Supply chain material data management method and data management system

Abstract

The invention discloses a supply chain material data management method and a data management system, and particularly relates to the technical field of data management. Through collecting and centrally storing material data of each link of a supply chain, an inventory turnover rate and residence time are analyzed, a safe inventory level is evaluated, and a purchase plan and inventory control are optimized. In order to solve the possible problem that inventory data is not updated in time or input errors occur, interference of the inventory data on inventory turnover rate calculation is evaluated through real-time analysis and abnormal input monitoring, and the accuracy of the inventory data is ensured; according to the method, the accuracy of the inventory turnover rate calculation result is divided into high, medium and low accuracy levels, and corresponding processing is carried out, so that purchase, production and supply chain decisions of enterprises are optimized, strategic decision errors are reduced, and the overall efficiency and response capability of a supply chain are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data management, and in particular to a supply chain material data management method and a data management system. Background Art

[0002] Supply chain material data management refers to the process of collecting, storing, analyzing and optimizing data related to materials in the supply chain (such as raw materials, semi-finished products, finished products, etc.). These data include inventory, material turnover speed, supplier information, transportation status, etc. The purpose is to ensure that each link in the supply chain has accurate and real-time material data support, so as to optimize production, reduce costs and reduce delays.

[0003] In the existing technology, the first step of supply chain management is to collect and centrally store the material data related to each link (such as procurement, production, warehousing and logistics) in a unified data platform to ensure data integration and real-time update. Subsequently, by analyzing the inventory turnover rate and the retention time of materials, the safety inventory level of various materials is evaluated, thereby optimizing the procurement plan and inventory control, ensuring that materials can be replenished in time, reducing inventory backlogs, and improving the overall efficiency and responsiveness of the supply chain. However, if the inventory data is not updated in real time, or if data entry errors are caused by system problems, the inventory turnover rate may be distorted. At the same time, due to incorrect data on inventory turnover, management may make long-term strategic decisions on procurement, production and supply chain adjustments based on inaccurate information. For example, a company may significantly expand its production capacity or adjust its supply chain structure, but in fact demand has not increased. These strategic decision errors may not be apparent in the short term, but in the long run, they will have irreversible effects on the company's resource allocation, financial health and market position. Summary of the invention

[0004] The purpose of the present invention is to provide a supply chain material data management method and a data management system to solve the deficiencies in the background technology.

[0005] In order to achieve the above object, the present invention provides the following technical solution: a supply chain material data management method, comprising the following steps:

[0006] S1: Obtain basic inventory data, including inventory quantity, sales volume, purchase quantity and inventory turnover days of each material, and calculate the actual inventory turnover rate based on the basic data;

[0007] S2: Analyze the fluctuation of the real-time update frequency of inventory data during the inventory management process, determine the deviation between inventory data and actual inventory status, and evaluate the real-time nature of data updates during the inventory management process;

[0008] S3: Monitor the integrity of data entry during inventory management, analyze the changes in the abnormal data entry ratio during inventory management, and determine the degree of interference of the abnormal data entry ratio changes on the inventory turnover rate calculation;

[0009] S4: Evaluate the accuracy of the actual inventory turnover rate calculation results based on the real-time nature of data updates during inventory management and the degree of interference of changes in the abnormal input ratio on the inventory turnover rate calculation;

[0010] S5: classify the accuracy of the actual inventory turnover rate calculation results into different levels, namely, high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results;

[0011] S6: The actual inventory turnover rate calculated with high accuracy is directly applied to long-term procurement and production planning to support management in making strategic decisions; the actual inventory turnover rate calculated with low accuracy is manually verified before being used for decision making;

[0012] S7: For actual inventory turnover rates with medium accuracy calculation results, the degree of abnormality of the accuracy of the actual inventory turnover rate calculation results within a fixed time period is predicted, and the calculation weights of the actual inventory turnover rate are dynamically adjusted based on the prediction results to optimize procurement and inventory control capabilities.

[0013] Preferably, in S2, after analyzing the fluctuation degree of the update frequency in different time periods, an abnormal update frequency fluctuation index is generated, and the method for obtaining the abnormal update frequency fluctuation index is:

[0014] Obtain the update frequency value calculated in the T time period in real time, determine a smoothing parameter α, set the initial value, that is, at time s = 0, set the initial update frequency, the expression is: EWMA 0 =Update frequency 0 ; Starting from time s = 1, the exponentially weighted moving average EWMA at each time point is calculated by the EWMA at the previous time point and the current update frequency. The expression is: EWMA s =α·Update frequency s+(1-α)·EWMA s-1 ; After obtaining the EWMA value at each time point, calculate the abnormal fluctuation index of the update frequency, the expression is: Where ER is the abnormal fluctuation index of update frequency.

[0015] Preferably, in S3, after analyzing the change of abnormal input ratio in different materials, an abnormal input ratio drift index is generated, and the method for obtaining the abnormal input ratio drift index is:

[0016] Construct abnormal input ratio dataset {x1 ,x 2 ,...,x r}, where x r Represents the abnormal input ratio at time r; divide the abnormal input ratio value in the data set into data in N time periods, calculate the mean μ and standard deviation σ of the abnormal input ratio value in the kth time period; for each time point r, calculate the posterior probability of it being a change point, the expression is: In the formula, P(change point|x 1:r ) is the posterior probability of a change point occurring at time r, P(x r | Change point) is the likelihood function value of the abnormal input ratio change at time r, P(change point) is the prior probability, indicating the prior hypothesis of the change point, P(x r ) is the total observed probability at time point r; the posterior probability P(change point|x 1:r ) is compared with the benchmark value of the change point posterior probability set under normal conditions in historical data. When P(change point|x 1:r ) is greater than or equal to the change point posterior probability benchmark value, it is considered that time r has drifted. When a change point is detected, the change point strength and influence are calculated, and the abnormal input ratio drift index is generated by the drift strength. The calculation expression of the drift strength is: In the formula, μ after is the mean of the new data segment after the change point, μ before is the mean of the old data segment before the change point; the drift intensity and the posterior probability are combined to generate the abnormal input ratio drift index, the expression is: KG = drift intensity r ×P(change point|x 1:r ), where KG is the abnormal input ratio drift index.

[0017] Preferably, in S4, the accuracy of the actual inventory turnover rate calculation result is evaluated according to the real-time nature of data update in the inventory management process and the degree of interference of abnormal input ratio changes on the inventory turnover rate calculation;

[0018] The abnormal fluctuation index of update frequency and the drift index of abnormal entry ratio are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses the accuracy value label of the actual inventory turnover calculation result predicted by each group of first eigenvectors as the prediction target, and takes minimizing the sum of prediction errors of the accuracy value labels of all actual inventory turnover calculation results as the training target. The machine learning model is trained until the sum of prediction errors converges, then the model training is stopped, and the accuracy value of the actual inventory turnover calculation result is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0019] Preferably, in S5, the accuracy of the actual inventory turnover rate calculation result is divided into different levels, namely, a high accuracy calculation result, a medium accuracy calculation result and a low accuracy calculation result, specifically:

[0020] Compare the accuracy value of the actual inventory turnover rate calculation result obtained with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and compare the accuracy value of the actual inventory turnover rate calculation result with the first standard threshold and the second standard threshold respectively;

[0021] If the accuracy value of the actual inventory turnover rate calculation result is greater than the second standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is high, and a high accuracy calculation signal is generated, and it is classified as a high accuracy calculation result;

[0022] If the accuracy value of the actual inventory turnover rate calculation result is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is average, and a medium accuracy calculation signal is generated at this time, and it is classified as a medium accuracy calculation result;

[0023] If the accuracy value of the actual inventory turnover rate calculation result is less than the first standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is low. At this time, a low accuracy calculation signal is generated and it is classified as a low accuracy calculation result.

[0024] Preferably, in S7, for the actual inventory turnover rate calculated with medium accuracy, the degree of abnormality of the accuracy of the actual inventory turnover rate calculation result within a fixed time period is predicted, and the calculation weight of the actual inventory turnover rate is dynamically adjusted according to the prediction result, specifically:

[0025] Set a fixed time period M to predict the accuracy of the inventory turnover rate calculation results, define the accuracy abnormality value of the inventory turnover rate calculation as E(t), and the calculation expression is: In the formula, AR actual (t) is the actual inventory turnover rate calculation result, ARpredicted(t) is the inventory turnover rate predicted based on historical data and the model;

[0026] By performing time series prediction on the abnormality degree E(t) in several time periods, the abnormality degree in the next time period is predicted. The expression is: is the predicted abnormality level in the next time period, f is the prediction function, and E(th) is the abnormality level in the past h time periods;

[0027] According to the predicted abnormality Dynamically adjust the calculation weight of inventory turnover rate to optimize inventory and purchasing decisions; set an initial weight W 0 , and correct it according to the prediction results. The expression is: Where W(t+1) is the calculation weight of inventory turnover rate at time t+1, W 0 is the initial weight, which indicates the basic weight of inventory turnover rate in decision making, and β is the adjustment coefficient.

[0028] Preferably, combined with the dynamically adjusted inventory turnover rate weight W(t+1), when making purchase and inventory decisions, the actual inventory turnover rate used will be corrected according to the calculated weight, thereby improving the accuracy of purchase and inventory control; the actual inventory turnover rate used in the final purchase and inventory decision is: AR adjusted (t+1)=W(t+1)·AR(t+1); AR adjusted (t+1) is the adjusted inventory turnover rate, which is used for actual purchasing and inventory decisions, and AR(t+1) is the original inventory turnover rate calculation result at time t+1.

[0029] The present invention also provides a supply chain material data management system, including an actual inventory turnover rate calculation module, an inventory data update real-time evaluation module, an abnormal input ratio change module, an accuracy evaluation module, an accuracy classification module, a decision application module and a dynamic adjustment module;

[0030] Actual inventory turnover rate calculation module: obtains basic inventory data, including the inventory quantity, sales volume, purchase quantity and inventory turnover days of each material, and calculates the actual inventory turnover rate based on the basic data;

[0031] Inventory data update real-time evaluation module: Analyze the fluctuation of the real-time update frequency of inventory data during the inventory management process, determine the deviation between inventory data and actual inventory status, and evaluate the real-time update of data during the inventory management process;

[0032] Abnormal input ratio change module: monitors the integrity of data input during inventory management, analyzes the abnormal input ratio change of data during inventory management, and determines the degree of interference of abnormal input ratio change on inventory turnover rate calculation;

[0033] Accuracy evaluation module: evaluates the accuracy of actual inventory turnover calculation results based on the real-time nature of data updates during inventory management and the degree of interference of abnormal input ratio changes on inventory turnover calculation;

[0034] Accuracy classification module: divides the accuracy of the actual inventory turnover rate calculation results into different levels, namely, high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results;

[0035] Decision-making application module: The actual inventory turnover rate calculated with high accuracy is directly applied to long-term procurement and production planning to support management in making strategic decisions; the actual inventory turnover rate calculated with low accuracy is manually verified before being used for decision-making;

[0036] Dynamic adjustment module: For actual inventory turnover rates with medium accuracy calculation results, the module predicts the degree of abnormality of the accuracy of the actual inventory turnover rate calculation results within a fixed time period, and dynamically adjusts the calculation weights of the actual inventory turnover rate based on the prediction results to optimize procurement and inventory control capabilities.

[0037] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0038] 1. The present invention collects and stores material data of each link in real time, and comprehensively evaluates the real-time and accuracy of the data with the help of the abnormal fluctuation index of update frequency and the abnormal input ratio drift index, thus solving the problem of inventory turnover rate distortion caused by untimely data update or input errors in the prior art. By introducing a machine learning model, the accuracy of the actual inventory turnover rate is evaluated according to the real-time nature of data update and the changes in input anomalies, and it is divided into different levels such as high, medium and low, so that the management can flexibly adjust the procurement plan and production strategy according to the accuracy of the data, effectively improving the overall efficiency and responsiveness of the supply chain.

[0039] 2. The present invention also uses a dynamic adjustment mechanism for medium-accuracy calculation results to dynamically adjust the weight of inventory turnover rate according to the forecast of abnormal degree of inventory turnover rate within a fixed time period, and optimizes procurement and inventory control decisions. This mechanism can effectively reduce decision-making errors caused by data deviations, ensure that reasonable adjustments and judgments can still be made when data is not completely reliable, thereby improving the efficiency of enterprise resource allocation, reducing inventory backlogs and shortages, and providing solid data support for the long-term strategic decision-making of enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0041] Figure 1 The figure is a flow chart of the method of the present invention.

[0042] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] Example 1, please refer to Figure 1 and Figure 2 As shown, a supply chain material data management method described in this embodiment includes the following steps:

[0045] S1: Obtain basic inventory data, including inventory quantity, sales volume, purchase quantity and inventory turnover days of each material, and calculate the actual inventory turnover rate based on the basic data;

[0046] S2: Analyze the fluctuation of the real-time update frequency of inventory data during the inventory management process, determine the deviation between inventory data and actual inventory status, and evaluate the real-time nature of data updates during the inventory management process;

[0047] S3: Monitor the integrity of data entry during inventory management, analyze the changes in the abnormal data entry ratio during inventory management, and determine the degree of interference of the abnormal data entry ratio changes on the inventory turnover rate calculation;

[0048] S4: Evaluate the accuracy of the actual inventory turnover rate calculation results based on the real-time nature of data updates during inventory management and the degree of interference of changes in the abnormal input ratio on the inventory turnover rate calculation;

[0049] S5: classify the accuracy of the actual inventory turnover rate calculation results into different levels, namely, high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results;

[0050] S6: The actual inventory turnover rate calculated with high accuracy is directly applied to long-term procurement and production planning to support management in making strategic decisions; the actual inventory turnover rate calculated with low accuracy is manually verified before being used for decision making;

[0051] S7: For actual inventory turnover rates with medium accuracy calculation results, the degree of abnormality of the accuracy of the actual inventory turnover rate calculation results within a fixed time period is predicted, and the calculation weights of the actual inventory turnover rate are dynamically adjusted based on the prediction results to optimize procurement and inventory control capabilities.

[0052] In S1, basic inventory data is obtained, including the inventory quantity, sales volume, purchase quantity and inventory turnover days of each material, and the actual inventory turnover rate is calculated based on the basic data, specifically:

[0053] Inventory refers to the actual quantity of a certain material in the current warehouse. It reflects the total amount of available materials that the company can use for production or sales. It is usually updated in real time through the enterprise resource planning (ERP) system, warehouse management system (WMS) or inventory management software. Inventory is the basic data for inventory management. It directly affects procurement and production plans. If the inventory is insufficient, the company may not be able to meet production needs or customer orders. Including: sales volume, sales volume refers to the total amount of materials sold by the company in a certain period of time. It reflects market demand and the speed of inventory consumption. It is obtained through the sales management system or ERP system, and can generally be extracted from sales records, customer orders or outbound documents. Sales volume data can help companies predict future demand, adjust inventory levels, and prevent excessive or insufficient inventory. Incoming volume, incoming volume refers to the amount of materials purchased by the company from suppliers and put into storage in a certain period of time. It reflects the replenishment of materials and ensures that the company can maintain sufficient inventory to meet production and sales needs. It is obtained through the procurement management system, ERP system or supply chain management system. It is usually recorded through purchase orders, supplier delivery notes and inbound documents. The purchase volume directly affects the inventory turnover rate. If the purchase volume is insufficient, the inventory may not be able to meet the demand; if the purchase volume is too much, it may lead to inventory backlog and capital occupation. Inventory turnover days refers to the average time from the purchase of materials to their sale. It reflects the length of time the inventory stays in the warehouse and is an auxiliary indicator for calculating inventory turnover. It is automatically generated through the inventory management system or ERP system. The average turnover days can be calculated based on historical data. Inventory turnover days can help companies optimize their inventory management strategies. Shortening the turnover time means better inventory liquidity, reducing capital occupation and storage costs.

[0054] The actual inventory turnover rate is an important indicator calculated based on the basic data of inventory. It is used to measure the flow efficiency of inventory materials. The expression is: Sales volume: the quantity of goods sold in a certain period (such as a month, a quarter, or a year). Average inventory: the average inventory level in a certain period, which can be calculated based on the average of the beginning inventory and the ending inventory; high turnover rate means efficient inventory management, and the company can quickly sell and replenish inventory, reducing inventory costs and storage pressure. Low turnover rate means inventory backlog, which may lead to excessive capital occupation or product expiration and deterioration. Companies may need to adjust their purchasing plans or promotion strategies to improve inventory liquidity.

[0055] In this application, by obtaining basic data on inventory, sales, purchase volume, and inventory turnover days, the actual inventory turnover rate can be accurately calculated. Inventory turnover rate helps companies optimize supply chain and inventory management, improve operational efficiency, and reduce the risk of inventory backlog and capital occupation. By continuously monitoring and adjusting inventory turnover rate, companies can achieve more efficient material mobility to meet production and market needs.

[0056] S2: Analyze the fluctuation of the real-time update frequency of inventory data during the inventory management process, determine the deviation between the inventory data and the actual inventory status, and evaluate the real-time nature of data updates during the inventory management process.

[0057] The frequency of real-time inventory data updates refers to the frequency of inventory data updates in the system after inventory changes (such as warehousing, outbound, transfer, inventory, etc.). Ideally, the system should reflect the latest inventory status in real time after inventory changes. Extract inventory change records from the inventory management system or ERP system, including the actual time of each warehousing, outbound, adjustment and other operations and the time when the system records are updated.

[0058] Get the actual time of inventory changes (such as the time when inventory is received or shipped out). Get the timestamp of the system's internal data update for inventory changes and calculate the update frequency. The expression is: The update frequency of each time period can be expressed by dividing the number of data updates by the length of the time period (for example, in hours, days, or weeks). The higher the frequency, the more timely and frequent the data updates.

[0059] The fluctuation of the inventory data update frequency measures the stability of the system update and reflects whether there is a significant delay or instability in the data update. If the data update frequency fluctuates greatly, it means that the system fails to update the data in a timely manner during certain time periods, resulting in a lag in inventory data, which may deviate significantly from the actual inventory status.

[0060] After analyzing the fluctuation degree of update frequency in different time periods, an abnormal update frequency fluctuation index is generated. The method for obtaining the abnormal update frequency fluctuation index is as follows:

[0061] Obtain the update frequency value calculated in the T time period in real time, and determine a smoothing parameter α, with a value range of 0<α<1. The larger α is, the more dependent it is on the latest update frequency value, usually between 0.1 and 0.3; set the initial value, that is, at time s=0, set the initial update frequency, the expression is: EWMA 0 =Update frequency 0 ; Starting from time s = 1, the exponentially weighted moving average EWMA at each time point is calculated by the EWMA at the previous time point and the current update frequency. The expression is: EWMA s =α·Update frequency s+(1-α)·EWMA s-1 ; After obtaining the EWMA value at each time point, calculate the abnormal fluctuation index of the update frequency, the expression is: Where ER is the abnormal fluctuation index of update frequency.

[0062] When the abnormal fluctuation index of update frequency is larger, it means that the current inventory update frequency has fluctuated significantly compared with the historical smoothed data. At this time, the deviation between the inventory data and the actual inventory status is large, indicating that the inventory data in the system fails to timely or accurately reflect the actual changes in inventory. The sudden increase or decrease in update frequency may indicate that the system fails to update inventory data at the expected frequency during certain periods, resulting in data lags or omissions, which in turn affects the accuracy of inventory management decisions. At this time, the real-time performance of data updates is poor, and timely attention and optimization are required.

[0063] On the contrary, when the update frequency abnormal fluctuation index is small, it means that the current inventory update frequency is relatively stable and close to the historical smoothed data. The deviation between the inventory data and the actual inventory status is small, which means that the system can reflect the changes in inventory in a timely manner, and the data update in the inventory management process is relatively timely and accurate. At this time, the system has better real-time performance and less data lag, which can provide management with reliable inventory information and ensure the effective operation of the supply chain.

[0064] S3: Monitor the integrity of data entry during the inventory management process, analyze the changes in the proportion of abnormal data entry during the inventory management process, and determine the degree of interference of the changes in the proportion of abnormal data entry on the calculation of inventory turnover rate.

[0065] The integrity of data entry refers to whether the inventory data in the inventory management system is complete and without omissions, and whether all inventory changes (such as warehousing, outbound, inventory counting, etc.) are truthfully entered into the system. The integrity of inventory data is the basis for ensuring the accuracy of inventory turnover calculation. If the data is incomplete, the calculation of inventory turnover will be disrupted, causing management to make decisions based on inaccurate information.

[0066] Collect relevant records for each inventory change, including the time, quantity, operator, and other information for each entry, exit, and transfer operation, and check whether there are any missing records or operations that have not been entered into the system. Compare the actual operation records (such as manual records or equipment logs in warehouse management) with the data in the system to identify records that have not been entered or entered incorrectly. The calculation expression for completeness is: If the completeness is close to 100%, it means that the inventory data entry is relatively complete.

[0067] The abnormal input ratio refers to the ratio of data input errors, omissions or duplications to the total data input during the inventory management process. Changes in the abnormal input ratio may interfere with the calculation of inventory turnover. Frequent abnormal inputs may cause distortion of inventory turnover. For example, missing or duplicate data inputs will affect inventory and sales information, causing the calculation of inventory turnover to deviate from the actual situation.

[0068] Identify abnormal data entry in inventory management through inventory counting, system logs or anomaly detection tools. Abnormalities include: Missed entry: Inventory change records that should be entered but are not entered into the system. Duplicate entry: The same inventory change is recorded multiple times. Incorrect entry: Information such as material quantity, entry / exit time, etc. is entered incorrectly. Calculate the abnormal entry ratio, the expression is: The abnormal entry ratio reflects the frequency of abnormal data entry.

[0069] After analyzing the changes in the abnormal input ratios of different materials, the abnormal input ratio drift index is generated. The method for obtaining the abnormal input ratio drift index is as follows:

[0070] Construct abnormal input ratio dataset {x 1 ,x 2 ,...,x r}, where x r Represents the abnormal input ratio at time r; divide the abnormal input ratio value in the data set into data in N time periods, calculate the mean μ and standard deviation σ of the abnormal input ratio value in the kth time period; for each time point r, calculate the posterior probability of it being a change point, the expression is: In the formula, P(change point|x 1:r ) is the posterior probability of a change point occurring at time r, P(x r | Change point) is the likelihood function value of the abnormal input ratio change at time r, P(change point) is the prior probability, indicating the prior hypothesis of the change point, P(x r ) is the total observed probability at time point r; the posterior probability P(change point|x 1:r ) is compared with the benchmark value of the change point posterior probability set under normal conditions in historical data. When P(change point|x1:r ) is greater than or equal to the change point posterior probability benchmark value, then it is considered that time r has drifted significantly. Usually, the benchmark value is set to 0.5 or 0.7 to ensure that significant changes are detected. When a change point is detected, the change point strength and impact are calculated, and the abnormal input ratio drift index is generated by the drift strength, where the calculation expression of the drift strength is: In the formula, μ after is the mean of the new data segment after the change point, μ before is the mean of the old data segment before the change point; the drift intensity and the posterior probability are combined to generate the abnormal input ratio drift index, the expression is: KG = drift intensity r ×P(change point|x 1:r ), where KG is the abnormal input ratio drift index.

[0071] When the abnormal input ratio drift index is larger, it means that the abnormal input ratio has changed significantly in a certain period of time, and the data input quality in the system has fluctuated greatly. This large drift usually has a great interference with the calculation of inventory turnover rate, resulting in a decrease in the accuracy of inventory data. Due to the increase in errors or omissions in inventory data, the errors in inventory turnover days and sales information are increased, and the inventory turnover rate may deviate from the actual value, causing management to make wrong purchasing, inventory or production decisions. Therefore, the larger the drift index, the more serious the interference of the change in the abnormal input ratio on the inventory turnover rate.

[0072] On the contrary, when the drift index of the abnormal input ratio is smaller, it means that the abnormal input ratio has remained relatively stable for a period of time and has not changed significantly. At this time, the input quality of inventory data is relatively stable, there are fewer errors or omissions, and the calculation of inventory turnover rate can better reflect the actual inventory status. The reliability of data input is high, and the error of inventory turnover rate calculation is small, so management can rely on these data to make more accurate decisions. The smaller the drift index, the less interference the change in the abnormal input ratio has on the inventory turnover rate calculation.

[0073] S4: Evaluate the accuracy of actual inventory turnover calculation results based on the real-time nature of data updates during inventory management and the degree of interference of changes in the proportion of abnormal entries on inventory turnover calculation.

[0074] The abnormal fluctuation index of update frequency and the drift index of abnormal entry ratio are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses the accuracy value label of the actual inventory turnover calculation result predicted by each group of first eigenvectors as the prediction target, and takes minimizing the sum of prediction errors of the accuracy value labels of all actual inventory turnover calculation results as the training target. The machine learning model is trained until the sum of prediction errors converges, then the model training is stopped, and the accuracy value of the actual inventory turnover calculation result is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

[0075] The method for obtaining the accuracy value of the actual inventory turnover rate calculation result is: from the first eigenvector training data of the trained machine learning model, obtain the corresponding function expression: AR=F(ER, KG); where F is the output function of the model, ER is the abnormal fluctuation index of the update frequency, KG is the abnormal entry ratio drift index, and AR is the accuracy value of the actual inventory turnover rate calculation result.

[0076] S5: The accuracy of the actual inventory turnover rate calculation results is divided into different levels, which are divided into high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results.

[0077] Compare the accuracy value of the actual inventory turnover rate calculation result obtained with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and compare the accuracy value of the actual inventory turnover rate calculation result with the first standard threshold and the second standard threshold respectively;

[0078] If the accuracy value of the actual inventory turnover rate calculation result is greater than the second standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is high, and a high accuracy calculation signal is generated, and it is classified as a high accuracy calculation result;

[0079] If the accuracy value of the actual inventory turnover rate calculation result is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is average, and a medium accuracy calculation signal is generated at this time, and it is classified as a medium accuracy calculation result;

[0080] If the accuracy value of the actual inventory turnover rate calculation result is less than the first standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is low. At this time, a low accuracy calculation signal is generated and it is classified as a low accuracy calculation result.

[0081] S6: The actual inventory turnover rate calculated with high accuracy is directly applied to long-term procurement and production planning to support management in making strategic decisions; the actual inventory turnover rate calculated with low accuracy is manually verified before being used for decision making;

[0082] When the system identifies a high-accuracy actual inventory turnover calculation result, it means that the real-time update of data and the accuracy of input have reached a high level, and the calculated inventory turnover rate has a high credibility. In this case, the actual inventory turnover rate can be directly applied to long-term procurement and production planning to help management make strategic decisions. Specific applications include:

[0083] Procurement forecast: With high-accuracy inventory turnover, companies can accurately predict inventory demand in the future to ensure that there will be no shortage of materials or over-purchase during the production cycle. For materials that are consumed faster or have less demand fluctuations, materials with high turnover rates usually require more frequent procurement arrangements.

[0084] Supply chain optimization: Highly accurate inventory turnover can help optimize supplier management. Enterprises can negotiate with suppliers for more reasonable delivery cycles and purchase batches based on the turnover speed of materials, thereby reducing inventory holding costs and improving supply chain efficiency.

[0085] Optimization of production cycle: Highly accurate inventory turnover rate helps enterprises to accurately arrange production rhythm, avoid production stagnation due to insufficient raw materials, or capital occupation due to inventory backlog. Based on the efficient feedback of turnover rate, management can reasonably arrange various production tasks to ensure on-time delivery.

[0086] Inventory capital management: Materials with high inventory turnover rates indicate that they are consumed quickly. Enterprises can reasonably allocate funds based on the turnover rate of these materials to improve the efficiency of production lines. For high-turnover materials, enterprises can reduce excessive inventory backlogs and release more working capital for other strategic investments.

[0087] Long-term strategy formulation: High-accuracy inventory turnover can provide data support for management to formulate long-term procurement, inventory and production strategies. By optimizing inventory management, companies can reduce supply chain risks, improve market response speed, and gain advantages in cost control.

[0088] Cost optimization and inventory management: Efficient inventory turnover can help management optimize inventory holding costs. For high-turnover materials, management can reduce storage costs by reducing inventory, further improving the company's capital turnover efficiency.

[0089] For inventory turnover rates with low accuracy identified by the system, it usually means that there are certain problems in the update or entry of inventory data, which may lead to large errors in the calculation results. Such calculation results should not be directly used for the company's procurement and production decisions without further verification and confirmation. In order to ensure the accuracy of the decision, these turnover rates need to be manually verified. The specific processing methods are as follows:

[0090] Data verification: For inventory turnover rates with low accuracy, manual verification is required, especially manual review and verification of key data such as inventory records, purchase volume, sales volume, etc., to ensure the integrity and accuracy of the data. For example, check whether there are problems such as missing data, duplicate entry, or system lag.

[0091] Inventory verification: If necessary, offline inventory can be arranged to compare with system records to confirm whether the actual inventory status is consistent with the system inventory. Inventory can correct possible inventory data deviations in a timely manner to ensure that the data truly reflects the inventory situation.

[0092] Exception handling: After manual verification, if significant deviations or anomalies are found in the data, data corrections need to be made within the system. For example, incorrect inventory records need to be adjusted, and purchase and sales quantities need to be recalculated to ensure the accuracy of inventory turnover.

[0093] Feedback and optimization: For low-accuracy calculation results, the verification results need to be fed back to the system after manual verification. The system can optimize the data update process and improve the accuracy of data entry based on the feedback information, thereby reducing future data anomalies.

[0094] Application after review: After manual verification and correction, the inventory turnover rate can be safely applied to short-term procurement or production arrangements. For example, for production lines or material procurement that need to be adjusted urgently, the procurement and production plans can be re-formulated based on the adjusted inventory turnover rate to avoid material shortages or backlogs caused by data errors.

[0095] Risk control: Manual verification can effectively avoid wrong decisions caused by data errors. Multiple checks on the turnover rate data of high-risk or important materials can ensure that the company will not make wrong purchases or production adjustments due to data distortion.

[0096] S7: For actual inventory turnover rates with medium accuracy calculation results, the degree of abnormality of the accuracy of the actual inventory turnover rate calculation results within a fixed time period is predicted, and the calculation weights of the actual inventory turnover rate are dynamically adjusted based on the prediction results to optimize procurement and inventory control capabilities.

[0097] Set a fixed time period M (such as a week or a month) to predict the accuracy of the inventory turnover rate calculation results and identify whether there is an abnormal degree. Define the accuracy abnormality value of the inventory turnover rate calculation as E(t), and the calculation expression is: In the formula, AR actual (t) is the actual inventory turnover rate calculation result, ARpredicted(t) is the inventory turnover rate predicted based on historical data and models; if the accuracy anomaly value of the inventory turnover rate calculation is high, it means that the accuracy of the inventory turnover rate calculation result is poor; conversely, if the accuracy anomaly value of the inventory turnover rate calculation is low, it means that the actual inventory turnover rate is more accurate.

[0098] By performing time series prediction on the abnormality degree E(t) in several time periods, the abnormality degree in the next time period can be predicted. The expression is: is the degree of abnormality predicted in the next time period, f is the prediction function, which can be a linear regression model, an ARIMA model, an exponentially weighted moving average (EWMA) or other methods. E(th) is the degree of abnormality in the past h time periods. Dynamically adjust the calculation weight of inventory turnover rate to optimize inventory and purchasing decisions. Set an initial weight W 0 , and correct it according to the prediction results. The expression is: Where W(t+1) is the calculation weight of inventory turnover rate at time t+1, W 0 is the initial weight, which indicates the basic weight of inventory turnover rate in decision making, and β is the adjustment coefficient, which is used to control the impact of abnormality on the weight, β>0. The larger β is, the greater the impact of abnormality on the weight is.

[0099] if If the value is higher (i.e., the degree of abnormality is predicted to be greater in the next time period), the weight W(t+1) will decrease accordingly, indicating that the current inventory turnover rate calculation result may not be accurate, and the weight given to it in purchasing and inventory decisions should be reduced.

[0100] if If the value of W(t+1) is lower (i.e., the degree of abnormality is predicted to be smaller in the next time period), the weight W(t+1) is larger, indicating that the inventory turnover rate is more reliable and its role in purchasing and inventory decisions can be increased.

[0101] Combined with the dynamically adjusted inventory turnover rate weight W(t+1), when making purchase and inventory decisions, the actual inventory turnover rate used will be corrected according to the calculated weight, thereby improving the accuracy of purchase and inventory control. The actual inventory turnover rate used in the final purchase and inventory decision is: ARadjusted (t+1)=W(t+1)·AR(t+1); AR adjusted (t+1) is the adjusted inventory turnover rate, which is used for actual purchasing and inventory decisions. AR(t+1) is the original inventory turnover rate calculation result at time t+1.

[0102] In this application, when the calculation result of the inventory turnover rate is of medium accuracy, by predicting the degree of accuracy abnormality within a fixed time period, future risks can be identified and the calculation weight of the inventory turnover rate can be dynamically adjusted. When the degree of abnormality of the prediction is large, the weight of the inventory turnover rate in the decision is reduced; when the degree of abnormality is small, the weight is increased. Ultimately, this mechanism is used to optimize procurement and inventory control capabilities to ensure more accurate and robust decisions.

[0103] In this embodiment, first, the inventory, sales, purchase volume and inventory turnover days of each material are obtained to calculate the actual inventory turnover rate. Next, the fluctuation of the inventory data update frequency is analyzed, the deviation between the inventory data and the actual inventory status is determined, and the real-time nature of the data update is evaluated. At the same time, the integrity of data entry is monitored, the changes in the abnormal entry ratio are analyzed, and the degree of interference with the inventory turnover rate calculation is determined. Then, based on the real-time nature of data update and the changes in the abnormal entry ratio, the accuracy of the actual inventory turnover rate calculation results is evaluated and divided into high, medium and low accuracy levels. For high accuracy results, they are directly applied to long-term procurement and production planning to support management in making strategic decisions; for low accuracy results, decisions need to be made after manual verification. For medium accuracy results, based on the prediction of the degree of abnormality within a fixed time period, the calculation weight of the turnover rate is dynamically adjusted to optimize procurement and inventory control capabilities.

[0104] Embodiment 2, a supply chain material data management system described in this embodiment includes an actual inventory turnover rate calculation module, an inventory data update real-time evaluation module, an abnormal input ratio change module, an accuracy evaluation module, an accuracy classification module, a decision application module and a dynamic adjustment module;

[0105] Actual inventory turnover rate calculation module: obtains basic inventory data, including the inventory quantity, sales volume, purchase quantity and inventory turnover days of each material, and calculates the actual inventory turnover rate based on the basic data;

[0106] Inventory data update real-time evaluation module: Analyze the fluctuation of the real-time update frequency of inventory data during the inventory management process, determine the deviation between inventory data and actual inventory status, and evaluate the real-time update of data during the inventory management process;

[0107] Abnormal input ratio change module: monitors the integrity of data input during inventory management, analyzes the abnormal input ratio change of data during inventory management, and determines the degree of interference of abnormal input ratio change on inventory turnover rate calculation;

[0108] Accuracy evaluation module: evaluates the accuracy of actual inventory turnover calculation results based on the real-time nature of data updates during inventory management and the degree of interference of abnormal input ratio changes on inventory turnover calculation;

[0109] Accuracy classification module: divides the accuracy of the actual inventory turnover rate calculation results into different levels, namely, high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results;

[0110] Decision-making application module: The actual inventory turnover rate calculated with high accuracy is directly applied to long-term procurement and production planning to support management in making strategic decisions; the actual inventory turnover rate calculated with low accuracy is manually verified before being used for decision-making;

[0111] Dynamic adjustment module: For actual inventory turnover rates with medium accuracy calculation results, the module predicts the degree of abnormality of the accuracy of the actual inventory turnover rate calculation results within a fixed time period, and dynamically adjusts the calculation weights of the actual inventory turnover rate based on the prediction results to optimize procurement and inventory control capabilities.

[0112] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0113] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0114] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0115] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A supply chain material data management method, characterized by: The steps include: S1: Obtain basic inventory data, including inventory quantity, sales volume, purchase quantity and inventory turnover days of each material, and calculate the actual inventory turnover rate based on the basic data; S2: Analyze the fluctuation of the real-time update frequency of inventory data during the inventory management process, determine the deviation between inventory data and actual inventory status, and evaluate the real-time nature of data updates during the inventory management process; S3: Monitor the integrity of data entry during inventory management, analyze the changes in the abnormal data entry ratio during inventory management, and determine the degree of interference of the abnormal data entry ratio changes on the inventory turnover rate calculation; S4: Evaluate the accuracy of the actual inventory turnover rate calculation results based on the real-time nature of data updates during inventory management and the degree of interference of changes in the abnormal input ratio on the inventory turnover rate calculation; S5: classify the accuracy of the actual inventory turnover rate calculation results into different levels, namely, high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results; S6: Apply the actual inventory turnover rate calculated with high accuracy directly to long-term procurement and production planning to support management in making strategic decisions; The actual inventory turnover rate of low-accuracy calculation results is manually verified before being used for decision-making; S7: For actual inventory turnover rates with medium accuracy calculation results, the degree of abnormality of the accuracy of the actual inventory turnover rate calculation results within a fixed time period is predicted, and the calculation weights of the actual inventory turnover rate are dynamically adjusted based on the prediction results to optimize procurement and inventory control capabilities.

2. A supply chain material data management method according to claim 1, characterized in that: In S2, the update frequency abnormal fluctuation index is generated after analyzing the fluctuation degree of the update frequency in different time periods. The method for obtaining the update frequency abnormal fluctuation index is: Get the update frequency value calculated in the T time period in real time, determine a smoothing parameter α, set the initial value, that is, at time s = 0, set the initial update frequency, the expression is: EWMA0 = update frequency 0; starting from time s = 1, the exponentially weighted moving average EWMA at each time point is calculated by the EWMA of the previous time point and the current update frequency, the expression is: EWMA s =α·Update frequency s+(1-α)·EWMA s-1 ; After obtaining the EWMA value at each time point, calculate the abnormal fluctuation index of the update frequency, the expression is: Where ER is the abnormal fluctuation index of update frequency.

3. A supply chain material data management method according to claim 2, characterized in that: In S3, the abnormal input ratio drift index is generated after analyzing the change of abnormal input ratio in different materials. The method for obtaining the abnormal input ratio drift index is: Construct an abnormal input ratio data set {x1,x2,...,x r }, where x r Represents the abnormal entry ratio at time r; divide the abnormal entry ratio value in the data set into data in N time periods, calculate the mean μ and standard deviation σ of the abnormal entry ratio value in the kth time period; for each time point r, calculate the posterior probability of it being a change point, the expression is: In the formula, P(change point|x 1:r ) is the posterior probability of a change point occurring at time r, P(x r | Change point) is the likelihood function value of the abnormal input ratio change at time r, P(change point) is the prior probability, indicating the prior hypothesis of the change point, P(x r ) is the total observed probability at time point r; the posterior probability P(change point|x 1:r ) is compared with the benchmark value of the change point posterior probability set under normal conditions in historical data. When P(change point|x 1:r ) is greater than or equal to the benchmark value of the posterior probability of the change point, it is considered that time r has drifted. When a change point is detected, the change point strength and influence are calculated, and the abnormal input ratio drift index is generated by the drift strength. The calculation expression of the drift strength is: In the formula, μ after is the mean of the new data segment after the change point, μ before is the mean of the old data segment before the change point; the drift intensity and the posterior probability are combined to generate the abnormal input ratio drift index, the expression is: KG = drift intensity r ×P(change point|x 1:r ), where KG is the abnormal input ratio drift index.

4. A supply chain material data management method according to claim 3, characterized in that: In S4, the accuracy of the actual inventory turnover rate calculation results is evaluated based on the real-time nature of data update in the inventory management process and the degree of interference of abnormal input ratio changes on the inventory turnover rate calculation; The abnormal fluctuation index of update frequency and the drift index of abnormal entry ratio are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses the accuracy value label of the actual inventory turnover calculation result predicted by each group of first eigenvectors as the prediction target, and takes minimizing the sum of prediction errors of the accuracy value labels of all actual inventory turnover calculation results as the training target. The machine learning model is trained until the sum of prediction errors converges, then the model training is stopped, and the accuracy value of the actual inventory turnover calculation result is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

5. A supply chain material data management method according to claim 4, characterized in that: In S5, the accuracy of the actual inventory turnover rate calculation results is divided into different levels, namely, high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results, specifically: Compare the accuracy value of the actual inventory turnover rate calculation result obtained with the gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and compare the accuracy value of the actual inventory turnover rate calculation result with the first standard threshold and the second standard threshold respectively; If the accuracy value of the actual inventory turnover rate calculation result is greater than the second standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is high, and a high accuracy calculation signal is generated, and it is classified as a high accuracy calculation result; If the accuracy value of the actual inventory turnover rate calculation result is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is average, and a medium accuracy calculation signal is generated at this time, and it is classified as a medium accuracy calculation result; If the accuracy value of the actual inventory turnover rate calculation result is less than the first standard threshold, it means that the accuracy of the actual inventory turnover rate calculation result is low. At this time, a low accuracy calculation signal is generated and it is classified as a low accuracy calculation result.

6. A supply chain material data management method according to claim 1, characterized in that: In S7, for the actual inventory turnover rate calculated with medium accuracy, the degree of abnormality of the accuracy of the actual inventory turnover rate calculation result within a fixed time period is predicted, and the calculation weight of the actual inventory turnover rate is dynamically adjusted according to the prediction result, specifically: Set a fixed time period M to predict the accuracy of the inventory turnover rate calculation results, define the accuracy abnormality value of the inventory turnover rate calculation as E(t), and the calculation expression is: In the formula, AR actual (t) is the actual inventory turnover rate calculation result, AR predicted (t) is the inventory turnover rate predicted based on historical data and models; By performing time series prediction on the abnormality degree E(t) in several time periods, the abnormality degree in the next time period is predicted. The expression is: is the predicted abnormality level in the next time period, f is the prediction function, and E(th) is the abnormality level in the past h time periods; According to the predicted abnormality Dynamically adjust the calculation weight of inventory turnover rate to optimize inventory and purchasing decisions; set an initial weight W0 and modify it according to the forecast results. The expression is: Where W(t+1) is the calculation weight of inventory turnover rate at time t+1, W0 is the initial weight, which represents the basic weight of inventory turnover rate in decision-making, and β is the adjustment coefficient.

7. A supply chain material data management method according to claim 6, characterized in that: Combined with the dynamically adjusted inventory turnover weight W(t+1), when making purchasing and inventory decisions, the actual inventory turnover rate used will be corrected according to the calculated weight, thereby improving the accuracy of purchasing and inventory control; Inventory turnover ratio actually used in final purchasing and inventory decisions: AR adjusted (t+1)=W(t+1)·AR(t+1); AR adjusted (t+1) is the adjusted inventory turnover rate, which is used for actual purchasing and inventory decisions, and AR(t+1) is the original inventory turnover rate calculation result at time t+1.

8. A supply chain material data management system, used to implement a supply chain material data management method according to any one of claims 1 to 7, characterized in that: It includes actual inventory turnover rate calculation module, inventory data update real-time evaluation module, abnormal input ratio change module, accuracy evaluation module, accuracy classification module, decision application module and dynamic adjustment module; Actual inventory turnover rate calculation module: obtains basic inventory data, including the inventory quantity, sales volume, purchase quantity and inventory turnover days of each material, and calculates the actual inventory turnover rate based on the basic data; Inventory data update real-time evaluation module: Analyze the fluctuation of the real-time update frequency of inventory data during the inventory management process, determine the deviation between inventory data and actual inventory status, and evaluate the real-time update of data during the inventory management process; Abnormal input ratio change module: monitors the integrity of data input during inventory management, analyzes the abnormal input ratio change of data during inventory management, and determines the degree of interference of abnormal input ratio change on inventory turnover rate calculation; Accuracy evaluation module: evaluates the accuracy of actual inventory turnover calculation results based on the real-time nature of data updates during inventory management and the degree of interference of abnormal input ratio changes on inventory turnover calculation; Accuracy classification module: divides the accuracy of the actual inventory turnover rate calculation results into different levels, namely, high accuracy calculation results, medium accuracy calculation results and low accuracy calculation results; Decision-making application module: The actual inventory turnover rate calculated with high accuracy is directly applied to long-term procurement and production planning to support management in making strategic decisions; The actual inventory turnover rate of low-accuracy calculation results is manually verified before being used for decision-making; Dynamic adjustment module: For actual inventory turnover rates with medium accuracy calculation results, the module predicts the degree of abnormality of the accuracy of the actual inventory turnover rate calculation results within a fixed time period, and dynamically adjusts the calculation weights of the actual inventory turnover rate based on the prediction results to optimize procurement and inventory control capabilities.

Images