Enterprise supply chain management system based on big data
By using big data analysis and linear regression models, inventory and order quantities are dynamically adjusted, solving the problems of inaccurate inventory forecasting and insufficient cost optimization in traditional supply chain management systems, and achieving optimal inventory levels and minimum costs.
Patent Information
- Application Number
- CN202510985028.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-17
- Publication Date
- 2025-10-31
AI Technical Summary
Traditional supply chain management systems lack scientific basis, resulting in inaccurate inventory forecasting, inability to make timely adjustments, and a lack of cost optimization, leading to unnecessary capital occupation and waste.
By adopting a big data-based enterprise supply chain management system, and through data collection, predictive models, optimal inventory units, and cost optimization units, combined with linear regression models and cost objective functions, inventory and order quantities are dynamically adjusted to optimize supply chain inventory preparation strategies.
It enables accurate market trend analysis and inventory forecasting, dynamically adjusts inventory and order quantities, reduces capital occupation, improves inventory turnover, and ensures that the supply chain meets market demand at the lowest possible cost.
Smart Images
Figure CN120875751A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of supply chain management technology, and more specifically to an enterprise supply chain management system based on big data. Background Technology
[0002] Traditional supply chain management systems often rely on experience and human judgment to make sales forecasts and inventory decisions. This approach lacks scientific basis and is prone to inaccurate forecasts, especially in the case of rapidly changing markets. It may be unable to adjust inventory and order quantities in a timely manner. Traditional supply chain management systems usually lack cost optimization processes, and inventory and order quantity decisions are often not based on scientific cost analysis. The lack of effective assessment of inventory and ordering costs may lead to unnecessary capital occupation and waste. The inventory management methods of traditional supply chain management systems usually lack dynamic adjustment mechanisms, and inventory levels may be set too high or too low, resulting in overstocking or understocking, which in turn affects cash flow and customer satisfaction. Summary of the Invention
[0003] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide an enterprise supply chain management system based on big data.
[0004] The technical solution adopted to solve the above-mentioned technical problems is: a big data-based enterprise supply chain management system, including: A data acquisition unit is used to acquire historical sales data of the target enterprise, wherein the historical sales data includes: sales volume data, inventory cost and ordering cost, and market trend analysis is performed based on the sales volume data to obtain market trend data; The prediction model unit is used to extract features from the market trend data to obtain a feature vector set of the market trend data, and to train the feature vector set based on a linear regression model to obtain a demand prediction model. The optimal inventory unit is used to predict and analyze the inventory products of the target enterprise based on the demand forecasting model to obtain inventory products that conform to market trends, and to perform dynamic calculations on safety stock and ordering for the inventory products to obtain the optimal inventory objective function. The cost optimization unit is used to obtain the optimal inventory level and the optimal order quantity based on the optimal inventory objective function, and to construct a cost objective function for the optimal inventory level and the optimal order quantity based on the inventory cost and the order cost, so as to obtain the cost minimization objective function; The inventory preparation strategy unit is used to optimize the inventory preparation of the target enterprise's supply chain based on the objective function of minimizing costs, so as to obtain a supply chain inventory preparation optimization strategy.
[0005] Preferably, market trend analysis is performed based on the sales data to obtain market trend data, including: The sales data is clustered over time to obtain sales clusters for different time periods; Feature extraction is performed on the sales clusters to obtain the sales feature vectors corresponding to the sales clusters; The market trend data is obtained by calculating a quarterly market trend index on the sales feature vector, wherein the formula for calculating the quarterly market trend index is as follows: ; in, Indicates the first Market trend data for the quarter, This indicates the size of the sales cluster. Indicates quarter, Indicates a quarterly cycle. Indicates the first Sales feature vector for each quarter Indicates the first The fitting trend coefficient for each quarter.
[0006] Preferably, feature extraction is performed on the market trend data to obtain a feature vector set of the market trend data, including: The variance of the market trend data is calculated to obtain the variance of the market trend data; the standard deviation of the market trend data is calculated based on the standard deviation formula to obtain the standard deviation of the market trend data. The market trend data is standardized using the Z-score standardization method and the standard deviation of the market trend data to obtain standardized data. Feature extraction is then performed on the standardized data to obtain a set of feature vectors for the market trend data.
[0007] Preferably, the feature vector set is trained based on a linear regression model to obtain a demand prediction model, including: The feature vector set is trained based on the mean squared error objective function to obtain the predicted feature vector corresponding to the minimum mean squared error. The mean squared error objective function is formulated as follows: MSE ; Where MSE represents the mean squared error, and N represents the size of the feature vector set. This represents the feature vector in the set of feature vectors. Represents the first eigenvector of the set of features One true feature vector, Represents the first eigenvector of the set of features One predicted feature vector; The linear relationship between the predicted feature vectors is calculated based on the linear regression model to obtain the demand prediction model.
[0008] Preferably, the inventory products of the target enterprise are predicted and analyzed based on the demand forecasting model to obtain inventory products that conform to market trends, including: Obtain the product types corresponding to the inventory products of the target company; Based on the demand forecasting model, market demand is forecasted for the product type to obtain product types that conform to market trends, and the inventory products are obtained based on the product types.
[0009] Preferably, the inventory products are dynamically calculated for safety stock and ordering to obtain the optimal inventory objective function, including: Obtain the inventory quantity corresponding to the inventory product, and predict the sales volume of the inventory product based on the demand forecasting model to obtain the maximum sales volume of the inventory product under the condition of conforming to market trends.
[0010] The maximum sales volume and the inventory quantity are used to perform dynamic calculations on safety stock and ordering to obtain the optimal inventory objective function, wherein the formula for the optimal inventory objective function is as follows: ; Where S represents the optimal inventory level for the product in stock. This indicates the ordering timeframe for the inventory products. Indicates quarter, Indicates that the inventory product is in the [number]th [year]. The maximum standard deviation of sales in a quarter This indicates that the inventory products are in The maximum standard deviation of sales volume within an ordering period Indicates that the inventory product is in the [number]th [year]. The average maximum sales volume for each quarter.
[0011] Preferably, obtaining the optimal inventory level and optimal order quantity based on the optimal inventory objective function includes: The optimal inventory level is calculated based on the optimal inventory objective function to obtain the optimal inventory level for the inventory product. Based on the demand forecasting model, the maximum sales volume corresponding to the inventory products is obtained, and the maximum sales volume is compared with the optimal inventory volume to obtain the optimal order quantity.
[0012] Preferably, a cost objective function is constructed based on the inventory cost and the ordering cost for the optimal inventory level and the optimal ordering level to obtain a cost minimization objective function, including: Based on the inventory cost, obtain the minimum inventory cost corresponding to the optimal inventory quantity; based on the ordering cost, obtain the minimum ordering cost corresponding to the optimal order quantity; Construct a cost objective function based on the minimum inventory cost and the minimum ordering cost. To obtain the minimum cost objective function, wherein the minimum cost objective function is... The formula is as follows: ; in, Indicates minimizing cost. Indicates a quarterly cycle. Indicates quarter, This represents the inventory cost. Indicates the first The optimal inventory level for each quarter. This represents the ordering cost. Indicates the first The optimal order quantity for each quarter.
[0013] The constraints of the objective function are as follows: ; ; in, This represents the minimum inventory cost. This represents the minimum ordering cost.
[0014] Preferably, the target enterprise's supply chain inventory optimization is performed based on a cost minimization objective function to obtain a supply chain inventory optimization strategy, including: Based on the minimum cost objective function, the order quantity corresponding to the minimum cost is obtained. Based on the order quantity, the target enterprise's supply chain is stocked to obtain the supply chain stocking optimization strategy.
[0015] The beneficial effects of the present invention are as follows: (1) The present invention obtains historical sales data through big data analysis and mining, and combines information such as sales volume, inventory cost and ordering cost to analyze market trends and make predictions. This trend analysis based on historical data can help enterprises anticipate market changes and make adjustments in advance to avoid excessive or insufficient inventory backlog; (2) The present invention predicts inventory through a demand forecasting model and calculates the optimal inventory target that conforms to market trends. According to changes in market demand, the safety stock and ordering quantity are dynamically adjusted to ensure the optimization of inventory; (3) Through the accurate calculation and analysis of the cost optimization unit, the present invention enables enterprises to improve inventory turnover while minimizing inventory and ordering costs and reducing unnecessary capital occupation; (4) The present invention proposes a stock preparation strategy through the stock preparation strategy unit to ensure that the supply chain can meet market demand while minimizing costs. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the overall system architecture in one embodiment of the present invention; Attached labels: 1. Data acquisition unit; 2. Predictive model unit; 3. Optimal inventory unit; 4. Cost optimization unit; 5. Inventory preparation strategy unit. Detailed Implementation
[0017] Example 1, as Figure 1 As shown, the present invention proposes a big data-based enterprise supply chain management system, comprising: Data acquisition unit 1 is used to acquire the historical sales data of the target enterprise. The historical sales data includes: sales volume data, inventory cost and ordering cost. Based on the sales volume data, market trends are analyzed to obtain market trend data. Prediction model unit 2 is used to extract features from market trend data to obtain a set of feature vectors of market trend data. The feature vector set is trained based on a linear regression model to obtain a demand prediction model. Optimal Inventory Unit 3 is used to predict and analyze the inventory products of the target enterprise based on the demand forecasting model in order to obtain inventory products that conform to market trends. It performs dynamic calculations on safety stock and ordering for inventory products to obtain the optimal inventory objective function. Cost optimization unit 4 is used to obtain the optimal inventory level and optimal order quantity based on the optimal inventory objective function, and to construct a cost objective function based on the optimal inventory level and optimal order quantity based on inventory cost and order cost, so as to obtain the minimized cost objective function. Inventory Optimization Strategy Unit 5 is used to optimize the inventory of the target enterprise's supply chain based on the objective function of minimizing cost, so as to obtain the supply chain inventory optimization strategy.
[0018] In this invention, market trend data refers to data reflecting the changing patterns of market demand obtained through the analysis of historical sales data. Feature extraction refers to selecting representative features from market trend data that can reveal the changing patterns of market demand. The feature vector set refers to a set of multiple feature vectors obtained from the feature extraction step. Each feature vector contains relevant features of a data point. These vectors are used as input variables for further analysis and modeling. The linear regression model is a statistical model that finds a linear relationship between independent variables (such as market characteristics) and dependent variables (such as future demand). Linear regression is used to predict future demand based on historical data and market trend characteristics. The demand forecasting model is a mathematical model that can predict the demand for goods or services in a future period by training on historical data and market trend data. Based on market trend characteristics, it helps companies make more accurate inventory management and ordering decisions. Safety stock refers to the extra inventory set by a company to cope with demand fluctuations or supply delays. The purpose of safety stock is to avoid stockouts caused by sudden events or demand fluctuations. The supply chain inventory optimization strategy refers to the optimization plan formulated by adjusting inventory management and ordering strategies in the supply chain. The aim of this strategy is to ensure the smooth operation of the supply chain, reduce inventory backlog, and ensure that companies can meet market demand.
[0019] Example 2: The enterprise supply chain management system based on big data proposed in this invention, compared with Example 1, further includes: Market trends are analyzed based on sales data to obtain market trend data, including: Perform time-based clustering on sales data to obtain sales clusters for different time periods; Features are extracted from the sales clusters to obtain the corresponding sales feature vectors. A quarterly market trend index is calculated from the sales feature vector to obtain market trend data. The formula for calculating the quarterly market trend index is as follows: ; in, Indicates the first Market trend data for the quarter, Indicates the size of the sales cluster. Indicates quarter, Indicates a quarterly cycle. Indicates the first Sales feature vector for each quarter Indicates the first The fitting trend coefficient for each quarter.
[0020] In this embodiment, time clustering refers to the process of grouping sales data according to time periods. Sales clusters refer to the results of grouping data points with similar sales characteristics using time clustering methods. Sales feature vectors are mathematical vectors that quantify each sales cluster. Feature vectors can help reveal the main characteristics of clusters and provide support for further analysis. Fitting trend coefficients are coefficients obtained by fitting historical data, reflecting the market trend of the quarter. These coefficients are used to adjust the calculation of sales feature vectors to more accurately reflect market changes in the quarter.
[0021] In an optional embodiment, feature extraction is performed on the market trend data to obtain a set of feature vectors for the market trend data, including: Variance is calculated on market trend data to obtain the market trend data variance; standard deviation is calculated on the market trend data variance based on the standard deviation formula to obtain the market trend data standard deviation. The market trend data is standardized using the Z-score standardization method and the standard deviation of the market trend data to obtain the standardized data corresponding to the market trend data. Feature extraction is then performed on the standardized data to obtain the feature vector set of the market trend data.
[0022] It should be noted that Z-score standardization is a data standardization method that transforms data into a standard normal distribution by calculating the deviation of the data from its mean.
[0023] In an optional embodiment, a demand prediction model is obtained by training the feature vector set based on a linear regression model, including: The feature vector set is trained based on the mean squared error objective function to obtain the predicted feature vector corresponding to the minimum mean squared error. The formula for the mean squared error objective function is as follows: MSE ; Where MSE represents the mean squared error, and N represents the size of the eigenvector set. This represents the eigenvectors in the eigenvector set. Represents the first eigenvector of the eigenvector set. One true feature vector, Represents the first eigenvector of the eigenvector set. One predicted feature vector; The linear relationship between the predicted feature vectors is calculated based on the linear regression model to obtain the demand prediction model.
[0024] It should be noted that mean squared error is a loss function commonly used in regression analysis. It is often used to measure the error between the model's predicted value and the true value. The goal of mean squared error is to minimize the error so that the model's prediction is as close to the true value as possible.
[0025] In an optional embodiment, a forecasting analysis of the target company's inventory products is performed based on a demand forecasting model to obtain inventory products that conform to market trends, including: Obtain the product types corresponding to the target company's inventory products; Market demand is forecasted for product types based on demand forecasting models to identify product types that align with market trends, and inventory is then acquired based on these product types.
[0026] In an optional embodiment, safety stock and order dynamic calculations are performed on the inventory products to obtain the optimal inventory objective function, including: Obtain the inventory quantity corresponding to the inventory products, and predict the sales volume of the inventory products based on the demand forecasting model to obtain the maximum sales volume of the inventory products under the condition of conforming to market trends.
[0027] The maximum sales volume and inventory quantity are dynamically calculated for safety stock and ordering to obtain the optimal inventory objective function, which is formulated as follows: ; Where S represents the optimal inventory level for the product. Indicates the ordering timeframe for inventory products. Indicates quarter, Indicates the inventory products in the first... The maximum standard deviation of sales in a quarter Indicates that the inventory products are in The maximum standard deviation of sales volume within an ordering period Indicates the inventory products in the first... The average maximum sales volume for each quarter.
[0028] It should be noted that the optimal inventory level refers to the amount of inventory needed to meet future demand while avoiding excessive or insufficient inventory. The calculation of the optimal inventory level is determined based on multiple factors such as sales forecasts, market trends, and inventory costs. The ordering time period refers to the time from when the company places an order for products to when the inventory products arrive and can be sold or used. The maximum sales standard deviation refers to the degree of fluctuation in the sales volume of inventory products within a certain period. The maximum sales mean refers to the average sales volume of inventory products within a certain time period. Dynamic ordering calculation refers to the process of dynamically adjusting the order quantity based on real-time market demand, inventory levels, and supply chain information. The optimal inventory objective function is calculated through a mathematical model and is used to determine the best inventory level that a company should maintain under given conditions.
[0029] In an optional embodiment, obtaining the optimal inventory level and optimal order quantity based on the optimal inventory objective function includes: The optimal inventory level is calculated based on the optimal inventory objective function to obtain the optimal inventory level for each product. The maximum sales volume corresponding to the inventory products is obtained based on the demand forecasting model. The maximum sales volume is compared with the optimal inventory level to obtain the optimal order quantity.
[0030] In an optional embodiment, a cost objective function is constructed based on inventory cost and ordering cost for the optimal inventory level and optimal ordering level to obtain a cost minimization objective function, including: Obtain the minimum inventory cost corresponding to the optimal inventory level based on inventory cost; obtain the minimum order cost corresponding to the optimal order quantity based on ordering cost; Construct a cost objective function based on minimum inventory cost and minimum ordering cost. The cost minimization objective function is obtained, and its formula is as follows: ; in, Indicates minimizing cost. Indicates a quarterly cycle. Indicates quarter, Indicates inventory cost, Indicates the first Optimal inventory level for the quarter This indicates the cost of ordering. Indicates the first The optimal order quantity for the quarter.
[0031] The constraints of the objective function are as follows: ; ; in, Indicates minimum inventory cost. This indicates the minimum order cost.
[0032] It should be noted that minimum inventory cost refers to the lowest inventory holding cost calculated based on the optimal inventory level, minimum ordering cost refers to the lowest ordering cost calculated based on the optimal order quantity, and the minimum cost objective function is a mathematical expression used to represent total cost in supply chain inventory management. This function usually includes two parts: inventory cost and ordering cost, and the goal is to minimize the total supply chain cost.
[0033] In an optional embodiment, the target enterprise's supply chain inventory optimization is performed based on a cost minimization objective function to obtain a supply chain inventory optimization strategy, including: Based on the objective function of minimizing cost, the order quantity corresponding to the minimum cost is obtained. Based on the order quantity, the target company's supply chain is stocked to obtain a supply chain stocking optimization strategy.
[0034] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.
Claims
1. A big data-based enterprise supply chain management system, comprising, characterized in that: Data acquisition unit (1), the data acquisition unit (1) is used to acquire the historical sales data of the target enterprise, wherein the historical sales data includes: sales volume data, inventory cost and ordering cost, and the market trend is analyzed based on the sales volume data to obtain market trend data; Prediction model unit (2), the prediction model unit (2) is used to extract features from the market trend data to obtain a feature vector set of the market trend data, and to train the feature vector set based on the linear regression model to obtain a demand prediction model; The optimal inventory unit (3) is used to predict and analyze the inventory products of the target enterprise based on the demand forecasting model, so as to obtain inventory products that conform to market trends, and to perform dynamic calculation of safety stock and ordering for the inventory products to obtain the optimal inventory objective function. Cost optimization unit (4), the cost optimization unit (4) is used to obtain the optimal inventory quantity and the optimal order quantity based on the optimal inventory objective function, and to construct a cost objective function for the optimal inventory quantity and the optimal order quantity based on the inventory cost and the order cost, so as to obtain the minimized cost objective function; The inventory preparation strategy unit (5) is used to optimize the inventory preparation of the target enterprise's supply chain based on the cost minimization objective function, so as to obtain the supply chain inventory preparation optimization strategy.
2. The enterprise supply chain management system based on big data according to claim 1, characterized in that, Based on the sales data, market trends are analyzed to obtain market trend data, including: The sales data is clustered over time to obtain sales clusters for different time periods; Feature extraction is performed on the sales clusters to obtain the sales feature vectors corresponding to the sales clusters; The market trend data is obtained by calculating a quarterly market trend index on the sales feature vector, wherein the formula for calculating the quarterly market trend index is as follows: ； in, Indicates the first Market trend data for the quarter, This indicates the size of the sales cluster. Indicates quarter, Indicates a quarterly cycle. Indicates the first Sales feature vector for each quarter Indicates the first The fitting trend coefficient for each quarter.
3. The enterprise supply chain management system based on big data according to claim 2, characterized in that, Feature extraction is performed on the market trend data to obtain a set of feature vectors for the market trend data, including: The variance of the market trend data is calculated to obtain the variance of the market trend data; the standard deviation of the market trend data is calculated based on the standard deviation formula to obtain the standard deviation of the market trend data. The market trend data is standardized using the Z-score standardization method and the standard deviation of the market trend data to obtain standardized data. Feature extraction is then performed on the standardized data to obtain a set of feature vectors for the market trend data.
4. The enterprise supply chain management system based on big data according to claim 3, characterized in that, The feature vector set is trained based on a linear regression model to obtain a demand prediction model, including: The feature vector set is trained based on the mean squared error objective function to obtain the predicted feature vector corresponding to the minimum mean squared error. The mean squared error objective function is formulated as follows: MSE ; Where MSE represents the mean squared error, and N represents the size of the feature vector set. This represents the feature vector in the set of feature vectors. Represents the first eigenvector of the set of features One true feature vector, Represents the first eigenvector of the set of features One predicted feature vector; The linear relationship between the predicted feature vectors is calculated based on the linear regression model to obtain the demand prediction model.
5. The enterprise supply chain management system based on big data according to claim 4, characterized in that, Based on the aforementioned demand forecasting model, the target company's inventory products are predicted and analyzed to obtain inventory products that conform to market trends, including: Obtain the product types corresponding to the inventory products of the target company; Based on the demand forecasting model, market demand is forecasted for the product type to obtain product types that conform to market trends, and the inventory products are obtained based on the product types.
6. The enterprise supply chain management system based on big data according to claim 5, characterized in that, The inventory products are dynamically calculated for safety stock and ordering to obtain the optimal inventory objective function, including: Obtain the inventory quantity corresponding to the inventory product, and predict the sales volume of the inventory product based on the demand forecasting model to obtain the maximum sales volume of the inventory product under the condition of conforming to market trends. The maximum sales volume and the inventory quantity are used to perform dynamic calculations on safety stock and ordering to obtain the optimal inventory objective function, wherein the formula for the optimal inventory objective function is as follows: ； Where S represents the optimal inventory level for the product in stock. This indicates the ordering timeframe for the inventory products. Indicates quarter, Indicates that the inventory product is in the [number]th [year]. The maximum standard deviation of sales in a quarter This indicates that the inventory products are in The maximum standard deviation of sales volume within an ordering period Indicates that the inventory product is in the [number]th [year]. The average maximum sales volume for each quarter.
7. A big data-based enterprise supply chain management system according to claim 6, characterized in that, Obtaining the optimal inventory level and optimal order quantity based on the aforementioned optimal inventory objective function includes: The optimal inventory level is calculated based on the optimal inventory objective function to obtain the optimal inventory level for the inventory product. Based on the demand forecasting model, the maximum sales volume corresponding to the inventory products is obtained, and the maximum sales volume is compared with the optimal inventory volume to obtain the optimal order quantity.
8. A big data-based enterprise supply chain management system according to claim 7, characterized in that, Based on the inventory cost and ordering cost, a cost objective function is constructed for the optimal inventory level and the optimal ordering level to obtain a cost minimization objective function, including: Based on the inventory cost, obtain the minimum inventory cost corresponding to the optimal inventory quantity; based on the ordering cost, obtain the minimum ordering cost corresponding to the optimal order quantity; Construct a cost objective function based on the minimum inventory cost and the minimum ordering cost. To obtain the minimum cost objective function, wherein the minimum cost objective function is... The formula is as follows: ； in, Indicates minimizing cost. Indicates a quarterly cycle. Indicates quarter, This represents the inventory cost. Indicates the first The optimal inventory level for each quarter. This represents the ordering cost. Indicates the first The optimal order quantity for the quarter; The constraints of the objective function are as follows: ; ; in, This represents the minimum inventory cost. This represents the minimum ordering cost.
9. A big data-based enterprise supply chain management system according to claim 8, characterized in that, Based on the objective function of minimizing cost, the supply chain of the target enterprise is optimized for inventory preparation to obtain a supply chain inventory preparation optimization strategy, including: Based on the minimum cost objective function, the order quantity corresponding to the minimum cost is obtained. Based on the order quantity, the target enterprise's supply chain is stocked to obtain the supply chain stocking optimization strategy.