An artificial intelligence-based logistics scheduling management system and method

Abstract

The present invention discloses a logistics scheduling management system and method based on artificial intelligence, belonging to the technical field of artificial intelligence. The present invention collects multi-source data from various online links of international trade and preprocesses the collected multi-source data; generates a merchant order dataset, a type order dataset, a merchant order relationship chain, and a merchant similar order relationship chain; analyzes the merchant order dataset using the support vector machine algorithm, calculates the logistics priority of each order, and generates a logistics priority order dataset; optimizes the logistics transportation path in combination with the logistics priority of each order, and gives priority to delivering high-priority orders; based on the merchant similar order relationship chain, combines with the type order dataset, and performs clustering analysis on similar goods to form a similar goods set; uses the merchant order relationship chain to form a merchant relationship set; based on the similar goods set and the merchant relationship set, generates a logistics bundling plan to schedule logistics vehicles or other transportation tools.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an artificial intelligence-based logistics scheduling management system and method. Background Art

[0002] With the rapid development of globalization and e-commerce, logistics scheduling management in the field of international trade has become increasingly complex, especially in terms of processing multi-source data and dynamic logistics scheduling. International trade usually involves multiple links, including order generation, logistics tracking, and management of merchant partnerships, which generate massive amounts of heterogeneous data. How to effectively collect, clean, analyze and optimize this data to improve logistics efficiency is an important challenge for current technology.

[0003] In terms of logistics priority and route optimization, most existing systems rely on static rules or manual judgment, and fail to dynamically adjust transportation plans based on real-time logistics data, which can easily lead to transportation delays or waste of resources. Existing technologies usually rely on manual or simple fixed algorithms for batch transportation, and cannot perform intelligent bundling based on cargo type, order relationship, and cooperative relationship between merchants, which can easily lead to waste of transportation resources and increase in logistics costs. Existing logistics systems rarely consider long-term cooperative relationships or transaction frequencies between merchants, resulting in the inability to optimize logistics collaboration between merchants and missed opportunities to improve transportation efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a logistics scheduling management system and method based on artificial intelligence to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A logistics scheduling management method based on artificial intelligence, the method comprising the following steps:

[0007] S100, collecting multi-source data from various online links of international trade, wherein the multi-source data includes: transaction orders, transaction logistics, and transaction merchant relationship chains; cleaning and formatting the collected multi-source data, removing redundant information and abnormal data, and standardizing the multi-source data;

[0008] S200, dividing the transaction order into a merchant order data set and a type order data set according to the transaction merchant and the type of goods; generating a merchant order relationship chain based on the merchant order data set and the transaction merchant relationship chain; generating a merchant same-type order relationship chain based on the type order data set and the merchant order relationship chain;

[0009] S300. Analyze the order urgency, goods transportation distance, and destination in the merchant order dataset using the support vector machine algorithm, calculate the logistics priority of each order, and generate a logistics priority order dataset; combine the logistics priority of each order, optimize the logistics transportation route through the Q-learning algorithm of reinforcement learning, dynamically adjust the transportation plan according to the destination of the goods, real-time traffic conditions, and logistics costs, and give priority to delivering high-priority orders.

[0010] S400. Based on the merchant similar order relationship chain, combine the type order dataset, perform clustering analysis on similar goods, identify similar goods that can be transported together, and form a similar goods set; utilize the merchant order relationship chain to identify merchants with order relationships, analyze the transaction frequency and geographical location relationship between merchants through the Dijkstra algorithm, and form a merchant relationship set; based on the similar goods set and the merchant relationship set, generate a logistics bundling plan, bundle merchants in the same region or logistics node with similar goods for transportation; according to the generated bundling transportation plan, dispatch logistics vehicles or other transportation tools to arrange for the batch transportation of goods.

[0011] According to step S100, automatically collect transaction orders from e-commerce platforms, B2B trading websites, and enterprise management systems, including order numbers, transaction times, goods details, transaction quantities, transaction prices, and buyer and seller information; obtain transaction logistics from the logistics management system and third-party logistics service providers, including shipping times, transportation status, delivery routes, and estimated arrival times; collect the contact information and transaction history of each trading merchant to form a merchant relationship chain.

[0012] Perform deduplication processing on the collected multi-source data to ensure that each data record is unique; identify and remove outliers through the Z-score analysis method; perform unified format processing on data from different sources to ensure that field names and data types are consistent; convert numerical data into numerical formats and convert categorical data into corresponding encoding formats; perform multi-source data fusion and integrate the cleaned and formatted data into the data warehouse.

[0013] According to step S200, extract merchant information from the buyer and seller information in the transaction order, group the orders by merchant ID to form a merchant order dataset; extract goods type information from the goods details in the transaction order, group the orders by goods type to generate a type order dataset; based on the merchant order dataset, combine the trading merchant relationship chain to establish a merchant order relationship chain; the merchant order relationship chain records the trading relationships between different merchants and shows the mutual influence of each merchant's orders; define the structure of the merchant order relationship chain, where nodes represent merchants and edges represent trading relationships, and each edge is appended with information from the transaction order.

[0014] Using the type order dataset, identify the orders of the same type of goods in each merchant order dataset. By analyzing the merchant order relationship chain, determine the mutual connections between the same type of orders; construct the merchant same-type order relationship chain to record the relationship between the orders of the same type of goods and the merchants to which they belong.

[0015] Extract the details of each good from the type order dataset and associate it with the merchant order dataset; use the good classification labels to compare the corresponding goods in each merchant order dataset with the goods in the type order dataset, and identify the same type of goods according to the preset classification criteria; extract the transaction records between merchants from the merchant order dataset and establish a preliminary merchant order relationship chain, with each merchant regarded as a node and the transaction record regarded as the edge connecting the nodes;

[0016] Evaluate the relationship strength between merchants according to the transaction quantity and transaction price, which is represented in the form of a weighted graph, and the weight of the edge represents the closeness of the transaction between merchants; combine the identified same type of goods with the merchant order relationship chain to construct the merchant same-type order relationship chain; each edge represents the order of the same type of goods and the relationship between the order and the merchant to which it belongs; each node of the merchant same-type order relationship chain represents a merchant, the edge represents the order relationship of the same type of goods, and each edge is attached with the information in the transaction order.

[0017] According to step S300, calculate the urgency score of the order based on the difference between the order creation time and the expected delivery time, and generate the order urgency. The smaller the difference, the higher the urgency; use the calculated transportation distance data as the goods transportation distance, with the unit of kilometers; encode the destination, specifically convert the destination into a numerical feature using one-hot encoding; define the logistics priority label for each order according to the actual business requirements.

[0018] Divide the order dataset into a training set and a test set. Select the support vector machine model and use the training set for model training. Take the order urgency, goods transportation distance, and destination as feature inputs and the priority label as the target output. Adjust the model parameters to optimize the classification accuracy; use the trained support vector machine model to predict the test set, generate the logistics priority score for each order, and evaluate the prediction effect of the model using accuracy, recall rate, and F1-score; combine the prediction results with the original order data to generate the logistics priority order dataset, including: order ID, order urgency score, goods transportation distance, destination encoding, and predicted logistics priority.

[0019] Analyze the order urgency, goods transportation distance, and destination through a scatter plot to view the relationships and distributions among the features, obtain the visualization results, and based on the visualization results, determine whether the data is linearly separable; when the features show a trend of linear separability, select the linear SVM as the support vector machine model, and the linear SVM uses a linear decision boundary to divide different classes; when the features show a non-linear distribution, select the RBF kernel as the support vector machine model, and the RBF kernel can map the data to a high-dimensional space and find a linear boundary in the high-dimensional space.

[0020] According to step S300, in combination with the logistics priority of each order, optimize the logistics transportation path through the reinforcement learning algorithm Q-learning. The specific steps are as follows:

[0021] S301. Define the state space. Each state represents a specific transportation environment, including the current transportation path, the destination of the goods, the real-time traffic conditions, the transportation cost, and the logistics priority of the current order. Each state is represented by a vector.

[0022] S302. Define the action space. The action represents the selectable transportation path or transportation method; set up a reward mechanism to prompt Q-learning to optimize the path.

[0023] S303. Create a Q-table and initialize the state-action value function Q(s, a) to 0, indicating the expected reward for taking a specific action a in a specific state s; use the ε-greedy strategy to select actions. In the initial stage of training, randomly select actions to explore the environment. As learning progresses, gradually reduce the randomness and select the action with the highest current Q value; after taking an action, observe the changes in the environment, obtain the new state and reward, and update the Q value. The update formula is:

[0024] ;

[0025] where is the learning rate, r is the immediate reward, is the discount factor, is the new state, is the new action; after several iterations, by continuously updating the Q value, enhance the model's strategy for transportation path selection and gradually find the optimal path.

[0026] According to step S400, select the K-means algorithm as the clustering algorithm to cluster similar goods; execute the clustering algorithm to divide similar goods into several clusters, and each cluster represents a group of similar goods that can be transported together to form a set of similar goods; during the clustering process, determine the number of clusters k through the elbow method.

[0027] Extract the transaction records of merchants from the merchant order relationship chain. Consider merchants as nodes in a network and transaction records as edges connecting the nodes. The weight of the edge is weighted based on the transaction frequency to form a weighted undirected graph. Use the Dijkstra algorithm to analyze the transaction frequency and geographical location relationship between merchants, and set the distance between merchants as the weight to optimize the transportation route. Integrate the obtained relationships between merchants to form a merchant relationship set.

[0028] Extract information from the same type of goods set and the merchant relationship set to determine the merchants in the same region or logistics node and their corresponding same type of goods set. Determine the rules for bundled transportation, including: geographical location, goods characteristics, and transportation cost. For geographical location, select merchants in the same city or logistics node for bundling. For goods characteristics, ensure that the bundled goods types are the same or compatible. For transportation cost, give priority to the bundled combination with low transportation cost.

[0029] Based on the rules of the bundled transportation, generate a logistics bundling plan. Each plan includes: a list of merchants, a list of goods, and a transportation route. Among them, the list of merchants is the merchant IDs participating in the bundled transportation and their locations. The list of goods is the same type of goods and quantities for the bundled transportation. The transportation route is the optimal transportation route from the starting point to each destination. According to the bundling plan, evaluate the transportation tools and conduct batch transportation. Develop a transportation plan, including the departure time, route, and arrival time.

[0030] According to the merchant order relationship chain, identify the merchants with order connections, select the merchants with relatively close geographical locations, and add them to the list of merchants. The merchant IDs are extracted from the order data, and the location information of the merchants can be obtained through the registration information of the merchants or the location data in the logistics system. Analyze the transaction frequency and cooperation intensity between merchants through the Dijkstra algorithm, and give priority to selecting the merchants with more order transactions or cooperation relationships to participate in the logistics bundled transportation.

[0031] During the clustering analysis process, the system will aggregate the same type of goods from different merchants to form a same type of goods set. The type of goods is determined by the classification label in the order data, and the quantity of goods is determined by the detailed information in each order. The list of goods includes the specific types of these goods and their corresponding transportation quantities. The same type of goods usually has similar transportation requirements, such as temperature, humidity, volume, packaging method, etc. Therefore, ensure that all bundled goods in the list of goods meet the same transportation conditions to ensure safety and efficiency during transportation.

[0032] Generate the optimal transportation route using path planning algorithms such as Dijkstra's shortest path algorithm or A* based on the merchant's geographical location and the destination location. This route takes into account traffic conditions, transportation distance, and the cargo requirements of each merchant to ensure the minimization of transportation time and cost. Combine the logistics priority of each order to ensure that high-priority orders can be delivered as soon as possible. When planning the route, visit the destinations of high-priority orders first and arrange the delivery locations of low-priority orders in sequence.

[0033] Evaluate the capacity of the transportation vehicle according to the quantity and type of goods in the goods list. Common transportation vehicles include trucks, vans, air freight containers, etc. When selecting, ensure that the volume of the vehicle is sufficient to accommodate all bundled goods and meet transportation conditions such as temperature control equipment and humidity control. Select a suitable transportation vehicle according to the distance of the transportation route and the geographical location of the destination. For example, vans can be selected for short-distance transportation, while air or sea transportation is required for long-distance cross-border transportation.

[0034] An artificial intelligence-based logistics scheduling and management system, including:

[0035] Multi-source data collection and processing module: including a data collection unit and a data preprocessing unit; wherein, the data collection unit collects multi-source data from various online links of international trade, and the multi-source data includes: transaction orders, transaction logistics, and transaction merchant relationship chains. The data preprocessing unit cleans and formats the collected multi-source data, eliminates redundant information and abnormal data, and standardizes the multi-source data;

[0036] Data classification and relationship chain generation module: including a data classification unit, a merchant order relationship chain generation unit, and a merchant similar order relationship chain generation unit; wherein, the data classification unit divides the transaction orders into a merchant order data set and a type order data set based on the transaction merchant and the goods type respectively. The merchant order relationship chain generation unit generates a merchant order relationship chain based on the merchant order data set and the transaction merchant relationship chain. The merchant similar order relationship chain generation unit generates a merchant similar order relationship chain based on the type order data set and the merchant order relationship chain;

[0037] Logistics priority calculation and path optimization module: including a logistics priority calculation unit and a logistics path optimization unit; wherein, the logistics priority calculation unit analyzes the order urgency, cargo transportation distance, and destination in the merchant order data set using the support vector machine algorithm, calculates the logistics priority of each order, and generates a logistics priority order data set. The logistics path optimization unit combines the logistics priority of each order, optimizes the logistics transportation path through the Q-learning reinforcement learning algorithm, dynamically adjusts the transportation plan according to the destination of the goods, real-time traffic conditions, and logistics costs, and gives priority to delivering high-priority orders;

[0038] Logistics bundling scheme generation module: including: similar goods clustering analysis unit, merchant relationship analysis unit, logistics bundling scheme generation unit, and transportation vehicle scheduling unit; among them, the similar goods clustering analysis unit, based on the merchant similar order relationship chain and combined with the type order data set, conducts clustering analysis on similar goods, identifies similar goods that can be transported together, and forms a set of similar goods. The merchant relationship analysis unit uses the merchant order relationship chain to identify merchants with order relationships, analyzes the transaction frequency and geographical location relationship between merchants through the Dijkstra algorithm, and forms a set of merchant relationships. The logistics bundling scheme generation unit generates a logistics bundling scheme based on the set of similar goods and the set of merchant relationships, bundles merchants in the same region or logistics node with similar goods for transportation, and the transportation vehicle scheduling unit schedules logistics vehicles or other transportation tools according to the generated bundling transportation scheme to arrange the batch transportation of goods.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] 1. The present invention introduces the support vector machine algorithm, and intelligently evaluates the logistics priority of orders according to multi-dimensional features such as the urgency of orders, transportation distance, destination, etc.; compared with traditional static rules, the present invention can dynamically adjust the priority according to real-time data to ensure the timely delivery of key orders.

[0041] 2. The present invention uses the reinforcement learning algorithm to optimize the path according to the logistics priority and real-time traffic conditions, dynamically adjusts the transportation plan, improves the transportation efficiency, and reduces the transportation cost; compared with static planning, the present invention can adapt to complex transportation environments and achieve more flexible scheduling.

[0042] 3. By analyzing similar goods through the clustering algorithm and combining with the merchant order relationship chain, the present invention can intelligently identify similar goods that can be transported together and merchants with cooperative relationships, generate a bundling transportation plan, maximize the utilization of logistics resources, and improve the batch transportation efficiency. Brief Description of the Drawings

[0043] Figure 1 is a schematic diagram of the steps of a logistics scheduling management method based on artificial intelligence of the present invention;

[0044] Figure 2 is a system structure diagram of a logistics scheduling management system based on artificial intelligence of the present invention. Detailed Embodiments

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] Example: Figure 1 - Figure 2 As shown, the present invention provides a technical solution.

[0047] According to one embodiment of the present invention, Figure 1 A step diagram of a logistics scheduling management method based on artificial intelligence is shown in the figure, a logistics scheduling management method based on artificial intelligence, the method comprises the following steps:

[0048] S100, collecting multi-source data from various online links of international trade, wherein the multi-source data includes: transaction orders, transaction logistics, and transaction merchant relationship chains; cleaning and formatting the collected multi-source data, removing redundant information and abnormal data, and standardizing the multi-source data;

[0049] S200, dividing the transaction order into a merchant order data set and a type order data set according to the transaction merchant and the type of goods; generating a merchant order relationship chain based on the merchant order data set and the transaction merchant relationship chain; generating a merchant same-type order relationship chain based on the type order data set and the merchant order relationship chain;

[0050] S300, using the support vector machine algorithm to analyze the order urgency, cargo transportation distance and destination in the merchant order data set, calculate the logistics priority of each order, and generate a logistics priority order data set; combining the logistics priority of each order, optimizing the logistics transportation path through the reinforcement learning algorithm Q-learning, dynamically adjusting the transportation plan according to the destination of the cargo, real-time traffic conditions and logistics costs, and giving priority to the delivery of high-priority orders;

[0051] S400. Based on the merchant order relationship chain of the same type and in combination with the type order data set, cluster analysis is performed on the same type of goods to identify the same type of goods that can be transported together and form a set of the same type of goods; using the merchant order relationship chain, merchants with order relationships are identified, and the transaction frequency and geographical location relationship between merchants are analyzed through the Dijkstra algorithm to form a merchant relationship set; based on the set of the same type of goods and the merchant relationship set, a logistics bundling plan is generated to bundle merchants in the same region or logistics node with the same type of goods for transportation; according to the generated bundling transportation plan, logistics vehicles or other means of transportation are dispatched to arrange the batch transportation of goods.

[0052] According to step S100, automatically collect transaction orders from e-commerce platforms, B2B trading websites, and enterprise management systems, including order numbers, transaction times, goods details, transaction quantities, transaction prices, and buyer and seller information; obtain transaction logistics through the logistics management system and third-party logistics service providers, including shipping times, transportation status, delivery routes, and estimated arrival times; collect contact information and transaction histories of each trading merchant to form a merchant relationship chain.

[0053] Perform deduplication processing on the collected multi-source data to ensure that each data record is unique; identify and remove outliers through the Z-score analysis method; perform unified format processing on data from different sources to ensure that field names and data types are consistent; convert numerical data into numerical formats and convert categorical data into corresponding encoding formats; perform multi-source data fusion and integrate the cleaned and formatted data into a data warehouse.

[0054] According to step S200, extract merchant information from the buyer and seller information in the transaction order, group the orders by merchant ID to form a merchant order dataset; extract goods type information from the goods details in the transaction order and group the orders by goods type to generate a type order dataset; based on the merchant order dataset, combine the trading merchant relationship chain to establish a merchant order relationship chain; the merchant order relationship chain records the trading relationships between different merchants and shows the mutual influence of each merchant's orders; define the structure of the merchant order relationship chain, where nodes represent merchants and edges represent trading relationships, and each edge is appended with information in the transaction order.

[0055] Use the type order dataset to identify orders for the same type of goods in each merchant order dataset, and determine the mutual connections between orders of the same type by analyzing the merchant order relationship chain; construct a merchant same-type order relationship chain to record the relationships between orders of the same type of goods and the merchants to which they belong.

[0056] Extract the goods details of each good from the type order dataset and associate it with the merchant order dataset; use the goods classification labels to compare the corresponding goods in each merchant order dataset with the goods in the type order dataset, and identify goods of the same type according to the preset classification criteria; extract the trading records between merchants from the merchant order dataset and establish a preliminary merchant order relationship chain, with each merchant regarded as a node and the trading records regarded as edges connecting the nodes.

[0057] Evaluate the relationship strength among merchants based on the transaction quantity and price, and represent it in the form of a weighted graph. The weight of an edge represents the closeness of transactions between merchants. Combine the identified same-category goods with the merchant order relationship chain to construct the merchant same-category order relationship chain. Each edge represents an order of the same-category goods and the relationship between the merchants to which the order belongs. Each node of the merchant same-category order relationship chain represents a merchant, and the edge represents the order relationship of the same-category goods. Each edge is appended with the information in the transaction order.

[0058] In this embodiment, a transaction order data set is used, which contains the following information:

[0059] Merchant information: The buyer and seller in each order have unique merchant IDs, respectively representing the merchants participating in the transaction.

[0060] Goods details: Each order details information such as the goods type, quantity, unit price, etc.

[0061] Transaction record: The order includes the quantity and amount of the transaction, representing the transaction relationship between merchants.

[0062] Using the merchant order data set, summarize the transaction records between merchants. Each merchant is used as a node, and the transaction relationship is used as an edge. Determine the relationship strength between merchants according to the transaction quantity and amount. According to the transaction quantity and total amount, the generated merchant order relationship chain is as follows: There are two transactions between A and B, with a total transaction amount of 11,240 yuan, and the transaction relationship strength is 2 transactions; there is 1 transaction between A and C, with a total transaction amount of 4,000 yuan; there is 1 transaction between B and C, with a total transaction amount of 4,400 yuan.

[0063] The merchant order relationship chain can be regarded as the following weighted graph: A↔B (weight = 2 transactions, amount = 11,240 yuan), A↔C (weight = 1 transaction, amount = 4,000 yuan), B↔C (weight = 1 transaction, amount = 4,400 yuan).

[0064] Using the goods classification labels, compare the goods in the merchant orders, identify the same-category goods, and generate the same-category goods set. For example, the goods in the electronic product category are grouped into the same category to form the electronic product same-category goods set. Combine the same-category goods set with the merchant order relationship chain to construct the merchant same-category order relationship chain. Each edge represents the order of the same-category goods and the relationship between the merchants to which it belongs. For example, there are two orders of electronic products between merchants A and B, and there is one order of electronic products between merchants B and C. The merchant same-category order relationship chain is as follows: A↔B (goods: electronic products, number of transactions = 2, quantity = 22 pieces), B↔C (goods: electronic products, number of transactions = 1, quantity = 8 pieces).

[0065] According to step S300, calculate the urgency score of the order based on the difference between the order creation time and the expected delivery time, and generate the order urgency. The smaller the difference, the higher the urgency. Use the calculated transportation distance data as the goods transportation distance, with the unit being kilometers. Encode the destination. Specifically, use one-hot encoding to convert the destination into numerical features. Define the logistics priority label for each order according to the actual business requirements.

[0066] Divide the order dataset into a training set and a test set. Select the support vector machine model and use the training set for model training. Take the order urgency, goods transportation distance, and destination as feature inputs, and the priority label as the target output. Adjust the model parameters to optimize the classification accuracy. Use the trained support vector machine model to predict the test set, generate the logistics priority score for each order, and evaluate the prediction effect of the model using accuracy, recall rate, and F1-score. Combine the prediction results with the original order data to generate a logistics priority order dataset, including: order ID, order urgency score, goods transportation distance, destination encoding, and predicted logistics priority.

[0067] Analyze the order urgency, goods transportation distance, and destination through a scatter plot to view the relationships and distributions between the features, and obtain the visualization results. According to the visualization results, determine whether the data is linearly separable. When the features show a trend of being linearly separable, select the linear SVM as the support vector machine model. The linear SVM uses a linear decision boundary to divide different classes. When the features show a non-linear distribution, select the RBF kernel as the support vector machine model. The RBF kernel can map the data to a high-dimensional space and find a linear boundary in the high-dimensional space.

[0068] In this embodiment, there are 1000 order data. Each order contains the order creation time, expected delivery time, goods transportation distance, destination information, and logistics priority label. The destination of the order is converted into numerical features and processed by one-hot encoding. The urgency score of the order is based on the difference between the order creation time and the expected delivery time. The smaller the difference, the higher the urgency. For example, for some orders, the difference is 2 days and the urgency score is 8 (out of 10), while for orders with a difference of 5 days, the urgency score is 4. In the orders, the range of goods transportation distance data is from 10 kilometers to 1500 kilometers. For example, the transportation distance of some orders is 100 kilometers, while that of some other orders is 1200 kilometers. The destination of the order is converted into numerical features using one-hot encoding. For example, for 3 different destinations, the encoding is: Destination A: [1, 0, 0], Destination B: [0, 1, 0], Destination C: [0, 0, 1]. The logistics priority label is defined according to the actual business requirements and is divided into "high priority" and "low priority". In the data, orders with high urgency and short transportation distance are marked as high priority, while orders with low urgency and long transportation distance are marked as low priority.

[0069] The order dataset is divided into a training set and a test set in a ratio of 8:2. Among them, 800 orders are used for model training and 200 orders are used for model testing. Through scatter plot analysis of features such as order urgency, goods transportation distance, and destination encoding, the relationships between features are observed. The experiment finds that the order data presents a non-linear distribution, so a support vector machine model with an RBF kernel is selected. The RBF kernel can map the data to a high-dimensional space to find a linear decision boundary in the high-dimensional space. Taking the order urgency score, transportation distance, and destination encoding as input features and the logistics priority as the target output, the SVM model is trained using the training set. The parameters of the SVM model (such as the C value and γ value) are adjusted by the grid search method to optimize the classification accuracy.

[0070] The test set is used for model prediction to generate the logistics priority score for each order, and the accuracy, recall rate, and F1-score are used to evaluate the prediction effect of the model. The experimental results show that: accuracy: 85%, recall rate: 82%, F1-score: 83%.

[0071] According to step S300, combined with the logistics priority of each order, the logistics transportation path is optimized through the Q-learning reinforcement learning algorithm. The specific steps are as follows:

[0072] In this embodiment, there is a logistics transportation environment with five main destinations (A, B, C, D, E) and multiple optional transportation routes. Each route has different transportation times, transportation costs, and dynamic changes affected by real-time traffic conditions. The task of the logistics transportation vehicle is to select the optimal route to deliver the goods to the destination according to the priority of the current order. Based on the priority data determined in the previous steps, each order is assigned one of three priorities: high, medium, or low, according to the urgency, transportation distance, and destination. There are 5 optional routes, namely P1, P2, P3, P4, and P5, and the timeliness, transportation cost, and traffic conditions of each route are different. According to the real-time traffic data, the traffic conditions on the transportation route are divided into three types: smooth, medium, and congested.

[0073] S301. Define the state space. Each state represents a specific transportation environment, including the current transportation route, the destination of the goods, the real-time traffic conditions, the transportation cost, and the logistics priority of the current order. Each state is represented by a vector.

[0074] Each state represents a snapshot of the logistics environment. The state vector includes: the current transportation route (P1 to P5), the destination (A to E), the real-time traffic conditions (smooth, medium, congested), the transportation cost (unit: yuan), and the logistics priority of the current order (high, medium, low). For example, a certain state can be represented as: [P1, B, medium, 100 yuan, high priority].

[0075] S302. Define the action space. An action represents an optional transportation route or transportation method. Set up a reward mechanism to encourage Q-learning to optimize the route. Select different transportation routes (P1, P2, P3, P4, P5) and different transportation methods.

[0076] S303. Create a Q-table and initialize the state-action value function Q(s, a) to 0, which represents the expected reward for taking a specific action a in a specific state s. Use the ε-greedy strategy to select actions. In the initial stage of training, randomly select actions to explore the environment. As learning progresses, gradually reduce the randomness and select the action with the highest current Q value. After taking an action, observe the changes in the environment, obtain the new state and reward, and update the Q value. The update formula is:

[0077] ;

[0078] Among them, is the learning rate, r is the immediate reward, is the discount factor, is the new state, is the new action. After several iterations, by continuously updating the Q value, enhance the model's strategy for transportation route selection and gradually find the optimal route.

[0079] Set the transportation timeliness and transportation cost as the main reward mechanism. The higher the timeliness and the lower the cost, the higher the reward. If the goods are delivered to the destination in time, a high reward is given; if the transportation is delayed due to a poor route selection, a punitive negative reward is given. The Q(s, a) value is initially set to 0, representing the initial expected reward for selecting a certain action a in a specific state s.

[0080] In the initial stage of training, a high ε value (e.g., 0.9) is adopted to randomly select transportation routes and transportation methods for exploration, in order to collect more data of state-action pairs. As the training progresses, ε gradually decreases (e.g., reduced to 0.1), and the path with the highest current Q value is selected more often to utilize the experience accumulated by the model.

[0081] According to step S400, select the K-means algorithm as the clustering algorithm to cluster similar goods; execute the clustering algorithm to divide similar goods into several clusters, and each cluster represents a group of similar goods that can be transported together, forming a set of similar goods; during the clustering process, determine the number of clusters k through the elbow method;

[0082] Extract the transaction records of merchants from the merchant order relationship chain, regard the merchants as nodes in the network, regard the transaction records as the edges connecting the nodes, and the weights of the edges are weighted based on the transaction frequency to form a weighted undirected graph; use the Dijkstra algorithm to analyze the transaction frequency and geographical location relationship between merchants, set the distance between merchants as the weight to optimize the transportation route; integrate the obtained relationships between merchants to form a set of merchant relationships.

[0083] Extract information from the set of similar goods and the set of merchant relationships to determine the merchants in the same region or logistics node and their corresponding sets of similar goods; determine the rules for bundled transportation, including: geographical location, goods characteristics, and transportation cost; for geographical location, select merchants in the same city or logistics node for bundling, for goods characteristics, ensure that the bundled goods types are the same or compatible, and for transportation cost, give priority to the bundled combination with low transportation cost;

[0084] Based on the rules of the bundled transportation, generate a logistics bundling plan, and each plan includes: a list of merchants, a list of goods, and a transportation route; wherein, the list of merchants is the merchant IDs and their locations participating in the bundled transportation, the list of goods is the similar goods and quantities for bundled transportation, and the transportation route is the optimal transportation route from the starting point to each destination; according to the bundling plan, evaluate the transportation tools for batch transportation; formulate a transportation plan, including the departure time, route, and arrival time.

[0085] According to an embodiment of the present invention, as Figure 2As shown in the system structure diagram of an artificial intelligence-based logistics scheduling management system, an artificial intelligence-based logistics scheduling management system includes:

[0086] Multi-source data acquisition and processing module: including a data acquisition unit and a data preprocessing unit; among them, the data acquisition unit collects multi-source data from various online links of international trade, and the multi-source data includes: transaction orders, transaction logistics, and transaction merchant relationship chains. The data preprocessing unit cleans and formats the collected multi-source data, eliminates redundant information and abnormal data, and standardizes the multi-source data;

[0087] Data classification and relationship chain generation module: including a data classification unit, a merchant order relationship chain generation unit, and a merchant similar order relationship chain generation unit; among them, the data classification unit divides the transaction orders into a merchant order data set and a type order data set based on the transaction merchants and goods types respectively. The merchant order relationship chain generation unit generates a merchant order relationship chain based on the merchant order data set and the transaction merchant relationship chain. The merchant similar order relationship chain generation unit generates a merchant similar order relationship chain based on the type order data set and the merchant order relationship chain;

[0088] Logistics priority calculation and path optimization module: including a logistics priority calculation unit and a logistics path optimization unit; among them, the logistics priority calculation unit uses the support vector machine algorithm to analyze the order urgency, goods transportation distance, and destination in the merchant order data set, calculates the logistics priority of each order, and generates a logistics priority order data set. The logistics path optimization unit combines the logistics priority of each order, optimizes the logistics transportation path through the Q-learning reinforcement learning algorithm, dynamically adjusts the transportation plan according to the destination of the goods, real-time traffic conditions, and logistics costs, and gives priority to delivering high-priority orders;

[0089] Logistics bundling scheme generation module: including a similar goods clustering analysis unit, a merchant relationship analysis unit, a logistics bundling scheme generation unit, and a transportation tool scheduling unit; among them, the similar goods clustering analysis unit clusters similar goods based on the merchant similar order relationship chain and combines the type order data set, identifies similar goods that can be transported together, and forms a similar goods set. The merchant relationship analysis unit uses the merchant order relationship chain to identify merchants with order relationships, analyzes the transaction frequency and geographical location relationship between merchants through the Dijkstra algorithm, and forms a merchant relationship set. The logistics bundling scheme generation unit generates a logistics bundling scheme based on the similar goods set and the merchant relationship set, bundles merchants in the same region or logistics node with similar goods for transportation. The transportation tool scheduling unit schedules logistics vehicles or other transportation tools according to the generated bundled transportation plan and arranges the batch transportation of goods.

[0090] According to this embodiment, there is a logistics network covering 10 cities, with a total of 50 merchants and 200 different types of goods. These merchants are distributed in different cities, have certain trading relationships, and a certain quantity of goods needs to be transported to different logistics nodes or destinations. The trading frequency and order relationships of the merchants are determined by order data, and there are varying degrees of trading among merchants such as merchant A, merchant B, merchant C, etc. The goods are classified according to types, mainly including food, household items, electronic products, etc., and each type of goods has different attributes (such as volume, weight, special transportation requirements, etc.).

[0091] Use the clustering analysis unit for similar goods to cluster the 200 different types of goods. Based on the merchant similar order relationship chain and type order data set, combined with the volume, weight, and transportation requirements of the goods, identify the similar goods that can be transported together to form a set of similar goods. For example:

[0092] Food goods are clustered into cluster 1, mainly including frozen foods, ambient temperature foods, etc.; household goods are clustered into cluster 2, mainly including furniture and household appliances; electronic goods are clustered into cluster 3, mainly including electronic products, mobile phones, computers, etc. Finally, 5 clusters of similar goods are formed, namely food, household items, electronic products, clothing, and other small commodities.

[0093] Utilize the merchant order relationship chain and analyze the trading frequency and geographical location relationship among merchants through the Dijkstra algorithm. Merchant A and merchant B have frequent trading and are close in geographical location, and the system identifies them as high-frequency trading merchants. The Dijkstra algorithm generates a set of merchant relationships based on trading frequency, geographical location, and order volume. For example: Merchant A and merchant B have a high trading frequency and are close in distance, forming a set of merchant relationships; merchant C and merchant D have less trading and are far in geographical location, so they are not bundled. After analysis, a total of 4 sets of merchant relationships are formed, and the merchants in each relationship set have a high trading frequency and are close in geographical location.

[0094] Based on the set of similar goods and the set of merchant relationships, generate a logistics bundling plan. The logistics bundling plan generation unit identifies the merchants and goods that can be bundled for transportation, and bundles the merchants in the same region or logistics node with similar goods for transportation. For example: Merchants A and B in cluster 1 (food category) and merchants E and F in cluster 2 (household category) can be bundled together in the same area and shipped to the same logistics node. Merchants G and H in cluster 3 (electronic products) are bundled together for transportation due to their high trading frequency and close geographical location.

[0095] The transportation vehicle scheduling unit arranges appropriate transportation vehicles according to the bundling plan. Two trucks are scheduled to handle the transportation of the same type of cargo cluster. The first truck is responsible for the transportation of food and household items, and the second truck is responsible for the transportation of electronic products. The transportation vehicle scheduling takes into account the transportation requirements of the cargo, such as the need for cold chain transportation for food and shockproof measures for electronic products.

[0096] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be encompassed within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. A logistics scheduling management method based on artificial intelligence, characterized in that, The method includes the following steps: S100. Collect multi-source data from various online links of international trade. The multi-source data includes: transaction orders, transaction logistics, and transaction merchant relationship chains; clean and format the collected multi-source data, eliminate redundant information and abnormal data, and standardize the multi-source data; S200. Based on the transaction orders, divide them into merchant order data sets and type order data sets according to transaction merchants and goods types respectively; based on the merchant order data sets and transaction merchant relationship chains, generate merchant order relationship chains; based on the type order data sets and merchant order relationship chains, generate merchant similar order relationship chains; S300. Use the support vector machine algorithm to analyze the order urgency, goods transportation distance, and destination in the merchant order data set, calculate the logistics priority of each order, and generate a logistics priority order data set; combine the logistics priority of each order, and optimize the logistics transportation path through the Q-learning algorithm of reinforcement learning, dynamically adjust the transportation plan according to the destination of the goods, real-time traffic conditions, and logistics costs, and give priority to delivering high-priority orders; S400. Based on the merchant similar order relationship chains, combine with the type order data sets, perform clustering analysis on similar goods, identify similar goods that can be transported together, and form a set of similar goods; use the merchant order relationship chains to identify merchants with order relationships, analyze the transaction frequency and geographical location relationships between merchants through the Dijkstra algorithm, and form a set of merchant relationships; based on the set of similar goods and the set of merchant relationships, generate a logistics bundling plan, bundle merchants in the same region or logistics node with similar goods for transportation; according to the generated bundling transportation plan, dispatch logistics vehicles or other transportation tools, and arrange for the batch transportation of goods.

2. The logistics scheduling management method based on artificial intelligence according to claim 1, characterized in that: According to step S100, automatically collect transaction orders from e-commerce platforms, B2B trading websites, and enterprise management systems, including order numbers, transaction times, goods details, transaction quantities, transaction prices, and buyer and seller information; obtain transaction logistics from the logistics management system and third-party logistics service providers, including shipping times, transportation status, delivery routes, and estimated arrival times; collect the contact information and transaction history of each transaction merchant to form a merchant relationship chain; Perform duplicate removal processing on the collected multi-source data to ensure that each data record is unique; identify and eliminate outliers through the Z-score analysis method; perform unified format processing on data from different sources to ensure that field names and data types are consistent; convert numerical data into numerical formats and convert categorical data into corresponding encoding formats; perform multi-source data fusion, and integrate the cleaned and formatted data into the data warehouse.

3. The logistics scheduling management method based on artificial intelligence according to claim 1 is characterized in that: According to step S200, extract merchant information from the buyer and seller information in the transaction orders, group the orders according to merchant IDs, and form a merchant order data set; Extract goods type information from the goods details in the transaction orders, group the orders according to the goods types, and generate a type order data set; Based on the merchant order data set, combine with the transaction merchant relationship chain, and establish a merchant order relationship chain; The merchant order relationship chain records the transaction relationships between different merchants and shows the mutual influence of each merchant's orders; defines the structure of the merchant order relationship chain, where nodes represent merchants and edges represent transaction relationships, and each edge is attached with information in the transaction order. Using the type order dataset, identify the orders of the same type of goods in each merchant order dataset. By analyzing the merchant order relationship chain, determine the mutual connections between the same type of orders; construct the merchant same-type order relationship chain to record the relationship between the orders of the same type of goods and the merchants to which they belong.

4. A logistics scheduling management method based on artificial intelligence according to claim 3, characterized in that: Extract the goods details of each item from the type order dataset and associate them with the merchant order dataset. Using the goods classification labels, compare the corresponding goods in each merchant order dataset with the goods in the type order dataset, and identify the same type of goods according to the preset classification criteria. Extract the transaction records between merchants from the merchant order dataset and establish a preliminary merchant order relationship chain, with each merchant regarded as a node and the transaction record regarded as an edge connecting the nodes. Evaluate the relationship strength between merchants according to the transaction quantity and transaction price, represented in the form of a weighted graph, where the weight of the edge represents the closeness of the transaction between merchants; combine the identified same type of goods with the merchant order relationship chain to construct the merchant same-type order relationship chain. Each edge represents the order of the same type of goods and the relationship between the order and the merchant to which it belongs; the nodes of each merchant same-type order relationship chain represent merchants, the edges represent the order relationship of the same type of goods, and each edge is attached with information in the transaction order.

5. A logistics scheduling management method based on artificial intelligence according to claim 1, characterized in that: According to step S300, calculate the urgency score of the order based on the difference between the order creation time and the expected delivery time, and generate the order urgency. The smaller the difference, the higher the urgency. Use the calculated transportation distance data as the goods transportation distance, with the unit of kilometers. Encode the destination, specifically convert the destination into numerical features using one-hot encoding; according to the actual business requirements, define the logistics priority label for each order. Divide the order dataset into a training set and a test set, select the support vector machine model, use the training set for model training, take the order urgency, goods transportation distance, and destination as feature inputs, and the priority label as the target output, and adjust the model parameters to optimize the classification accuracy. Use the trained support vector machine model to predict the test set, generate the logistics priority score for each order, and evaluate the prediction effect of the model using accuracy, recall rate, and F1-score; combine the prediction results with the original order data to generate the logistics priority order dataset, including: order ID, order urgency score, goods transportation distance, destination encoding, and predicted logistics priority.

6. The logistics scheduling management method based on artificial intelligence according to claim 5, characterized in that: Analyze the order urgency, goods transportation distance, and destination through a scatter plot to view the relationships and distributions among the features, obtain the visualization results, and based on the visualization results, determine whether the data is linearly separable; when the features show a trend of being linearly separable, select the linear SVM as the support vector machine model, and the linear SVM uses a linear decision boundary to divide different classes; when the features show a non-linear distribution, select the RBF kernel as the support vector machine model, and the RBF kernel can map the data to a high-dimensional space and find a linear boundary in the high-dimensional space.

7. An artificial intelligence-based logistics scheduling management method according to claim 6, characterized in that: According to step S300, in combination with the logistics priority of each order, optimize the logistics transportation path through the reinforcement learning algorithm Q-learning. The specific steps are as follows: S301. Define the state space. Each state represents a specific transportation environment, including the current transportation path, the goods destination, the real-time traffic conditions, the transportation cost, and the logistics priority of the current order. Each state is represented by a vector. S302. Define the action space. An action represents an optional transportation path or transportation method. Set up a reward mechanism to prompt Q-learning to optimize the path. S303. Create a Q-table and initialize the state-action value function Q(s, a) to 0, indicating the expected reward for taking a specific action a in a specific state s; use the ε-greedy strategy to select actions. In the initial stage of training, randomly select actions to explore the environment. As learning progresses, gradually reduce the randomness and select the action with the highest current Q value; after taking an action, observe the changes in the environment, obtain the new state and reward, and update the Q value. The update formula is: ； Among them, is the learning rate, r is the immediate reward, is the discount factor, is the new state, is the new action; After several iterations, by continuously updating the Q value, the strategy of the model for transportation path selection is enhanced, and the optimal path is gradually found.

8. A logistics scheduling management method based on artificial intelligence according to claim 1, characterized in that: According to step S400, select the K-means algorithm as the clustering algorithm to cluster similar goods; execute the clustering algorithm to divide similar goods into several clusters. Each cluster represents a group of similar goods that can be transported together, forming a set of similar goods. During the clustering process, determine the number of clusters k through the elbow method. Extract the transaction records of merchants from the merchant order relationship chain. Consider the merchants as nodes in the network and the transaction records as the edges connecting the nodes. The weights of the edges are weighted based on the transaction frequency to form a weighted undirected graph; use the Dijkstra algorithm to analyze the transaction frequency and geographical location relationships between merchants, and set the distance between merchants as the weight to optimize the transportation path. Integrate the obtained relationships between merchants to form a set of merchant relationships.

9. The logistics scheduling management method based on artificial intelligence according to claim 8, characterized in that: Extract information from the set of similar goods and the set of merchant relationships to determine the merchants in the same region or logistics node and their corresponding sets of similar goods; determine the rules for bundled transportation, including: geographical location, goods characteristics, and transportation cost; for geographical location, select merchants in the same city or logistics node for bundling; for goods characteristics, ensure that the bundled goods types are the same or compatible; for transportation cost, give priority to the bundled combination with a low transportation cost. Based on the rules of bundled transportation, generate logistics bundling solutions, each solution including: a merchant list, a goods list, and a transportation route; wherein, the merchant list is the merchant IDs participating in the bundled transportation and their locations, the goods list is the same type of goods and quantities for bundled transportation, and the transportation route is the optimal transportation route from the starting point to each destination; according to the bundling solution, evaluate transportation tools and conduct batch transportation; formulate a transportation plan, including departure time, route, and arrival time.

10. A logistics scheduling management system based on artificial intelligence, which uses a logistics scheduling management method based on artificial intelligence described in any one of claims 1-9, characterized in that, Including: Multi-source data collection and processing module: including: a data collection unit and a data preprocessing unit; wherein, the data collection unit collects multi-source data from various online links of international trade, and the multi-source data includes: transaction orders, transaction logistics, and transaction merchant relationship chains, and the data preprocessing unit cleans and formats the collected multi-source data, eliminates redundant information and abnormal data, and standardizes the multi-source data; Data classification and relationship chain generation module: including: a data classification unit, a merchant order relationship chain generation unit, and a merchant similar order relationship chain generation unit; wherein, the data classification unit divides the transaction orders into a merchant order data set and a type order data set based on the transaction merchants and goods types respectively, the merchant order relationship chain generation unit generates a merchant order relationship chain based on the merchant order data set and the transaction merchant relationship chain, and the merchant similar order relationship chain generation unit generates a merchant similar order relationship chain based on the type order data set and the merchant order relationship chain; Logistics priority calculation and path optimization module: including: a logistics priority calculation unit and a logistics path optimization unit; wherein, the logistics priority calculation unit analyzes the order urgency, goods transportation distance, and destination in the merchant order data set using the support vector machine algorithm, calculates the logistics priority of each order, generates a logistics priority order data set, and the logistics path optimization unit combines the logistics priority of each order, optimizes the logistics transportation path through the Q-learning algorithm of reinforcement learning, dynamically adjusts the transportation plan according to the destination of the goods, real-time traffic conditions, and logistics costs, and gives priority to delivering high-priority orders; Logistics bundling solution generation module: including: a similar goods clustering analysis unit, a merchant relationship analysis unit, a logistics bundling solution generation unit, and a transportation tool scheduling unit; wherein, the similar goods clustering analysis unit clusters similar goods based on the merchant similar order relationship chain and combines the type order data set, identifies similar goods that can be transported together, and forms a similar goods set, the merchant relationship analysis unit uses the merchant order relationship chain to identify merchants with order relationships, analyzes the transaction frequency and geographical location relationship between merchants through the Dijkstra algorithm, and forms a merchant relationship set, the logistics bundling solution generation unit generates a logistics bundling solution based on the similar goods set and the merchant relationship set, bundles merchants in the same region or logistics node with similar goods for transportation, and the transportation tool scheduling unit schedules logistics vehicles or other transportation tools according to the generated bundled transportation solution and arranges batch transportation of goods.

Images