Goods source collaborative delivery system based on dynamic planning and multi-model collaboration

Abstract

The invention discloses a goods source collaborative delivery system based on dynamic planning and multi-model collaboration, and the system comprises a retailer sales volume prediction module which is used for predicting the sales volume of a retailer; the market demand estimation module is used for predicting the market demand and outputting a prediction result; the production dynamic planning module is used for dynamically adjusting production plans for different products per month on the basis of annual production capacity requirements, finished production progress and market requirements of the State Tobacco Administration; and the goods source distribution planning module is based on a production plan result output by the production dynamic planning module, and the goods source collaborative delivery system based on dynamic planning and multi-model collaboration can effectively realize accurate planning of production, marketing and supply of tobacco goods sources and improve operation benefits.

Description

Technical Field

[0001] The present invention relates to a delivery system, and more particularly to a coordinated delivery system for goods sources based on dynamic programming and multi-model collaboration. Background Art

[0002] China's tobacco industry has a special management model and follows the national tobacco monopoly system. According to this system, the production and sales of cigarette products are subject to strict planning management and distribution control. In terms of process, commercial tobacco companies across the country submit procurement requirements to the industrial enterprises responsible for the brand based on market demand forecasts. Industrial enterprises will arrange specific production tasks based on overall planning and production capacity, as well as procurement needs. After the production is completed, the products will be distributed to commercial tobacco companies in various places through professional logistics companies according to the pre-determined quotas, and then distributed to retailers by commercial tobacco companies.

[0003] Although the cigarette launch model has introduced a variety of strategies to adapt to market changes, these models still have some shortcomings: (1) Data lag: Cigarette sales data and market demand feedback often have a certain lag, which may lead to launch decisions that cannot respond to rapid market changes in a timely manner. For example, some hot products may quickly become popular or outdated, and the traditional launch mechanism based on historical data may not be able to capture these trends in a timely manner. (2) Lack of personalized services: The existing launch model relies more on unified standards (such as gear and price range) and fails to fully consider the unique needs of different regions and different consumer groups. This may result in the specific needs of some retailers or consumers not being met, affecting consumer experience and satisfaction. (3) Limited flexibility: Although there are ways to flexibly adjust the launch strategy based on retailer feedback and market demand, in actual operation, this flexibility may be greatly limited due to factors such as approval processes and policy restrictions. In addition, focused launch may lead to excessive concentration of resources on a few retailers, which is not conducive to the healthy development of the overall market. (4) Poor matching between supply and demand: Although the method of releasing goods according to a fixed purchase quantity limit helps to control inventory risks, it may also cause some retailers to face a situation of insufficient supply or oversupply. In particular, for those retailers with strong management capabilities and good sales, a fixed purchase quantity limit may restrict their development potential.

[0004] In order to solve the technical problems existing in the above-mentioned background technology, the patent (CN115713289A) proposes a supply delivery method and system based on long short-term memory neural network, which solves the problem of insufficient data utilization and low value of existing supply delivery technology. Patent (CN117952716A) proposes a supply distribution method and system to ensure retailers' independent ordering options, improve customer satisfaction, and increase tobacco companies' sales. Patent (CN118469188A) A cigarette supply precision delivery system can accurately predict and analyze market demand by collecting and analyzing information such as retailer behavior and geographic location, and combine artificial intelligence technology to establish an intelligent cigarette supply delivery system. Using machine learning technology, the market consumption potential is predicted and analyzed, thereby providing accurate cigarette supply delivery recommendations.

[0005] However, although new delivery mechanisms are constantly being proposed, which can meet market demand to a certain extent, they are mainly delivery models that are adjusted according to the characteristics of cigarette products on the basis of conventional product delivery. They often focus on the optimization of a single node, fail to pay attention to the particularity of the tobacco industry itself, and ignore the global optimality of the entire delivery network, failing to achieve full-process coverage of the coordinated delivery of cigarette product sources.

[0006] According to the above-mentioned particularities of the tobacco industry, tobacco industrial enterprises, commercial tobacco companies and retailers need to coordinate with each other in the entire launch strategy to achieve efficient use of resources. Current research on launch strategies mainly focuses on the distribution of cigarette products by commercial enterprises to tobacco retailers, ignoring the allocation of goods from industrial enterprises to commercial enterprises, which in turn affects the adjustment of commercial enterprises' launch strategies for retailers. Due to the particularities of the tobacco industry, there is an information island phenomenon between industrial enterprises and commercial enterprises, a lack of effective communication channels between departments, and low efficiency in collaboration between upstream and downstream links, which leads to unreasonable inventory management. Summary of the invention

[0007] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a coordinated supply delivery system based on dynamic programming and multi-model collaboration, which corresponds to the replenishment link on the supply side and the sales link on the channel side respectively. The two links are promoted in parallel and cooperate with each other to balance inventory turnover and out-of-stock.

[0008] To achieve the above object, the present invention provides the following technical solution: a supply source collaborative delivery system based on dynamic programming and multi-model collaboration, comprising: Retailer sales forecasting module, which is used by commercial enterprises to forecast the sales of retailers and, based on the forecast results, to preset a reasonable purchase ceiling standard for each retailer as the basis for subsequent allocation; Market demand prediction module: This module is used to predict market demand and output the prediction results to guide the formulation of production plans; A production dynamic planning module, which is in communication with the market demand estimation module to receive the prediction results as input and then output the production plan results to guide the supply and distribution of products; The supply allocation planning module is based on the production plan results output by the production dynamic planning module, combined with the historical sales data, market demand forecasts and seasonal fluctuations of commercial enterprises in each province, to conduct in-depth supply and demand matching analysis to determine the brand, quantity and time nodes to be supplied to each province.

[0009] Commercial companies adjust their distribution strategies toward retailers and their purchasing restrictions on retailers based on supply plans.

[0010] As a further improvement of the present invention, the specific steps of the retailer sales volume prediction module for predicting the retailer sales volume are as follows: Step 1.1, collect basic data, including historical sales data, retail information and market factors; Step 1.2, use a combination of time series analysis and regression analysis to forecast sales; Step 1.3, build a sales forecasting model, and then use the sales forecasting model to predict retail sales.

[0011] As a further improvement of the present invention, the specific method of using a method combining time series analysis and regression analysis to predict sales in step 1.2 is: using historical sales data, trend analysis and seasonal decomposition are performed on the sales time series of each product of each retailer to identify long-term trends, seasonal fluctuations and random fluctuations, and then the time series prediction model is applied to predict the basic sales to obtain the predicted value without considering the influence of external factors, and then a multivariate linear regression model is constructed to perform regression analysis using retailer information and market factors as independent variables and historical sales as dependent variables; through regression analysis, the influence coefficient of each factor on sales is obtained, which is used to adjust the basic sales predicted by the time series.

[0012] As a further improvement of the present invention, the specific method of constructing the sales forecasting model in step 1.3 is: set up For the Retailers in Forecast sales for the period are the relevant influencing factors for the retailer. is the regression coefficient, For the time trend, is the seasonal factor, is a random error term. Then the prediction model can be expressed as: in, and Through time series analysis, we find that Determined by regression analysis.

[0013] As a specific implementation of the improvement, the specific steps of the market demand estimation module for predicting market demand are as follows: Step 2.1: Collect and integrate historical transaction data between industrial and commercial enterprises, covering order volume, delivery cycle, inventory level, return rate, seasonal sales fluctuations, and clean the collected data, remove outliers, fill missing values, and perform normalization or standardization to ensure data quality and analysis accuracy. Then, based on business understanding and data analysis, extract features that have an important impact on demand forecasting, laying the foundation for model construction; Step 2.2, use the long short-term memory neural network to build a demand forecasting model to predict the demand for different cigarette products by various commercial enterprises in the future.

[0014] As a further improvement of the present invention, the specific method of predicting the demand for different cigarette products by various commercial enterprises in the future period in step 2.2 is: set up is the forecast demand of the jth commercial company in period t, is the past demand series of the commercial company, Other factors that affect demand: is the algorithm mapping function, is the model parameter, the prediction model can be expressed as: in, and Trained through machine learning algorithms.

[0015] As a further improvement of the present invention, the specific steps of the production dynamic planning module receiving the prediction result as input and then outputting the production plan result are as follows: Step 3.1, in-depth analysis of the annual production volume indicators and product specification standards set by the State Tobacco Administration to ensure that the production plan complies with industry norms, comprehensively consider the inventory, supply chain stability and price fluctuations of key raw materials such as tobacco leaves, filter rods and packaging materials, determine the availability and cost constraints of raw materials, evaluate the production capacity, production efficiency, equipment maintenance plan and worker scheduling of each production line, and ensure that the production plan matches the actual capacity of the production line; Step 3.2, based on the future market demand forecast results provided by the market demand estimation module, determine the total production volume and time series distribution of each brand of cigarettes and complete the dynamic production planning.

[0016] As a further improvement of the present invention, the specific method of producing dynamic programming in step 3.2 is as follows: set up For the Tobacco products in Monthly production plan, For the The product is in The total demand application volume per month, For the The annual production volume of the product is limited. is the total number of production lines, For the The production capacity of the production line, For the The product is in The production cycle on the production line, For the The inventory of the product, For the Safety stock levels for various products, and calculate monthly production plans for various products , the production optimization objectives are as follows: Use linear programming, integer programming or heuristic algorithms to solve the above model and obtain the optimal production plan for each product every month. .

[0017] As a further improvement of the present invention, the specific steps of the supply allocation planning module for supply allocation are as follows: Step 4.1, based on the production plan results output by the production dynamic planning module, combined with the historical sales data, market demand forecasts and seasonal fluctuations of commercial enterprises in each province, conduct in-depth supply and demand matching analysis to determine the brand, quantity and time nodes to be supplied to each province and formulate precise supply strategies; Step 4.2, formulate a supply plan based on the precise supply strategy formulated in step 4.1; Step 4.3, notify each commercial company of the supply results as a basis for the commercial company to adjust its retail distribution strategy. Arrange the logistics distribution of the product according to the supply plan to ensure timely delivery to the commercial company to complete the coordinated distribution.

[0018] As a further improvement of the present invention, the specific method of formulating the supply plan in step 4.2 is as follows: set up For the Industrial Company of the month The first commercial company The number of tobacco products For the Industrial Company of the month The production plan quantity of the product, For the Month A commercial company The demand for the product For the Month The inventory of the product, For the supply strategy Product for The allocation ratio or priority of each commercial company is expressed as follows: For all products and month , the supply to commercial firms should be less than or equal to the production: Wherever possible, try to meet the needs of commercial companies, namely: .

[0019] The beneficial effects of the present invention, compared with the methods in the prior art: (1) Multi-model fusion and dynamic programming: The present invention combines multiple machine learning models such as time series analysis, regression analysis, and long short-term memory neural network (LSTM) to accurately predict sales and market demand. At the same time, the dynamic programming method is used to optimize the production plan to ensure that the production capacity and cost are optimized while meeting the demand. This multi-model fusion and dynamic programming method improves the accuracy of the prediction and the scientific nature of the decision.

[0020] (2) Refined management and differentiated strategy: The present invention collects and analyzes a large amount of basic data to make a refined forecast of the sales volume of each retailer and set a reasonable upper limit standard for sales volume. At the same time, in the supply allocation stage, a differentiated supply strategy is formulated according to the market characteristics, consumer preferences and competition pattern of each province. This refined management and differentiated strategy helps to meet the specific needs of different markets and improve customer satisfaction and sales volume of tobacco companies.

[0021] (3) Collaborative optimization of the entire process: The present invention realizes the collaborative optimization of the whole process from production to sales. Through accurate sales forecast and market demand estimation, the production plan is guided; through optimized production plan and supply allocation planning, the timely supply and reasonable distribution of products are ensured; finally, through collaborative delivery strategy, the optimal configuration of cigarette products at the retail level is achieved. This method of collaborative optimization of the whole process improves the overall operational efficiency and market response speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the execution flow of the supply collaborative delivery system based on dynamic programming and multi-model collaboration of the present invention. DETAILED DESCRIPTION

[0023] The present invention will be further described below in detail with reference to the embodiments shown in the accompanying drawings.

[0024] Reference Figure 1 As shown, a supply source collaborative delivery system based on dynamic programming and multi-model collaboration in this embodiment includes: S1. Retail sales forecast module S1.1. Collect basic data, including historical sales data, retail information and market factors.

[0025] Collect daily / weekly / monthly tobacco product sales records of each retailer over a period of time (e.g., the past year), including sales volume, sales revenue, product categories, etc.; characteristics of the retailer such as location, scale, business type (e.g., convenience stores, supermarkets, tobacco and alcohol specialty stores, etc.), customer base, etc.; consider external factors that may affect sales, such as holidays, seasonal changes, promotional activities, policy adjustments, etc.

[0026] S1.2. Use a combination of time series analysis and regression analysis to forecast sales.

[0027] Using historical sales data, we can analyze the sales time series of each retailer and each product and perform seasonal decomposition to identify long-term trends, seasonal fluctuations, and random fluctuations. We can also use time series forecasting models (such as the ARIMA model) to forecast basic sales and obtain forecast values ​​without considering the impact of external factors.

[0028] A multivariate linear regression model was constructed, with retailer information (such as store size, business type) and market factors (such as holiday effects and seasonal coefficients) as independent variables and historical sales as dependent variables for regression analysis. Through regression analysis, the coefficients of influence of each factor on sales were obtained, which were used to adjust the basic sales of time series forecasts to make the forecast closer to the actual situation.

[0029] S1.3. Build a sales forecast model set up For the Retailers in Forecast sales for the period are the relevant influencing factors of the retailer (such as store size, seasonal coefficient, etc.), is the regression coefficient, For the time trend, is the seasonal factor, is a random error term. Then the prediction model can be expressed as: in, and Through time series analysis, we find that Determined by regression analysis.

[0030] According to the forecast results, set a reasonable sales ceiling for each retailer as the basis for subsequent purchases and allocations. Regularly update the forecast model to adapt to market changes and ensure the timeliness and accuracy of the forecast results.

[0031] S2. Market demand estimation module S2.1. Collect and integrate historical transaction data between industrial and commercial enterprises, covering key indicators such as order volume, delivery cycle, inventory level, return rate, seasonal sales fluctuations, etc. Clean the collected data, remove outliers, fill in missing values, and perform normalization or standardization to ensure data quality and accuracy of analysis. Based on business understanding and data analysis, extract features that have an important impact on demand forecasting, such as historical sales trends, holiday effects, promotional activities, etc., to lay the foundation for model construction.

[0032] S2.2. Use the long short-term memory neural network (LSTM) to build a demand forecasting model to predict the demand for different cigarette products by various commercial enterprises in the future.

[0033] set up is the forecast demand of the jth commercial company in period t, is the past demand series of the commercial company, Other factors that affect demand (such as market trends, retail sales forecasts, etc.), is the algorithm mapping function, is the model parameter. Then the prediction model can be expressed as: in, and Trained through machine learning algorithms.

[0034] S2.3 Application of prediction results The forecast results are used as input to the industrial company's production planning model (part S3) to guide the formulation of production plans. Based on the forecast results, demand confirmation and negotiation are carried out with commercial companies to ensure that both supply and demand sides agree on the forecast results. The forecast model is updated regularly to adapt to market changes and ensure the timeliness and accuracy of the forecast results.

[0035] S3, production dynamic planning module S3.1. Construction of production constraints: In-depth analysis of the annual production volume indicators and product specification standards set by the State Tobacco Administration to ensure that the production plan meets industry standards. Comprehensively consider the inventory, supply chain stability and price fluctuations of key raw materials such as tobacco leaves, filter rods, and packaging materials to determine the availability and cost constraints of raw materials. Evaluate the production capacity, production efficiency, equipment maintenance plan and worker scheduling of each production line to ensure that the production plan matches the actual capacity of the production line.

[0036] S3.2, Dynamic Production Planning: Based on the future market demand forecast results provided by S2, determine the total production volume and time series distribution of each brand of cigarettes. For the Tobacco products in Monthly production plan, For the The product is in The total demand application volume per month, For the The annual production volume of the product is limited. is the total number of production lines, For the The production capacity of the production line, For the The product is in The production cycle on the production line, For the The inventory of the product, For the Safety stock level for a product.

[0037] The goal of the production planning model is to determine the monthly production plan for each product based on the principles of meeting demand, optimizing capacity and cost control. The production optimization objectives are as follows: Use linear programming, integer programming or heuristic algorithm to solve the above model and get the optimal production plan for each product every month. .

[0038] The production plan results are used as input to the supply model (S4) of the industrial company to the commercial enterprise to guide the supply and distribution of products. According to the production plan, production resources such as raw material procurement, production line arrangement, and staffing are adjusted to ensure the smooth implementation of the production plan.

[0039] S4, Supply Allocation Planning Module S4.1、Precise supply strategy formulation: Based on the production plan results output by S3, combined with the historical sales data, market demand forecasts and seasonal fluctuations of commercial enterprises in each province, an in-depth supply and demand matching analysis is conducted to determine the brand, quantity and time nodes to be supplied to each province.

[0040] Based on the market characteristics, consumer preferences and competitive landscape of each province, we formulate differentiated supply strategies, such as priority supply to key markets, new product promotion strategies, inventory adjustment strategies, etc., to meet the specific needs of different markets.

[0041] S4.2, Supply plan formulation: set up For the Industrial Company of the month The first commercial company The number of tobacco products For the Industrial Company of the month The production plan quantity of the product, For the Month A commercial company The demand for the product For the Month The inventory of the product, For the supply strategy Product for The supply plan aims to determine the quantity and category of tobacco products to be supplied to each commercial company each month on the basis of meeting the supply principle. It can be expressed in the following form: For all products and month , the supply to commercial firms should be less than or equal to the production: Wherever possible, try to meet the needs of commercial companies, namely: S4.3. Collaborative delivery The supply results will be notified to each commercial company as a basis for the commercial company to adjust its retail distribution strategy. According to the supply plan, the logistics distribution of the products will be arranged to ensure timely delivery to the commercial company.

[0042] Commercial companies adjust their scheduled targets for retailers based on the types and quantities of goods they can obtain, thereby achieving collaborative optimization of the entire process of cigarette products.

[0043] In summary, the supply collaborative delivery system based on dynamic programming and multi-model collaboration in this embodiment achieves: (1) Dynamic programming and multi-model collaboration: The present invention proposes a comprehensive system that combines dynamic programming methods and multiple forecasting models (retailer sales forecasting model, market demand estimation model, production dynamic programming model, supply allocation planning model) to achieve coordinated delivery of tobacco supply. These models are not isolated, but are coordinated and promoted in parallel to ensure the balance between inventory turnover and shortage, thereby optimizing inventory management and sales efficiency.

[0044] (2) Collaborative optimization of the entire process: The present invention not only focuses on a single link (such as sales forecast or production plan), but also realizes the collaborative optimization of the whole process from the supply end to the channel end. Through precise supply strategies and supply plans, it ensures seamless connection between production, supply and sales, and improves overall operational efficiency.

[0045] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A supply source collaborative delivery system based on dynamic programming and multi-model collaboration, characterized by: include: Retailer sales forecasting module, which is used by commercial enterprises to forecast the sales of retailers and, based on the forecast results, to preset a reasonable purchase ceiling standard for each retailer as the basis for subsequent allocation; Market demand prediction module: This module is used to predict market demand and output the prediction results to guide the formulation of production plans; A production dynamic planning module, which is in communication with the market demand estimation module to receive the prediction results as input and then output the production plan results to guide the supply and distribution of products; The supply allocation planning module is based on the production plan results output by the production dynamic planning module, combined with the historical sales data, market demand forecasts and seasonal fluctuations of commercial enterprises in each province, to conduct in-depth supply and demand matching analysis to determine the brand, quantity and time nodes supplied to each province; based on the supply plan, commercial companies adjust their retail sales strategy and purchase restrictions on retailers.

2. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to claim 1 is characterized in that: The specific steps of the retail sales volume forecasting module for forecasting retail sales volume are as follows: Step 1.1, collect basic data, including historical sales data, retail information and market factors; Step 1.2, use a combination of time series analysis and regression analysis to forecast sales; Step 1.3, build a sales forecasting model, and then use the sales forecasting model to predict retail sales.

3. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to claim 2 is characterized by: The specific method of using a combination of time series analysis and regression analysis to predict sales in step 1.2 is: using historical sales data, performing trend analysis and seasonal decomposition on the sales time series of each product of each retailer, identifying long-term trends, seasonal fluctuations and random fluctuations, and then applying the time series prediction model to predict the basic sales to obtain the predicted value without considering the impact of external factors, and then constructing a multivariate linear regression model, taking retailer information and market factors as independent variables and historical sales as dependent variables, and performing regression analysis; through regression analysis, the influence coefficient of each factor on sales is obtained, which is used to adjust the basic sales predicted by the time series.

4. The tobacco supply collaborative delivery system based on dynamic programming and multi-model collaboration according to claim 2 or 3, characterized in that: The specific method of constructing the sales forecast model in step 1.3 is: set up For the Retailers in Forecast sales for the period are the relevant influencing factors for the retailer. is the regression coefficient, For the time trend, is the seasonal factor, is a random error term; the prediction model can be expressed as: ； in, and Through time series analysis, we find that Determined by regression analysis.

5. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to any one of claims 1 to 3, characterized in that: The specific steps of the market demand estimation module for predicting market demand are as follows: Step 2.1: Collect and integrate historical transaction data between industrial and commercial enterprises, covering order volume, delivery cycle, inventory level, return rate, seasonal sales fluctuations, and clean the collected data, remove outliers, fill missing values, and perform normalization or standardization to ensure data quality and analysis accuracy. Then, based on business understanding and data analysis, extract features that have an important impact on demand forecasting, laying the foundation for model construction; Step 2.2, use the long short-term memory neural network to build a demand forecasting model to predict the demand for different cigarette products by various commercial enterprises in the future.

6. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to claim 5 is characterized in that: The specific method of predicting the demand for different cigarette products by various commercial enterprises in the future period in step 2.2 is: set up is the forecast demand of the jth commercial company in period t, is the past demand series of the commercial company, Other factors that affect demand: is the algorithm mapping function, is the model parameter, the prediction model can be expressed as: ； in, and Trained through machine learning algorithms.

7. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to any one of claims 1 to 3, characterized in that: The specific steps of the production dynamic planning module receiving the prediction results as input and then outputting the production plan results are as follows: Step 3.1, in-depth analysis of the annual production volume indicators and product specification standards set by the State Tobacco Administration to ensure that the production plan complies with industry norms, comprehensively consider the inventory, supply chain stability and price fluctuations of key raw materials such as tobacco leaves, filter rods and packaging materials, determine the availability and cost constraints of raw materials, evaluate the production capacity, production efficiency, equipment maintenance plan and worker scheduling of each production line, and ensure that the production plan matches the actual capacity of the production line; Step 3.2, based on the future market demand forecast results provided by the market demand estimation module, determine the total production volume and time series distribution of each brand of cigarettes and complete the dynamic production planning.

8. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to claim 7 is characterized in that: The specific method of producing dynamic programming in step 3.2 is as follows: set up For the Tobacco products in Monthly production plan, For the The product is in The total demand application volume per month, For the The annual production volume of the product is limited. is the total number of production lines, For the The production capacity of the production line, For the The product is in The production cycle on the production line, For the The inventory of the product, For the Safety stock levels for various products, and calculate monthly production plans for various products , the production optimization objectives are as follows: ； ； Use linear programming, integer programming or heuristic algorithms to solve the above model and obtain the optimal production plan for each product every month. .

9. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to any one of claims 1 to 3, characterized in that: The specific steps of the supply allocation planning module for supply allocation are as follows: Step 4.1, based on the production plan results output by the production dynamic planning module, combined with the historical sales data, market demand forecasts and seasonal fluctuations of commercial enterprises in each province, conduct in-depth supply and demand matching analysis to determine the brand, quantity and time nodes to be supplied to each province and formulate precise supply strategies; Step 4.2, formulate a supply plan based on the precise supply strategy formulated in step 4.1; Step 4.3, notify each commercial company of the supply results, which will serve as a basis for the commercial company to adjust its retail distribution strategy; arrange product logistics distribution according to the supply plan to ensure timely delivery to the commercial company to complete coordinated distribution.

10. The supply source collaborative delivery system based on dynamic programming and multi-model collaboration according to claim 9 is characterized in that: The specific method of formulating the supply plan in step 4.2 is as follows: set up For the Industrial Company of the month The first commercial company The number of tobacco products For the Industrial Company of the month The production plan quantity of the product, For the Month A commercial company The demand for the product For the Month The inventory of the product, For the supply strategy Product for The allocation ratio or priority of each commercial company is expressed as follows: For all products and month , the supply to commercial firms should be less than or equal to the production: ； Wherever possible, try to meet the needs of commercial companies, namely: 。

Images