in the multiple production scheduling combination paths are traversed to obtain an optimal entity node set;

[0022] All the optimal entity nodes in the optimal entity node set are connected to obtain the optimal production scheduling combination path.

[0023] Preferably, the first objective function is constructed based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, including:

[0024] The total processing time of the target automotive parts to be produced is calculated based on the total number of sub-part types, the processes and processing equipment required for the sub-parts, wherein the calculation formula of the total processing time is as follows:

[0025] ;

[0026] in, represents the total processing time of the target automotive part to be produced, Indicates sub-assembly Process Is the equipment being processed? Upper processing, if the sub-accessories Process In processing equipment On processing, , otherwise, , Indicates sub-assembly Process In processing equipment The completion time of the processing, Indicates sub-assembly Process In processing equipment The start time of processing, Indicates the total number of sub-assembly types. Represents the total number of processes, Indicates the total number of processing equipment;

[0027] The total waiting time of the target automotive parts to be produced is calculated based on the total number of sub-part types, the processes and processing equipment required for the sub-parts, wherein the calculation formula of the total waiting time is as follows:

[0028] ;

[0029] in, represents the total waiting time of the target automotive parts to be produced, Indicates sub-assembly Process In processing equipment Before you finish the waiting time in the waiting area, Indicates sub-assembly Process In processing equipment Start waiting time in the waiting area before;

[0030] The total queue time of the target auto parts to be produced is calculated based on the total number of sub-parts, the processes and processing equipment required for the sub-parts, wherein the calculation formula of the total queue time is as follows:

[0031] ;

[0032] in, represents the total queue time of the target auto parts to be produced, Indicates sub-assembly Process In processing equipment After completion, complete the queue time in the queue area. Indicates sub-assembly Process In processing equipment After completion, the queue time will start in the queue area;

[0033] A first objective function is constructed based on the total processing time, total waiting time and total queuing time of the target automotive parts to be produced, wherein the first objective function is as follows:

[0034] ;

[0035] in, represents the first objective function, Represents the maximum value function.

[0036] Preferably, the first constraint condition is as follows:

[0037] ;

[0038] in, Indicates the process Optional processing equipment set, Indicates sub-assembly Process In processing equipment processing time.

[0039] Preferably, constructing a second objective function based on the type of processing equipment required for the sub-assembly includes:

[0040] The startup carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the startup carbon emissions of the processing equipment is as follows:

[0041] ;

[0042] in, Indicates the carbon emissions from the start-up of the processing equipment. Indicates the starting power of the processing equipment, Indicates the start-up time of the processing equipment, represents the carbon emission factor of electric energy;

[0043] The carbon emissions of fixture installation of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula of the carbon emissions of fixture installation of the processing equipment is as follows:

[0044] ;

[0045] in, Indicates the carbon emissions of fixture installation of processing equipment, Indicates the fixture installation power of the processing equipment, Indicates the fixture installation time of the processing equipment;

[0046] The processing carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the processing carbon emissions of the processing equipment is as follows:

[0047] ;

[0048] in, represents the processing carbon emissions of the processing equipment, Indicates the processing power of the processing equipment, Indicates the processing time of the processing equipment;

[0049] The carbon emissions from fixture removal of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for carbon emissions from fixture removal of the processing equipment is as follows:

[0050] ;

[0051] in, Indicates the carbon emissions from fixture removal of processing equipment, Indicates the fixture removal power of the processing equipment, Indicates the fixture disassembly time of the processing equipment;

[0052] The standby carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the standby carbon emissions of the processing equipment is as follows:

[0053] ;

[0054] in, represents the standby carbon emissions of processing equipment, Indicates the standby power of the processing equipment, Indicates the standby time of processing equipment;

[0055] The shutdown carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the shutdown carbon emissions of the processing equipment is as follows:

[0056] ;

[0057] in, Indicates the shutdown carbon emissions of processing equipment, Indicates the shutdown power of the processing equipment, Indicates the shutdown time of the processing equipment;

[0058] A second objective function is constructed based on the carbon emissions from starting up the processing equipment, the carbon emissions from installing the fixture, the carbon emissions from processing, the carbon emissions from removing the fixture, the carbon emissions from standby, and the carbon emissions from shutting down, wherein the second objective function is as follows:

[0059] ;

[0060] in, represents the second objective function, Represents the minimum value function.

[0061] Preferably, the second constraint condition is as follows:

[0062] .

[0063] Preferably, solving the production scheduling optimization model based on an improved ant algorithm to obtain a production scheduling solution includes:

[0064] Based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, a plurality of processing orders are obtained, the plurality of processing orders are used as mobile nodes of ants, and the parallel processing equipment of each process is used as a virtual node of the ants;

[0065] Initializing the ant colony, wherein the initialization includes setting the number of ants, setting the initial position of each ant to a randomly selected processing order, and assigning an initial pheromone value to each path, wherein the number of ants is proportional to the total number of processing orders, and the path is an assignment of a process of each processing order to a processing device;

[0066] Based on the parallel processing equipment available in each process of each processing order, a candidate equipment list is constructed, and the path is selected based on the pheromone strength and heuristic information to obtain the process sequence and processing equipment of each processing order, wherein the path selection formula is as follows:

[0067] ;

[0068] in, Indicates the path chosen by the ant The probability of Indicates the path The pheromone strength, Indicates the path Heuristic information, and represents the weight factor;

[0069] After completing one iteration, the pheromone is updated according to the first objective function value and the second objective function value of each ant, and a path is selected. The above operation is repeated until the maximum number of iterations is reached, and the global optimal solution is output. Based on the global optimal solution, a production scheduling plan is obtained.

[0070] The technical solution adopted to solve the above technical problems is: an automobile parts production scheduling system based on big data analysis, which is applicable to the automobile parts production scheduling method based on big data analysis, including:

[0071] A graph construction unit, the graph construction unit is used to construct an automobile knowledge graph based on the relationship between automobile parts, and to construct a relationship triple based on the automobile knowledge graph;

[0072] A production scheduling combination unit, the production scheduling combination unit is used to obtain target automobile parts to be produced, match the target automobile parts to be produced with the automobile knowledge graph to obtain multiple production scheduling combination paths, and select the multiple production scheduling combination paths based on the inventory of the automobile parts to obtain the optimal production scheduling combination path;

[0073] A component determination unit, the component determination unit is used to determine the sub-components required for the target automobile component to be produced based on the optimal production scheduling combination path, determine the process required for the sub-component based on the total number of sub-component types, and determine the processing equipment required for the sub-component based on the process required for the sub-component;

[0074] A target construction unit, the target construction unit is used to construct a first objective function and a first constraint condition based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assembly, and to construct a second objective function and a second constraint condition based on the types of processing equipment required for the sub-assembly;

[0075] A production scheduling unit, the production scheduling unit is used to construct a production scheduling optimization model based on the first objective function, the first constraint condition, the second objective function and the second constraint condition, and solve the production scheduling optimization model based on the improved ant algorithm to obtain a production scheduling plan.

[0076] The beneficial effects of the present invention are as follows: (1) The present invention presents various types of automobile parts and their relationships in a graphical manner by constructing an automobile knowledge graph, which can more clearly reflect the interdependence between the parts. This structured information helps the production scheduling system to quickly identify the optimal production scheduling combination path when facing a complex production environment, significantly reducing possible resource waste and production bottlenecks; (2) The present invention can flexibly adjust the production scheduling strategy during the scheduling process by considering the inventory of parts, which means that the production line can respond to changes in market demand more promptly and avoid economic losses caused by excess or shortage of inventory. The optimized production scheduling path not only ensures the continuity of production, but also effectively controls inventory costs and improves the overall operational efficiency of the enterprise; (3) The present invention takes environmental factors into consideration when constructing the production scheduling optimization model. By quantifying the carbon emissions of various types of equipment, it achieves effective control of the environmental impact of the production process. Moreover, by matching with the automobile knowledge graph, the target automobile parts to be produced can be quickly located to the most suitable production path. This intelligent matching greatly improves the automation of scheduling, reduces the need for manual intervention, and makes the production process smoother. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A schematic flow chart of the steps of an overall method in an embodiment of the present invention;

[0078] Figure 2 A schematic diagram of the system architecture of an overall system in an embodiment of the present invention.

[0079] Figure numerals: 1. Graph construction unit; 2. Production scheduling combination unit; 3. Accessory determination unit; 4. Target construction unit; 5. Production scheduling unit. DETAILED DESCRIPTION

[0080] Embodiment 1, as Figure 1 As shown, the present invention proposes a method for automobile parts production scheduling based on big data analysis, comprising:

[0081] S1. Build an automobile knowledge graph based on the relationship between automobile parts, and build relationship triples based on the automobile knowledge graph;

[0082] S2. Obtain target auto parts to be produced, match the target auto parts to be produced with the automobile knowledge graph to obtain multiple production scheduling combination paths, and select the multiple production scheduling combination paths based on the inventory of auto parts to obtain the optimal production scheduling combination path;

[0083] S3. Determine the sub-parts required for the target automobile parts to be produced based on the optimal production scheduling combination path, determine the processes required for the sub-parts based on the total number of sub-part types, and determine the processing equipment required for the sub-parts based on the processes required for the sub-parts;

[0084] S4, constructing a first objective function and a first constraint condition based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, and constructing a second objective function and a second constraint condition based on the types of processing equipment required for the sub-assemblies;

[0085] S5. Construct a production scheduling optimization model based on the first objective function, the first constraint condition, the second objective function and the second constraint condition, and solve the production scheduling optimization model based on the improved ant algorithm to obtain a production scheduling solution.

[0086] In the present invention, the automobile knowledge graph is a graphical data structure used to represent the relationship between automobile parts. It consists of multiple entity nodes (representing parts) and relationship edges (representing the association between parts); the relationship triple refers to the basic structure in the knowledge graph, which consists of three parts: the head entity node, the relationship edge and the tail entity node. For example, in automobile parts, a triple may represent "part A contains part B"; the production scheduling combination path refers to multiple feasible production sequences and combinations formed according to the relationship between parts in production scheduling; the optimal production scheduling combination path refers to the optimal production sequence obtained after screening the inventory in multiple production scheduling combination paths to improve production efficiency and reduce costs; sub-parts constitute smaller components of the target automobile parts, and are the parts that need to be produced first when producing the target parts; the ant algorithm is a heuristic algorithm that imitates the foraging behavior of ants and simulates multiple "ants" to find the optimal path in the solution space, which is suitable for solving complex optimization problems.

[0087] Embodiment 2: A method for automobile parts production scheduling based on big data analysis proposed in the present invention, compared with embodiment 1, this embodiment also includes: the automobile knowledge graph includes multiple entity nodes and multiple relationship edges, wherein the entity nodes correspond to automobile parts, and the relationship edges correspond to the relationships between automobile parts, wherein the relationships between automobile parts include inclusion relationships and category relationships, the relationship triples include a head entity node, a relationship edge and a tail entity node, and the production scheduling combination path consists of multiple relationship triplets.

[0088] In an optional embodiment, the target automobile parts to be produced are matched with the automobile knowledge graph to obtain multiple production scheduling combination paths, including:

[0089] A1. Match the target automobile parts to be produced with multiple entity nodes in the automobile knowledge graph to obtain target entity nodes that match the multiple entity nodes in the automobile knowledge graph;

[0090] A2. Taking the target entity node as the starting node, traverse all entity nodes connected to the starting node in the automobile knowledge graph based on the depth-first search algorithm, and record the traversal path to obtain multiple recorded paths;

[0091] A3. Link multiple record paths based on the relationship triples to obtain multiple relationship triple sequences, wherein the multiple relationship triple sequences correspond to multiple production scheduling combination paths.

[0092] It should be noted that the depth-first search algorithm is a graph traversal algorithm that starts from a node and goes as deep as possible into each branch until it reaches the end or there are no more accessible nodes, and then backtracks to search other branches.

[0093] In an optional embodiment, multiple production scheduling combination paths are selected based on the inventory of automobile parts to obtain the optimal production scheduling combination path, including:

[0094] B1. The first entity node of multiple production scheduling combination paths is taken as the central entity node, and the next entity node of the central entity node is obtained;

[0095] B2. Calculate the weight coefficient from the central entity node to the next entity node to obtain a weight coefficient set, and take the next entity node corresponding to the maximum weight coefficient in the weight coefficient set as the optimal entity node, where the calculation formula of the weight coefficient is as follows:

[0096] ;

[0097] in, Represents the weight coefficient from the central entity node to the next entity node, represents the rectified linear unit activation function, A function representing the correlation between two entity nodes. The weight matrix represents the feature transformation of the entity node, and the weight matrix is ​​related to the inventory of auto parts. Represents the central entity node The embedding vector of Indicates the next entity node The embedding vector of

[0098] B3. The optimal entity node is used as a new central entity node, and the above operation is repeated until all entity nodes in multiple production scheduling combination paths are traversed to obtain the optimal entity node set;

[0099] B4. Connect all the optimal entity nodes in the optimal entity node set to obtain the optimal production scheduling combination path.

[0100] It should be noted that the weight matrix of feature transformation refers to the matrix used to perform linear transformation on input data (features of entity nodes) in machine learning. The specific value of this matrix may be associated with the inventory of the entity; the embedding vector refers to a representation method that maps entity nodes to a continuous vector space, so that similar nodes are closer in the vector space, which is helpful for subsequent calculations; the rectified linear unit activation function is a mathematical function used to adjust the output of nodes in a neural network, and is often used for nonlinear transformations to improve the expressiveness of the model.

[0101] In an optional embodiment, a first objective function is constructed based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, including:

[0102] C1. Calculate the total processing time of the target automotive parts to be produced based on the total number of sub-part types, the processes and processing equipment required for the sub-parts, where the calculation formula for the total processing time is as follows:

[0103] ;

[0104] in, represents the total processing time of the target automotive part to be produced, Indicates sub-assembly Process Is the equipment being processed? Upper processing, if the sub-accessories Process In processing equipment On processing, , otherwise, , Indicates sub-assembly Process In processing equipment The completion time of the processing, Indicates sub-assembly Process In processing equipment The start time of processing, Indicates the total number of sub-assembly types. Represents the total number of processes, Indicates the total number of processing equipment;

[0105] C2. Calculate the total waiting time of the target auto parts to be produced based on the total number of sub-parts, the processes and processing equipment required for the sub-parts, where the calculation formula for the total waiting time is as follows:

[0106] ;

[0107] in, represents the total waiting time of the target automotive parts to be produced, Indicates sub-assembly Process In processing equipment Before you finish the waiting time in the waiting area, Indicates sub-assembly Process In processing equipment Start waiting time in the waiting area before;

[0108] C3. Calculate the total queue time of the target auto parts to be produced based on the total number of sub-parts, the processes and processing equipment required for the sub-parts, where the calculation formula for the total queue time is as follows:

[0109] ;

[0110] in, represents the total queue time of the target auto parts to be produced, Indicates sub-assembly Process In processing equipment After completion, complete the queue time in the queue area. Indicates sub-assembly Process In processing equipment After completion, the queue time will start in the queue area;

[0111] C4. Constructing a first objective function based on the total processing time, total waiting time and total queuing time of the target automobile parts to be produced, wherein the first objective function is as follows:

[0112] ;

[0113] in, represents the first objective function, Represents the maximum value function.

[0114] In an optional embodiment, the first constraint is as follows:

[0115] ;

[0116] in, Indicates the process Optional processing equipment set, Indicates sub-assembly Process In processing equipment processing time.

[0117] In an optional embodiment, a second objective function is constructed based on the type of processing equipment required for the sub-assembly, including:

[0118] D1. Calculate the startup carbon emissions of the processing equipment based on the types of processing equipment required for the sub-assemblies. The calculation formula for the startup carbon emissions of the processing equipment is as follows:

[0119] ;

[0120] in, Indicates the carbon emissions from the start-up of the processing equipment. Indicates the starting power of the processing equipment, Indicates the start-up time of the processing equipment, represents the carbon emission factor of electric energy;

[0121] D2. Calculate the carbon emissions from fixture installation of processing equipment based on the types of processing equipment required for the sub-assemblies. The calculation formula for carbon emissions from fixture installation of processing equipment is as follows:

[0122] ;

[0123] in, Indicates the carbon emissions of fixture installation of processing equipment, Indicates the fixture installation power of the processing equipment, Indicates the fixture installation time of the processing equipment;

[0124] D3. Calculate the processing carbon emissions of processing equipment based on the types of processing equipment required for the sub-assemblies. The calculation formula for processing carbon emissions of processing equipment is as follows:

[0125] ;

[0126] in, represents the processing carbon emissions of the processing equipment, Indicates the processing power of the processing equipment, Indicates the processing time of the processing equipment;

[0127] D4. Calculate the carbon emissions from fixture removal of processing equipment based on the types of processing equipment required for the sub-assemblies. The calculation formula for carbon emissions from fixture removal of processing equipment is as follows:

[0128] ;

[0129] in, Indicates the carbon emissions from fixture removal of processing equipment, Indicates the fixture removal power of the processing equipment, Indicates the fixture disassembly time of the processing equipment;

[0130] D5. Calculate the standby carbon emissions of processing equipment based on the types of processing equipment required for the sub-assemblies. The calculation formula for the standby carbon emissions of processing equipment is as follows:

[0131] ;

[0132] in, represents the standby carbon emissions of processing equipment, Indicates the standby power of the processing equipment, Indicates the standby time of processing equipment;

[0133] D6. Calculate the shutdown carbon emissions of the processing equipment based on the types of processing equipment required for the sub-assembly. The calculation formula for the shutdown carbon emissions of the processing equipment is as follows:

[0134] ;

[0135] in, Indicates the shutdown carbon emissions of processing equipment, Indicates the shutdown power of the processing equipment, Indicates the shutdown time of the processing equipment;

[0136] D7. Construct a second objective function based on the carbon emissions from starting up the processing equipment, the carbon emissions from installing the fixture, the carbon emissions from processing, the carbon emissions from removing the fixture, the carbon emissions from standby, and the carbon emissions from shutting down. The second objective function is as follows:

[0137] ;

[0138] in, represents the second objective function, Represents the minimum value function.

[0139] In an optional embodiment, the second constraint is as follows:

[0140] .

[0141] In an optional embodiment, the production scheduling optimization model is solved based on the improved ant algorithm to obtain a production scheduling solution, including:

[0142] E1. Based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, multiple processing orders are obtained, and the multiple processing orders are used as mobile nodes of ants, and the parallel processing equipment of each process is used as the virtual node of the ant;

[0143] E2. Initializing the ant colony, wherein the initialization includes setting the number of ants, setting the initial position of each ant to a randomly selected processing order, and assigning an initial pheromone value to each path, wherein the number of ants is proportional to the total number of processing orders, and the path is the assignment of a process of each processing order to a processing device;

[0144] E3. Based on the parallel processing equipment available in each process for each processing order, a list of candidate equipment is constructed, and the path is selected based on the pheromone strength and heuristic information to obtain the process sequence and processing equipment for each processing order. The path selection formula is as follows:

[0145] ;

[0146] in, Indicates the path chosen by the ant The probability of Indicates the path The pheromone strength, Indicates the path Heuristic information, and represents the weight factor;

[0147] E4. After completing one iteration, update the pheromone according to the first objective function value and the second objective function value of each ant, select the path, repeat the above operations until the maximum number of iterations is reached, output the global optimal solution, and obtain the production scheduling plan based on the global optimal solution.

[0148] It should be noted that pheromone refers to the numerical value used to indicate the quality of a path in the ant algorithm. The strength of the pheromone will affect the ant's path selection; the candidate equipment list refers to the set of optional processing equipment that can be used for each processing order, and the ant will select from this list; iteration refers to the core process of the ant algorithm. Each iteration will update the pheromone according to the ant's path selection and objective function value to gradually approach the optimal solution; the global optimal solution refers to the best path or production scheduling plan found after all iterations are completed, which can achieve the optimal goal under constraints.

[0149] Embodiment three, as Figure 2 As shown, the present invention proposes an automobile parts production scheduling system based on big data analysis, which is applicable to the automobile parts production scheduling method based on big data analysis, including:

[0150] A graph construction unit 1, which is used to construct an automobile knowledge graph based on the relationship between automobile parts, and to construct a relationship triple based on the automobile knowledge graph;

[0151] The production scheduling combination unit 2 is used to obtain the target automobile parts to be produced, match the target automobile parts to be produced with the automobile knowledge graph to obtain multiple production scheduling combination paths, and select the multiple production scheduling combination paths based on the inventory of automobile parts to obtain the optimal production scheduling combination path;

[0152] The component determination unit 3 is used to determine the sub-components required for the target automobile component to be produced based on the optimal production scheduling combination path, determine the process required for the sub-component based on the total number of sub-component types, and determine the processing equipment required for the sub-component based on the process required for the sub-component;

[0153] A target construction unit 4, the target construction unit 4 is used to construct a first target function and a first constraint condition based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assembly, and to construct a second target function and a second constraint condition based on the types of processing equipment required for the sub-assembly;

[0154] The production scheduling unit 5 is used to construct a production scheduling optimization model based on the first objective function, the first constraint condition, the second objective function and the second constraint condition, and solve the production scheduling optimization model based on the improved ant algorithm to obtain a production scheduling solution.

[0155] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge scope of technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A method for scheduling production of automobile parts based on big data analysis, characterized in that: include: Building an automobile knowledge graph based on the relationships between automobile parts, and building a relationship triple based on the automobile knowledge graph; Acquire target auto parts to be produced, match the target auto parts to be produced with the auto knowledge graph to obtain multiple production scheduling combination paths, and select the multiple production scheduling combination paths based on the inventory of the auto parts to obtain the optimal production scheduling combination path; Determine the sub-parts required for the target automobile part to be produced based on the optimal production scheduling combination path, determine the processes required for the sub-parts based on the total number of sub-part types, and determine the processing equipment required for the sub-parts based on the processes required for the sub-parts; Constructing a first objective function and a first constraint condition based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, and constructing a second objective function and a second constraint condition based on the types of processing equipment required for the sub-assemblies; Building a production scheduling optimization model based on the first objective function, the first constraint condition, the second objective function and the second constraint condition, and solving the production scheduling optimization model based on an improved ant algorithm to obtain a production scheduling solution; The target automobile parts to be produced are matched with the automobile knowledge graph to obtain multiple production scheduling combination paths, including: Matching the target automobile parts to be produced with multiple entity nodes in the automobile knowledge graph to obtain target entity nodes that match the multiple entity nodes in the automobile knowledge graph; Taking the target entity node as the starting node, traversing all entity nodes connected to the starting node in the automobile knowledge graph based on a depth-first search algorithm, and recording the traversal paths to obtain multiple recorded paths; Linking the multiple record paths based on the relationship triples to obtain multiple relationship triple sequences, wherein the multiple relationship triple sequences correspond to multiple production scheduling combination paths; The multiple production scheduling combination paths are selected based on the inventory of the automobile parts to obtain the optimal production scheduling combination path, including: Taking the first entity node of the plurality of production scheduling combination paths as a central entity node, and obtaining the next entity node of the central entity node; The weight coefficient from the central entity node to the next entity node is calculated to obtain a weight coefficient set, and the next entity node corresponding to the maximum weight coefficient in the weight coefficient set is taken as the optimal entity node, wherein the calculation formula of the weight coefficient is as follows: ； in, Represents the weight coefficient from the central entity node to the next entity node, represents the rectified linear unit activation function, A function representing the correlation between two entity nodes. The weight matrix represents the feature transformation of the entity node, and the weight matrix is ​​related to the inventory of auto parts. Represents the central entity node The embedding vector of Indicates the next entity node The embedding vector of The optimal entity node is used as a new central entity node, and the above operation is repeated until all entity nodes in the multiple production scheduling combination paths are traversed to obtain an optimal entity node set; Connecting all the optimal entity nodes in the optimal entity node set to obtain an optimal production scheduling combination path; A first objective function is constructed based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, including: The total processing time of the target automotive parts to be produced is calculated based on the total number of sub-part types, the processes and processing equipment required for the sub-parts, wherein the calculation formula of the total processing time is as follows: ； in, represents the total processing time of the target automotive part to be produced, Indicates sub-assembly Process Is the equipment being processed? Upper processing, if the sub-accessories Process In processing equipment On processing, , otherwise, , Indicates sub-assembly Process In processing equipment The completion time of the processing, Indicates sub-assembly Process In processing equipment The start time of processing, Indicates the total number of sub-assembly types. Represents the total number of processes, Indicates the total number of processing equipment; The total waiting time of the target automotive parts to be produced is calculated based on the total number of sub-part types, the processes and processing equipment required for the sub-parts, wherein the calculation formula of the total waiting time is as follows: ； in, represents the total waiting time of the target automotive parts to be produced, Indicates sub-assembly Process In processing equipment Before you finish the waiting time in the waiting area, Indicates sub-assembly Process In processing equipment Start waiting time in the waiting area before; The total queue time of the target auto parts to be produced is calculated based on the total number of sub-parts, the processes and processing equipment required for the sub-parts, wherein the calculation formula of the total queue time is as follows: ； in, represents the total queue time of the target auto parts to be produced, Indicates sub-assembly Process In processing equipment After completion, complete the queue time in the queue area. Indicates sub-assembly Process In processing equipment After completion, the queue time will start in the queue area; A first objective function is constructed based on the total processing time, total waiting time and total queuing time of the target automotive parts to be produced, wherein the first objective function is as follows: ； in, represents the first objective function, represents the maximum value function; The first constraint is as follows: ； in, Indicates the process Optional processing equipment set, Indicates sub-assembly Process In processing equipment processing time; A second objective function is constructed based on the type of processing equipment required for the sub-assembly, including: The startup carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the startup carbon emissions of the processing equipment is as follows: ； in, Indicates the carbon emissions from the start-up of the processing equipment. Indicates the starting power of the processing equipment, Indicates the start-up time of the processing equipment, represents the carbon emission factor of electric energy; The carbon emissions of fixture installation of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula of the carbon emissions of fixture installation of the processing equipment is as follows: ； in, Indicates the carbon emissions of fixture installation of processing equipment, Indicates the fixture installation power of the processing equipment, Indicates the fixture installation time of the processing equipment; The processing carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the processing carbon emissions of the processing equipment is as follows: ； in, represents the processing carbon emissions of the processing equipment, Indicates the processing power of the processing equipment, Indicates the processing time of the processing equipment; The carbon emissions from fixture removal of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for carbon emissions from fixture removal of the processing equipment is as follows: ； in, Indicates the carbon emissions from fixture removal of processing equipment, Indicates the fixture removal power of the processing equipment, Indicates the fixture disassembly time of the processing equipment; The standby carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the standby carbon emissions of the processing equipment is as follows: ； in, represents the standby carbon emissions of processing equipment, Indicates the standby power of the processing equipment, Indicates the standby time of processing equipment; The shutdown carbon emissions of the processing equipment are calculated based on the type of processing equipment required for the sub-assembly, wherein the calculation formula for the shutdown carbon emissions of the processing equipment is as follows: ； in, Indicates the shutdown carbon emissions of processing equipment, Indicates the shutdown power of the processing equipment, Indicates the shutdown time of the processing equipment; A second objective function is constructed based on the carbon emissions from starting up the processing equipment, the carbon emissions from installing the fixture, the carbon emissions from processing, the carbon emissions from removing the fixture, the carbon emissions from standby, and the carbon emissions from shutting down, wherein the second objective function is as follows: ； in, represents the second objective function, Represents the minimum value function; The second constraint is as follows: ； The production scheduling optimization model is solved based on the improved ant algorithm to obtain a production scheduling solution, including: Based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assemblies, a plurality of processing orders are obtained, the plurality of processing orders are used as mobile nodes of ants, and the parallel processing equipment of each process is used as a virtual node of the ants; Initializing the ant colony, wherein the initialization includes setting the number of ants, setting the initial position of each ant to a randomly selected processing order, and assigning an initial pheromone value to each path, wherein the number of ants is proportional to the total number of processing orders, and the path is an assignment of a process of each processing order to a processing device; Based on the parallel processing equipment available in each process of each processing order, a candidate equipment list is constructed, and the path is selected based on the pheromone strength and heuristic information to obtain the process sequence and processing equipment of each processing order, wherein the path selection formula is as follows: ； in, Indicates the path chosen by the ant The probability of Indicates the path The pheromone strength, Indicates the path Heuristic information, and represents the weight factor; After completing one iteration, the pheromone is updated according to the first objective function value and the second objective function value of each ant, and a path is selected. The above operation is repeated until the maximum number of iterations is reached, and the global optimal solution is output. Based on the global optimal solution, a production scheduling plan is obtained.

2. The method for automobile parts production scheduling based on big data analysis according to claim 1 is characterized in that: The automobile knowledge graph includes multiple entity nodes and multiple relationship edges, wherein the entity nodes correspond to the automobile parts, and the relationship edges correspond to the relationships between the automobile parts, wherein the relationships between the automobile parts include inclusion relationships and category relationships, the relationship triples include a head entity node, a relationship edge and a tail entity node, and the production scheduling combination path is composed of multiple relationship triplets.

3. An automobile parts production scheduling system based on big data analysis, which is applicable to an automobile parts production scheduling method based on big data analysis as described in any one of claims 1-2, characterized in that: include: A graph construction unit (1), the graph construction unit (1) is used to construct an automobile knowledge graph based on the relationship between automobile parts, and to construct a relationship triple based on the automobile knowledge graph; A production scheduling combination unit (2), the production scheduling combination unit (2) being used to obtain a target automobile part to be produced, match the target automobile part to be produced with the automobile knowledge graph to obtain a plurality of production scheduling combination paths, and select the plurality of production scheduling combination paths based on the inventory of the automobile part to obtain an optimal production scheduling combination path; A component determination unit (3), the component determination unit (3) being used to determine the sub-components required for the target automobile component to be produced based on the optimal production scheduling combination path, determine the process required for the sub-component based on the total number of sub-component types, and determine the processing equipment required for the sub-component based on the process required for the sub-component; A target construction unit (4), the target construction unit (4) being used to construct a first target function and a first constraint condition based on the total number of sub-assembly types, the processes and processing equipment required for the sub-assembly, and to construct a second target function and a second constraint condition based on the types of processing equipment required for the sub-assembly; A production scheduling unit (5), the production scheduling unit (5) is used to construct a production scheduling optimization model based on the first objective function, the first constraint condition, the second objective function and the second constraint condition, and solve the production scheduling optimization model based on an improved ant algorithm to obtain a production scheduling solution.

Images