Durian logistics optimization method and device considering preservation constraint

Abstract

The invention relates to a durian logistics optimization method and device considering preservation constraints, and the method comprises the steps: determining a center set and a dynamic preservation feature set of each logistics vehicle according to the distribution center of the logistics vehicles of the durian and the preservation state features of the durian loaded on each logistics vehicle; a logistics route map is constructed by combining the introduction of a durian quality attenuation model, the information interaction training of the logistics route map is carried out, the model optimization output is used as an optimal transportation route map, and the optimal combination time and the optimal combination object of logistics combination of each logistics vehicle are determined. And the logistics combination range is limited and shrunk based on insurance constraints, so that the logistics time, the logistics cost and the logistics damage of large-scale logistics vehicles are effectively reduced, and the quality of the durian is effectively ensured.

Description

Technical Field

[0001] This invention relates to the field of fresh durian transportation technology, and in particular to a method and apparatus for optimizing durian logistics considering preservation constraints. Background Technology

[0002] As a high-value tropical fruit, durian is extremely sensitive to time and temperature during transportation, making it crucial for logistics teams to ensure its freshness and quality. To guarantee durian quality, road logistics often employs rapid, individual transport methods.

[0003] However, with the increasing prevalence of durian and logistics, the volume of durian transported to remote areas via road logistics is gradually increasing. Traditional, fast, individual transportation methods are insufficient to handle large-volume durian shipments. Therefore, large durian traders typically dispatch hundreds of logistics vehicles simultaneously during the harvest season, transporting durians from multiple plantations to distribution centers across the country. Compared to individual transportation, large-scale logistics vehicle transport effectively reduces the logistics costs of durians. However, because durian's freshness is extremely sensitive, and large-scale logistics vehicles suffer from long dispatch times and numerous unforeseen events, resulting in prolonged delivery times, increased pressure on the freshness of durians, and a higher risk of damage during transport, ultimately reducing durian quality. Summary of the Invention

[0004] Therefore, it is necessary to provide a durian logistics optimization method and device that takes into account preservation constraints, addressing the problems of long dispatching time and frequent emergencies in large-scale logistics vehicles, which result in long logistics time, high pressure on the preservation quality of durians, and easy increase in transportation damage and reduction of durian quality.

[0005] The durian logistics optimization method considering preservation constraints in this application includes the following steps:

[0006] Based on the distribution centers of durian logistics vehicles and the preservation status characteristics of durians loaded on each logistics vehicle, a central set and a dynamic preservation feature set for each logistics vehicle are determined. The dynamic preservation feature set is then substituted into a pre-defined durian quality decay model, and the combination time set for each logistics vehicle and its logistics combination is calculated based on the central set and the dynamic preservation feature set. For each logistics vehicle at each distribution center, a logistics route map is constructed with the logistics vehicle as the central node, based on its logistics combination and combination time set. The node features of the logistics route map include preservation status characteristics, and the edge features include quality risk weights caused by combination time. A pre-trained neural network model is used to process the logistics route map, outputting the optimal combination time and optimal combination object for the logistics vehicle at the distribution center, as the optimal transportation route map. The neural network model is trained with the objective of minimizing the overall transportation cost under the loss of preservation status.

[0007] The aforementioned durian logistics optimization method considering preservation constraints determines the central set and dynamic preservation characteristic set of each logistics vehicle based on the distribution centers of the durian logistics vehicles and the preservation status characteristics of the durians loaded on each vehicle. By introducing a durian quality decay model, a logistics route map is constructed. Through information interaction training of the logistics route map, the model optimization output is used as the optimal transportation route map. This clarifies the optimal combination time and optimal combination objects for each logistics vehicle, and narrows the scope of logistics combination based on insurance constraints. This effectively reduces the logistics time, cost, and cargo damage of large-scale logistics vehicles, while effectively ensuring the quality of durians.

[0008] In one embodiment, a durian transport logistics vehicle dynamic preservation feature set Includes: Used to represent logistics vehicles Initial quality value when durians are loaded at the point of origin. ;

[0009] Used to indicate logistics vehicles At any moment Durian quality value ;

[0010] Remaining shelf life Used to represent logistics vehicles At any moment The durian quality value decays from its current value to a preset acceptable quality threshold. ;

[0011] Used to indicate logistics vehicles The cargo damage cost coefficient representing the economic loss per unit of reduced quality when loading durians. ;

[0012] Used to indicate the minimum delivery quality required by the destination of logistics vehicles. .

[0013] In one embodiment, the durian quality degradation model is as follows:

[0014] ;

[0015] in, Indicates logistics vehicles The rate of quality degradation of loaded durians Indicates logistics vehicles From the point of departure to the time The time elapsed;

[0016] Based on the durian quality degradation model, the logistics vehicle can be obtained. Remaining shelf life of durian As shown in the following formula:

[0017] / .

[0018] In one embodiment, the step of calculating the combination time set of each logistics vehicle and its logistics combination based on the central set and the dynamic preservation feature set includes:

[0019] For logistics vehicles At the distribution center Its logistics vehicles combined with logistics Base waiting time between As shown in the following formula:

[0020] ;

[0021] in, express Arrival at the distribution center At that moment, express Arrival at the distribution center The moment;

[0022] By constraining the base waiting time using the durian quality decay model, a feasible waiting time is obtained. As shown in the following formula:

[0023] ;

[0024] in, express From the distribution center Time to reach the destination;

[0025] Combined time set As shown in the following formula:

[0026] ;

[0027] in, This refers to the logistics combination.

[0028] In one embodiment, the steps for constructing a logistics route map with each logistics vehicle as the central node at each distribution center, based on its logistics combinations and combination time sets, are as follows:

[0029] The logistics route map is as follows:

[0030] ;

[0031] in, This represents a logistics combination that serves as a node. Represents the set of edges consisting of nodes;

[0032] This represents the node feature matrix for the nodes of a logistics vehicle. Its eigenvectors As shown in the following formula:

[0033] ;

[0034] in, Indicates logistics vehicles One-hot encoded identifier vector; Indicates logistics vehicles Distribution Center Order encoding vectors in each distribution center; Indicates the remaining driving distance from the distribution center to the destination;

[0035] The edge feature matrix represents the nodes of a logistics vehicle. and nodes edge Its eigenvectors As shown in the following formula:

[0036] ;

[0037] in, Indicates feasible waiting time One-hot encoded vector;

[0038] The weight of quality risk caused by waiting is expressed as follows:

[0039] ;

[0040] The estimated weights of the logistics combination are expressed as follows:

[0041] ;

[0042] in, Indicates logistics vehicles Fuel cost per unit distance for solo driving; Indicates logistics vehicles Unit distance fuel consumption cost for logistics combination; Indicates logistics vehicles With logistics vehicles The length of the shared path; This indicates the total number of logistics vehicles in a logistics combination.

[0043] In one embodiment, the training process of the neural network model is as follows:

[0044] Building a training dataset ;in, Indicates distribution center The optimal scheduling decision label includes the optimal combination time and the optimal combination object;

[0045] The training dataset is trained using a neural network model with an encoder-decoder structure; wherein the encoder of the neural network model includes L layers of graph convolutional layers, and the node features and edge features of the l-th layer are updated: the node feature update is as follows:

[0046] ;

[0047] The edge features are updated as follows:

[0048] ;

[0049] in, Represents the aggregation function of a node. Aggregate functions representing edges;

[0050] The decoder output is used as the optimal transportation route map.

[0051] In one embodiment, the output process of the decoder includes the following steps:

[0052] The prediction of the softmax classifier is used as the optimal waiting time;

[0053] The output of an attention-based multi-label classifier is used as the optimal combination object.

[0054] This application also provides a durian logistics optimization device that takes into account preservation constraints, including:

[0055] The set processing module is configured to determine the central set and dynamic preservation feature set of each logistics vehicle based on the distribution centers of the durian logistics vehicles and the preservation status characteristics of the durians loaded on each logistics vehicle. The model input module is configured to input the dynamic preservation feature set into a preset durian quality decay model, and calculate the combination time set of each logistics vehicle and its logistics combination based on the central set and the dynamic preservation feature set. The data combination module is configured to construct a logistics route map with each logistics vehicle as the central node at each distribution center, based on its logistics combination and combination time set. The node features of the logistics route map include preservation status characteristics, and the edge features include quality risk weights caused by combination time. The prediction optimization module is configured to process the logistics route map using a pre-trained neural network model, and output the optimal combination time and optimal combination object of the logistics vehicle at the distribution center as the optimal transportation route map. The neural network model is trained with the objective of minimizing the comprehensive transportation cost under the loss of preservation status.

[0056] The aforementioned durian logistics optimization device, which considers preservation constraints, determines the central set and dynamic preservation characteristic set of each logistics vehicle based on the distribution centers of the durian logistics vehicles and the preservation status characteristics of the durians loaded on each vehicle. By introducing a durian quality decay model, a logistics route map is constructed. Through information interaction training of the logistics route map, the model optimization output serves as the optimal transportation route map. This clarifies the optimal combination time and optimal combination objects for each logistics vehicle, and narrows the scope of logistics combination based on insurance constraints. This effectively reduces the logistics time, cost, and cargo damage of large-scale logistics vehicles, while effectively ensuring the quality of durians.

[0057] A computer storage medium storing computer instructions, which, when executed by a processor, implement the durian logistics optimization method considering preservation constraints of any of the above embodiments.

[0058] The aforementioned computer storage medium determines the central set and dynamic preservation characteristic set of each logistics vehicle based on the distribution centers of durian logistics vehicles and the preservation status characteristics of durians loaded on each logistics vehicle. By introducing a durian quality decay model, a logistics route map is constructed. Through information interaction training of the logistics route map, the model optimization output is used as the optimal transportation route map. This clarifies the optimal combination time and optimal combination objects for each logistics vehicle to combine logistics, and narrows the scope of logistics combination based on insurance constraints. This effectively reduces the logistics time, cost, and cargo damage of large-scale logistics vehicles, and effectively ensures the quality of durians.

[0059] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the durian logistics optimization method considering preservation constraints of any of the above embodiments.

[0060] The aforementioned computer equipment determines the central set and dynamic preservation characteristic set of each logistics vehicle based on the distribution centers of durian logistics vehicles and the preservation status characteristics of the durians loaded on each logistics vehicle. By introducing a durian quality decay model, a logistics route map is constructed. Through information interaction training of the logistics route map, the model optimization output is used as the optimal transportation route map. This clarifies the optimal combination time and optimal combination objects for each logistics vehicle to combine logistics, and narrows the scope of logistics combination based on insurance constraints. This effectively reduces the logistics time, cost, and cargo damage of large-scale logistics vehicles, and effectively ensures the quality of durians. Attached Figure Description

[0061] Figure 1 A flowchart of a method for optimizing durian logistics considering preservation constraints, as one embodiment;

[0062] Figure 2 A structural diagram of a durian logistics optimization device module considering preservation constraints in one embodiment;

[0063] Figure 3 This is a schematic diagram of the internal structure of a computer according to one embodiment. Detailed Implementation

[0064] To better understand the purpose, technical solution, and technical effects of this invention, the invention will be further explained and described below in conjunction with the accompanying drawings and embodiments. It should also be stated that the embodiments described below are only for explaining this invention and are not intended to limit this invention.

[0065] This invention provides a method for optimizing durian logistics while considering preservation constraints.

[0066] Figure 1 A flowchart illustrating a method for optimizing durian logistics considering preservation constraints in one embodiment is shown below. Figure 1 As shown, one embodiment of the durian logistics optimization method considering preservation constraints includes steps S100 to S102:

[0067] S100, Based on the distribution centers of durian logistics vehicles and the preservation status characteristics of durians loaded on each logistics vehicle, determine the center set and dynamic preservation feature set of each logistics vehicle; S101, Substitute the dynamic preservation feature set into a preset durian quality decay model, and calculate the combination time set of each logistics vehicle and its logistics combination based on the center set and the dynamic preservation feature set; S102, For each logistics vehicle at each distribution center, construct a logistics route map with the logistics vehicle as the center node based on its logistics combination and combination time set; wherein, the node features of the logistics route map include preservation status characteristics, and the edge features include quality risk weights caused by combination time; S103, Process the logistics route map using a pre-trained neural network model, and output the optimal combination time and optimal combination object of the logistics vehicle at the distribution center as the optimal transportation route map; wherein, the neural network model is trained with the objective of minimizing the comprehensive transportation cost under the loss preservation status.

[0068] The dynamic preservation feature set includes multiple features, including durian quality values. In this embodiment, the freshness of the durian is normalized and categorized, and converted into quality values. The quality values ​​range from 0 to 1, with 1 representing the highest quality.

[0069] In one embodiment, a durian transport logistics vehicle dynamic preservation feature set Includes: Used to represent logistics vehicles Initial quality value when durians are loaded at the point of origin. ;

[0070] Used to indicate logistics vehicles At any moment Durian quality value ;

[0071] Remaining shelf life Used to represent logistics vehicles At any moment The durian quality value decays from its current value to a preset acceptable quality threshold. ;

[0072] Used to indicate logistics vehicles The cargo damage cost coefficient representing the economic loss per unit of reduced quality when loading durians. ;

[0073] Used to indicate the minimum delivery quality required by the destination of logistics vehicles. .

[0074] Among them, the acceptable quality threshold Based on empirical values, a value of 0.6-0.7 is preferred. (Cost coefficient for cargo damage) The cost can be determined based on the quality and unit price of durian, and is expressed as the economic cost per unit weight, generally ranging from 100-300 yuan / ton. For example:

[0075] Musang King durian, =0.025, =200 yuan / ton;

[0076] D24 Durian =0.018, =150 yuan / ton;

[0077] Golden Phoenix Durian =0.022, =180 yuan / t.

[0078] Meanwhile, in the embodiments of this application, the logistics vehicles include vehicles numbered 1, 2, ..., i, ..., j, ..., n, and the distribution centers include vehicles numbered 1, 2, ..., k, ..., m.

[0079] In one embodiment, the durian quality degradation model is as follows:

[0080] ;

[0081] in, Indicates logistics vehicles The rate of quality degradation of loaded durians Indicates logistics vehicles From the point of departure to the time The time elapsed;

[0082] Based on the durian quality degradation model, the logistics vehicle can be obtained. Remaining shelf life of durian As shown in the following formula:

[0083] / .

[0084] Among them, based on the normalization of durian quality, the quality degradation rate It is 0.01-0.03 / h.

[0085] In one embodiment, the step of calculating the combination time set of each logistics vehicle and its logistics combination based on the central set and the dynamic preservation feature set includes:

[0086] For logistics vehicles At the distribution center Its logistics vehicles combined with logistics Base waiting time between As shown in the following formula:

[0087] ;

[0088] in, express Arrival at the distribution center At that moment, express Arrival at the distribution center The moment;

[0089] By constraining the base waiting time using the durian quality decay model, a feasible waiting time is obtained. As shown in the following formula:

[0090] ;

[0091] in, express From the distribution center Time to reach the destination;

[0092] Combined time set As shown in the following formula:

[0093] ;

[0094] in, This refers to the logistics combination.

[0095] In one embodiment, the steps for constructing a logistics route map with each logistics vehicle as the central node at each distribution center, based on its logistics combinations and combination time set, are as follows:

[0096] The logistics route map is as follows:

[0097] ;

[0098] in, This represents a logistics combination that serves as a node. Represents the set of edges consisting of nodes;

[0099] This represents the node feature matrix for the nodes of a logistics vehicle. Its eigenvectors As shown in the following formula:

[0100] ;

[0101] in, Indicates logistics vehicles One-hot encoded identifier vector; Indicates logistics vehicles Distribution Center Order encoding vectors in each distribution center; Indicates the remaining driving distance from the distribution center to the destination;

[0102] The edge feature matrix represents the nodes of a logistics vehicle. and nodes edge Its eigenvectors As shown in the following formula:

[0103] ;

[0104] in, Indicates feasible waiting time One-hot encoded vector;

[0105] The weight of quality risk caused by waiting is expressed as follows:

[0106] ;

[0107] The estimated weights of the logistics combination are expressed as follows:

[0108] ;

[0109] in, Indicates logistics vehicles Fuel cost per unit distance for solo driving; Indicates logistics vehicles Unit distance fuel consumption cost for logistics combination; Indicates logistics vehicles With logistics vehicles The length of the shared path; This indicates the total number of logistics vehicles in a logistics combination.

[0110] Among them, fuel consumption cost per unit distance The cost is 0.8-1.2 yuan per kilometer, representing the fuel cost per unit distance. The cost is 0.65-0.9 yuan per kilometer. Total number of logistics vehicles. Typically, there are 2-5 vehicles.

[0111] In one embodiment, the training process of the neural network model is as follows:

[0112] Building a training dataset ;in, Indicates distribution center The optimal scheduling decision label includes the optimal combination time and the optimal combination object;

[0113] The training dataset is trained using a neural network model with an encoder-decoder structure; wherein the encoder of the neural network model includes L layers of graph convolutional layers, and the node features and edge features of the l-th layer are updated: the node feature update is as follows:

[0114] ;

[0115] The edge features are updated as follows:

[0116] ;

[0117] in, Represents the aggregation function of a node. Aggregate functions representing edges;

[0118] The decoder output is used as the optimal transportation route map.

[0119] The decoder's output process includes the following steps:

[0120] The prediction of the softmax classifier is used as the optimal waiting time;

[0121] The output of an attention-based multi-label classifier is used as the optimal combination object.

[0122] The following example illustrates this point.

[0123] Application settings:

[0124] A typical transportation network includes 15 distribution centers;

[0125] Durian parameters: =0.02 / h, acceptable quality threshold It is 0.65.

[0126] Compared with traditional logistics combinations, the optimized logistics combination results of the embodiments of this application are shown in Table 1 below:

[0127] Table 1. Schematic diagram of logistics combination optimization effect in the embodiments of this application.

[0128] index MILP Rule Heuristics GNN without preservation Examples of this application Solution time (s) 45.2 0.8 0.3 0.4 Quality pass rate 98.7% 92.3% 85.1% 99.1% Fuel savings 14.2% 11.5% 16.8% 15.6% Comprehensive cost benchmark +12.3% +8.7% -5.2%

[0129] The aforementioned durian logistics optimization method considering preservation constraints determines the central set and dynamic preservation characteristic set of each logistics vehicle based on the distribution centers of the durian logistics vehicles and the preservation status characteristics of the durians loaded on each vehicle. By introducing a durian quality decay model, a logistics route map is constructed. Through information interaction training of the logistics route map, the model optimization output is used as the optimal transportation route map. This clarifies the optimal combination time and optimal combination objects for each logistics vehicle, and narrows the scope of logistics combination based on insurance constraints. This effectively reduces the logistics time, cost, and cargo damage of large-scale logistics vehicles, while effectively ensuring the quality of durians.

[0130] This invention also provides a durian logistics optimization device that takes into account preservation constraints.

[0131] Figure 2 This is a structural diagram of a durian logistics optimization device module considering preservation constraints in one embodiment, as shown below. Figure 2 As shown, a durian logistics optimization device considering preservation constraints according to one embodiment includes modules 100, 101, 102, and 103:

[0132] The set processing module 100 is configured to determine the central set and dynamic preservation feature set of each logistics vehicle based on the distribution centers of the durian logistics vehicles and the preservation status characteristics of the durians loaded on each logistics vehicle; the model input module 101 is configured to input the dynamic preservation feature set into a preset durian quality decay model, and calculate the combination time set of each logistics vehicle and its logistics combination based on the central set and the dynamic preservation feature set; the data combination module 102 is configured to construct a logistics route map with each logistics vehicle as the central node at each distribution center, based on its logistics combination and combination time set; wherein the node features of the logistics route map include preservation status features, and the edge features include quality risk weights caused by combination time; the prediction optimization module 103 is configured to process the logistics route map using a pre-trained neural network model, and output the optimal combination time and optimal combination object of the logistics vehicle at the distribution center as the optimal transportation route map; wherein the neural network model is trained with the objective of minimizing the comprehensive transportation cost under the loss preservation status.

[0133] The aforementioned durian logistics optimization device, which considers preservation constraints, determines the central set and dynamic preservation characteristic set of each logistics vehicle based on the distribution centers of the durian logistics vehicles and the preservation status characteristics of the durians loaded on each vehicle. By introducing a durian quality decay model, a logistics route map is constructed. Through information interaction training of the logistics route map, the model optimization output serves as the optimal transportation route map. This clarifies the optimal combination time and optimal combination objects for each logistics vehicle, and narrows the scope of logistics combination based on insurance constraints. This effectively reduces the logistics time, cost, and cargo damage of large-scale logistics vehicles, while effectively ensuring the quality of durians.

[0134] This invention also provides a computer storage medium storing computer instructions that, when executed by a processor, implement the durian logistics optimization method considering preservation constraints of any of the above embodiments.

[0135] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0136] Alternatively, if the integrated units of the present invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the parts that contribute to related technologies, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, terminal, or network device, etc.) to execute all or part of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, RAM, ROM, magnetic disks, or optical disks.

[0137] Corresponding to the computer storage medium described above, one embodiment also provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement any of the durian logistics optimization methods considering preservation constraints as described in the above embodiments.

[0138] This computer device can be a terminal, and its internal structure diagram can be as follows: Figure 3 As shown, the computer device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for optimizing durian logistics considering preservation constraints. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device casing, or an external keyboard, touchpad, or mouse.

[0139] The aforementioned computer equipment determines the central set and dynamic preservation characteristic set of each logistics vehicle based on the distribution centers of durian logistics vehicles and the preservation status characteristics of the durians loaded on each logistics vehicle. By introducing a durian quality decay model, a logistics route map is constructed. Through information interaction training of the logistics route map, the model optimization output is used as the optimal transportation route map. This clarifies the optimal combination time and optimal combination objects for each logistics vehicle to combine logistics, and narrows the scope of logistics combination based on insurance constraints. This effectively reduces the logistics time, cost, and cargo damage of large-scale logistics vehicles, and effectively ensures the quality of durians.

[0140] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0141] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for optimizing durian logistics considering preservation constraints, characterized in that, Including the following steps: Based on the distribution centers of durian logistics vehicles and the preservation status characteristics of durians loaded on each logistics vehicle, a central set and a dynamic preservation feature set for each logistics vehicle are determined. The dynamic preservation feature set is then substituted into a pre-defined durian quality decay model, and the combination time set for each logistics vehicle and its logistics combination is calculated based on the central set and the dynamic preservation feature set. For each logistics vehicle at each distribution center, a logistics route map is constructed with the logistics vehicle as the central node, based on its logistics combination and combination time set. The node features of the logistics route map include preservation status characteristics, and the edge features include quality risk weights caused by combination time. A pre-trained neural network model is used to process the logistics route map, outputting the optimal combination time and optimal combination object for the logistics vehicle at the distribution center, as the optimal transportation route map. The neural network model is trained with the objective of minimizing the overall transportation cost under the loss of preservation status.

2. The durian logistics optimization method considering preservation constraints according to claim 1, characterized in that, Durian transport logistics vehicles dynamic preservation feature set Includes: Used to represent logistics vehicles Initial quality value when durians are loaded at the point of origin. ; Used to indicate logistics vehicles At any moment Durian quality value ; Remaining shelf life Used to represent logistics vehicles At any moment The durian quality value decays from its current value to a preset acceptable quality threshold. ; Used to indicate logistics vehicles The cargo damage cost coefficient representing the economic loss per unit of reduced quality when loading durians. ; Used to indicate the minimum delivery quality required by the destination of logistics vehicles. .

3. The durian logistics optimization method considering preservation constraints according to claim 2, characterized in that, The durian quality degradation model is as follows: ； in, Indicates logistics vehicles The rate of quality degradation of loaded durians Indicates logistics vehicles From the point of departure to the time The time elapsed; Based on the durian quality degradation model, the logistics vehicle can be obtained. Remaining shelf life of durian As shown in the following formula: / 。 4. The durian logistics optimization method considering preservation constraints according to claim 3, characterized in that, The step of calculating the combination time set of each logistics vehicle and its logistics combination based on the central set and the dynamic preservation feature set includes: For logistics vehicles At the distribution center Its logistics vehicles combined with logistics Base waiting time between As shown in the following formula: ； in, express Arrival at the distribution center At that moment, express Arrival at the distribution center The moment; By constraining the base waiting time using the durian quality decay model, a feasible waiting time is obtained. As shown in the following formula: ； in, express From the distribution center Time to reach the destination; Combined time set As shown in the following formula: ； in, This refers to the logistics combination.

5. The durian logistics optimization method considering preservation constraints according to claim 4, characterized in that, The steps for constructing a logistics route map with each logistics vehicle as the central node at each distribution center, based on its logistics combinations and combination time sets, are as follows: The logistics route map is as follows: ； in, This represents a logistics combination that serves as a node. Represents the set of edges consisting of nodes; This represents the node feature matrix for the nodes of a logistics vehicle. Its eigenvectors As shown in the following formula: ； in, Indicates logistics vehicles One-hot encoded identifier vector; Indicates logistics vehicles Distribution Center Order encoding vectors in each distribution center; Indicates the remaining driving distance from the distribution center to the destination; The edge feature matrix represents the nodes of a logistics vehicle. and nodes edge Its eigenvectors As shown in the following formula: ； in, Indicates feasible waiting time One-hot encoded vector; The weight of quality risk caused by waiting is expressed as follows: ； The estimated weights of the logistics combination are expressed as follows: ； in, Indicates logistics vehicles Fuel cost per unit distance for solo driving; Indicates logistics vehicles Unit distance fuel consumption cost for logistics combination; Indicates logistics vehicles With logistics vehicles The length of the shared path; This indicates the total number of logistics vehicles in a logistics combination.

6. The durian logistics optimization method considering preservation constraints according to claim 5, characterized in that, The training process of the neural network model is as follows: Building a training dataset ;in, Indicates distribution center The optimal scheduling decision label includes the optimal combination time and the optimal combination object; The training dataset is trained using a neural network model with an encoder-decoder structure; wherein the encoder of the neural network model includes L layers of graph convolutional layers, and the node features and edge features of the l-th layer are updated as follows: ； The edge features are updated as follows: ； in, Represents the aggregation function of a node. Aggregate functions representing edges; The decoder output is used as the optimal transportation route map.

7. The durian logistics optimization method considering preservation constraints according to claim 6, characterized in that, The output process of the decoder includes the following steps: The prediction of the softmax classifier is used as the optimal waiting time; The output of an attention-based multi-label classifier is used as the optimal combination object.

8. A durian logistics optimization device considering preservation constraints, characterized in that, include: The set processing module is configured to determine the central set and dynamic preservation feature set of each logistics vehicle based on the distribution center of the durian logistics vehicle and the preservation status characteristics of the durian loaded on each logistics vehicle. The model input module is configured to input the dynamic preservation feature set into a preset durian quality decay model, and calculate the combination time set of each logistics vehicle and its logistics combination based on the central set and the dynamic preservation feature set. The data combination module is configured to construct a logistics route map with each logistics vehicle as the central node at each distribution center, based on its logistics combination and combination time set. The node features of the logistics route map include preservation status features, and the edge features include the quality risk weight caused by the combination time. The prediction optimization module is configured to process the logistics route map using a pre-trained neural network model, and output the optimal combination time and optimal combination object of the logistics vehicle at the distribution center as the optimal transportation route map. The neural network model is trained with the objective of minimizing the comprehensive transportation cost under the loss of preservation status.

9. A computer storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed by the processor, they implement the durian logistics optimization method considering preservation constraints as described in any one of claims 1 to 7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the durian logistics optimization method considering preservation constraints as described in any one of claims 1 to 7.

Images