Supply chain logistics provider multi-task cooperation method for intelligent manufacturing

Abstract

The present invention relates to the technical field of industrial supply chains. Disclosed is a supply chain logistics provider multi-task cooperation method for intelligent manufacturing. The method comprises: acquiring logistics provider task initial assignment information, calculating a logistics task initial assignment total cost, and then using a sealed-bid auction to complete logistics task reassignment in intelligent manufacturing, so as to achieve supply chain logistics provider multi-task cooperation. According to the present invention, by establishing an optimal logistics task cost model and an autonomous logistics task assignment model, horizontal and efficient cooperation of supply chain logistics providers for intelligent manufacturing is achieved, logistics provider horizontal cooperation with low cost and high satisfaction is achieved under the condition that logistics information is opaque and no third-party agent is introduced, and rational and efficient configuration of resources between supply chain demanders and suppliers is achieved.

Description

A multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing Technical Field

[0001] The present invention belongs to the field of industrial supply chain technology, and specifically relates to a multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing. Background Art

[0002] With the implementation of ambitious manufacturing transformation and upgrading plans across the globe, such as the Industrial Internet, Industry 5.0, and Smart Manufacturing 2025, smart manufacturing, as a new model and approach to new manufacturing technologies and industrial transformation, has garnered significant attention from both academia and industry. Supply chain collaboration, an essential component of smart manufacturing, can agilely and efficiently facilitate the transformation and upgrading of traditional manufacturing to smart manufacturing, and has become a research hotspot within both academia and industry. Supply chain collaboration for smart manufacturing aims to reduce logistics costs, improve production efficiency, and enhance customer satisfaction through collaboration and information sharing among manufacturing enterprises, thereby enabling intelligent and flexible collaborative scheduling of supply chain logistics. However, the decentralized resource management and illicit competition among logistics enterprises in traditional manufacturing supply chains lead to low logistics efficiency, high logistics costs, and irrational resource allocation, making it difficult to meet the demands of flexible collaborative resource allocation for smart manufacturing supply chains. Therefore, integrating decentralized supply chain resources to achieve efficient collaboration among logistics providers and realize low-cost, green, and smart logistics poses a significant challenge in this field.

[0003] In the manufacturing industry, logistics services are generally provided through a general contracting model, while traditional supply chain logistics collaboration primarily involves collaboration between logistics service integrators and functional logistics service providers. Logistics service integrators integrate clients' logistics tasks and providers' resources, centrally allocate and schedule logistics tasks, and then contract them to individual logistics service providers. However, optimal centralized allocation is achieved by integrators under the premise of complete transparency of provider information. In actual applications, competition among providers leads to a lack of horizontal coordination in task allocation, making it difficult to maximize profits and resulting in irrational resource allocation. For example, patent application CN114926128A discloses a logistics task allocation method, apparatus, device, and storage medium. This method uses genetic cross-optimization to optimize the task delivery chain based on total delivery time, thereby generating an optimal task list. However, this technique only considers the optimal total delivery time, i.e., the lowest total delivery cost, and does not consider user satisfaction with each logistics task. Furthermore, this method performs tasks in series, making logistics task information completely transparent, which may not guarantee data security. This method uses a genetic algorithm to optimize the task delivery chain, resulting in low efficiency and potentially ineffective allocation of multiple logistics tasks. Therefore, horizontal collaboration among logistics providers, as an important way to rationally and efficiently allocate resources between demanders and suppliers in the supply chain, has important research significance. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention proposes a multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing. By establishing an optimal logistics task cost model and an autonomous logistics task allocation model, efficient horizontal collaboration among logistics providers oriented to intelligent manufacturing supply chain is achieved. By integrating the logistics task auction based on sealed-bid auction and the logistics task cost method based on the improved BSO algorithm, low-cost and high-satisfaction horizontal collaboration among logistics providers is achieved under the condition of opaque logistics information and without introducing third-party agents, thereby realizing reasonable and efficient resource allocation between supply chain demanders and suppliers.

[0005] The multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing described in the present invention includes the following steps:

[0006] Step 1: Obtain the initial assignment information of the logistics provider's tasks and calculate the total cost of the initial assignment of logistics tasks;

[0007] Step 2: Use a sealed-bid auction method to redistribute smart manufacturing logistics tasks. Each logistics provider, acting as the auctioneer of logistics tasks, selects the most costly logistics tasks from their respective logistics task sets as tasks to be auctioned, and publishes the auction information corresponding to the tasks in the auction process.

[0008] Step 3: Other logistics providers, i.e., bidders, add the tasks to be auctioned to their own task set and use the improved brainstorming algorithm, i.e., the L-BSO algorithm, to independently calculate the minimum total cost to complete the task. ;

[0009] Step 4: After calculating the lowest cost, each logistics provider will determine a bid for the task based on the cost of completing the task within the specified time and submit the bid to the auctioneer in a sealed envelope.

[0010] Step 5: The auctioneer compares the task costs and timeouts of each bidder and selects the best bidder to complete the task transfer;

[0011] Step 6: After all tasks are auctioned, each logistics provider updates its own task subset based on the auction results, thus achieving multi-task collaboration among supply chain logistics providers.

[0012] Furthermore, in step 1, initially, each logistics provider is responsible for the logistics tasks in the area it contracts. Each logistics provider has its own logistics task set, which includes the starting point, end point, and deadline for completion of each task. The minimum total cost of completing the initial logistics tasks is calculated based on the initial task allocation of each logistics provider. , including logistics and transportation costs and overtime penalty costs The specific process is as follows:

[0013] Step 1.1: Calculate logistics and transportation costs :

[0014] (1)

[0015] (2)

[0016] in, is the number of logistics providers, Represents a task set The task set of the jth logistics provider in ; Since the present invention uses auction to complete the logistics task transaction, it is necessary to calculate the transportation cost of a single task, which is: (3)

[0017] in, Indicates a task transportation costs, Indicates that there is a task The set of tasks contained in the vehicle, represents the transportation cost calculation function; the transportation cost of a single task is primarily used in auction bidding to measure the differences in transportation costs of different logistics providers completing the same task. Since each logistics provider needs to assign tasks in its subtask set to its own transportation vehicles, and the logistics transportation costs considered in this invention are mainly affected by path length, the cost calculation here is for the task set owned by a single vehicle. By optimally planning the execution order of tasks in the task set, the cost of completing all tasks in the task set can be calculated.

[0018] Step 1.2: Calculate the logistics overtime penalty cost :

[0019] (4)

[0020] (5)

[0021] (6)

[0022] in, represents the sum of the overtime penalty costs of each logistics provider completing the task subset, Indicates a task The overtime penalty cost is determined by the timeout of each task. If the actual completion time of a task exceeds the user's expected completion time, the task needs to have an overtime penalty cost. Otherwise, no overtime penalty cost is added. and Represents tasks The actual completion time and the expected completion time of the task, Indicates that the logistics provider has The quotation of the task; Formula (5) represents the situation when the actual completion time of the task exceeds the expected completion time of the task; at this time, the overtime penalty cost is proportional to the task price and is affected by the penalty coefficient The influence of size affects the state of task allocation by changing the size of the coefficient. Formula (6) represents the situation where the actual completion time of a task is less than the expected completion time of the task. At this time, the overtime penalty cost is zero and the task allocation state is not affected.

[0023] The constraints of the logistics task cost model are: (7)

[0024] Furthermore, in step 3, the L-BSO algorithm is used to calculate the minimum total cost of the task to be auctioned. The specific steps are as follows:

[0025] Step 3.1: Other logistics providers obtain the current auction logistics tasks Detailed information, including the starting point of the logistics task, the end point of the logistics task and the deadline for completion of the logistics task, will be Join its logistics task subset middle;

[0026] Step 3.2: Calculate the logistics task subset The distance between each task in the game is randomly generated. Individual composition , the position of each individual is represented as a current solution , calculate the fitness value of each individual, that is, the total cost of completing each logistics task in the task subset ;

[0027] Step 3.3: Use K-means algorithm to Individuals aggregate into Class, calculate the fitness of each individual, select the individual with the best fitness value in each class to generate Cluster centers ; Generate a random number ;if Less than probability , then randomly select a cluster center and randomly generate a new individual to replace the cluster center. is a given constant value;

[0028] Step 3.4: Generate a random number ,if Less than probability , then a class is randomly selected to generate a new individual; when a class is randomly selected, a random number is generated ,if Less than probability Select the class center of the class and add random perturbations to generate new individuals; otherwise, randomly select an individual in the class and use the single individual update strategy to generate new individuals. 、 is a given constant;

[0029] Step 3.5: If Greater than probability , then two classes are randomly selected to generate new individuals; when two classes are randomly selected, a random number is generated ,if Less than probability Select the class centers of the two classes, otherwise randomly select two individuals from the two classes, and use the heuristic crossover operator method based on adaptive transfer to perform multiple individual fusion updates. is a given constant;

[0030] Step 3.6: Compare the fitness value of the newly generated individual with that of the current individual (i.e., the lowest total cost of the logistics task), and select the one with the smaller fitness value as the new individual for the next iteration.

[0031] Step 3.7: If If the individual update is completed, go to step 3.8, otherwise go to step 3.4;

[0032] Step 3.8: Adaptively adjust the cluster k value based on the current population's optimal fitness. The adaptive adjustment formula is as follows: (9)

[0033] in is the current k value, is the initial set k value; is the optimal fitness of the current population, The optimal fitness of the initial population; represents rounding up;

[0034] Step 3.9: If the maximum number of iterations is reached, stop the iteration, find the optimal path, and generate the lowest logistics cost for the logistics provider to complete the task. Otherwise, return to step 3.3.

[0035] Furthermore, in step 3.4, the update strategy for a single individual is: Assume is the currently randomly selected individual, and the cluster center of the class to which the individual belongs is ,exist Randomly select a position in ,if middle Then search Zhongyu Equal position, and then compare that position with Exchange rows and eventually generate new individuals ,use replace .

[0036] Furthermore, the specific steps of using the heuristic crossover operator method based on adaptive transfer to perform multiple individual fusion updates are as follows:

[0037] The two selected individuals are taken as two parent individuals. Assume that the two parent individuals are and , transfer the first m positions of the two parent individuals to the two child individuals respectively, delete the first m child nodes of the two parent individuals; select the mth position as the starting point of the two child individuals, and find the parent individual position. Right position and ,as well as Left position and ;Compare and and distance, select the position with the closer distance as the next position of child individual 1, and update all the positions of child individual 1; compare and and The position with the closer distance is selected as the next position of child individual 2, and all positions of child individual 2 are updated accordingly; the definition formula of m is as follows: (8)

[0038] in is the maximum number of iterations, is the current iteration number, is the dimension of the problem solution, is the adaptive coefficient, To round down.

[0039] Furthermore, based on an analysis of actual task scenarios, it is concluded that the recipient of the auctioned task can be other service vehicles of the same logistics provider or other logistics providers. Therefore, the basic unit of logistics task allocation in this invention is set as a single service vehicle, and the basic unit of task auction is also a service vehicle. This ensures that the relatively optimal logistics provider can be selected for each task. The optimal bidding strategy selected by this invention is:

[0040] (10)

[0041] (11)

[0042] in, Logistics vehicles Complete the task The required cost, Represents a collection of logistics vehicles; Indicates transport vehicle Will the task Transfer to transport vehicle , which contains two values, 0 means not transferred, 1 means transferred, and each logistics task can only be transferred to one logistics vehicle.

[0043] The beneficial effects of the present invention are as follows: the method described in the present invention not only considers economic benefits but also fully considers time factors, achieving a reasonable allocation of logistics tasks; while calculating logistics transportation costs, it also calculates overtime penalty costs, allowing for dynamic adjustment of penalty weights, effectively improving user satisfaction. The present invention utilizes a multi-task collaboration method among multiple logistics providers based on a sealed-bid auction, enabling autonomous logistics task scheduling among logistics providers without the need for information transparency and the introduction of third-party agents; and simultaneously employs a logistics task cost calculation method based on an improved brainstorming algorithm, which can reduce logistics task transportation costs, effectively improve computational efficiency, and shorten the final model solution time.

[0044] BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a schematic diagram of multi-task collaborative allocation for logistics providers;

[0046] Figure 2 is a schematic diagram of a single individual update strategy;

[0047] FIG3 is a flow chart of the method of the present invention;

[0048] Figure 4 is a flow chart of calculating the minimum cost of logistics tasks using the L-BSO algorithm;

[0049] Figure 5 is a diagram of the initial distribution execution of logistics tasks;

[0050] Figure 6 is the execution graph after task redistribution when the penalty weight is 0.1;

[0051] Figure 7 is the execution diagram after task reallocation when the penalty weight is 0.9.

[0052] DETAILED DESCRIPTION

[0053] In order to make the contents of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments in conjunction with the accompanying drawings.

[0054] This paper proposes a multi-task collaboration method for supply chain logistics providers in intelligent manufacturing, which is aimed at the scenario where upstream enterprises in the intelligent manufacturing supply chain supply products to downstream enterprises through third-party logistics providers; each product supply and transportation task is recorded as a logistics task. , upstream companies entrust multiple logistics companies to handle multiple tasks.

[0055] As shown in the small figure (a) of Figure 1, under the initial task allocation, the cost of task 3 handled by logistics provider A is obviously too high. This is because there is no collaborative relationship between logistics providers and they can only complete the tasks independently, resulting in suboptimal logistics costs and high logistics costs. Assume that horizontal collaboration is achieved among logistics providers, and certain high-cost tasks in the task set are handed over to other logistics providers to complete. As shown in the small figure (b) of Figure 1, if the logistics tasks are redistributed among logistics providers, and logistics provider B handles logistics task 3, the overall logistics cost can be reduced. The present invention aims to implement collaborative allocation of multiple tasks among logistics providers to achieve the lowest overall logistics cost and the highest user satisfaction, where user satisfaction is reflected in the form of overtime penalty costs. The ultimate collaborative goal is to minimize the total cost: (1)

[0056] in, Represents the total cost, including logistics transportation cost and overtime penalty cost.

[0057] As shown in FIG3 , the multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing according to the present invention includes the following steps:

[0058] Step 1: Obtain the initial task allocation information of the logistics providers and calculate the total cost of the initial logistics task allocation. Initially, each logistics provider is responsible for the logistics tasks in the area it contracts. Each logistics provider has its own logistics task set, which includes the starting point, end point, and deadline for each task. Calculate the minimum total cost of completing the initial logistics tasks based on the initial task allocation of each logistics provider. , including logistics and transportation costs and overtime penalty costs ;

[0059] Step 2: A sealed-bid auction is used to redistribute smart manufacturing logistics tasks. Each logistics provider, acting as the auctioneer of logistics tasks, selects the logistics tasks with the highest costs from their respective logistics task sets as tasks to be auctioned and publishes the auction information corresponding to these tasks in the auction process.

[0060] Step 3: Other logistics providers, i.e., bidders, add the task to their own task set and use the improved brainstorming algorithm to independently calculate the lowest total cost to complete the task. ,The algorithm is referred to as L-BSO algorithm, and the cost calculation process is shown in Figure 4;

[0061] Step 3.1: Other logistics providers obtain the current auction logistics tasks Detailed information, including the starting point of the logistics task, the end point of the logistics task and the deadline for completion of the logistics task, will be Join its logistics task subset middle;

[0062] Step 3.2: Calculate the logistics task subset The distance between each task in the game is randomly generated. Individual composition , the position of each individual is represented as a current solution , calculate the fitness value of each individual, that is, the total cost of completing each logistics task in the task subset ;

[0063] Using K-means algorithm Individuals aggregate into Class, calculate the fitness of each individual, select the individual with the best fitness value in each class to generate Cluster centers ; Generate a random number ;if Less than probability , then randomly select a cluster center and randomly generate a new individual to replace the cluster center. is a given constant value;

[0064] Step 3.4: Generate a random number ,if Less than probability , then a class is randomly selected to generate a new individual; when a class is randomly selected, a random number is generated ,if Less than probability Select the class center of the class and add random perturbations to generate new individuals; otherwise, randomly select an individual in the class and use the single individual update strategy to generate new individuals. 、 is a given constant;

[0065] Among them, the single individual update strategy is shown in Figure 2, specifically: Assume is the currently randomly selected individual, and the cluster center of the class to which the individual belongs is ,exist Randomly select a position in ,if middle Then search Zhongyu Equal position, and then compare that position with Exchange and eventually generate new individuals ,use replace ;

[0066] Step 3.5: If Greater than probability , then two classes are randomly selected to generate new individuals; when two classes are randomly selected, a random number is generated ,if Less than probability Select the class centers of the two classes, otherwise randomly select two individuals from the two classes, and use the heuristic crossover operator method based on adaptive transfer to perform multiple individual fusion updates. is a given constant;

[0067] Step 3.6: Compare the fitness value of the newly generated individual with that of the current individual (i.e., the lowest total cost of the logistics task), and select the one with the smaller fitness value as the new individual for the next iteration.

[0068] Step 3.7: If If the individual update is completed, go to step 3.8, otherwise go to step 3.4;

[0069] Step 3.8: Adaptively adjust the cluster k value based on the current population's optimal fitness. The adaptive adjustment formula is as follows: (9)

[0070] in For the current value, is the initial set k value; is the optimal fitness of the current population, is the optimal fitness of the initial population; represents rounding up;

[0071] Step 3.9: If the maximum number of iterations is reached, stop the iteration, find the optimal path, and generate the lowest logistics cost for the logistics provider to complete the task. Otherwise, return to step 3.3.

[0072] Further,

[0073] Step 4: After calculating the lowest cost, each logistics provider shall determine the bid for the task based on the cost of completing the task within the specified time and submit the bid for the task in a sealed form to the auctioneer.

[0074] Step 5: The auctioneer compares the task costs and timeouts of each bidder and selects the best bidder to complete the task transfer. The present invention sets the basic unit of logistics task allocation to a single service vehicle, so the basic unit of task auctions is also a service vehicle. This ensures that the best logistics provider is selected for each task.

[0075] Step 6: After all tasks are auctioned, each logistics provider updates its own task subset based on the auction results, thus achieving multi-task collaboration among supply chain logistics providers.

[0076] In summary, to verify the effectiveness of the proposed method, we considered that in real-world situations, there are multiple cooperating logistics providers, each of which has multiple logistics vehicles. Therefore, we implemented the method on a dataset with 30 logistics tasks, two logistics providers, and two transport vehicles per provider. As shown in Figure 5, the two logistics providers are located at coordinates (0, 0) and (30, 0), respectively.

[0077] Figure 5 shows the optimal task execution order and optimal task execution path for each logistics provider on their respective logistics task subsets before the initial allocation of logistics tasks. The closed-loop directed path graph formed by the points of different shapes in the figure represents the task execution paths of different vehicles. Subsequently, through the sealed-bid auction proposed by this invention, each logistics provider auctions off its own high-cost tasks, and logistics tasks are collaboratively allocated. Figures 6 and 7 show the reallocation of logistics tasks under different timeout penalty weights, respectively. Different timeout penalty weights have a certain impact on logistics task reallocation.

[0078] Figures 6 and 7 show the logistics task allocation when the timeout penalty weight values ​​are 0.1 and 0.9, respectively. The figures also show the logistics providers to which each task belongs and the optimal task execution order and execution path of each logistics provider on its updated logistics task subset.

[0079] Table 1 Cost reduction rate and number of overtime tasks under different overtime penalty weights:

[0080]

[0081] Table 1 shows the impact of different overtime penalty weights on logistics provider costs and the number of overtime tasks. As can be seen from the table, when the penalty weight is 0 (no penalty for logistics overtime), the number of overtime tasks is 6. When the overtime penalty weight is 0.3, the number of overtime tasks is reduced to 4, improving user satisfaction.

[0082] When the timeout penalty weight is not set, the multi-task collaboration method for logistics providers proposed in the present invention can promote horizontal collaboration among various logistics providers and reduce the total logistics cost by 13%. When the timeout penalty weight is set, the setting of the penalty cost increases the overall logistics cost. The multi-task collaboration method for logistics providers proposed in the present invention can reduce transportation costs and penalty costs, resulting in an overall total cost reduction rate that is higher than the total cost reduction rate when the penalty weight is 0. During the entire process of experimental verification of the change in the penalty weight value, although the reduction rate of the total logistics cost fluctuated slightly, it was maintained at more than 16%.

[0083] Experimental simulations confirm the feasibility of the proposed multi-task collaboration method for logistics providers, demonstrating its effectiveness. The proposed multi-task collaboration method can effectively reduce the number of overdue tasks, improve the quality of logistics services, and significantly reduce the overall cost of the logistics transportation network.

[0084] The above description is only a preferred embodiment of the present invention and is not intended to further limit the present invention. All equivalent changes made using the contents of the present invention description and drawings are within the scope of protection of the present invention.

Claims

1. A multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing, characterized in that: The following steps are involved: Step 1: Obtain the initial assignment information of logistics providers' tasks and calculate the total cost of initial assignment of logistics tasks; Step 2: Each logistics provider, as the auctioneer of logistics tasks, selects logistics tasks with high costs in its own logistics task set as tasks to be auctioned, and publishes auction information corresponding to the tasks to be auctioned in the auction process; Step 3: Other logistics providers, i.e. bidders, add the tasks to be auctioned to their own task set and use the improved L-BSO algorithm to independently calculate the minimum total cost to complete the task. ； Step 4: Each logistics provider calculates the lowest total cost After that, the bidder shall determine the bid for the task according to the cost of completing the task within the specified time, and submit the bid for the task in a sealed form to the auctioneer; Step 5: The auctioneer compares the task costs and timeouts of the bidders and selects the best bidder to complete the task transfer; Step 6: After all task auctions are completed, each logistics provider updates its own task subset based on the auction results, thus achieving multi-task collaboration among supply chain logistics providers.

2. The multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing according to claim 1, characterized in that: In step 1, initially, each logistics provider is responsible for the logistics tasks in the area it contracts. Each logistics provider has its own logistics task set, which includes the starting point, end point and deadline of each task. The minimum total cost of completing the initial logistics tasks is calculated based on the initial task allocation of each logistics provider. Including logistics and transportation costs and overtime penalty cost ; The specific process is as follows: Step 1.1: Calculate logistics and transportation costs ： (1) (2) in, is the number of logistics providers, Represents a collection of tasks The task set of the jth logistics provider in the present invention; since the present invention adopts the auction method to complete the logistics task transaction, it is necessary to calculate the transportation cost of a single task, and the transportation cost of a single task is: (3) in, Indicates the task transportation costs, Indicates that it has a task The set of tasks contained in the vehicle, represents the transportation cost calculation function; Step 1.2: Calculate the logistics overtime penalty cost ： (4) (5) (6) in, represents the sum of the overtime penalty costs of each logistics provider to complete the task subset, Indicates the task The overtime penalty cost is determined by the overtime situation of each task. If the actual completion time of a task exceeds the user's expected completion time, the task needs to have an overtime penalty cost, otherwise no overtime penalty cost needs to be added. and Respectively represent tasks The actual completion time and the expected completion time of the task, Indicates the logistics provider's The quotation of the task; Formula (5) represents the situation when the actual completion time of the task exceeds the expected completion time of the task; at this time, the overtime penalty cost is proportional to the task price and is affected by the penalty coefficient The influence of size, by changing the size of the coefficient, affects the state of task allocation; Formula (6) represents the situation where the actual completion time of a task is less than the expected completion time of the task. At this time, the overtime penalty cost is zero, and the task allocation state is not affected; The constraints of the logistics task cost model are: (7)。 3. The multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing according to claim 2 is characterized in that: The overtime penalty cost is determined by the overtime situation of each task. If the actual completion time of a task exceeds the user's expected completion time, the task needs to have an overtime penalty cost. Otherwise, no overtime penalty cost needs to be added.

4. The multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing according to claim 2 is characterized in that: In step 3, the L-BSO algorithm is used to calculate the minimum total cost of the task to be auctioned. The specific steps are: Step 3.1: Other logistics providers obtain the current auction logistics tasks The detailed information of the logistics task, including the starting point, the end point and the deadline for completion of the logistics task, will be Join its logistics mission subset middle; Step 3.2: Calculate the logistics task subset The distance between each task in the game is randomly generated. Individual composition , the position of each individual is represented as a solution , calculate the fitness value of each individual, that is, the total cost of completing each logistics task in the task subset ； Step 3.3: Use K-means algorithm to Aggregation of individuals class, calculate the fitness of each individual, select the individual with the best fitness value in each class to generate Cluster Centers ; Generate a random number ;if Less than probability , then randomly select a cluster center and randomly generate a new individual to replace the cluster center. is a given constant value; Step 3.4: Generate a random number ,if Less than probability , then a class is randomly selected to generate a new individual; when a class is randomly selected, a random number is generated ,if Less than probability Select the class center of the class and add random perturbations to generate new individuals; Otherwise, randomly select an individual in the class and use the single individual update strategy to generate a new individual. 、 is a given constant; Step 3.5: If Greater than probability , then two classes are randomly selected to generate new individuals; when two classes are randomly selected, a random number is generated ,if Less than probability Select the class centers of the two classes, otherwise randomly select two individuals from the two classes, and use the heuristic crossover operator method based on adaptive transfer to perform multiple individual fusion updates. is a given constant; Step 3.6: Compare the fitness values ​​of the newly generated individual with the current individual, and select the one with the smaller fitness value as the new individual for the next iteration; Step 3.7: If If the update of each individual is completed, go to step 3.8, otherwise go to step 3.4; Step 3.8: Adaptively adjust the cluster k value based on the current population's optimal fitness. The adaptive adjustment formula is as follows: (8) in is the current k value, is the initial set k value; is the optimal fitness of the current population, The optimal fitness of the initial population; represents rounding up; Step 3.9: If the maximum number of iterations is reached, stop the iteration, find the optimal path, and generate the minimum logistics cost for the logistics provider to complete the task, otherwise return to step 3.

3.

5. The multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing according to claim 4 is characterized in that: In step 3.4, the update strategy for a single individual is: Assume is the currently randomly selected individual, and the cluster center of the class to which the individual belongs is ,exist Randomly select a position in ,if middle Then search for Zhongyu equal position, and then compare that position with Exchange rows and finally generate new individuals ,use replace .

6. The multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing according to claim 4 is characterized in that: The specific steps of using the heuristic crossover operator method based on adaptive transfer to perform multiple individual fusion updates are as follows: the two selected individuals are used as two parent individuals, assuming that the two parent individuals are and , pass the first m positions of the two parent individuals to the two child individuals respectively, delete the first m child nodes of the two parent individuals; select the mth position as the starting point of the two child individuals, and find the parent individual position Right position and ,as well as Left position and ;Compare and and distance, select the position with a closer distance as the next position of sub-individual 1, and update all the positions of sub-individual 1; compare and and The position with a closer distance is selected as the next position of sub-individual 2, so as to update all the positions of sub-individual 2; the definition formula of m is as follows: (9) in is the maximum number of iterations, is the current iteration number, is the dimension of the problem solution, is the adaptive coefficient, To round down.

7. The multi-task collaboration method for supply chain logistics providers oriented to intelligent manufacturing according to claim 1 is characterized in that: The recipient of the auctioned task is other service vehicles of the same logistics provider or other logistics providers. The basic unit of logistics task allocation is set to a single service vehicle, and the basic unit of task auction is also a service vehicle. The optimal bidding strategy is: (10) (11) in, Indicates logistics vehicles Complete the task The cost required, Represents a collection of logistics vehicles; Indicates transport vehicle Will the task Transfer to transport vehicle , which contains two values, 0 means not to transfer, 1 means to transfer, and each logistics task can only be transferred to one logistics vehicle.

Images