Logistics distribution optimization system based on multi-source data analysis

The logistics and distribution optimization system, which uses multi-source data analysis, integrates data from multiple channels, establishes mapping relationships, mines logistics and distribution elements, and formulates efficient scheduling plans. This solves the problem of inaccurate inventory layout and distribution strategies, improves logistics and distribution efficiency, and reduces costs.

CN118504777BActive Publication Date: 2024-11-01WUXI GUANYUN INFORMATION TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410918129.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-10
Publication Date
2024-11-01
Estimated Expiration
2044-07-10

AI Technical Summary

Technical Problem

Existing logistics and distribution systems struggle to integrate and analyze logistics data from different channels, leading to inaccurate inventory layout and distribution strategies, resulting in low efficiency and high costs.

Method used

A logistics and distribution optimization system based on multi-source data analysis is adopted. Through parameter decomposition calculation, mapping relationship establishment and backtracking search optimization, multi-channel data is integrated to establish the mapping relationship between logistics parameters and inventory distribution information, explore logistics and distribution elements, and formulate efficient scheduling plans.

Benefits of technology

It has achieved comprehensive optimization and precise control of logistics and distribution, improving delivery efficiency and reducing costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118504777B_ABST
    Figure CN118504777B_ABST
Patent Text Reader

Abstract

The present invention discloses a logistics distribution optimization system based on multi-source data analysis, which relates to the field of data processing technology. The system includes: obtaining multi-source data of users; extracting logistics parameters, obtaining logistics orders, commodity codes, order commodity quantities, and order demand time; obtaining inventory distribution information; establishing a demand mapping relationship between the logistics parameters and inventory distribution information; mining logistics distribution elements according to the demand mapping relationship, and determining logistics distribution optimization variables; optimizing logistics distribution with the goal of maximizing the comprehensive results of logistics distribution elements, and obtaining a logistics distribution scheduling plan. The system solves the technical problem that it is difficult to integrate and analyze logistics data from different channels in existing logistics distribution, which leads to inaccurate inventory layout and distribution strategies, poor logistics distribution efficiency, and high distribution costs, and realizes comprehensive optimization and precise control of logistics distribution, achieving the technical effect of improving logistics distribution efficiency and reducing distribution costs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field related to data processing, and specifically to a logistics distribution optimization system based on multi-source data analysis. Background Art

[0002] Under the wave of digitalization and globalization, the logistics and distribution industry is facing unprecedented challenges and opportunities. With the vigorous development of e-commerce, consumers have an increasing demand for fast, accurate and low-cost delivery services, hoping that orders can be delivered faster and more accurately. At the same time, with the popularization of multi-channel sales, the information of online and offline stores is becoming more and more abundant, providing more and more complex data sources for logistics and distribution. The logistics industry is also facing multiple problems such as rising costs, traffic congestion, and environmental pressure. The optimization of logistics and distribution is particularly important. However, traditional logistics and distribution optimization data sharing is not timely, information flow is slow, and often only relies on a single data source for decision-making. The various modules of logistics and distribution lack the ability to monitor and make decisions in real time, making it difficult to grasp the storage distribution of goods in real time, including warehouse attributes, location and corresponding inventory. This leads to inaccurate formulation of logistics and distribution plans, low efficiency of the entire logistics and distribution, and increased distribution costs. Summary of the invention

[0003] This application provides a logistics distribution optimization system based on multi-source data analysis, and adopts technical means such as parameter decomposition operation, establishment of mapping relationship, backtracking search optimization, etc., to solve the technical problem that it is difficult to integrate and analyze logistics data from different channels in existing logistics distribution, which leads to inaccurate inventory layout and distribution strategy, resulting in poor logistics distribution efficiency and high distribution costs. It realizes comprehensive optimization and precise control of logistics distribution, and achieves the technical effect of improving logistics distribution efficiency and reducing distribution costs.

[0004] The present application provides a logistics distribution optimization system based on multi-source data analysis, the system comprising: a user multi-source data acquisition module, the user multi-source data acquisition module is used to acquire the user's multi-source data, the multi-source data comes from the user's multi-channel store information, the multi-channel store information includes online attribute tags and offline attribute tags; a logistics parameter extraction module, the logistics parameter extraction module is used to extract logistics parameters from the multi-source data, and acquire logistics orders, commodity codes, order commodity quantities, and order demand time; an inventory distribution information acquisition module, the inventory distribution information acquisition module is used to acquire inventory distribution information, the inventory distribution information is used to describe the storage distribution of the user's commodities, including warehouse attributes, warehouse locations, commodity codes, and corresponding inventory quantities, wherein the warehouse attributes include location Domain distribution warehouse, offline store configuration warehouse; mapping relationship establishment module, the mapping relationship establishment module is used to align the time of logistics orders with the order demand time according to the logistics parameters of the multi-source data, and establish the demand mapping relationship between the logistics parameters and the inventory distribution information with the commodity code, commodity quantity and the delivery address of the logistics order; logistics distribution optimization variable determination module, the logistics distribution optimization variable determination module mines the logistics distribution elements based on the online attribute labels and offline attribute labels according to the demand mapping relationship, and determines the logistics distribution optimization variables; logistics distribution scheduling plan acquisition module, the logistics distribution scheduling plan acquisition module is based on the logistics distribution optimization variables, takes the maximization of the comprehensive result of logistics distribution elements as the goal, optimizes the logistics distribution, and obtains the logistics distribution scheduling plan.

[0005] In a possible implementation, the mapping relationship establishment module also performs the following processing: setting the order start time node according to the order demand time, aligning the order start time node according to time, and outputting the order timing chain node block, the order timing chain node block includes all order parameters of each time node and the online attribute labels and offline attribute labels; based on the order timing chain node block, for all order parameters of the time node, the first demand matching condition is that the commodity code and the corresponding commodity quantity of the logistics order and the warehouse distribution quantity are satisfied, and the second demand matching condition is that the distance between the delivery address of the logistics order and the warehouse location is the minimum; according to the first demand matching condition and the second demand matching condition, the demand matching relationship is screened in turn, and the demand mapping relationship between the logistics parameters and the inventory distribution information is established.

[0006] In a possible implementation, the mapping relationship establishment module also performs the following processing: respectively obtaining the historical logistics data of the multi-channel store information; performing logistics peak period forecasting based on the historical logistics data to determine the predicted logistics peak information of each channel store information; based on the predicted logistics peak information of each channel store information, performing inventory constraints on the inventory distribution information according to peak time, store location, and predicted logistics volume, and constructing an inventory distribution plan timing chain; and adding the inventory distribution plan timing chain as demand constraint information to the demand mapping relationship.

[0007] In a possible implementation, the mapping relationship establishment module also performs the following processing: setting the delivery speed tolerance of the logistics order for online attributes and offline attributes; extracting the attribute tags of the goods, the attribute tags including product category, popularity, and seasonal restrictions; configuring the evaluation coefficient of the attribute tags, evaluating the delivery speed of the attribute tags according to the evaluation coefficient, and obtaining the delivery constraints of the commodity attributes; integrating the delivery speed tolerance of the logistics order and the delivery constraints of the commodity attributes to obtain the delivery time threshold of each order in the demand mapping relationship; when the delivery time threshold reaches the preset condition, taking the minimum delivery time as the logistics delivery element; when the delivery time threshold does not reach the preset condition, taking the minimum delivery cost as the logistics delivery element.

[0008] In a possible implementation, the inventory distribution information acquisition module also performs the following processing: obtaining the warehouse location of the regional distribution warehouse to construct a warehouse distribution map; obtaining the in-stock inventory information and in-transit inventory information of the regional distribution warehouse, and using the in-stock inventory information and in-transit inventory information to generate an inventory dynamic time series chain; fitting the inventory dynamic time series chain to the warehouse distribution map; obtaining the commodity inventory information of the offline store configuration warehouse and the store's inventory digestion information; configuring unsaleable inventory according to the inventory digestion information and commodity inventory information, sending an inventory sharing request based on the unsaleable inventory, and according to the inventory sharing request, adding the offline store configuration warehouse and its unsaleable inventory to the warehouse distribution map to obtain the inventory distribution information.

[0009] In a possible implementation, the inventory distribution information acquisition module further performs the following processing: obtaining historical cases of slow-selling goods, and using the historical cases of slow-selling goods to train a time series evaluation model, wherein the time series evaluation model includes the time series sales information of goods, the time series characteristics of goods attribute labels, the probability of slow-selling and the time series relationship of stagnation; obtaining the time of goods entering the warehouse, the inventory quantity of goods, and the attribute labels of goods according to the inventory information of goods; constructing the time series sales information of goods according to the time of goods entering the warehouse and the inventory digestion information; inputting the time series sales information of goods and the time series characteristics of the attribute labels of goods into the time series evaluation model for evaluation, and obtaining the predicted probability of slow-selling and the predicted time series relationship of stagnation; determining the slow-selling goods and their slow-selling inventory quantity according to the predicted probability of slow-selling and the time series relationship of stagnation, the time of goods entering the warehouse, and the inventory quantity of goods, and configuring the slow-selling inventory.

[0010] In a possible implementation, the logistics distribution scheduling plan acquisition module also performs the following processing: taking the order timing chain node block as the evaluation target, and taking the maximum comprehensive result of the logistics distribution elements of the node as the target, constructing an evaluation function; taking the cost of changing the demand constraint information as a constraint condition; based on the constraint condition and the evaluation function, constructing an optimization space, and optimizing the node distribution strategy through the optimization space according to the preset optimization rules to obtain a set of optimal solutions; taking the multi-channel store information as a participant, establishing a benefit function for each participant; based on the benefit function, using the Nash equilibrium algorithm, searching the set of optimal solutions to obtain the logistics distribution scheduling plan.

[0011] In a possible implementation, the decomposition parameter operation module further performs the following processing: the expression for constructing the evaluation function is:

[0012] , ,

[0013] in, is the deviation adjustment factor, , q is the number of nodes in the order timing chain node area, H is the evaluation value of any node, m is the number of orders with the minimum delivery time as the logistics distribution factor in a node in the order timing chain node block, n is the number of orders with the minimum delivery cost as the logistics distribution factor in a node in the order timing chain node block, is the delivery time evaluation function of order i based on decision variable x, is the delivery cost evaluation function of order j according to decision variable x, and , The output result is a normalized value. is the number of nodes in the order timing chain node block, The mean of the Euclidean distances of decision variables between orders with the minimum delivery time as the logistics distribution factor.

[0014] Through the logistics distribution optimization system based on multi-source data analysis proposed in this application, it is planned to obtain multi-source data of users; extract logistics parameters, obtain logistics orders, commodity codes, order commodity quantities, and order demand time; obtain inventory distribution information; establish a demand mapping relationship between the logistics parameters and inventory distribution information; mine logistics distribution elements according to the demand mapping relationship, and determine logistics distribution optimization variables; based on the logistics distribution optimization variables, optimize logistics distribution with the goal of maximizing the comprehensive results of logistics distribution elements, and obtain logistics distribution scheduling plans. It solves the technical problems of the existing logistics distribution that it is difficult to integrate and analyze logistics data from different channels, which leads to inaccurate inventory layout and distribution strategies, resulting in poor logistics distribution efficiency and high distribution costs, realizes comprehensive optimization and precise control of logistics distribution, and achieves the technical effect of improving logistics distribution efficiency and reducing distribution costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0016] Figure 1 A schematic diagram of the structure of a logistics distribution optimization system based on multi-source data analysis provided in an embodiment of the present application;

[0017] Figure 2 A schematic diagram of the execution process of an inventory distribution information acquisition module in a logistics distribution optimization system based on multi-source data analysis provided in an embodiment of the present application.

[0018] Explanation of the reference numerals: user multi-source data acquisition module 10, logistics parameter extraction module 20, inventory distribution information acquisition module 30, mapping relationship establishment module 40, logistics distribution optimization variable determination module 50, logistics distribution scheduling plan acquisition module 60. DETAILED DESCRIPTION

[0019] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0020] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0021] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0022] The present application embodiment provides a logistics distribution optimization system based on multi-source data analysis, such as Figure 1 As shown, the system comprises:

[0023] The user multi-source data acquisition module 10 is used to acquire the user's multi-source data, and the multi-source data comes from the user's multi-channel store information, and the multi-channel store information includes online attribute tags and offline attribute tags. The user's multi-source data refers to all relevant information of the purchased order and the multi-channel logistics information of the order delivery, etc. The online attribute tag refers to the tag set for the goods, stores or user behavior on the e-commerce platform, which helps the platform to classify and recommend goods, and conduct user portrait analysis, including but not limited to product attribute tags, store attribute tags, user behavior tags, etc., among which, for example, product attribute tags such as color, size, material, brand, price range, etc. are used to describe the physical or market characteristics of the goods; such as store level, praise rate, sales ranking, store opening time, etc., are used to evaluate the comprehensive strength and credibility of the store; such as browsing history, purchase record, search keywords, etc. User behavior tags are used to analyze the user's shopping preferences and habits. Offline attribute labels are more related to the operation and management of physical stores, and are of great value for understanding user behavior and market trends. They include but are not limited to store location labels, store image labels, marketing activity labels, etc. For example, business district, street, floor, etc., are used to describe the geographical location and surrounding environment of the store; such as decoration style, display method, service attitude, etc., are used to evaluate the physical image of the store and customer experience; such as promotional activities, new product launches, membership days, etc., are used to describe the store's market strategy and marketing activities.

[0024] The logistics parameter extraction module 20 is used to extract logistics parameters from the multi-source data to obtain logistics orders, commodity codes, order commodity quantities, and order demand time. Logistics parameters are extracted from multi-source data to obtain logistics parameters such as logistics orders, product codes, order product quantity, and order demand time. Specifically, a logistics order is an instruction containing product delivery information, which is usually initiated by a buyer or merchant, indicating the products to be delivered, the destination, the delivery method, etc., including the order number, sender information, recipient information, delivery address, delivery method (such as express delivery, ordinary mail, self-pickup, etc.), freight, etc.; the product code (or product barcode) is a barcode composed of Arabic numerals or letters, which is used to identify the brand, origin, specifications, purpose and other unique information of the product, including the product code itself, which is usually used for product identification, inventory management, sales statistics, etc.; the order product quantity refers to the number of products that need to be delivered in a certain logistics order, including the quantity of each product and the total number of all products in the order; the order demand time refers to the length of time required for the customer's appointment order or expedited order, and is also the deadline for logistics services, including the latest delivery time of the order and the customer's expected delivery time.

[0025] The inventory distribution information acquisition module 30 is used to acquire inventory distribution information. The inventory distribution information is used to describe the storage distribution of user goods, including warehouse attributes, warehouse locations, product codes and corresponding inventory quantities, wherein the warehouse attributes include regional distribution warehouses and offline store configuration warehouses. The inventory distribution information describes the storage distribution of user goods, and usually contains a series of detailed data related to the storage and location of goods, including warehouse attributes, warehouse locations, product codes and corresponding inventory quantities, wherein the warehouse attributes include regional distribution warehouses and offline store configuration warehouses. Specifically, regional distribution warehouses are usually used for large-scale logistics distribution and may be located in central areas with convenient transportation, such as logistics hubs or locations close to major transportation routes. Regional distribution warehouses usually store a large number of goods in order to quickly respond to orders from different regions and efficiently transport them over long distances. The main goal is to optimize logistics costs and improve distribution efficiency; offline store configuration warehouses are usually associated with specific retail stores and are used to store the goods required by the store. Goods may be located near or at the same location as a store in order to quickly replenish the store. The inventory volume of a warehouse configured for an offline store may vary depending on factors such as the size of the store, sales, seasonal demand, etc. Warehouse location refers to the specific geographical coordinates or address of the warehouse, which is crucial for logistics management and transportation planning, and determines the transportation route and cost of goods from the warehouse to the end consumer; the product code is a code used to uniquely identify a product, usually a number, letter, or a combination of numbers and letters, used to track goods in inventory management systems, sales systems and other related systems; the corresponding inventory volume refers to the storage quantity of each product in each warehouse, which determines whether the company can meet customer needs, maintain production operations and optimize costs.

[0026] A mapping relationship establishment module 40 is used to perform time alignment of logistics orders based on the logistics parameters of the multi-source data and the order demand time, and to establish a demand mapping relationship between the logistics parameters and the inventory distribution information based on the product code, the product quantity, and the delivery address of the logistics order. In the logistics and distribution process, by integrating and analyzing multi-source data, it is ensured that the order is processed at the correct time point, and the matching inventory distribution information is found according to the product code, the product quantity and the delivery address of the logistics order, so as to effectively meet customer needs. Specifically, when processing logistics orders, the demand time of the order is a key factor. This time point determines when the order needs to be processed, when it is delivered, and when it reaches the customer. Time alignment means adjusting and optimizing the logistics process according to the demand time of the order to ensure that the order can be processed and delivered on time, such as warehouse operations, transportation arrangements, distribution routes, and other links. adjustment and optimization to ensure the coordination and efficiency of the entire logistics chain; by establishing the demand mapping relationship between logistics parameters and inventory distribution information based on commodity codes, commodity quantities, and the delivery address of logistics orders, we can accurately understand which commodities need to be shipped from where, how many, and where the delivery destination is, so as to achieve accurate inventory management and rapid response. When an order is generated, it can automatically find the corresponding warehouse and inventory level in the inventory distribution information according to the commodity code and quantity, and generate the corresponding distribution plan. At the same time, according to the delivery address, the optimal distribution route and transportation method can be calculated, thereby further improving distribution efficiency and reducing transportation costs.

[0027] The logistics distribution optimization variable determination module 50 performs logistics distribution element mining based on the online attribute tags and offline attribute tags and the demand mapping relationship to determine the logistics distribution optimization variables. Based on the analysis of online and offline attribute tags and the demand mapping relationship, the logistics distribution element mining is performed to determine a series of logistics distribution optimization variables, such as vehicle scheduling, path planning, distribution strategy, inventory allocation, etc. Specifically, vehicle scheduling refers to determining the number and type of required vehicles and the optimal vehicle allocation plan according to order requirements, transportation routes and distribution methods; vehicle scheduling refers to determining the number and type of required vehicles and the optimal vehicle allocation plan according to order requirements, transportation routes and distribution methods; distribution strategy refers to strategies including distribution time windows, distribution quantity restrictions, batch distribution, etc., to optimize distribution efficiency and cost; inventory allocation refers to determining which warehouse to ship from, the quantity and priority of shipments according to order requirements and inventory distribution. Optimization of logistics distribution is achieved, distribution efficiency is improved, costs are reduced and customer satisfaction is improved. By integrating multi-source data, customer needs can be more accurately predicted and met, and intelligent and precise logistics distribution can be achieved.

[0028] The logistics distribution scheduling scheme acquisition module 60 is based on the logistics distribution optimization variables, takes the maximum comprehensive result of logistics distribution elements as the goal, performs logistics distribution optimization, and obtains a logistics distribution scheduling scheme. By integrating and utilizing multi-source data, combining the elements of logistics distribution and optimization variables, an efficient and cost-effective logistics distribution scheduling scheme is formulated. Specifically, the maximum comprehensive result of logistics distribution elements refers to the optimization goal of taking the comprehensive results of multiple aspects such as the lowest cost, the highest efficiency, and the highest customer satisfaction. By setting these goals, it can be ensured that the logistics distribution scheduling scheme can maximize the economic and social benefits of the enterprise while meeting customer needs, and optimize the route planning, vehicle scheduling, and storage point layout of logistics distribution. Specifically, route planning optimization refers to combining real-time traffic information and data analysis to formulate the best distribution route, including the shortest path , least traffic congestion and least number of transfers, etc.; vehicle scheduling optimization refers to the reasonable arrangement of vehicle departure time, route and loading to minimize empty load rate and mileage, and reduce transportation costs; storage point layout optimization refers to the reasonable planning of the location and number of storage points according to product sales volume, transportation network and customer needs, so that the distribution distance is the shortest and the service coverage is the widest, reducing the number and time of transfers; finally, the logistics distribution scheduling plan is obtained, including logistics delivery time, distribution vehicles, distribution routes, etc. During the actual implementation process, through real-time data analysis, necessary adjustments and optimizations are made to the distribution scheduling plan to ensure the smooth progress of the distribution process and achieve the set optimization goals.

[0029] The logistics distribution optimization system based on multi-source data analysis according to the embodiment of the present invention is used to solve the technical problem that it is difficult to integrate and analyze logistics data from different channels in existing logistics distribution, which leads to inaccurate inventory layout and distribution strategy, resulting in poor logistics distribution efficiency and high distribution costs, and realizes comprehensive optimization and precise control of logistics distribution, achieving the technical effect of improving logistics distribution efficiency and reducing distribution costs. The logistics distribution optimization system based on multi-source data analysis includes: a user multi-source data acquisition module 10, a logistics parameter extraction module 20, an inventory distribution information acquisition module 30, a mapping relationship establishment module 40, a logistics distribution optimization variable determination module 50, and a logistics distribution scheduling solution acquisition module 60.

[0030] The specific configuration of the mapping relationship establishment module 40 will be described in detail below. The mapping relationship establishment module 40 may further include: setting the order start time node according to the order demand time, aligning the order start time node according to time, and outputting the order timing chain node block, the order timing chain node block including all order parameters of each time node and the online attribute label and offline attribute label. Setting the order start time node according to the order demand time and aligning the order start time node according to time is to build an orderly, time-arranged order sequence, so as to better manage and track the order processing progress and status. Specifically, the order start time node is used to mark the time point when the order starts to be processed or delivered. The setting of the start time node can be adjusted according to the actual situation. For example, considering the order processing time, transportation time and other factors, all orders are sorted according to their start time nodes to ensure that the orders are arranged in chronological order. If multiple orders have the same start time node, they can be further sorted according to other factors (such as order priority, customer importance, etc.) to obtain an order timing chain, which is an order arranged in chronological order. Sequence, each order is associated with its starting time node. A node block refers to one or more orders corresponding to each time node in the order timing chain. These orders and their related parameters (such as product code, quantity, delivery address, etc.) and attribute tags (online attribute tags, offline attribute tags) are combined into a node block. The node block contains all order parameters at each time node, such as order number, product code, quantity, delivery address and other basic information; online attribute tags, such as product attribute tags (type, size, weight, etc.), store attribute tags (level, reputation, etc.), user behavior tags (purchase records, search habits, etc.); offline attribute tags, such as warehouse attribute tags (geographic location, storage capacity, etc.), marketing activity tags (promotional activities, new product releases, etc.).

[0031] The mapping relationship establishment module 40 also includes, based on the order timing chain node block, for all order parameters of the time node, the first demand matching condition is that the commodity code and the corresponding commodity quantity of the logistics order meet the warehouse distribution quantity, and the second demand matching condition is that the distance between the delivery address of the logistics order and the warehouse location is the minimum. Extract the order parameters of each time node in the order timing chain node block, including the commodity code and the corresponding quantity, consult the inventory data of each warehouse, understand the current inventory quantity of each commodity code, compare, confirm whether there is a warehouse that can meet the commodity code and quantity requirements of the order, if there are multiple warehouses that can meet the requirements, enter the second demand matching condition for further screening; extract the delivery address of each order in the order timing chain node block, use geographic location data or map API, calculate the actual distance between each delivery address and each warehouse, sort according to the distance, select the warehouse with the smallest distance as the distribution warehouse, if the distance is the same or the difference is not large, other factors (such as the distribution capacity of the warehouse, transportation cost, etc.) can be considered for further screening. The method further includes screening the demand matching relationship in accordance with the first demand matching condition and the second demand matching condition in turn, and establishing a demand mapping relationship between the logistics parameters and the inventory distribution information. Screening the demand matching relationship in accordance with the first demand matching condition and the second demand matching condition in turn, determining the optimal distribution warehouse corresponding to each order, and then establishing a demand mapping relationship between the logistics parameters (including order number, product code, quantity, delivery address, etc.) and the inventory distribution information (including warehouse number, product code, inventory quantity, warehouse location, etc.), clearly indicating which warehouse each order needs to be shipped from, the quantity shipped, the delivery address, and other information.

[0032] The specific configuration of the mapping relationship establishment module 40 will be described in detail below. The mapping relationship establishment module 40 may further include: obtaining the historical logistics data of the multi-channel store information respectively. Obtain the historical logistics data of the store information of multiple channels such as online e-commerce platforms and offline physical stores, for example, order information, shipment information, distribution information, etc. It also includes predicting the logistics peak period according to the historical logistics data, and determining the predicted logistics peak information of the store information of each channel. There are different logistics peak periods for online and offline promotions. For the peak period, in order to ensure the normal logistics, it is necessary to make arrangements and reservations for inventory distribution in advance. Therefore, when optimizing logistics distribution, the time period should be fully considered to ensure the reserved amount during the peak period. It also includes, based on the predicted logistics peak information of the store information of each channel, the inventory constraints of the inventory distribution information are performed according to the peak time, store location, and predicted logistics volume, and the inventory distribution plan time chain is constructed. By collecting and analyzing the historical logistics data of multi-channel stores, the time, store location and predicted logistics volume of future logistics peaks are predicted, and then inventory constraints are imposed on the inventory distribution information based on these prediction results to ensure that customer needs can be efficiently and accurately met during the logistics peak period. Specifically, peak time refers to the logistics peak period, such as holidays, promotional activities, etc.; according to the geographical location and radiation range of each channel store, combined with historical sales data and logistics data, analyze which stores will have a significant increase in logistics volume during the logistics peak period. Special attention is paid to those stores located in densely populated areas, commercial centers or transportation hubs; based on historical logistics data, sales data and market trends, the logistics volume of each store during the logistics peak period is predicted. According to the predicted logistics peak time, store location and predicted logistics volume, inventory constraints are imposed on the inventory distribution information, including determining the inventory level, replenishment plan and distribution strategy of each warehouse, to ensure that customer needs can be met in a timely and accurate manner during the logistics peak period. After completing the inventory constraints, constructing the inventory distribution plan time chain includes arranging the inventory distribution plan in chronological order to form an orderly time chain, the starting point of the time chain is the start time of the inventory distribution plan, and the end point is the end time. In the time chain, each time node corresponds to one or more inventory distribution plan tasks, such as replenishment, transfer, and delivery. It also includes adding the inventory distribution plan time chain as demand constraint information to the demand mapping relationship.

[0033] The specific configuration of the mapping relationship establishment module 40 will be described in detail below. The mapping relationship establishment module 40 may further include: setting the delivery speed tolerance of logistics orders with online attributes and offline attributes. Setting the delivery speed tolerance of logistics orders with online attributes and offline attributes refers to setting an acceptable tolerance range or time flexibility for the delivery speed of logistics orders according to the characteristics of different online and offline sales channels and customer needs, aiming to ensure the efficiency and accuracy of logistics services, while taking into account the specific needs of different sales channels and customer groups. Specifically, the delivery speed tolerance of online attribute logistics orders refers to setting an acceptable tolerance range for the delivery speed of orders placed through e-commerce platforms or other online platforms in online sales channels; the delivery speed tolerance of offline attribute logistics orders refers to setting an acceptable tolerance range for the delivery speed of orders placed by customers in offline sales channels, such as physical stores, wholesale markets, etc. It also includes extracting attribute tags of commodities, wherein the attribute tags include product categories, popularity, and seasonal restrictions. Product category refers to the general classification or type of a product. For example, a product may belong to the smartphone subcategory under the electronics category, or the coat subcategory under the clothing category. Popularity is usually used to describe the sales performance of a product in the market. It can be specific sales data (such as sales volume, sales revenue) or a rating based on sales data (such as hot-selling, ordinary, slow-selling, etc.). Seasonal restriction means that the sales or use of a product is affected by seasonal changes. For example, sales of products such as swimsuits and ice cream are usually higher in summer, while products such as down jackets and hot packs are more popular in winter.

[0034] The mapping relationship establishment module 40 also includes configuring the evaluation coefficient of the attribute label, evaluating the delivery speed of the attribute label according to the evaluation coefficient, and obtaining the commodity attribute delivery constraint. The evaluation coefficient is used to measure the influence of commodity attributes (such as product category, popularity, and seasonal restrictions) on the delivery speed. The attribute label (product category, popularity, and seasonal restrictions) of each commodity is matched with the evaluation coefficient. According to the specific attributes of the commodity and the corresponding evaluation coefficient, a comprehensive delivery speed evaluation score is calculated to reflect the degree of demand for the delivery speed of the commodity; the commodity attribute delivery constraint refers to the restriction set on the delivery speed under the premise of meeting the commodity attributes and customer needs. Obtaining the constraint includes setting the corresponding delivery speed threshold or range according to the comprehensive delivery speed evaluation score of the commodity; for commodities exceeding the delivery speed threshold or range, corresponding delivery priority, delivery method or delivery time and other constraints can be set to ensure that the commodity can be delivered to the customer on time and according to quality. It also includes fusing the delivery speed tolerance of the logistics order and the commodity attribute delivery constraint to obtain the delivery time threshold of each order in the demand mapping relationship. A preliminary evaluation is conducted on each order based on the tolerance of logistics order delivery speed and the delivery constraints of commodity attributes. Based on the preliminary evaluation results, an initial delivery time threshold is set for each order, which can both meet customer needs and ensure logistics efficiency and cost control. The delivery time threshold refers to the maximum limit on the time from order confirmation to the delivery of goods to the customer, while satisfying the tolerance of logistics order delivery speed and the delivery constraints of commodity attributes.

[0035] The mapping relationship establishment module 40 also includes that when the delivery time threshold reaches the preset condition, the delivery time is minimized as a logistics distribution element. In a logistics distribution decision or optimization process, if the delivery time threshold of an order is triggered or meets a certain preset condition (such as time is tight, the customer has strict time requirements, etc.), then the delivery time will become the primary or key factor in determining the logistics distribution plan or strategy. Specifically, when planning or executing logistics distribution, priority is given to minimizing the delivery time to ensure that the order can be delivered as soon as possible. It also includes that when the delivery time threshold does not meet the preset condition, the delivery cost is minimized as a logistics distribution element. In the decision or optimization process of logistics distribution, if the delivery time of an order is not so urgent that it needs to be given priority (that is, the preset condition of the delivery time threshold is not met), then the system will take reducing the delivery cost as the main optimization goal. Specifically, when planning or executing logistics distribution, the system will give priority to minimizing the delivery cost, including but not limited to transportation costs, warehousing costs, human resource costs, etc.

[0036] The specific configuration of the inventory distribution information acquisition module 30 will be described in detail below. Figure 2As shown, the inventory distribution information acquisition module 30 may further include: obtaining the warehouse location of the regional distribution warehouse to construct a warehouse distribution map; fitting the inventory dynamic time series chain to the warehouse distribution map. The specific location (latitude and longitude information) of the warehouse is obtained through a geographic information system (GIS) or a related map service, and the obtained warehouse location information is displayed in the form of a map, which is convenient for intuitively understanding the distribution of the warehouse, clearly showing the geographical characteristics of the region where the warehouse is located, and integrating the inventory dynamic time series chain data with the warehouse location data to ensure that the inventory data of each warehouse corresponds to its location on the map. It also includes obtaining the in-stock inventory information and in-transit inventory information of the regional distribution warehouse, and generating an inventory dynamic time series chain using the in-stock inventory information and in-transit inventory information. The in-stock inventory information refers to the remaining inventory dynamically adjusted according to the delivery order of the real-time logistics distribution scheduling plan, integrating the collected in-stock and in-transit inventory information, and using time series analysis technology to arrange the integrated inventory data in chronological order to obtain an inventory dynamic time series chain.

[0037] The inventory distribution information acquisition module 30 also includes obtaining the commodity inventory information of the offline store configuration warehouse and the store's inventory digestion information; configuring the unsaleable inventory according to the inventory digestion information and the commodity inventory information, sending an inventory sharing request based on the unsaleable inventory, and according to the inventory sharing request, adding the offline store configuration warehouse and its unsaleable inventory to the warehouse distribution map to obtain the inventory distribution information. Inventory digestion information reflects the sales speed and inventory turnover rate of goods, which is usually obtained by analyzing sales data, including product sales volume, sales frequency, inventory turnover rate, etc. The product inventory information is collected from the warehouse system of each offline store, usually including product name, product quantity, storage location, etc. By comparing the sales data and inventory volume of different products, it is possible to identify which products sell quickly (best-selling products) and which products sell slowly (slow-moving products). Configuring slow-moving inventory means classifying and marking these products. For slow-moving inventory, inventory costs can be reduced and inventory turnover rate can be improved by sharing inventory with other stores or warehouses. Based on the information of slow-moving inventory, an inventory sharing request can be generated and sent to potential sharing parties through an internal system or an external platform. After identifying the slow-moving inventory and sending the inventory sharing request, the warehouse location of these offline stores and the slow-moving inventory information can be added to the warehouse distribution map, and finally the inventory distribution information is obtained, including the geographical location of each warehouse, inventory volume, inventory type (such as best-selling products, slow-moving products), and inventory status (such as sharing, waiting to be shared, etc.).

[0038] The specific configuration of the inventory distribution information acquisition module 30 will be described in detail below. The inventory distribution information acquisition module 30 may further include: obtaining historical cases of unsalable goods, and using the historical cases of unsalable goods to train a time series evaluation model, wherein the time series evaluation model includes time series sales information of goods, time series characteristics of goods attribute labels, probability of unsalability, and time series relationship of stagnation. Historical cases of unsalable goods refer to cases of goods that have poor sales and inventory backlogs due to various reasons in the past period of time. Specifically, historical sales data is extracted from a sales database or an inventory management system to identify unsalable goods, such as goods whose sales volume is below a certain threshold for a long time (such as three consecutive months or more). The time series evaluation model is a prediction model based on time series analysis, which is used to evaluate the unsalable risk of goods and predict the unsalable trend. The historical cases of unsalable goods and related sales data are divided into training sets and test sets. According to the requirements of the model, the original data is preprocessed and feature extracted. The model is constructed based on a suitable time series analysis algorithm (such as ARIMA, LSTM, etc.) to capture the time series dependency and periodic changes in the data. The model is trained using training data, and the model parameters are adjusted to optimize the prediction performance. Use the test set to evaluate the trained model; the product time series sales information reflects the sales data of the product over time, such as historical sales volume, sales revenue, sales growth rate, etc., which can help the model understand the sales trend and cyclical changes of the product; the product attribute label time series characteristics refer to the characteristics of the product attribute labels (such as brand, category, price, quality, etc.) changing over time, which can reveal the impact of changes in product attributes on sales, thereby helping the model to more accurately predict the risk of unsalable products; the probability of unsalable products refers to the main prediction result output by the model, indicating the possibility of unsalable products in a certain period of time in the future. The calculation of the probability of unsalable products is usually based on the statistical analysis of historical unsalable cases and current sales data; the stagnation time series relationship describes the time series pattern of stagnant or declining product sales. By identifying these patterns, the model can better predict the unsalable trend and timing of products.

[0039] The inventory distribution information acquisition module 30 also includes obtaining the commodity entry time, commodity inventory, and commodity attribute tags according to the commodity inventory information; and constructing the commodity time series sales information according to the commodity entry time and inventory digestion information. Key data is obtained from the commodity inventory information, including the commodity entry time, commodity inventory, and commodity attribute tags, and the commodity time series sales information is constructed, wherein the commodity time series sales information refers to the sequence of commodity sales data changing over time, reflecting the sales dynamics of the commodity. Specifically, first, a time range (such as the past year, a quarter, a month, etc.) needs to be determined, and this time range is divided into several time points (such as every day, every week, every month, etc.), and the sales data of the commodity is mapped to the corresponding time point according to the commodity entry time and inventory digestion information (such as sales volume, sales frequency, etc.). When constructing the time series sales information, the commodity attribute tags can also be considered as additional information. For example, the time series sales information can be constructed separately according to the different attributes of the commodity (such as brand, model, etc.).

[0040] The inventory distribution information acquisition module 30 also includes inputting the time series sales information of the commodity and the time series characteristics of the commodity attribute label into the time series evaluation model for evaluation to obtain the time series relationship between the prediction probability of slow sales and the prediction of stagnation; determining the slow-selling commodities and their slow-selling inventory according to the time series relationship between the prediction probability of slow sales and the prediction of stagnation, the commodity entry time, and the commodity inventory, and configuring the slow-selling inventory. The prediction probability of unsalable refers to the possibility that the model calculates the commodity in a certain period of time in the future based on the input data. The stagnation prediction time series relationship refers to the time series pattern of the model predicting the stagnation or decline of commodity sales, that is, predicting how long the unsalable state may last or when it will begin. Specifically, a threshold is set according to the prediction probability of unsalable output by the model. When the prediction probability of unsalable of a commodity exceeds this threshold, the commodity is considered to be a potential unsalable commodity. For commodities considered to be potential unsalable commodities, the unsalable inventory is further determined in combination with the commodity entry time and the current inventory, including the time division of the inventory, identifying those parts that have not been sold for a long time or the sales speed is extremely slow as unsalable inventory. After determining the unsalable commodities and their unsalable inventory, they are configured and managed, such as adjusting sales strategies (such as price reduction promotions, bundling sales, etc.), optimizing inventory placement (such as placing unsalable commodities in a more conspicuous position), or negotiating with suppliers for returns or exchanges, etc., in order to reduce inventory backlogs as soon as possible, improve inventory turnover, and reduce inventory costs.

[0041] The specific configuration of the logistics distribution scheduling scheme acquisition module 60 will be described in detail below. The logistics distribution scheduling scheme acquisition module 60 may further include: taking the order timing chain node block as the evaluation target, and taking the maximum comprehensive result of the node's logistics distribution elements as the target, to construct an evaluation function. Taking the maximization of the comprehensive result of the node's logistics distribution elements as the goal, that is, hoping that the performance of each node block in the logistics distribution elements is as good as possible, an evaluation function is constructed to quantitatively evaluate the comprehensive performance of the node block in the logistics distribution elements. It also includes taking the cost of changing the demand constraint information as a constraint condition; based on the constraint condition and the evaluation function, constructing an optimization space, optimizing the node distribution strategy through the optimization space according to the preset optimization rules, and obtaining the optimal solution to be selected. The optimization space is a set of all feasible solutions. In the optimization of logistics distribution strategies, the optimization space contains all distribution strategies that meet the constraints. First, the feasible domain of the solution is determined according to the constraints. Then, according to the characteristics and needs of the problem, all distribution strategies that meet the constraints are listed, which together constitute the optimization space. The preset optimization rules are a set of rules and strategies used to guide the search for the optimal solution in the optimization space. According to the preset optimization rules, the optimal distribution strategy is searched in the optimization space until the optimal solution that meets the conditions is found or the preset termination conditions are reached. Finally, a set of potential optimal solutions is obtained, that is, the optimal solution to be selected.

[0042] The logistics distribution scheduling scheme obtaining module 60 also includes taking the multi-channel store information as a participant and establishing a benefit function for each participant; based on the benefit function, using the Nash equilibrium algorithm, searching the optimal solution set to be selected to obtain the logistics distribution scheduling scheme. The multi-channel store information is regarded as a participant in the game, including different stores, suppliers, logistics service providers, etc., each of which has different resources and goals. A benefit function is defined for each participant to describe the benefit or utility of the participant under a specific strategy combination. The benefit function may be based on multiple factors, such as transportation cost, time efficiency, customer satisfaction, etc. Nash equilibrium is a strategy combination in which each participant's strategy is the best response to other participants' strategies. Nash equilibrium means that each participant has chosen a strategy that maximizes his or her own benefits, and no participant is willing to unilaterally change his or her strategy. Using the Nash equilibrium algorithm to search the optimal solution set means finding a solution that satisfies the Nash equilibrium conditions among the existing feasible solutions. Specifically, a strategy combination is selected from the optimal solution set as the initial solution. For each participant, its benefits under the current strategy combination are evaluated, and new strategies that can improve its benefits are found. If all participants cannot improve their benefits by unilaterally changing their strategies, the current strategy combination reaches Nash equilibrium, which is the desired logistics distribution scheduling solution; otherwise, the strategy combination is updated and the search continues.

[0043] The specific configuration of the logistics distribution scheduling scheme obtaining module 60 will be described in detail below. The logistics distribution scheduling scheme obtaining module 60 may further include: the expression for constructing the evaluation function is:

[0044] , ,

[0045] in, is the deviation adjustment factor, , q is the number of nodes in the order timing chain node area, H is the evaluation value of any node, m is the number of orders with the minimum delivery time as the logistics distribution factor in a node in the order timing chain node block, n is the number of orders with the minimum delivery cost as the logistics distribution factor in a node in the order timing chain node block, is the delivery time evaluation function of order i based on decision variable x, is the delivery cost evaluation function of order j according to decision variable x, and , The output result is a normalized value. is the number of nodes in the order timing chain node block, The mean of the Euclidean distances of decision variables between orders with the minimum delivery time as the logistics distribution factor. is the calculation formula of Euclidean distance, which is used to calculate the delivery time. Order and Adjacency Order distance, deviation adjustment factor It is the average silhouette coefficient when dividing orders. The average silhouette coefficient is a value between 0 and 1. When the average silhouette coefficient is 1, it means that the result of order division is more reliable, and the corresponding deviation adjustment coefficient is 1, that is, the order division does not need to be adjusted; on the contrary, when the average silhouette coefficient is close to 0, it means that the error of order division is the largest; when judging based on time, there may be a certain error, and the deviation adjustment coefficient is used Make adjustments.

[0046] In summary, the technical effects achieved by this application are:

[0047] 1. Through the integration of multi-source data and the extraction of logistics parameters, the system can more accurately predict and meet order requirements, reduce waiting and delays in the delivery process, and thus improve the overall logistics and delivery efficiency;

[0048] 2. Based on real-time knowledge of inventory distribution information and optimized logistics distribution scheduling, the system can reduce unnecessary transportation and storage costs and reduce overall logistics costs;

[0049] 3. Based on the logistics parameters and inventory distribution information of multi-source data, the system can establish a demand mapping relationship, mine logistics distribution elements according to online and offline attribute labels, and formulate logistics distribution scheduling plans. It has strong adaptability and flexibility, and can adapt to the needs of goods and orders from different channels and with different attributes, thereby meeting the ever-changing market demand.

[0050] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0051] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. Logistics distribution optimization system based on multi-source data analysis, characterized by: The system comprises: A user multi-source data acquisition module, which is used to acquire the user's multi-source data, the multi-source data comes from the user's multi-channel store information, and the multi-channel store information includes online attribute tags and offline attribute tags; A logistics parameter extraction module, which is used to extract logistics parameters from the multi-source data to obtain logistics orders, commodity codes, order commodity quantities, and order demand time; An inventory distribution information acquisition module, which is used to acquire inventory distribution information. The inventory distribution information is used to describe the storage distribution of user products, including warehouse attributes, warehouse locations, product codes, and corresponding inventory quantities. The warehouse attributes include regional distribution warehouses and offline store configuration warehouses. A mapping relationship establishment module, the mapping relationship establishment module is used to perform time alignment of logistics orders based on the logistics parameters of the multi-source data with the order demand time, and establish a demand mapping relationship between the logistics parameters and the inventory distribution information with the commodity code, commodity quantity, and the delivery address of the logistics order; A logistics distribution optimization variable determination module, which mines logistics distribution elements based on the online attribute labels and offline attribute labels and the demand mapping relationship to determine logistics distribution optimization variables; A logistics distribution scheduling scheme obtaining module, which performs logistics distribution optimization based on the logistics distribution optimization variables and takes the maximum comprehensive result of logistics distribution elements as the goal to obtain a logistics distribution scheduling scheme; It is characterized in that the mapping relationship establishment module performs the following steps: An order start time node is set according to the order demand time, and the order start time node is aligned according to time, and an order timing chain node block is output, wherein the order timing chain node block includes all order parameters of each time node and the online attribute label and offline attribute label; Based on the order time chain node block, for all order parameters of the time node, the first requirement matching condition is that the commodity code and the corresponding commodity quantity of the logistics order meet the warehouse distribution quantity, and the second requirement matching condition is that the distance between the delivery address of the logistics order and the warehouse location is the minimum; According to the first demand matching condition and the second demand matching condition, the demand matching relationship is screened in sequence to establish a demand mapping relationship between the logistics parameters and the inventory distribution information; The mapping relationship establishing module executes the following steps: Respectively obtain historical logistics data of the multi-channel store information; Perform logistics peak period forecasting based on the historical logistics data to determine forecast logistics peak period information for each channel store information; Based on the predicted logistics peak information of the store information of each channel, inventory constraints of inventory distribution information are performed according to peak time, store location, and predicted logistics volume, and a time series chain of inventory distribution plan is constructed; Adding the inventory distribution plan time sequence chain as demand constraint information into the demand mapping relationship; The mapping relationship establishing module executes the following steps: Set the delivery speed tolerance of logistics orders with online and offline attributes; Extracting attribute tags of commodities, wherein the attribute tags include product category, popularity, and seasonal restrictions; configuring an evaluation coefficient of the attribute tag, evaluating the delivery speed of the attribute tag according to the evaluation coefficient, and obtaining a delivery constraint condition of the commodity attribute; According to the delivery speed tolerance of the logistics order and the delivery constraint conditions of the commodity attributes, the delivery time threshold of each order in the demand mapping relationship is obtained; When the delivery time threshold reaches the preset condition, the minimum delivery time is used as the logistics delivery element; When the delivery time threshold does not meet the preset condition, the minimum delivery cost is used as the logistics delivery factor; The steps performed by the inventory distribution information acquisition module include: Obtain the warehouse locations of regional distribution warehouses to build a warehouse distribution map; Acquire the in-stock inventory information and in-transit inventory information of the regional distribution warehouse, and generate an inventory dynamic time sequence chain using the in-stock inventory information and in-transit inventory information; Fitting the inventory dynamic time series chain into the warehouse distribution map; Obtain product inventory information of offline store configuration warehouses and store inventory digestion information; configuring unsalable inventory according to the inventory digestion information and the commodity inventory information, sending an inventory sharing request based on the unsalable inventory, and adding the offline store configured warehouse and its unsalable inventory to the warehouse distribution map according to the inventory sharing request to obtain the inventory distribution information; The logistics distribution scheduling plan acquisition module performs the following steps: Taking the order time chain node block as the evaluation target, taking the maximum comprehensive result of the logistics distribution elements of the node as the target, constructing the evaluation function; The cost of changing the demand constraint information is used as a constraint condition; Based on the constraints and evaluation functions, an optimization space is constructed, and node distribution strategies are optimized through the optimization space according to preset optimization rules to obtain an optimal solution set; Taking the multi-channel store information as participants, establishing a profit function for each participant; Based on the profit function, the Nash equilibrium algorithm is used to search the optimal solution set to obtain the logistics distribution scheduling plan.

2. The logistics distribution optimization system based on multi-source data analysis according to claim 1, characterized in that: The steps performed by the inventory distribution information acquisition module include: Obtain historical slow-selling commodity cases, and use the historical slow-selling commodity cases to train a time series evaluation model, wherein the time series evaluation model includes commodity time series sales information, commodity attribute label time series characteristics, slow-selling probability, and stagnation time series relationship; Obtaining the product entry time, product inventory, and product attribute labels according to the product inventory information; Constructing commodity time series sales information according to the commodity entry time and inventory digestion information; Input the time series sales information of the commodity and the time series characteristics of the commodity attribute label into the time series evaluation model for evaluation, and obtain the time series relationship between the prediction probability of unsalable goods and the prediction of stagnation; According to the slow-moving goods prediction probability and the stagnation prediction time series relationship, the goods entry time, and the goods inventory, the slow-moving goods and their slow-moving inventory are determined, and the slow-moving inventory is allocated.

3. The logistics distribution optimization system based on multi-source data analysis according to claim 1, characterized in that: The expression for constructing the evaluation function is: , , in, is the deviation adjustment factor, , q is the number of nodes in the order timing chain node area, H is the evaluation value of any node, m is the number of orders with the minimum delivery time as the logistics distribution factor in a node in the order timing chain node block, n is the number of orders with the minimum delivery cost as the logistics distribution factor in a node in the order timing chain node block, is the delivery time evaluation function of order i based on decision variable x, is the delivery cost evaluation function of order j according to decision variable x, and , The output result is a normalized value. is the number of nodes in the order timing chain node block, The mean of the Euclidean distances of decision variables between orders with the minimum delivery time as the logistics distribution factor.