A lace fabric supply chain management method based on big data

Abstract

The present application relates to the lace fabric supply chain management technical field.The present application relates to a kind of lace fabric supply chain management method based on big data.It includes the following steps: S1, the inventory data of enterprise, sales data are acquired, while the lace type corresponding to the lace fabric of enterprise is acquired, and the sales customer list is extracted in sales data;The present application realizes accurate prediction by multidimensional data fusion, on the one hand, combined with customer group portrait and network heat trend data, extract accurate heat information and calculate fluctuation coefficient, clear market heat influence degree on customer order, on the other hand, combined with historical order difference analysis screening reference period, with the maximum quantity difference limit prediction range, avoid prediction data deviate from actual demand, can let enterprise master customer real demand in advance, both reduce the inventory accumulation caused by prediction too high, also avoid the shortage problem caused by prediction too low, greatly reduce supply chain resource waste.

Description

TECHNICAL FIELD

[0001] The application relates to the lace fabric supply chain management technical field, in particular to a lace fabric supply chain management method based on big data. BACKGROUND

[0002] The existing lace fabric supply chain management technology works in the scene of static data statistics combined with artificial subjective decision, that is, relies on historical inventory account books, simple sales records and other static data, and combines the subjective judgment of managers on the market to formulate procurement and inventory strategies.

[0003] At present, network sales account for most of the terminal sales market, and due to the lack of combination with network market trend data, the dynamic changes of lace fabric market demand cannot be captured in time, for example, when a certain type of lace suddenly becomes a popular element, the rising signal of heat cannot be quickly identified, leading to lagging demand prediction, on the other hand, the depth analysis of customer group portrait is lacking, and the procurement preferences of different customer groups cannot be accurately matched, which is easy to cause the problems of unsalable and frequent out-of-stock of part of the lace fabric, therefore, a lace fabric supply chain management method based on big data is proposed. SUMMARY

[0004] The purpose of the application is to provide a lace fabric supply chain management method based on big data to solve the problems raised in the background technology.

[0005] To achieve the above purpose, a lace fabric supply chain management method based on big data is provided, comprising the following steps:

[0006] S1, obtaining the inventory data and sales data of the enterprise, obtaining the lace type corresponding to the lace fabric of the enterprise, and extracting the sales customer list in the sales data;

[0007] S2, performing customer group portrait analysis according to the sales customer list, then combining the lace type with the customer group portrait to perform accurate analysis and overall analysis of the heat trend data in the network, and combining the sales data with the sales customer list to perform order screening to obtain the historical order data corresponding to the customer;

[0008] S3, combining the accurate heat trend data with the historical order data to analyze the fluctuation of the customer, obtaining the fluctuation coefficient of the heat trend data change to the customer, and performing development coefficient analysis through the overall heat trend data;

[0009] S4, extract the latest order data, combine the latest order data with the fluctuation coefficient and the development coefficient to predict the order data analysis, then analyze the difference of the order quantity in the historical order data, according to the difference analysis result, the reference period of historical order data is screened, and the quantity difference analysis is carried out according to the screened reference period combined with the historical prediction order data;

[0010] S5, extract the maximum quantity difference of S4 analysis result, limit the range of prediction order data through the maximum quantity difference, then compare the prediction order data with the inventory data, and send the order to the supply chain according to the comparison result.

[0011] As a further improvement of the technical solution, the S1 extracts the lace fabric related inventory data and sales data from the enterprise management end by establishing data connection with the enterprise management end.

[0012] As a further improvement of the technical solution, the steps of S1 are as follows:

[0013] S1.1, according to the inventory data and sales data, the lace type is extracted, and the lace type corresponding to the enterprise sales lace fabric is obtained;

[0014] S1.2, the sales customer list is extracted in the sales data, and the sales customer list corresponding to the lace type is obtained.

[0015] As a further improvement of the technical solution, the steps of S2 are as follows:

[0016] S2.1, according to the sales customer list, the customer group portrait analysis is carried out, and the customer group portrait corresponding to the sales customer list is obtained;

[0017] S2.2, according to the lace type, the overall analysis of the heat trend data in the network is carried out, and the overall heat trend data of the lace type in the network is obtained;

[0018] S2.3, in the overall heat trend data, according to the customer group portrait as the screening condition, the accurate analysis is carried out, and the accurate heat trend data corresponding to the customer group portrait is obtained;

[0019] S2.4, the sales data is combined with the sales customer list to carry out order screening, and the historical order data corresponding to the customer is obtained.

[0020] As a further improvement of the technical solution, the steps of S2.2 in the overall analysis of the heat trend data are as follows:

[0021] S2.2.1, the image extraction and introduction analysis of the lace type are carried out, and the image data and introduction data corresponding to the lace type are obtained;

[0022] S2.2.2, search for products in the network sales platform through image data and introduction data, and analyze the product heat trend of the searched products, obtain the product heat trend of each product, and aggregate the product heat trend as overall heat trend data.

[0023] As a further improvement of the technical solution, the steps of S3 are as follows:

[0024] S3.1, fluctuation analysis of customers is performed by combining accurate heat trend data with historical order data, and the fluctuation coefficient of the influence of accurate heat trend data on customer order data when heat trend fluctuates is obtained according to the analysis result;

[0025] S3.2, development analysis is performed according to the overall heat trend data, the development state of the lace type in the market is obtained, and the development coefficient is set according to the development state;

[0026] If the overall heat trend data continues to rise, the development coefficient is closer to 1;

[0027] If the overall heat trend data continues to decline, the development coefficient is farther away from 1;

[0028] 0 < development coefficient ≤ 1.

[0029] As a further improvement of the technical solution, the steps of S4 are as follows:

[0030] S4.1, extract the latest order data of the customer, perform initial analysis of the predicted order data by combining the latest order data with the fluctuation coefficient, obtain the initial analysis of the predicted order data, and then perform adjustment analysis of the initial analysis of the predicted order data by combining the development coefficient, obtain the final analysis of the predicted order data;

[0031] S4.2, difference analysis is performed on the customer according to the order quantity in the historical order data of the customer, reference period filtering is performed on the historical order data according to the difference analysis, and the corresponding reference period of the customer is obtained;

[0032] The greater the difference analysis result is, the longer the reference period is;

[0033] The smaller the difference analysis result is, the shorter the reference period is;

[0034] S4.3, extract the historical order data according to the reference period obtained in S4.2, and then perform quantity difference analysis on the historical order data and the historical predicted order data in the same period, and obtain the quantity difference value of the historical order data and the historical predicted order data in the same period.

[0035] As a further improvement of the technical solution, the steps of S5 are as follows:

[0036] S5.1, extract the maximum quantity difference value obtained in S4.3, and limit the range of the predicted order data by the maximum quantity difference value, so that the difference between the predicted order data and the latest order data does not exceed the maximum quantity difference value;

[0037] S5.2, compare the predicted order data with the inventory data, when the predicted order data is less than the inventory data, calculate the difference between the predicted order data and the inventory data, and send an order to the supply chain according to the difference calculation result, otherwise, when the predicted order data is greater than the inventory data, do not send an order to the supply chain.

[0038] Compared with the prior art, the beneficial effects of the present application are:

[0039] 1. In the lace fabric supply chain management method based on big data, accurate prediction is realized through multi-dimensional data fusion, on the one hand, the accurate heat information is extracted and the fluctuation coefficient is calculated by combining the customer group portrait and the network heat trend data, and the influence degree of market heat on customer order is determined, on the other hand, the reference period is selected by combining historical order difference analysis, and the maximum quantity difference is used to limit the prediction range, so as to avoid the deviation of prediction data from the actual demand, so that the enterprise can master the real demand of customers in advance, reduce the inventory accumulation caused by overestimation of prediction, and avoid the shortage problem caused by underestimation of prediction, and greatly reduce the waste of supply chain resources.

[0040] 2. In the lace fabric supply chain management method based on big data, the final predicted order data is directly compared with the real-time inventory data to form a clear replenishment decision logic, when the predicted order data is less than the inventory data, an order is sent to the supply chain according to the difference between the two; if the predicted data is greater than the inventory data, no order is sent, the inventory linkage mechanism based on data can accurately control the inventory level, reduce unnecessary inventory funds, and avoid additional costs caused by emergency procurement due to insufficient inventory, significantly improve the inventory turnover rate, and reduce the overall operating cost.

[0041] 3. In the lace fabric supply chain management method based on big data, the overall heat trend data of lace types in the network is obtained in real time, the market development state of the product is judged by the development coefficient analysis, and the order prediction is adjusted accordingly, when the market heat rises, the development coefficient is close to 1, the predicted order quantity can be timely increased to seize the market opportunity, when the heat decreases, the development coefficient deviates from 1, the prediction can be reduced to avoid overstock, and the ability to quickly respond to market changes enables the enterprise to flexibly adjust the supply chain strategy, seize the opportunity in market competition, and enhance the core competitiveness. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 The overall flowchart of the present application is shown in the figure;

[0043] Figure 2A flow chart of lace type extraction according to inventory data and sales data of the present application;

[0044] Figure 3 A flow chart of customer group portrait analysis according to the sales customer list of the present application;

[0045] Figure 4 A flow chart of development coefficient setting according to the development state of the present application;

[0046] Figure 5 A flow chart of extracting the latest order data of the customer of the present application;

[0047] Figure 6 A flow chart of range limiting of the predicted order data by the maximum quantity difference of the present application. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0049] Please refer to Figure 1 - Figure 6 As shown in the figure, the embodiment aims to provide a lace fabric supply chain management method based on big data, which comprises the following steps:

[0050] S1, obtaining inventory data and sales data of the enterprise, obtaining lace fabric corresponding lace types of the enterprise at the same time, and extracting sales customer list in the sales data; obtaining core basic data of supply chain management, laying the foundation for subsequent analysis;

[0051] S1 extracts lace fabric related inventory data and sales data in the enterprise management terminal by establishing data connection with the enterprise management terminal.

[0052] The steps of S1 are as follows:

[0053] S1.1, according to the inventory data (screening out all the inventory records of lace fabric categories) and the sales data (screening out all the sales records of lace fabric categories), extracting lace types, obtaining lace types (such as pattern style, process type, etc.) corresponding to the lace fabric sold by the enterprise, identifying and determining the lace types corresponding to the lace fabric sold by the enterprise, and forming a complete lace type list;

[0054] S1.2, sales customer list extraction is performed in the sales data, customer identification information is extracted from the filtered transaction records, and the customer identification information corresponding to the same lace type is de-duplicated to form a sales customer list corresponding to the lace type.

[0055] Thus, the association between the lace type and the corresponding sales customer list is established, and the data matching is completed.

[0056] S2, customer group portrait analysis is performed according to the sales customer list, then the lace type is combined with the customer group portrait for precise analysis and overall analysis of the heat trend data in the network, and the sales data is combined with the sales customer list for order screening to obtain the historical order data corresponding to the customer; combined with internal customer data and external market data, three key analyses of customer portrait, market heat, and historical order are completed;

[0057] The steps of S2 are as follows:

[0058] S2.1, customer group portrait analysis is performed according to the sales customer list to obtain the customer group portrait corresponding to the sales customer list as a whole, and the steps are as follows:

[0059] The sales customer list is sorted, the basic information (such as the region, enterprise type, and procurement scale) and transaction characteristics (such as procurement frequency, preferred product type, and order amount) of each customer in the list are collected, the collected customer information is classified and counted, for example, the customer distribution ratio is divided according to the region, and the customer level is divided according to the procurement scale, then the common characteristics of the customer group are refined, such as the mainstream procurement region, the main enterprise type, and the high-frequency lace type, and these common characteristics are integrated to form the customer group portrait corresponding to the sales customer list as a whole, covering the basic attributes and procurement behavior preferences of the customer.

[0060] S2.2, overall analysis of the heat trend data in the network is performed according to the lace type to obtain the overall heat trend data of the lace type in the network;

[0061] The steps of S2.2 in the overall analysis of the heat trend data are as follows:

[0062] S2.2.1, image extraction (collecting the corresponding physical image or design drawing, extracting clear image data) and introduction analysis (including process characteristics, application scenarios, style description, etc., forming introduction data) are performed on the lace type to obtain the image data and introduction data corresponding to the lace type;

[0063] S2.2.2, product search is performed in the network sales platform through the image data and introduction data, and product heat trend analysis is performed on the searched products to obtain the product heat trend of each product, and the product heat trend is summarized as overall heat trend data.

[0064] Wherein, the lace type image data and introduction data are taken as search keywords to search products on mainstream network sales platforms (such as e-commerce platforms, fabric procurement platforms, etc.), collect product information highly matched with the lace type in the search results, record the sales data, user interaction data and time dimension information of each product, then analyze the sales and interaction amount change trend of each matched product in different time periods to determine the heat trend of the product, meanwhile, integrate the heat trend data of all matched products, and weight and integrate the heat trends of similar products to form the overall heat trend data of the lace type in the network, and the formula is as follows:

[0065] ;

[0066] Wherein, H j (t) is the heat value of the jth product at time t, S j (t) is the standardized sales of the jth product at time t, I j (t) is the standardized interaction amount of the jth product at time t, and a and β are corresponding weight coefficients.

[0067] ;

[0068] Wherein, H total (t) is the overall heat trend value of the lace type at time t, n is the total number of matched products searched, w j is the weight of the jth product.

[0069] S2.3, in the overall heat trend data, according to the customer group portrait as a screening condition for accurate analysis, obtain the accurate heat trend data corresponding to the customer group portrait; based on the overall heat trend data, comb the target customer feature labels corresponding to the products in the data, set the screening condition according to the core features of the customer group portrait, screen out the product data highly matched with the portrait features in the overall heat trend data, and at the same time, analyze the heat trend of the screened product data, calculate the heat change in different time periods, then integrate the screened data and analysis results to form the accurate heat trend data corresponding to the customer group portrait, which reflects the product heat change law concerned by the target customer group;

[0070] S2.4, combine the sales data with the sales customer list to screen orders and obtain the historical order data corresponding to the customers.

[0071] S3, combine the accurate heat trend data with the historical order data to analyze the fluctuation of customers, obtain the fluctuation coefficient of the heat trend data change to customers, and analyze the development coefficient through the overall heat trend data; quantify the influence of market heat on customers and the market development potential of products;

[0072] The steps of S3 are as follows:

[0073] S3.1, combine the accurate heat trend data with the historical order data to analyze the fluctuation of the customer, and obtain the fluctuation coefficient of the accurate heat trend data on the order data of the customer when the heat appears trend fluctuation;

[0074] Among them, the historical order data of the customer is sorted, the changes of key indicators such as order quantity and order amount with time are extracted, the baseline fluctuation range of order data is determined, the time period with obvious fluctuation in the accurate heat trend data is extracted, then the heat fluctuation period and the corresponding period of customer order data change are compared and analyzed, the relevance of heat fluctuation and order data change (such as whether heat rise is accompanied by order quantity increase) is identified, then the degree of deviation of customer order data from the baseline fluctuation range when heat fluctuation occurs is calculated, the influence weight in different fluctuation scenarios is determined according to the deviation degree, the influence weights of various fluctuation scenarios are comprehensively considered, and the fluctuation coefficient of the influence of accurate heat trend data fluctuation on customer order data is obtained, and the formula is as follows:

[0075] ;

[0076] Among them, F is the fluctuation coefficient, n is the total number of heat fluctuation periods selected, O i is the actual order data in the i-th fluctuation period, is the baseline average value of order data, R i is the weight of the i-th fluctuation period, H i is the heat value in the i-th fluctuation period, is the baseline average value of heat data;

[0077] S3.2, according to the overall heat trend data, the development of the lace type in the market is analyzed, and the development coefficient is set according to the development state;

[0078] If the overall heat trend data continues to rise, the development coefficient is closer to 1;

[0079] If the overall heat trend data continues to decline, the development coefficient is farther away from 1;

[0080] 0 < development coefficient ≤ 1;

[0081] Wherein, select a certain time period (half a year) of the overall heat trend data, comb the data curve with time, analyze the overall trend of the change curve, and judge the market development state of the lace type: if the curve continues to rise, it indicates that the market is in the growth stage; if the curve continues to decline, it indicates that the market is in the recession stage; if the curve is smooth and fluctuates, it indicates that the market is in a stable stage, and then set the initial development coefficient benchmark value according to the development state (such as 0.5 in the stable stage, and adjust the benchmark value according to the heat trend change rate: the more obvious the upward trend, the closer the development coefficient to 1; the more obvious the downward trend, the closer the development coefficient to 0, finally determine the development coefficient of the lace type in the current market environment, the formula is as follows:

[0082] ;

[0083] Wherein, D is the development coefficient, e is the natural constant, k is the adjustment coefficient, and ΔH is the change rate of the overall heat trend.

[0084] S4, extract the latest order data, combine the fluctuation coefficient and the development coefficient to analyze the predicted order data, then analyze the difference of the order quantity in the historical order data, filter the reference period according to the difference analysis result, and analyze the quantity difference according to the filtered reference period and the historical predicted order data; combine the latest order, coefficient data to predict the demand, and verify the prediction accuracy through historical data;

[0085] The steps of S4 are as follows:

[0086] S4.1, extract the latest order data of the customer, combine the latest order data with the fluctuation coefficient to analyze the initial predicted order data, obtain the initial analysis of the predicted order data, then combine the initial analysis of the predicted order data with the development coefficient to adjust the analysis (appropriately increase the initial predicted value when the market rises), and obtain the final analysis of the predicted order data; the formula is as follows:

[0087] ;

[0088] Wherein, O init is the initial predicted order data, O new is the latest order data of the customer, sign (ΔH precise ) is the sign coefficient of the accurate heat trend change, if the accurate heat rises, ΔH precise > 0, the coefficient value is 1; if it decreases, ΔH precise < 0, the coefficient value is -1;

[0089] ;

[0090] Wherein, O finalFor the final prediction order data, (D-0.5) x 2 is the development state adjustment factor. When D = 1, the factor is 1, corresponding to the maximum adjustment range of market rise; when D = 0.5, the factor is 0, corresponding to market stability without adjustment; when D < 0.5, the factor is negative, corresponding to market decline adjustment.

[0091] S4.2, difference analysis on the number of orders in the historical order data of the customer, reference period screening of the historical order data according to the difference analysis, and obtaining the reference period corresponding to the customer;

[0092] The greater the difference analysis result is, the longer the reference period is;

[0093] The smaller the difference analysis result is, the shorter the reference period is;

[0094] Among them, the historical order data of the customer in the past period (1 year) is sorted, the maximum value and the minimum value difference of the order quantity in different periods (such as every month, every quarter) are calculated, and the difference degree of the order quantity is evaluated. According to the difference analysis result, the reference period length is determined - if the order quantity fluctuation difference is large (such as the maximum value and the minimum value difference is large), it means that the customer order stability is poor, and a longer reference period (such as 12 months) needs to be selected to cover more fluctuation scenarios; if the difference is small, it means that the order stability is good, and a shorter reference period (such as 6 months) can be selected;

[0095] S4.3, extracting historical order data according to the reference period obtained in S4.2, and then combining the historical order data in the same period with the historical prediction order data to perform quantity difference analysis, and obtaining the quantity difference value of the historical order data in the same period and the historical prediction order data;

[0096] S5, extracting the maximum quantity difference value of the analysis result of S4, and limiting the prediction order data through the maximum quantity difference value, and then combining the prediction order data with the inventory data for comparison, and sending orders to the supply chain according to the comparison result. Based on the prediction data and the inventory data, the final supply chain order sending decision is given, which is the landing link of the whole method;

[0097] The steps of S5 are as follows:

[0098] S5.1, extracting the maximum quantity difference value obtained in S4.3, limiting the prediction order data through the maximum quantity difference value, so that the difference value between the prediction order data and the latest order data does not exceed the maximum quantity difference value;

[0099] Among the set of quantity difference values between historical orders and historical predicted orders, the largest absolute value is filtered out to determine the maximum quantity difference value. Based on the latest order data, the upper and lower limit ranges of the predicted order data are set in combination with the maximum quantity difference value (i.e., the predicted order data cannot exceed the latest order data plus the maximum quantity difference value, nor can it be lower than the latest order data minus the maximum quantity difference value). Then, it is checked whether the final predicted order data obtained previously is within the range. If it is out of the range, it is adjusted to the nearest boundary value. If it is within the range, it remains unchanged, ensuring that the difference between the predicted order data and the latest order data does not exceed the maximum quantity difference value.

[0100] S5.2, compare the predicted order data with the inventory data. When the predicted order data is less than the inventory data, calculate the difference between the predicted order data and the inventory data, and send an order to the supply chain according to the difference calculation result. Conversely, when the predicted order data is greater than the inventory data, do not send an order to the supply chain.

[0101] Among the set of quantity difference values between historical orders and historical predicted orders, the largest absolute value is filtered out to determine the maximum quantity difference value. Based on the latest order data, the upper and lower limit ranges of the predicted order data are set in combination with the maximum quantity difference value (i.e., the predicted order data cannot exceed the latest order data plus the maximum quantity difference value, nor can it be lower than the latest order data minus the maximum quantity difference value). Then, it is checked whether the final predicted order data obtained previously is within the range. If it is out of the range, it is adjusted to the nearest boundary value. If it is within the range, it remains unchanged, ensuring that the difference between the predicted order data and the latest order data does not exceed the maximum quantity difference value.

[0102] If the predicted order data is greater than or equal to the inventory data, it means that the existing inventory is sufficient to meet the predicted demand, and there is no need to send an order to the supply chain, ending the process.

[0103] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above embodiments, and the above embodiments and descriptions in the specification are only preferred examples of the present application and are not intended to limit the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A big data-based supply chain management method for lace fabrics, characterized in that: Includes the following steps: S1. Obtain the company's inventory data and sales data, and at the same time obtain the lace type corresponding to the company's lace fabric, and extract the sales customer list from the sales data. S2. Based on the sales customer list, conduct customer group profile analysis, and then combine the lace type with the customer group profile to conduct precise and overall analysis of the popularity trend data on the network. At the same time, combine the sales data with the sales customer list to screen orders and obtain the corresponding historical order data of customers. S3. Combine accurate popularity trend data with historical order data to conduct fluctuation analysis on customers, obtain the fluctuation coefficient of popularity trend data changes on customers, and conduct development coefficient analysis through overall popularity trend data; The steps in S3 are as follows: S3.

1. Combine accurate popularity trend data with historical order data to conduct fluctuation analysis on customers, and obtain the fluctuation coefficient of the impact of accurate popularity trend data on customer order data when the popularity trend fluctuates based on the analysis results. ； Where F is the fluctuation coefficient, n is the total number of selected heat fluctuation periods, and O i This represents the actual order data for the i-th fluctuation period. R is the baseline average of the order data. i H represents the weight for the i-th fluctuation period. i The heat value during the i-th fluctuation period. This is the baseline average value of the heat index data; S3.

2. Conduct development analysis based on overall popularity trend data to obtain the development status of this type of lace in the market, and set the development coefficient based on the development status. ； Where D is the development coefficient, e is the natural constant, and k is the adjustment coefficient. H represents the rate of change of the overall popularity trend; S4. Extract the latest order data, combine the latest order data with the fluctuation coefficient and the development coefficient to conduct predictive order data analysis, then conduct difference analysis on the order quantity in the historical order data, filter the historical order data for reference periods based on the difference analysis results, and conduct quantity difference analysis based on the filtered reference periods and historical predicted order data. The steps in S4 are as follows: S4.1 Extract the latest order data from customers, combine the latest order data with the volatility coefficient to perform initial analysis of predicted order data, obtain the predicted order data of the initial analysis, and then combine the predicted order data of the initial analysis with the development coefficient to perform adjustment analysis, and obtain the predicted order data of the final analysis. ； Wherein, O init is the initial predicted order data, O new is the customer's latest order data, sign (ΔH precise ) is the sign coefficient of the accurate heat trend change, if the accurate heat rises, ΔH precise > 0, the coefficient value is 1; if it falls, ΔH precise < 0, the coefficient value is -1; ； Among them, O final For the final predicted order data, (D-0.5)×2 is the development status adjustment factor. When D=1, the factor is 1, corresponding to the maximum adjustment range when the market rises; when D=0.5, the factor is 0, corresponding to no adjustment when the market is stable; when D<0.5, the factor is negative, corresponding to adjustment when the market falls. S4.2 Analyze the difference in the number of orders in the customer's historical order data, filter the historical order data for reference time periods based on the difference analysis, and obtain the corresponding reference time period for the customer; The greater the difference in the analysis results, the longer the reference period. The smaller the difference analysis result, the shorter the reference period; S4.3 Extract historical order data based on the reference time period obtained in S4.2, and then perform quantity difference analysis by combining the historical order data of the same period with the historical predicted order data to obtain the quantity difference between the historical order data and the historical predicted order data of the same period. S5. Extract the maximum quantity difference from the analysis results of S4, and limit the range of the predicted order data by the maximum quantity difference. Then, compare the predicted order data with the inventory data, and send the order to the supply chain based on the comparison results.

2. The method for managing the lace fabric supply chain based on big data according to claim 1, characterized in that: S1 establishes a data connection with the enterprise management terminal to extract inventory and sales data related to lace fabrics.

3. The method for managing the lace fabric supply chain based on big data according to claim 1, characterized in that: The steps in S1 are as follows: S1.1 Extract lace types based on inventory and sales data to obtain the lace types corresponding to the lace fabrics sold by the company; S1.2 Extract the sales customer list from the sales data to obtain the sales customer list corresponding to the lace type.

4. The method for managing the lace fabric supply chain based on big data according to claim 1, characterized in that: The steps in S2 are as follows: S2.

1. Based on the sales customer list, conduct customer group profile analysis to obtain the overall customer group profile corresponding to the sales customer list; S2.2 Analyze the overall popularity trend data of lace type on the network to obtain the overall popularity trend data of lace type on the network; S2.

3. In the overall popularity trend data, conduct precise analysis based on customer group profiles as the filtering conditions to obtain the precise popularity trend data corresponding to the customer group profiles. S2.

4. Combine sales data with the sales customer list to screen orders and obtain the corresponding historical order data for each customer.

5. The method for managing the lace fabric supply chain based on big data according to claim 4, characterized in that: The steps in S2.2 for the overall analysis of popularity trend data are as follows: S2.2.

1. Perform image extraction and description analysis on the lace types to obtain the image data and description data corresponding to the lace types; S2.2.

2. Search for products on the online sales platform using image data and description data, analyze the product popularity trend of the searched products, obtain the product popularity trend of each product, and summarize them as overall popularity trend data.

6. The method for supply chain management of lace fabric based on big data according to claim 1, characterized in that: If the overall popularity trend data continues to rise, the development coefficient will be closer to 1. If the overall popularity trend continues to decline, the development coefficient will move further away from 1. 0 < development coefficient ≤ 1.

7. The method for supply chain management of lace fabric based on big data according to claim 1, characterized in that: The steps in S5 are as follows: S5.1 Extract the maximum quantity difference obtained in S4.3, and use the maximum quantity difference to limit the range of the predicted order data so that the difference between the predicted order data and the latest order data does not exceed the maximum quantity difference; S5.

2. Compare the predicted order data with the inventory data. When the predicted order data is less than the inventory data, calculate the difference between the predicted order data and the inventory data, and send the order to the supply chain based on the difference calculation result. Conversely, when the predicted order data is greater than the inventory data, do not send the order to the supply chain.

Images