Multi-party collaborative optimization method and system for urban logistics distribution

Abstract

This invention relates to the field of logistics optimization management technology, and proposes a multi-party collaborative optimization method and system for urban logistics distribution. The method includes: acquiring vehicle information for various types of vehicles in an urban logistics distribution fleet and order information for several logistics orders; extracting several distribution centers, outlets, and delivery destinations for urban logistics distribution and constructing a distribution map structure; obtaining the shortest delivery path for each logistics order; obtaining several order categories and the total weight of goods in each order category; obtaining optimal transportation factors; quantifying time-based transportation costs; constructing a transportation objective function based on the optimal transportation factors; and obtaining the optimal vehicle type for each segment of the shortest delivery path for each order category, thereby allocating logistics vehicles to each logistics order and achieving multi-party collaborative optimization of urban logistics distribution. This invention aims to solve the problem of inaccurate task allocation due to the complexity of vehicle types in logistics distribution fleets.

Description

Technical Field

[0001] This invention relates to the field of logistics optimization management technology, specifically to a multi-party collaborative optimization method and system for urban logistics distribution. Background Technology

[0002] Multi-party collaborative optimization for urban logistics distribution requires integrating factors such as vehicles, road networks, orders, transportation capacity, and time costs to achieve collaborative optimization of the urban logistics distribution network and improve logistics distribution efficiency. Due to the diversity of logistics transportation modes, the modes of transportation used to reach different cities vary, resulting in multiple distribution centers related to different transportation modes within cities. Furthermore, order types and delivery times necessitate collaborative optimization of distribution network nodes to address time costs.

[0003] Traditional scheduling algorithms in urban logistics delivery processes cannot accurately allocate tasks to various types of vehicles in a logistics delivery fleet. They do not consider vehicle type and function, which can easily lead to mismatches such as "light vehicles carrying heavy goods" or refrigerated trucks performing ordinary tasks, reducing resource utilization. At the same time, the right-of-way restrictions of different vehicles and the impact of the number of orders placed at different times on vehicle capacity and assembly time of delivery network nodes are not taken into account in traditional scheduling algorithms. As a result, the delivery network only allocates tasks based on vehicle availability, which cannot give full play to the corresponding functions of each type of vehicle. Furthermore, the assembly time is extended due to the number of orders, resulting in cargo backlog, which seriously affects the efficiency of urban logistics collaborative delivery. Summary of the Invention

[0004] This invention provides a multi-party collaborative optimization method and system for urban logistics distribution to solve the problem of the inability to accurately allocate tasks due to the complexity of existing logistics distribution fleets with diverse vehicle types. The specific technical solution adopted is as follows:

[0005] This invention proposes a multi-party collaborative optimization method for urban logistics distribution, which includes the following steps:

[0006] The system acquires vehicle information for various types of vehicles in the urban logistics delivery fleet, including vehicle functions, vehicle load capacity, vehicle speed, and road restriction information; it also acquires the origin and destination of several logistics orders, order type, and cargo weight.

[0007] Based on the origin and destination of logistics orders and the delivery process, several distribution centers, outlets and delivery destinations of urban logistics distribution are extracted and a delivery map structure is constructed. For each logistics order, the shortest delivery path is obtained in the delivery map structure according to the order type. Based on the shortest delivery path, logistics orders are divided into several order categories and the total weight of goods in each order category. Combining the vehicle load, vehicle function and road restriction information of each type of vehicle, the optimal transportation factor for each road segment in the shortest delivery path of each order category for each type of vehicle is obtained.

[0008] This study analyzes the impact of the number of logistics orders in each order category on the assembly time of each node in the shortest delivery path. Combining the vehicle speed of each type of vehicle and the distance of each road segment in the shortest delivery path, the study quantifies the time transportation cost of each road segment in the shortest delivery path for each order category under each type of vehicle. Combining the optimization of transportation factors, the study constructs the transportation objective function of each road segment in the shortest delivery path for each type of vehicle in each order category.

[0009] Based on the transportation objective function, the optimal transportation vehicle type is obtained for each segment of the shortest delivery path for each order category, and then logistics vehicles are allocated to each logistics order.

[0010] Optionally, the specific method for extracting several distribution centers, outlets, and delivery destinations of urban logistics distribution and constructing a distribution map structure includes:

[0011] Extract several distribution centers and outlets in the city, as well as the roads from each distribution center to each outlet. Treat each distribution center and outlet as a node, and the roads as edges between the distribution centers and the corresponding nodes of the outlets. The distances of the roads are used as edge values.

[0012] Based on the transportation destinations in a large number of logistics orders, several transportation destinations belonging to the same area are collectively used as the delivery destinations of the corresponding area. Each delivery destination is used as a node, and the driving routes and corresponding distances from each network point to each delivery destination are obtained. The edge values ​​between the network point corresponding node and the delivery destination corresponding node are obtained. Based on each node and the edges and edge values ​​between nodes, a graph structure is constructed and used as the delivery graph structure.

[0013] Optionally, the method for obtaining the shortest delivery path is as follows:

[0014] For any logistics order, obtain the node corresponding to the transportation destination in the delivery graph structure. Based on the node of the transportation destination and the order type of the logistics order, determine the network points through which the logistics order passes during transportation. Obtain the network points and corresponding nodes of the delivery destination in the transportation process of the logistics order. Based on the distribution center, network points and corresponding nodes of the delivery destination, obtain the shortest delivery path for the logistics order in the delivery graph structure.

[0015] Optionally, the specific method for obtaining several order categories and the total weight of goods in each order category includes:

[0016] Group several logistics orders with the same shortest delivery path into one order category;

[0017] The total weight of goods in any order category is the sum of the weights of all logistics orders within that order category.

[0018] Optionally, the optimal transportation factor for each road segment in the shortest delivery path for each order category and for each type of vehicle is obtained using the following method:

[0019] For any segment of the shortest delivery path for any order category and any type of vehicle, obtain the ratio of the total weight of goods for that order category to the vehicle load capacity for that type of vehicle. If the ratio is less than or equal to 1, use the ratio as the transportation matching factor between that order category and that type of vehicle. If the ratio is greater than 1, convert the ratio into a mixed number, and use the proper fractional part of the mixed number as the numerator and the sum of the integer part plus 1 as the denominator to obtain the ratio, which is used as the transportation matching factor between that order category and that type of vehicle.

[0020] Obtain the road restriction information corresponding to this road segment and this type of vehicle; for vehicle functions, obtain the word vector of the order type corresponding to this order category and the word vector of the vehicle function of this type of vehicle, and use the cosine similarity of the two word vectors as the type matching factor between this order category and this type of vehicle.

[0021] The product of the transportation matching factor, the type matching factor, and the road restriction information corresponding to the road segment and the vehicle type is used as the preferred transportation factor for the vehicle type in the shortest delivery path for the order category.

[0022] Optionally, the time transportation cost of each road segment in the shortest delivery path for each order category under each type of vehicle is obtained as follows:

[0023] Based on the number of logistics orders in each order category, and the number of logistics orders in other order categories under each road segment of the shortest delivery path, combined with the total weight of goods in each order category, the loading and unloading time cost of each road segment in the shortest delivery path of each order category is obtained.

[0024] For any segment of the shortest delivery path for any order category, obtain the speeds of several vehicles of any type on that segment of the road, as well as the departure times corresponding to each vehicle speed. Obtain the average time when all logistics orders in that order category arrive at the departure node of that segment of the road, and use this as the ideal departure time for that segment of the shortest delivery path for that order category.

[0025] The absolute value of the difference between the departure time corresponding to any vehicle speed and the ideal departure time is taken as the time deviation of the vehicle speed. The difference obtained by subtracting the time deviation from 24 hours is taken as the time proximity of the vehicle speed. The time proximity of all vehicle speeds on this road segment is weighted and normalized, and the result is taken as the time weight of each vehicle speed. Based on the time weight, the speeds of all vehicles on this road segment are weighted and summed, and the result is taken as the driving speed of this type of vehicle on this road segment in the shortest delivery path for this order category. The ratio of the distance of this road segment to the driving speed is taken as the transportation time cost of this type of vehicle on this road segment in the shortest delivery path for this order category.

[0026] The product of the loading and unloading time cost of that segment of the shortest delivery path for that order category and the transportation time cost of that type of vehicle on that segment of the path is taken as the time transportation cost of that type of vehicle on that segment of the shortest delivery path for that order category.

[0027] Optionally, the specific method for obtaining the loading and unloading time cost of each segment of the shortest delivery path for each order category includes:

[0028] For any segment of the shortest delivery path for any order category, obtain several other order categories that are the same as that segment of the shortest delivery path for that order category, and use them as similar categories to that order category;

[0029] The total weight of goods in the order category and all its similar categories is linearly normalized, and the result is used as the weight reference weight of the order category and its similar categories. Based on the weight reference weight, the number of logistics orders in the order category and its similar categories is weighted and averaged, and the result is used as the loading and unloading time cost of that segment of the road in the shortest delivery path of the order category.

[0030] Optionally, the specific method for constructing the transportation objective function for each road segment in the shortest delivery path for each order category and for each type of vehicle includes:

[0031] The expression formed by dividing the preferred transportation factor of any segment of the shortest delivery path for any order category by the time transportation cost of that segment of the road for that type of vehicle is used as the transportation objective function for that segment of the shortest delivery path for that order category.

[0032] Optionally, the method for obtaining the optimal transport vehicle type for each segment of the shortest delivery path for each order category based on the transport objective function includes:

[0033] For any segment of the shortest delivery path for any order category, the vehicle type corresponding to the maximum value of the transportation objective function output for that segment of the road is taken as the optimal transportation vehicle type for all logistics orders under that order category in the shortest delivery path.

[0034] The present invention also proposes a multi-party collaborative optimization system for urban logistics distribution, the system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above method.

[0035] The beneficial effects of this invention are as follows: This invention analyzes the distribution of distribution centers, outlets, and delivery destinations in the urban logistics distribution process, classifies logistics orders into order categories, and obtains the shortest delivery path. Combining the functions, load capacity, and road restrictions of various types of vehicles in the delivery fleet, it performs optimal matching analysis of roads and vehicles along the shortest delivery path, while also considering assembly time and vehicle speed. Furthermore, based on the optimal analysis of roads and vehicles, it considers time transportation costs, thereby achieving optimal matching of roads and vehicles. Specifically, by using the transportation origin and destination of logistics orders, combined with the distribution of distribution centers and outlets in the logistics distribution network, a delivery map structure is constructed, and the shortest delivery path is generated for each logistics order. Simultaneously, logistics orders with the same shortest delivery path are categorized to obtain order categories. The system obtains the total weight of goods for each category and conducts optimization analysis on the road segments and vehicle types within the shortest delivery path to optimize each order category from the perspective of vehicle capacity and functionality. Within each road segment of the shortest delivery path, the number of logistics orders within each order category directly affects the assembly time at each node. Therefore, it is necessary to consider the relationship between road distance and vehicle speed to reduce the impact of longer assembly times on the normal delivery process. Combining optimized transportation factors, the system comprehensively analyzes the optimal matching of road segments and vehicle types within the shortest delivery path from both capacity and time cost perspectives. This allows for the allocation of transportation vehicles for logistics orders, achieving collaborative optimization of distribution centers, network points, capacity, assembly time, and transportation vehicles in urban logistics delivery. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a schematic diagram of a multi-party collaborative optimization method for urban logistics distribution provided in an embodiment of the present invention. Detailed Implementation

[0038] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0039] Please see Figure 1 The diagram illustrates a flowchart of a multi-party collaborative optimization method for urban logistics distribution provided by an embodiment of the present invention. The method includes the following steps:

[0040] Step S001: Obtain vehicle information for various types of vehicles in the city's logistics delivery fleet; obtain the origin and destination of several logistics orders, order type, and cargo weight.

[0041] The purpose of this embodiment is to conduct a comprehensive analysis of the transportation capacity of urban logistics distribution networks, including distribution centers, outlets, and delivery stations, combined with various types of vehicles in the delivery fleet, such as large trucks, refrigerated trucks, medium-sized vans, new energy vans, unmanned delivery vehicles, and tricycles used by couriers for daily delivery. This analysis, combined with cost analysis of assembly and transportation time, aims to optimize the distribution network, transportation capacity, and time costs in urban logistics distribution, thereby achieving efficient use of various types of vehicles for logistics distribution. Therefore, it is first necessary to obtain vehicle information and relevant information about logistics orders.

[0042] Specifically, for a fleet of vehicles in an urban logistics delivery vehicle group, which corresponds to several vehicle types, vehicle information is obtained for each type of vehicle, including vehicle functions, load capacity, speed, and road restrictions. For each vehicle type, the vehicle functions are first obtained, which usually correspond to the order type. For example, air express orders typically require fast transportation in new energy vans, cold chain fresh food orders require refrigerated trucks, and large heavy goods orders require large trucks. Medium-sized vans are used to transport large quantities of ordinary express orders, while unmanned delivery vehicles and couriers' tricycles are mainly affected by road restrictions during the delivery process, and their corresponding vehicle functions are transportation between network points and delivery stations. Vehicle load capacity is obtained directly from the rated load capacity of each type of vehicle; road restriction information is the traffic restriction status of each type of vehicle between delivery network nodes, such as yellow-label trucks being prohibited from driving on certain road sections. The collected data is expressed as whether each type of vehicle is prohibited from driving on the roads between the distribution center and the outlet, and between the outlet and the delivery station. Prohibition is represented by 0, and not prohibited is represented by 1; vehicle speed is the average speed of each type of transport vehicle in each historical transport process on the corresponding road, that is, the average speed of the vehicle under any historical transport, and the departure time of each historical transport is also recorded (no date restriction, the corresponding time in a 24-hour day).

[0043] Furthermore, it acquires several logistics orders that arrive in the current city on the same day, and uses the distribution center where the logistics orders arrive as the starting point of the logistics order and the delivery address as the ending point of the logistics order, and records the weight of the goods in each logistics order; at the same time, it records the order type of each logistics order, such as air express, air cold chain fresh food, large heavy goods express, ordinary express, etc.

[0044] It should be noted that multi-party collaborative optimization of urban logistics distribution needs to consider the distribution network, including distribution centers, outlets, delivery stations (delivery destinations, such as door-to-door delivery, express stations, and express lockers), the transportation capacity of logistics vehicles, and the loading, unloading, and delivery time losses at each node. Since urban logistics distribution fleets are diverse in vehicle types, distribution centers involve multiple transportation modes such as air, port, land, and rail, and logistics orders correspond to different order types, it is necessary to optimize the logistics network based on order type and delivery process to achieve effective utilization of logistics vehicles, while also considering the time losses at each node, in order to achieve overall multi-party optimization of urban logistics distribution.

[0045] Step S002: Based on the transportation origin and destination of logistics orders and the delivery process, extract several distribution centers, outlets and delivery destinations of urban logistics distribution and construct a delivery map structure. For each logistics order, combine the order type to obtain the shortest delivery path in the delivery map structure. Based on the shortest delivery path, divide the logistics orders to obtain several order categories and the total weight of goods in each order category. Combine the vehicle load, vehicle function and road restriction information of each type of vehicle to obtain the optimal transportation factor for each type of vehicle in each segment of the shortest delivery path of each order category.

[0046] It should be noted that after goods arrive in the corresponding city via air, land, sea, and rail transport, they first arrive at various distribution centers in the city, such as land and air transport distribution centers. Then, the distribution centers determine the delivery points in the city based on the delivery address of the logistics order. After arriving at the distribution points, the goods are delivered by couriers or unmanned delivery fleets. The delivery destinations include express stations, express lockers, and specific addresses for door-to-door delivery. The distribution graph structure is constructed by using distribution centers, distribution points, and delivery destinations as nodes. At the same time, the road information of logistics vehicles during the delivery process between nodes is basically fixed, and the actual distance traveled is used as the boundary value. Furthermore, for each logistics order, the distribution point and delivery destination are determined based on its delivery address and logistics type, and the shortest path is extracted from the distribution graph structure.

[0047] Preferably, in one embodiment of the present invention, based on the origin and destination of the logistics order and the delivery process, several distribution centers, outlets and delivery destinations of urban logistics delivery are extracted and a delivery map structure is constructed. The shortest delivery path is obtained in the delivery map structure for each logistics order based on the order type. The specific method includes:

[0048] Extract several distribution centers and outlets in the city, as well as the routes from each distribution center to each outlet (distribution centers and outlets can be directly obtained based on the logistics distribution network; routes are mostly fixed, and if not fixed, the shortest distance route is used, which will not be elaborated in this embodiment). Treat each distribution center and outlet as a node, and the route as the edge between the corresponding node and the outlet, with the distance of the route as the edge value. Simultaneously, based on the delivery address, the delivery destination in a large number of logistics orders, several delivery destinations belonging to the same area (usually a residential area or community) for door-to-door delivery, parcel locker storage, and parcel station storage are collectively used as the delivery destination for the corresponding area, i.e., one area corresponds to one delivery destination. Treat each delivery destination as a node as well, and obtain the route and distance from each outlet to each delivery destination. Obtain the edge value between the node corresponding to the outlet and the node corresponding to the delivery destination. Based on each node and the edges and edge values ​​between nodes, construct a graph structure as the delivery graph structure.

[0049] Furthermore, for any logistics order, the node corresponding to the transportation destination in the delivery graph structure is obtained for its transportation endpoint. Based on the node of the transportation destination and the order type of the logistics order, the network points through which the logistics order passes in the transportation process are determined (such as air express and ordinary land transportation, even if the delivery destination is the same, the transportation network points are different). Thus, the network points and the corresponding nodes of the delivery destination in the transportation process of the logistics order are obtained. At the same time, based on the known order type, the distribution center, the network points, and the nodes corresponding to the delivery destination are used to obtain the shortest delivery path for the logistics order in the delivery graph structure.

[0050] It should be further explained that, since the shortest delivery path corresponds to a fixed route of a distribution center-outlet-delivery destination, logistics orders are classified based on the same shortest delivery path, so that the vehicle selection for each segment of the shortest delivery path can be optimized and evaluated in the future based on the same type of logistics orders.

[0051] Preferably, in one embodiment of the present invention, logistics orders are divided based on the shortest delivery path to obtain several order categories and the total weight of goods in each order category. The specific method includes:

[0052] Several logistics orders with the same shortest delivery path and the same order type are grouped into one order category; the sum of the weights of all logistics orders in any order category is taken as the total weight of the goods in that order category.

[0053] Preferably, in one embodiment of the present invention, the optimal transportation factor for each type of vehicle in the shortest delivery path for each order category is obtained by combining the vehicle load, vehicle function, and road restriction information of each type of vehicle. The specific method includes:

[0054] It should be noted that logistics orders within the same order category share the same order type, which determines the matching relationship between different types of vehicles and their functions. For example, orders for air-freighted cold chain fresh produce require refrigerated vehicles for transportation. This analysis examines the matching relationship between order types and vehicle functions corresponding to the order category. Simultaneously, the analysis considers the road conditions and vehicle restrictions along the shortest delivery path to determine which vehicles can operate on the corresponding roads. By combining the difference between the vehicle's load capacity and the total weight of the goods in the order category, and taking transportation costs into account, an optimal matching analysis is performed on each segment of the shortest delivery path corresponding to the order category and each type of vehicle.

[0055] Specifically, for any segment of the shortest delivery path for any order category and any type of vehicle, the ratio of the total weight of goods for that order category to the vehicle load capacity for that type of vehicle is obtained. If the ratio is less than or equal to 1, the ratio is used as the transportation matching factor between that order category and that type of vehicle. If the ratio is greater than 1, the ratio is converted into a mixed number, and the proper fractional part of the mixed number is used as the numerator, and the sum of the integer part plus 1 is used as the denominator to obtain the ratio, which is used as the transportation matching factor between that order category and that type of vehicle.

[0056] Furthermore, the road restriction information corresponding to the road segment and the type of vehicle is obtained, i.e., whether the corresponding type of vehicle is restricted from driving (0 for prohibited driving and 1 for permitted driving), which has already been obtained in step S001; for vehicle functions, the word vector of the order type corresponding to the order category and the word vector of the vehicle function of the type of vehicle are obtained, and the cosine similarity of the two word vectors is used as the type matching factor between the order category and the type of vehicle. The word vectors are obtained using the Word2vec model, and the number of vector dimensions in the word vectors is the same to ensure the calculation of cosine similarity; the product of the transportation matching factor, the type matching factor and the road restriction information corresponding to the road segment and the type of vehicle is used as the preferred transportation factor for the type of vehicle in the shortest delivery path of the order category.

[0057] It should be noted that the closer the ratio of the total weight of goods to the vehicle's load capacity is to 1, the closer its transport capacity is to meeting the transport needs of all goods in the corresponding order category, and the larger the transport matching factor. If the ratio exceeds 1, it indicates that one vehicle cannot transport all the goods for the corresponding type of vehicle. In this case, the proper fractional part of the mixed-number method needs to be used to reconsider whether the transport is matched. At the same time, adding 1 to the integer part indicates how many vehicles are needed to adjust the transport matching factor. The more vehicles needed and the smaller the proper fractional part, the less matched the transport is, and larger vehicles are needed for the corresponding transport. Combining road restriction information and the matching relationship between vehicle function and order type, since there are certain synonyms in the textual descriptions of vehicle function and order type to indicate relevance, word vector similarity is used for matching judgment to obtain the optimal transport factor.

[0058] Thus, by combining the origin and destination of logistics orders with the distribution centers and outlets of the logistics distribution network, a delivery map structure is constructed, and the shortest delivery path is generated for each logistics order. At the same time, logistics orders with the same shortest delivery path are classified to obtain order categories. The total weight of goods is obtained for each order category, and the optimal analysis of each road segment and each type of vehicle transportation in the shortest delivery path is performed to optimize each order category from the perspective of vehicle capacity and function.

[0059] Step S003: Analyze the impact of the number of logistics orders in each order category on the assembly time of each node in the shortest delivery path. Combine the vehicle speed of each type of vehicle and the distance of each road segment in the shortest delivery path to quantify the time transportation cost of each road segment in the shortest delivery path for each order category under each type of vehicle. Combine the optimized transportation factors to construct the transportation objective function of each road segment in the shortest delivery path for each order category for each type of vehicle.

[0060] It should be noted that after the capacity optimization and matching analysis, the corresponding time transportation cost needs to be analyzed. The larger the number of logistics orders in the same order category, the longer the loading and unloading time of the transport vehicle from leaving the node to arriving at the node will be, and the greater the impact on the overall logistics and delivery process time. Based on this, it is necessary to consider the time spent by the transport vehicle on the road according to the vehicle speed and road network conditions, so as to quantify the time transportation cost of each road segment under the corresponding vehicle.

[0061] Preferably, in one embodiment of the present invention, the impact of the number of logistics orders in each order category on the assembly time of each node in the shortest delivery path is analyzed. Combining the vehicle speed of each type of vehicle and the distance of each segment of the shortest delivery path, the time transportation cost of each segment of the shortest delivery path for each order category under each type of vehicle is quantified. The specific method includes:

[0062] For any segment of the shortest delivery path for any order category, obtain several other order categories that are identical to that segment of the shortest delivery path for that order category, and use them as similar categories for that order category; linearly normalize the total weight of goods for that order category and all its similar categories, and use the result as the weight reference weight for that order category and each of its similar categories; based on the weight reference weight, take a weighted average of the number of logistics orders in that order category and its similar categories, and use the result as the loading and unloading time cost for that segment of the shortest delivery path for that order category.

[0063] It should be noted that there are several routes that are the same from the distribution center to the outlet, but different order categories are due to different delivery destinations. Similarly, there are several routes that are the same from the outlet to the delivery destination, but different order categories are due to different distribution centers. Since the routes are the same, the goods will be loaded into the same or similar transport vehicles during loading and unloading. Therefore, the loading and unloading time cost needs to take all categories into account. At the same time, the larger the weight of the goods, the longer the time spent in the loading, unloading and sorting process. Therefore, the time spent in the loading and unloading process, which is directly determined by the number of logistics orders, is weighted and averaged using the total weight of the corresponding goods as a weight, in order to quantify the loading and unloading time cost.

[0064] It should be further explained that the actual vehicle speed in the vehicle information is matched with the road, which is the vehicle speed of the corresponding transport vehicle in the past multiple transports on the corresponding road (the average speed of traveling on that road segment). The transport speed of each type of vehicle on the current transport road is determined by referring to the historical vehicle speed. The transport time cost is quantified by combining the distance, and then the time transport cost is obtained by combining the loading and unloading time cost.

[0065] Furthermore, for this road segment, the speeds of several vehicles of any type corresponding to this road segment (historical vehicle speeds) and the departure times corresponding to each vehicle speed are obtained. The average time of arrival at the departure node of all logistics orders in this order category (average time of arrival at the distribution center or online store) is obtained as the ideal departure time for this road segment in the shortest delivery path for this order category. The absolute value of the difference between the departure time corresponding to any vehicle speed and the ideal departure time is taken as the time deviation of the vehicle speed. The difference obtained by subtracting the time deviation from 24 hours is taken as the time proximity of the vehicle speed. The time proximity of all vehicle speeds in this road segment is weighted and normalized, and the result is taken as the time weight of each vehicle speed. Based on the time weight, all vehicle speeds in this road segment are weighted and summed, and the result is taken as the driving speed of this type of vehicle in this road segment in the shortest delivery path for this order category. The ratio of the distance (actual length) of this road segment to the driving speed is taken as the transportation time cost of this type of vehicle in this road segment in the shortest delivery path for this order category.

[0066] Furthermore, the product of the loading and unloading time cost of that segment of the shortest delivery path for that order category and the transportation time cost of that type of vehicle on that segment of the road is taken as the time transportation cost of that type of vehicle on that segment of the shortest delivery path for that order category.

[0067] It should be further explained that the time-based transportation cost comprehensively considers the vehicle's transportation time and the loading and unloading time of goods at the network points during the transportation process. It also combines the optimal transportation factors to consider the optimal matching relationship between transportation capacity and vehicle functions and order categories. The transportation objective function is comprehensively constructed to achieve the optimal matching analysis between each road segment and each type of vehicle in the shortest delivery path of the order category.

[0068] Preferably, in one embodiment of the present invention, the objective function for the transportation of each road segment in the shortest delivery path for each order category for each type of vehicle is constructed by combining preferred transportation factors, including the following specific method:

[0069] The expression formed by dividing the preferred transportation factor of any segment of the shortest delivery path for any order category by the time transportation cost of that segment of the road for that type of vehicle is used as the transportation objective function for that segment of the shortest delivery path for that order category. The ratio of the preferred transportation factor to the time transportation cost is the output value of the transportation objective function for the corresponding type of vehicle.

[0070] Therefore, in each segment of the shortest delivery path, the number of logistics orders in each order category directly affects the assembly time at each node. Thus, it is necessary to consider the relationship between road distance and vehicle speed to reduce the impact of longer assembly times on the delivery process under normal time flow. Combining optimal transportation factors, a comprehensive analysis of the optimal matching between each segment of the shortest delivery path and the type of transport vehicle is conducted from both the perspectives of transport capacity and time cost to provide a foundation for subsequent optimization.

[0071] Step S004: Based on the transportation objective function, obtain the optimal transportation vehicle type for each segment of the shortest delivery path for each order category, and allocate logistics vehicles to each logistics order to achieve multi-party collaborative optimization of urban logistics delivery.

[0072] Specifically, for any segment of the shortest delivery path for any order category, the vehicle type corresponding to the maximum output value of the transportation objective function for that segment of the road is taken as the optimal transportation vehicle type for all logistics orders under that order category on that day. Thus, the corresponding transportation vehicles are obtained for each segment of the shortest delivery path for each logistics order under that order category. This achieves multi-party collaborative optimization of urban logistics delivery, realizing comprehensive and multi-faceted collaborative optimization from distribution centers, outlets and delivery destinations, as well as transportation capacity, assembly time and transportation vehicles.

[0073] Thus, by analyzing the distribution of distribution centers, outlets, and delivery destinations in the urban logistics distribution process, order categories are classified and the shortest delivery path is obtained. Combining the characteristics of various types of vehicles in the delivery fleet, such as their functions, load capacity, and road restrictions, the optimal matching analysis of each segment of the road and vehicle in the shortest delivery path is conducted. At the same time, assembly time and vehicle speed are considered. Furthermore, based on the optimal analysis of roads and vehicles, time transportation costs are considered to achieve the optimal matching of roads and vehicles, thereby allocating transportation vehicles for logistics orders. This achieves the collaborative optimization of multiple parties, including distribution centers, outlets, transportation capacity, assembly time, and transportation vehicles, in the urban logistics distribution process.

[0074] Another embodiment of the present invention provides a multi-party collaborative optimization system for urban logistics distribution. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the above-described method steps S001 to S004.

[0075] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-party collaborative optimization method for urban logistics distribution, characterized in that, The method includes the following steps: Obtain vehicle information for various types of vehicles in the urban logistics delivery fleet, including vehicle functions, vehicle load capacity, vehicle speed, and road restriction information; obtain the origin and destination of several logistics orders, order type, and cargo weight. Based on the origin and destination of logistics orders and the delivery process, several distribution centers, outlets and delivery destinations of urban logistics distribution are extracted and a delivery map structure is constructed. For each logistics order, the shortest delivery path is obtained in the delivery map structure in combination with the order type. Based on the shortest delivery path, the logistics orders are divided to obtain several order categories and the total weight of goods in each order category. For any segment of the shortest delivery path for any order category and any type of vehicle, obtain the ratio of the total weight of goods for that order category to the vehicle load capacity for that type of vehicle. If the ratio is less than or equal to 1, use the ratio as the transportation matching factor between that order category and that type of vehicle. If the ratio is greater than 1, convert the ratio into a mixed number, and use the proper fractional part of the mixed number as the numerator and the sum of the integer part plus 1 as the denominator to obtain the ratio, which is used as the transportation matching factor between that order category and that type of vehicle. Obtain the road restriction information corresponding to this road segment and this type of vehicle; for vehicle functions, obtain the word vector of the order type corresponding to this order category and the word vector of the vehicle function of this type of vehicle, and use the cosine similarity of the two word vectors as the type matching factor between this order category and this type of vehicle. The product of the transportation matching factor, the type matching factor, and the road restriction information corresponding to the road segment and the vehicle type is used as the preferred transportation factor for the vehicle type in the shortest delivery path for the order category. This study analyzes the impact of the number of logistics orders in each order category on the assembly time of each node in the shortest delivery path. Combining the vehicle speed of each type of vehicle and the distance of each road segment in the shortest delivery path, the study quantifies the time transportation cost of each road segment in the shortest delivery path for each order category under each type of vehicle. Combining the optimization of transportation factors, the study constructs the transportation objective function of each road segment in the shortest delivery path for each type of vehicle in each order category. Based on the transportation objective function, the optimal transportation vehicle type is obtained for each segment of the shortest delivery path for each order category, and then logistics vehicles are allocated to each logistics order. This paper extracts several distribution centers, outlets, and delivery destinations for urban logistics distribution and constructs a delivery graph structure. The specific methods include: extracting several distribution centers and outlets in the city, as well as the roads from each distribution center to each outlet. Each distribution center and outlet is treated as a node, and the roads are treated as edges between the corresponding nodes of the distribution centers and outlets, with the distances between the roads being the edge values. Based on the transportation destinations in a large number of logistics orders, several transportation destinations belonging to the same area are collectively treated as delivery destinations for the corresponding area. Each delivery destination is treated as a node. The roads from each outlet to each delivery destination and the corresponding distances are obtained, and the edge values ​​between the corresponding nodes of the outlets and the corresponding nodes of the delivery destinations are obtained. Based on each node and the edges and edge values ​​between nodes, a graph structure is constructed and used as the delivery graph structure. The time transportation cost of each road segment in the shortest delivery path for each order category, under each type of vehicle, is obtained as follows: Based on the number of logistics orders in each order category and the number of logistics orders in other order categories under each road segment in its shortest delivery path, combined with the total weight of goods in each order category, the loading and unloading time cost of each road segment in the shortest delivery path for each order category is obtained; for any road segment in the shortest delivery path of any order category, the speeds of several vehicles corresponding to each type of vehicle on that road segment are obtained, as well as the departure time corresponding to each vehicle speed. The average time when all logistics orders in that order category arrive at the departure node of that road segment is obtained as the ideal departure time for that road segment in the shortest delivery path of that order category; the absolute value of the difference between the departure time corresponding to any vehicle speed and the ideal departure time is taken as the ideal departure time for that vehicle. The time deviation of vehicle speed is calculated by subtracting the time deviation from 24 hours, and the difference is taken as the time proximity of the vehicle speed. The time proximity of all vehicle speeds on this road segment is weighted and normalized, and the result is taken as the time weight of each vehicle speed. The speeds of all vehicles on this road segment are weighted and summed based on the time weights, and the result is taken as the travel speed of this type of vehicle on this road segment in the shortest delivery path for this order category. The ratio of the distance of this road segment to the travel speed is taken as the transportation time cost of this type of vehicle on this road segment in the shortest delivery path for this order category. The product of the loading and unloading time cost of this road segment in the shortest delivery path for this order category and the transportation time cost of this type of vehicle on this road segment is taken as the time transportation cost of this type of vehicle on this road segment in the shortest delivery path for this order category.

2. The multi-party collaborative optimization method for urban logistics distribution according to claim 1, characterized in that, The shortest delivery path can be obtained as follows: For any logistics order, obtain the node corresponding to the transportation destination in the delivery graph structure. Based on the node of the transportation destination and the order type of the logistics order, determine the network points through which the logistics order passes during transportation. Obtain the network points and corresponding nodes of the delivery destination in the transportation process of the logistics order. Based on the distribution center, network points and corresponding nodes of the delivery destination, obtain the shortest delivery path for the logistics order in the delivery graph structure.

3. The multi-party collaborative optimization method for urban logistics distribution according to claim 1, characterized in that, The specific methods for obtaining several order categories and the total weight of goods in each order category are as follows: Group several logistics orders with the same shortest delivery path into one order category; The total weight of goods in any order category is the sum of the weights of all logistics orders within that order category.

4. The multi-party collaborative optimization method for urban logistics distribution according to claim 1, characterized in that, The specific methods for obtaining the loading and unloading time costs of each segment of the shortest delivery path for each order category are as follows: For any segment of the shortest delivery path for any order category, obtain several other order categories that are the same as that segment of the shortest delivery path for that order category, and use them as similar categories to that order category; The total weight of goods in the order category and all its similar categories is linearly normalized, and the result is used as the weight reference weight for the order category and its similar categories. Based on the weight reference weight, the number of logistics orders in the order category and its similar categories are weighted and averaged. The result is used as the loading and unloading time cost of that segment of the road in the shortest delivery path for that order category.

5. The multi-party collaborative optimization method for urban logistics distribution according to claim 1, characterized in that, The specific methods for constructing the transportation objective function for each road segment in the shortest delivery path for each order category and for each type of vehicle are as follows: The expression formed by dividing the preferred transportation factor of any segment of the shortest delivery path for any order category by the time transportation cost of that segment of the road for that type of vehicle is used as the transportation objective function for that segment of the shortest delivery path for that order category.

6. The multi-party collaborative optimization method for urban logistics distribution according to claim 1, characterized in that, The optimal vehicle type is obtained for each segment of the shortest delivery path for each order category based on the transportation objective function. The specific methods include: For any segment of the shortest delivery path for any order category, the vehicle type corresponding to the maximum value of the transportation objective function output for that segment of the road is taken as the optimal transportation vehicle type for all logistics orders under that order category in the shortest delivery path.

7. A multi-party collaborative optimization system for urban logistics distribution, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the multi-party collaborative optimization method for urban logistics distribution as described in any one of claims 1-6.

Images