Purchasing management method based on intelligent supply chain management platform

Abstract

The application relates to the field of information technology and provides a procurement management method based on an intelligent supply chain management platform, which comprises the following steps: obtaining material demand data submitted by each department, standardizing the time stamp and demand quantity of the material demand data, and forming standardized data containing an auxiliary classification dimension, a time dimension and a quantity dimension; generating an initial procurement plan table according to real-time demand data and preset procurement management rules, grouping procurement data in the initial procurement plan table according to preset grouping rules, and obtaining a grouped procurement plan table; obtaining response time prediction data and comparing the response time prediction data with a preset response time threshold value; if the response time prediction data is greater than the threshold value, optimizing procurement behavior by adjusting priority and resource allocation, and determining an optimized procurement strategy. The method realizes the standardization, transparency and high efficiency of the procurement process, effectively promotes the digital transformation and intelligent upgrading of enterprise procurement.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and in particular to a procurement management method based on an intelligent supply chain management platform. BACKGROUND

[0002] Supply chain management, as the core field of modern enterprise operation, plays a decisive role in improving resource allocation efficiency, reducing operating costs, and enhancing market competitiveness. With the acceleration of globalization and digitization, enterprises have higher requirements for the efficiency and transparency of the supply chain, especially in the procurement management link, which directly affects the timeliness and economy of material supply. However, many current supply chain management solutions have exposed significant defects in practical application. Traditional methods often rely on manual operation or decentralized systems, resulting in low efficiency of procurement processes and prominent data silo problems, making it impossible to achieve inter-departmental collaboration and real-time demand management. In addition, supplier management is mostly offline, lacking systematic online integration and data analysis support, making it difficult for enterprises to scientifically evaluate and optimize procurement decisions. These limitations point to the core challenges in the field of supply chain management that need to be broken through. First, the real-time and accuracy of material demand management is insufficient, as the lack of a unified platform to integrate departmental demands often leads to a disconnection between procurement plans and actual demand. Second, the electronicization and collaboration of procurement processes are limited, especially order processing and supplier integration, which still rely on manual or semi-automated methods, resulting in low efficiency and errors. Finally, the lack of data analysis capabilities prevents enterprises from extracting valuable insights from massive procurement data, hindering the intelligentization of decision support. These technical factors directly lead to the blindness of procurement plan development, the inefficiency of expert review, and the difficulty of implementing long-term cooperation models. Therefore, how to build an integrated and intelligent supply chain management platform to realize real-time demand management and data-driven decision support for procurement processes has become a key issue for improving enterprise procurement management efficiency. Solving this problem requires focusing on the integration of demand management, process optimization, and data analysis technologies to ensure efficient operation of the entire procurement management chain. SUMMARY

[0003] The present application provides a procurement management method based on an intelligent supply chain management platform, mainly including:

[0004] The material demand data submitted by each department is acquired, the timestamps and demand quantities of the material demand data are standardized, the standardized material demand data is classified and arranged according to department names and material types, and standardized data containing auxiliary classification dimensions, a time dimension and a quantity dimension are formed, wherein the auxiliary classification dimensions include departments and material types; the single purchase quantity of each department single order is extracted from the standardized data, and the purchase quantity fluctuation range is calculated respectively, abnormal purchase orders in the standardized data are removed according to a preset fluctuation threshold, and real-time demand data is obtained; an initial purchase plan table is generated according to the real-time demand data and a preset purchase management rule, the purchase data in the initial purchase plan table is grouped according to a preset grouping rule, and a grouped purchase plan table is obtained; order feature data in the grouped purchase plan table is combined with an access identifier and a blacklist identifier in a preset supplier qualification library to calculate a matching degree of the order feature data and supplier feature data, each purchase order is respectively published to a preset number of suppliers with the highest matching degree, the access identifier and no blacklist identifier to invite quotes, the order state is judged and tracked according to the feedback of the confirmation receipt of the supplier, and tracking logs are obtained; order processing data and supplier response time are extracted from the tracking logs, a pre-trained ARIMA model is used to calculate predicted supplier response time, and response time prediction data is obtained, wherein the supplier response time refers to the time interval from the initiation of the order to the supplier to the sending of the confirmation receipt by the supplier, and the response time prediction data contains the predicted supplier response time for future orders; the response time prediction data is compared with a preset response time threshold, if greater than the preset response time threshold, the order purchase behavior is optimized by adjusting the order priority and resource allocation, and the optimized order purchase strategy is determined; the order purchase task is executed through the optimized order purchase strategy, platform operation data is obtained from platform operation logs, and collaboration index data is calculated according to the platform operation data, wherein the collaboration index data includes order response speed and matching accuracy; the collaboration index data is analyzed through a decision tree algorithm to determine key influencing factors affecting the collaboration index data, procurement cycle length data including planned procurement time and actual arrival time is obtained from historical procurement data, Pearson correlation coefficients and significance levels of the key influencing factors and the procurement cycle length data are calculated, and final procurement decision data is generated according to the key influencing factors and the corresponding Pearson correlation coefficients and significance levels; configuration parameters are extracted from the final procurement decision data, the order purchase strategy and resource scheduling rules of the platform are iteratively updated according to the configuration parameters, and full-chain operation state data is obtained, wherein the configuration parameters include supplier priority weight, response time index weight, maximum supplier order carrying capacity and emergency order resource quota proportion.

[0005] The technical scheme provided by the embodiment of the application can have the following beneficial effects:

[0006] The application discloses a procurement management method based on an intelligent supply chain management platform. The method collects real-time material demand data from various departments and classifies and processes the data using clustering algorithms to determine real-time demand patterns. Based on the demand patterns, the application generates an initial procurement plan and has a dynamic adjustment mechanism to respond to demand changes. The procurement plan is distributed to suppliers through the platform, and the order status is monitored. The application also uses time series analysis algorithms to calculate process efficiency indicators and adjusts task priorities and resource allocation through optimization algorithms. In addition, the application uses decision tree algorithms to analyze key factors affecting procurement decisions and continuously iterates and updates the entire chain operation status. This comprehensive, efficient and flexible solution standardizes, transparentizes and streamlines the procurement process, effectively promoting the digital transformation and intelligent upgrade of enterprise procurement. BRIEF DESCRIPTION OF DRAWINGS

[0007] Figure 1 A flowchart of the procurement management method based on the intelligent supply chain management platform of the application. DETAILED DESCRIPTION

[0008] The technical solutions in the embodiments of the application will be described in detail below with reference to the accompanying drawings. The described embodiments are only a part of the embodiments of the application.

[0009] As Figure 1 , the procurement management method based on the intelligent supply chain management platform of the present embodiment can specifically include:

[0010] Step S101, obtain the material demand data submitted by each department, standardize the time stamp and demand quantity of the material demand data, classify and organize the standardized material demand data according to department name and material type, and form standardized data containing auxiliary classification dimension, time dimension and quantity dimension, wherein the auxiliary classification dimension includes department and material type.

[0011] Obtain the material demand data submitted by each department through the real-time acquisition system to obtain preliminary data containing time stamp, demand quantity, department and material type. Standardize the time stamp and demand quantity in the preliminary data according to the preset standardization rules to generate demand records with time identifiers. Classify and organize the demand records with time identifiers according to department name and material type to obtain standardized data containing auxiliary classification dimension, time dimension and quantity dimension.

[0012] The preliminary data including timestamp, demand quantity, department, material type and group bid label are obtained by acquiring the material demand data submitted by each department through the real-time acquisition system. The platform conducts penetrating management on the enterprise procurement process, and all demand reporting and approval processes are online, ensuring compliance with internal control requirements. For example, an enterprise material management system collects office supplies demand submitted by each department daily, including submission time such as 2025-05-15-08:00:00, demand quantity such as 100 A4 papers, department such as administrative department, and material type such as office supplies. Real-time acquisition can be achieved through online forms, and data is automatically uploaded to the central database after being filled in by each department, ensuring data timeliness and accuracy. This process relies on timestamp to record the submission time, facilitating subsequent tracing and analysis, and helping to quickly respond to demand. The timestamp and demand quantity in the preliminary data are standardized to generate demand records with time identifiers. For example, the timestamp may appear in different formats due to system differences, such as 2025-05-15-08:00:00 or 15 / 05 / 2025 / 08:00. Standardization processing unifies it into ISO format such as 2025-05-15T08:00:00Z, ensuring cross-system compatibility. Standardization of demand quantity unifies and standardizes the unit format. In one embodiment, the demand quantity is standardized. Standardization includes unified procurement units and merging of similar materials. For example, the administrative department purchases A4 paper in "order" units, and the finance department in "box" units. After standardization, they are unified into "order" units, 1 box equals 10 orders, and 60 boxes of the finance department are converted to 600 orders. After merging similar materials, the total demand for A4 paper is 710 orders. After unit unification, the format is standardized, such as "710 orders of A4 paper" to "710 units of A4 paper (unit: order)", avoiding misunderstandings due to unclear expression. The demand records generated after standardization contain unified time identifiers and quantified demand values, facilitating subsequent classification and statistics, and improving data processing efficiency. The demand records are classified and arranged according to department name and material type to obtain standardized data containing auxiliary classification dimensions, time dimensions and quantity dimensions. Specifically, classification can be achieved through database grouping query. For example, the administrative department submitted 600 units of A4 paper (unit: order) and 50 units of pen refill (unit: piece) on 2025-05-15, and the finance department submitted 110 units of A4 paper (unit: order), which formed a record table with department and material type as dimensions after classification. Auxiliary classification dimensions such as material priority can be further refined, such as A4 paper marked as high priority and pen refill as low priority. The time dimension retains daily demand snapshots, and the quantity dimension summarizes total demand. This standardized data supports multi-dimensional analysis, such as statistical analysis of consumable consumption by department or analysis of demand fluctuations by time, facilitating managers to optimize inventory allocation. It should be noted that the standardized data after classification and arrangement can be generated into reports through visualization tools, such as column chart to display A4 paper demand of each department, or line chart to reflect monthly demand trend.Such analysis helps to find the cause of the surge in demand for A4 paper, such as the onboarding of new employees, and thus to stock up in advance. Technically, real-time collection ensures data timeliness, standardized processing improves data consistency, classification and arrangement enhance data analyzability, and ultimately achieve efficient and accurate material management, reducing the risk of inventory overstock or shortage. This extension scheme complements the core classification and arrangement, and together supports intelligent decision-making for material management.

[0013] In step S102, the single purchase quantity of each department single order is extracted from the standardized data, and the fluctuation range of the purchase quantity is calculated respectively. According to the preset fluctuation threshold, the abnormal purchase order in the standardized data is removed, and the real-time demand data is obtained.

[0014] The single purchase quantity of each department single order is extracted from the standardized data, and the standard deviation of each department is calculated according to the sequence composed of multiple single purchase quantities of each department, as the fluctuation range value. The single purchase quantity of each department multiple single order includes the current order single purchase quantity and the historical order single purchase quantity. If the fluctuation range value exceeds the preset fluctuation threshold, the current purchase order is marked as an abnormal purchase order, and the abnormal purchase order is removed to obtain the real-time demand data.

[0015] In particular, based on the standardized data, the single purchase quantity of each department is extracted and the fluctuation range value is calculated, which can effectively identify abnormal procurement data and ensure the accuracy of real-time demand data. Specifically, the standardized data includes dimensions such as department, material type, timestamp and demand quantity, for example, the administrative department submits an order of 100 units of A4 paper on 2025-05-15. The single purchase quantity of each order of each department is extracted, that is, the demand quantity of each order is obtained, forming a sequence, for example, the single purchase quantity of A4 paper of the administrative department in the last five purchases is 100, 120, 110, 115 and 300 units. The standard deviation calculation reflects the fluctuation degree of the sequence. Assuming that the preset fluctuation threshold is 50 units, if the standard deviation of the A4 paper purchase sequence of the administrative department is 75.8 units, which exceeds the threshold, it is marked as abnormal procurement data. The single purchase quantity of the current order, 300 units, is a significantly high single purchase quantity and may be an outlier that needs to be removed. In an embodiment, the identification of abnormal procurement data can be achieved by database screening. The platform automatically extracts the recent order data of a certain department for a certain material type, generates a purchase quantity sequence, calculates the standard deviation and compares it with the threshold. For example, the purchase quantity of the pen refill of the finance department in the last five times is 50, 55, 60, 52 and 200 units, and the standard deviation is 58.4 units, which exceeds the threshold of 50 units, and the current order of 200 units is marked as abnormal. After removal, the remaining data forms real-time demand data, for example, the real-time demand of the pen refill of the finance department is calculated based on 50, 55, 60 and 52 units to obtain an average value, and a more stable demand trend is obtained. It can be understood that the special application mode of the abnormal procurement order provides flexibility for management. For example, the abnormal order of 300 units of A4 paper of the administrative department may be due to temporary demand for a large-scale meeting, and the department can submit an application for explanation, and after approval, it is purchased separately. This mechanism avoids the delay of emergency demand caused by abnormal removal. After removing the abnormal data from the real-time demand data, the daily demand is more accurately reflected, for example, the average demand of A4 paper of the administrative department is about 111 units, which facilitates inventory optimization. In an embodiment, the setting of the fluctuation threshold can be dynamically adjusted in combination with historical data. For example, by analyzing the A4 paper procurement data of the administrative department in the past six months, the average value of the standard deviation is 40 units, and the threshold can be set to 1.5 times the standard deviation, that is, 60 units, to ensure that the threshold is reasonable. The dynamic threshold adapts to the fluctuation characteristics of different departments and material types, improving the accuracy of abnormal identification. For example, the generation of real-time demand data supports subsequent analysis, such as aggregating monthly demand by department or predicting inventory consumption by material type. After removing the abnormal data, the demand for A4 paper of the administrative department is more stable, and the predicted demand for the next month is about 110 units, and the procurement plan is more accurate. The special application mechanism ensures that temporary large demand is met, for example, the abnormal order of 200 units of pen refill of the finance department is used for new employee training after the application is confirmed. This scheme improves the adaptability and accuracy of material management through abnormal identification, data cleaning and flexible approval.

[0016] In step S103, an initial procurement plan table is generated according to the real-time demand data and preset procurement management rules, and the procurement data in the initial procurement plan table is grouped according to preset grouping rules to obtain a grouped procurement plan table.

[0017] The material type, demand quantity, demand department, and timestamp field in the real-time demand data are matched with preset procurement management rules to generate an initial procurement plan table, wherein the preset management rules include a framework procurement proportion allocation rule and a supplier carrying capacity limitation rule, and the initial procurement plan table contains procurement material name, specification and model, recommended procurement quantity, minimum / maximum procurement quantity limit, demand department, and planned procurement time, and the recommended procurement quantity is calculated comprehensively according to the demand quantity, preset safety stock threshold, and supplier carrying capacity. According to the preset grouping rules, the procurement data in the initial procurement plan table is aggregated or split to form multiple procurement groups, wherein the grouping rules include classification by demand department, classification by material type, classification by supplier adaptation category, and classification by procurement urgency level. Data standardization processing is performed on each procurement group to generate a grouped procurement plan table containing group number, material category, demand department set, procurement quantity summary, planned procurement time interval, and corresponding supplier range, wherein the standardization processing includes uniform procurement unit, merging of similar materials, and labeling of group leader or interfacing supplier.

[0018] In an embodiment, the real-time demand data includes fields such as material type, demand quantity, demand department, timestamp, etc., which are matched with preset procurement management rules to generate an initial procurement plan table. The procurement management rules include framework procurement proportion allocation rules and supplier carrying capacity limitation rules. For example, the framework procurement proportion allocation rules stipulate that 55% of the procurement quantity of A4 paper for the administrative department is provided by the main supplier, and 45% is provided by the backup supplier; the supplier carrying capacity limitation rules set that the main supplier can supply a maximum of 5000 units of A4 paper per month. Based on the real-time demand data, the administrative department has a demand of 110 units of A4 paper on 2025-05-15, combined with a safety stock threshold of 50 units, the recommended procurement quantity is 160 units. The initial procurement plan table records the material name A4 paper, specification and model 70g, recommended procurement quantity 160 units, minimum procurement quantity 100 units, maximum procurement quantity 200 units, demand department administrative department, and planned procurement time 2025-05-20. The supplier carrying capacity ensures that the procurement quantity does not exceed the upper limit of the supplier, ensuring stable supply. For example, the initial procurement plan table is aggregated or split according to preset grouping rules, which include grouping rules such as demand department, material type, supplier adaptation category, and procurement urgency level classification. The administrative department and the finance department both purchase A4 paper, which can be aggregated according to the material type to generate an A4 paper procurement group; if the procurement urgency level is different, such as regular procurement for the administrative department and emergency procurement for the finance department, then it can be split according to the urgency level. After grouping, the A4 paper regular procurement group contains 110 units of demand from the administrative department and 600 units of regular demand from the finance department, totaling 710 units. The grouping rules flexibly adapt to different scenarios to ensure clear procurement plans. The grouped procurement plan table records the group number A4-001, the material category A4 paper, the demand department set administrative department and finance department, the procurement quantity 710 units of A4 paper (units: ream), the planned procurement time interval 2025-05-20 to 2025-05-25, and the corresponding supplier range main supplier A and backup supplier B. The group leader is marked as procurement manager Zhang, ensuring clear interface. It can be understood that the grouped procurement plan table after standardization provides clear guidance for subsequent procurement execution. For example, 710 reams of A4 paper are allocated according to the framework procurement proportion, with 391 reams provided by the main supplier A and 319 reams provided by the backup supplier B, which meets the supplier carrying capacity. The emergency procurement group is arranged with priority to supplier B to shorten the delivery time. The grouped procurement plan table supports aggregating demand by department or material type, facilitating procurement planning and inventory management, and ensuring procurement efficiency and resource optimization.

[0019] Step S104, according to the order feature data in the group procurement plan table, combined with the preset supplier qualification library access identification and blacklist identification, calculate the matching degree of order feature data and supplier feature data, respectively to the highest matching degree, with access identification and no blacklist identification of the preset number of suppliers to publicize the invitation of offer, according to the supplier feedback confirmation receipt to judge the order state and track the order state, get the tracking log.

[0020] The order feature data contains material type, procurement quantity and order urgency. Based on the access identification and blacklist identification in the supplier qualification library, the supplier feature data set is constructed, wherein the access identification is determined according to the supplier supply ability score and delivery punctuality rate, and the blacklist is determined according to the historical default record. The matching degree of order feature data and supplier feature data is calculated by weighted Euclidean distance algorithm, and each procurement order is respectively invited to publicize the invitation of offer to the preset number of suppliers with the highest matching degree, with access identification and no blacklist identification. According to the supplier feedback confirmation receipt, the order state is marked, the time of order to supplier to initiate the invitation of offer, the supplier response time and the state change information are recorded, and the tracking log containing order basic information, supplier information and state trajectory is generated, wherein the confirmation receipt includes one of the three state information, and the three state information is respectively confirmation, rejection and pending.

[0021] Specifically, the order characteristic data including the material type, the purchase quantity, and the order urgency level and the group bid label are extracted from the grouped procurement plan table to ensure that the data accurately reflects the procurement demand. The order characteristic data includes the material type, the purchase quantity, and the order urgency level. For example, based on the foregoing grouped procurement plan table, the A4 paper demand of the group number A4-001 is 710 reams, the material type is A4 paper, the purchase quantity is 710 reams, and the urgency level is regular procurement. When extracting, the system automatically parses the table fields to generate structured order characteristic data, such as material type: A4 paper, purchase quantity: 710 reams, and urgency level: regular. It should be noted that the urgency level can be determined according to the time requirement submitted by the demand department or the inventory emergency state, for example, when the inventory is lower than the safety threshold of 50 reams, it is automatically marked as urgent. For example, when constructing the supplier characteristic data set, the supplier qualification library provides access identification and blacklist identification. The access identification is based on the supply capacity score and the on-time delivery rate. For example, the supply capacity score of supplier A is 85 points, and the on-time delivery rate is 95%, and the access identification is “high quality”; supplier C is listed in the blacklist due to historical breach records, and the blacklist identification is “limited cooperation”. Specifically, the supply capacity score comprehensively evaluates the supplier's production capacity, logistics capacity, and financial stability, and the on-time delivery rate is calculated based on the performance data in the past 12 months. The supplier characteristic data set thus includes fields such as supplier name, access identification, blacklist identification, etc., to ensure that high-quality suppliers are preferred when purchasing. It can be understood that when constructing the supplier characteristic data set, it can also be based on multi-dimensional labels in the supplier qualification library (such as material type adaptation label, green supply chain label, and human carrying capacity score). For example, supplier A is marked as “high-priority emergency order supplier” and has a human carrying capacity score of 90 points (out of 100), and when the system calculates the matching degree by the weighted Euclidean distance algorithm, it synchronously checks the current human carrying capacity of the supplier (such as the current order quantity / maximum carrying capacity ≤ 80%), and if the threshold is exceeded, the procurement process for alternative suppliers is automatically triggered. In one embodiment, the matching degree of the order characteristic data and the supplier characteristic data is calculated by the weighted Euclidean distance algorithm. The order characteristic data such as the purchase quantity and the urgency level is matched with the supply capacity and the on-time delivery rate of the supplier. The core logic is to first standardize the characteristic data of different dimensions to a unified dimension, then quantify the “difference degree” of the order and the supplier by the weighted Euclidean distance formula, and finally convert the distance into the matching degree. For example, the regular order of 710 reams of A4 paper, first, the data is processed, and the purchase quantity is standardized, assuming that the maximum purchase quantity is 1000 reams, and 710 reams of A4 paper is standardized to 0.71; the regular order urgency level is set to 1 (assuming that the urgent order is 3 and the more urgent order is 2). The supply capacity score of supplier A is 4 (out of 5), and the standardized value is 0.8, and the on-time rate is 0.95 (100% on-time rate is 1); the supply capacity score of supplier B is 3, and the standardized value is =0.8, and the on-time rate is 0.95 (100% on-time rate is 1); the supply capacity score of supplier B is 3, and the standardized value is =0.6, the punctuality rate is 0.90. The calculation formula of the weighted Euclidean distance is: wherein X is the order characteristic data vector, Y is the supplier characteristic data vector, w i is the weight of the i-th characteristic, n is the number of characteristics, x i is the standardized value of the i-th characteristic of the order, y i is the standardized value of the i-th characteristic of the supplier. In this embodiment, the number of characteristics n = 2, the order characteristic data vector X = [0.71, 1], the characteristic data vector Y A of supplier A = [0.8, 0.95], the characteristic data vector Y B of supplier B = [0.6, 0.90], the supply capacity weight w1 = 0.6, and the punctuality rate weight w2 = 0.4. The weighted Euclidean distance between supplier A and the order is calculated as: ≈0.0765. The weighted Euclidean distance is converted into the matching degree, and the formula is: matching degree = Then the matching degree of supplier A is: ≈0.93. The weighted Euclidean distance between supplier B and the order is calculated as: ≈0.1061, and the matching degree of supplier B is: ≈0.90, the supplier A's supply capability score is high and the on-time rate is good, and the matching degree score is 0.93 after calculation; the supplier B's score is 0.90. The system selects the supplier A with no blacklist identification and the highest matching degree. It should be noted that in the weight setting, the supply capability accounts for 60%, and the on-time rate accounts for 40%, to balance the supply stability and delivery efficiency. Understandably, after the pre-set number of in-house suppliers meeting the procurement conditions with the highest matching degree, access identification and no blacklist identification are determined, the system automatically sends a quotation invitation and publicly pushes the quotation demand to the pre-set number of in-house suppliers meeting the procurement conditions. After receiving the invitation, the supplier can make an online quotation through the platform. At the same time, the platform triggers the expert extraction module to automatically extract the evaluation experts according to the pre-set rules (such as professional matching degree, expert avoidance mechanism, random extraction algorithm) to ensure the openness and fairness of the evaluation process. In the evaluation link, the transaction result takes price as the main decision factor, and the supplier's comprehensive score (such as supply capability score, historical default record, delivery time limit, etc.) is comprehensively judged. Among them, the number of suppliers can be determined according to the type of procurement, for example, only 5 suppliers are needed for regular procurement of daily consumables, and 15 suppliers are needed for special procurement of major matters. After the supplier feedback confirmation receipt, the system marks the order status and generates a tracking log. Supplier A feedbacks "confirmation" for the order of 710 reams of A4 paper, and the order status is updated to "confirmed". The tracking log records the order number A4-001-01, the material type A4 paper, the supplier A, the time of initiating the quotation invitation 2025-05-20-10:00, the response time 2025-05-20-12:00, and the status change to "confirmed". Specifically, if the supplier feedbacks "to be processed", the system will re-conduct public bidding for the second-best supplier after 24 hours. The log also includes the status change history, such as from "to be purchased" to "confirmed", which is convenient for tracing. It can be understood that the above process ensures efficient procurement of procurement orders, reasonable selection of suppliers, and clear status tracking. The extraction of order feature data improves the demand matching accuracy, the screening of supplier feature data set selects reliable partners, the matching degree calculation optimizes resource allocation, and the tracking log provides data support for subsequent audit and optimization.

[0022] In step S105, order processing data and supplier response time are extracted from the tracking log, a pre-trained ARIMA model is used to calculate the predicted supplier response time, and response time prediction data is obtained, wherein the supplier response time refers to the time interval from the initiation of the quotation invitation to the supplier to the sending of the confirmation receipt by the supplier, and the response time prediction data includes the predicted response time of the supplier for future orders.

[0023] According to the time of initiating the quotation invitation to the supplier according to the order, the confirmation receipt time and the corresponding supplier information, the actual response time is calculated and the abnormal order is removed, and the comprehensive response time data including the timestamp, the current response time and the historical response time are generated in combination with the historical response time of the supplier. The comprehensive response time data is cleaned by using a data preprocessing method, the abnormal values are removed and the missing values are filled, and the standardized time series data is obtained. The standardized time series data is input into a pre-trained ARIMA model, and response time prediction data including a prediction time point, a supplier identifier and a response time prediction value are generated, wherein the ARIMA model is pre-trained by historical time series data composed of historical response time of the supplier.

[0024] For example, in the scenario of purchase order procurement and supplier response management, calculating the actual response time and generating the integrated response time data is an important step to optimize supplier management. The actual response time refers to the time difference from the time of initiating the invitation for quotation to the time of confirming the receipt from the supplier. For example, order A4-002 initiated the invitation for quotation to supplier B on 2025-05-20-09:00, and the confirmation receipt time was 2025-05-20-11:30, with an actual response time of 2.5 hours. When removing abnormal orders, a threshold can be set, such as considering a response time exceeding 48 hours as abnormal. Assuming that the response time of supplier C for order A4-003 was 72 hours, the system automatically marked it as abnormal and removed it. The integrated response time data is generated in combination with the historical response time of the supplier, which is determined based on the average value of the past 12 months. For example, the historical response time of supplier B was 2.8 hours, and the response time of the current order A4-002 was 2.5 hours, with the integrated data recorded as: timestamp 2025-05-20-11:30, current response time 2.5 hours, historical response time 2.8 hours. In one embodiment, data preprocessing cleans the integrated response time data. Removing outliers can use the box plot method, setting the threshold at 1.5 times the interquartile range. For example, the historical time series of supplier A was 2.0, 2.5, 3.0, and 50.0 hours, with 50.0 hours removed due to exceeding the threshold. Filling missing values can use the mean filling method, assuming that the response time of a certain order of supplier D is missing, the system fills it with the historical mean of 3.2 hours. The standardized time series data obtained after cleaning has a uniform format and no outliers, such as timestamp accurate to the minute, response time unit in hours, and composed of timestamp, current response time, and historical response time, such as: timestamp 2025-05-20-11:30, current response time 2.5 hours, historical response time 2.8 hours. For example, ARIMA model training is based on standardized time series data. ARIMA model captures the trend and periodicity of response time by analyzing the autoregressive, difference, and moving average characteristics of time series. In training, the system takes the identification of supplier B, the prediction time point, and the historical time series (such as 2.5, 2.7, 2.4 hours) as input, automatically adjusts the model parameters, and generates a fitted model. After training is completed, the model can reflect the fluctuation rule of the supplier's response time. In one embodiment, the future response time is predicted based on the trained ARIMA model. The system inputs the identification of supplier B and the prediction time point, such as 2025-06-01-09:00, and the model outputs the predicted response time of 2.6 hours. The prediction data includes timestamp 2025-06-01-09:00, supplier B, and predicted response time 2.6 hours. The prediction result can guide the order procurement behavior, and preferentially select suppliers with short response times.It should be noted that the above process optimizes the efficiency of supplier selection and order management through response time calculation, data cleaning, model training and prediction. Abnormal order elimination ensures data reliability, data preprocessing improves model input quality, and ARIMA model prediction provides accurate reference for future procurement behavior. This method realizes the reasonable allocation of time resources in procurement management, improves the response speed and stability of the supply chain.

[0025] In step S106, the response time prediction data is compared with the preset response time threshold. If it is greater than the preset response time threshold, the order procurement behavior is optimized by adjusting the order priority and resource allocation, and the optimized order procurement strategy is determined.

[0026] The response time prediction data of each supplier is compared with the preset response time threshold. If the response time is greater than the preset response time threshold, the supplier is marked as an over-standard supplier. For the orders of the over-standard supplier that have successfully bid, the task priority is increased according to the urgency. The quota of the over-standard supplier is reduced, and the quota of the high-quality supplier is increased. The on-time delivery rate index weight is increased in the order procurement process, the matching degree of the order and the supplier is recalculated, and the optimized order procurement strategy is determined.

[0027] For example, in the context of order procurement and supplier response management, comparative analysis of supplier response time prediction data is a key step in optimizing supply chain management. The system compares the predicted response time with a pre-set threshold to filter out suppliers who exceed the threshold. Assuming the pre-set response time threshold is 4 hours, supplier A has a predicted response time of 3.5 hours, and supplier B has a predicted response time of 5 hours, supplier B is marked as an out-of-specification supplier. The threshold is set based on historical data analysis, usually taking the upper quartile of supplier response times to ensure reasonableness. After marking the out-of-specification supplier, the system can generate an exception report, recording the out-of-specification timestamp of supplier B, such as 2025-06-01-09:00, and the predicted value of 5 hours, to facilitate subsequent management decisions. In one embodiment, for orders successfully bid by out-of-specification suppliers, the system adjusts the order priority based on the order urgency. Order urgency can be determined by delivery deadline and customer priority. For example, order A4-005 needs to be delivered by 2025-06-03, and the customer rating is high. After successful bidding by out-of-specification supplier B, the system automatically upgrades its priority to "urgent" and notifies the logistics department to arrange transportation resources in advance. Conversely, non-urgent orders such as A4-006 have a more relaxed delivery deadline, and the priority remains unchanged. This dynamic adjustment ensures timely processing of critical orders. Specifically, reducing the order quota of out-of-specification suppliers is an important means of optimizing resource allocation. Assuming supplier B's monthly quota is 50 orders, due to the out-of-specification response time, the system reduces its quota to 30, while increasing the quota of high-quality supplier C from 50 to 70. The evaluation of high-quality suppliers is based on historical response times such as 2.8 hours and on-time delivery rates such as 98%. After adjusting the quota, the system generates a new quota table, recording the adjustment time 2025-06-01 and the quota changes of each supplier, ensuring transparency and traceability in the procurement process. For example, increasing the weight of the "on-time delivery rate" indicator in the order procurement process can optimize supplier matching. The original matching model may set the weights of response time, cost, and on-time rate to 4:3:3, and now adjust them to 3:3:4, highlighting the importance of on-time rate. Taking order A4-007 as an example, supplier C has an on-time rate of 98% and a response time of 2.8 hours, while supplier D has an on-time rate of 90% and a response time of 3.0 hours. After adjusting the weights, supplier C's matching score is higher than supplier D, and supplier C is preferred in the evaluation. The matching degree is calculated based on weighted average, considering various indicators to ensure that the procurement result better meets the efficiency requirements of the supply chain. In one embodiment, the optimized order procurement strategy is further improved through real-time monitoring and feedback. The system records the timestamp of each invitation to suppliers, such as 2025-06-02-10:00, order number A4-008, eligible supplier C, and predicted response time 2.7 hours. If the actual response time deviates from the predicted value, the system automatically adjusts the weights or quotas for the next procurement. For example, if the actual response time of eligible supplier C is 3.5 hours, the system may reduce its matching score.This closed-loop mechanism ensures that the procurement strategy is continuously optimized to adapt to the dynamic performance of suppliers. It is worth noting that the implementation of the above method relies on data-driven decision support. The labeling of over-standard suppliers, priority adjustment, and matching optimization form a complete order management closed loop. Each link is supported by specific data such as timestamps, predicted values, and on-time rates, ensuring transparent and quantifiable operations. This method effectively balances the matching of supplier resources and order demand, improving the overall response efficiency and stability of the supply chain.

[0028] In step S107, the order procurement task is executed through the optimized order procurement strategy, platform operation data is obtained from the platform operation log, and collaboration index data is calculated based on the platform operation data, wherein the collaboration index data includes order response speed and matching accuracy.

[0029] The order procurement task is executed through the optimized order procurement strategy, a quote invitation is sent to each supplier, and order procurement data is recorded in real time to obtain a platform operation log, wherein the platform operation log includes order number, time of sending a quote invitation to a supplier, receipt time, and actual successful bidder. From the platform operation log, operation data is obtained, for orders that have received a confirmation receipt, the average interval between the time of sending a quote invitation to a supplier and the receipt time is calculated as the order response speed. By comparing the actual successful bidder in the platform operation log with the target supplier preset by the strategy, the proportion of the number of matching orders is calculated as the matching accuracy. Collaboration index data including response speed and matching accuracy is generated.

[0030] In an embodiment, the optimized order procurement strategy executes procurement order procurement tasks through an automated system, ensuring efficient pushing of orders to supplier interfaces. The system transmits order information such as order number, procurement material, and quantity to suppliers in real-time through API interfaces. For example, Order B5-001 contains 100 parts and needs to be delivered by 2025-06-05, the system pushes it to the interface of Supplier C on 2025-06-02-14:00. After pushing, the system records platform operation logs, including order number B5-001, the time of initiating a quotation invitation to the supplier 2025-06-02-14:00, the time of receiving a reply 2025-06-02-14:30, and the actual successful bidder Supplier C. The logs are stored in the form of timestamps, ensuring traceability. This recording method facilitates subsequent analysis of order processing efficiency. For example, extract operation data from platform operation logs, filter orders that have received confirmation replies normally, and calculate the interval between the time of initiating a quotation invitation to the supplier and the time of receiving a reply. Suppose the log shows that Order B5-001 initiates a quotation invitation to the supplier at 2025-06-02-14:00 and receives a reply at 2025-06-02-14:30, with an interval of 0.5 hours; Order B5-002 has an interval of 0.4 hours. The system calculates the average response speed of 100 orders in the month as 0.45 hours. The response speed reflects the efficiency of the supplier in processing orders, providing data support for optimizing procurement strategies. Specifically, compare the actual successful bidder in the log with the target supplier preset by the strategy to calculate the prediction accuracy. Suppose the strategy presets Order B5-001 and B5-002 to face Suppliers C-L for bidding, and the log shows that the actual bidding faces the same suppliers. Among the 100 orders in the month, 90 orders have actual suppliers consistent with the preset, with an accuracy rate of 90%. A high accuracy rate indicates that the system effectively implements the optimization strategy and reduces human intervention. It should be noted that orders with low accuracy may be due to temporary adjustments by suppliers or system errors, and further analysis of timestamps and supplier data in the log is required. In an embodiment, the synergy indicator data is generated in combination with response speed and matching accuracy. For example, the system generates a report showing a response speed of 0.45 hours and a matching accuracy of 90%. The synergy indicator is evaluated comprehensively through weighted calculation, assuming that the response speed weight is 0.6 and the accuracy rate weight is 0.4, resulting in a synergy score of 0.78. The report is displayed in the form of a chart, making it easy for managers to intuitively understand the efficiency of the supply chain. Preferably, the system can adjust the strategy based on the synergy indicator, such as prioritizing suppliers with faster response speeds. For example, if Supplier D has a response speed of 0.6 hours, which is lower than the average, the system may reduce its quota and increase the orders of Supplier C. The log records the adjustment time 2025-06-03-09:00 and the details of the quota change, ensuring transparency. This closed-loop feedback mechanism optimizes procurement through real-time data analysis, improving the efficiency of supply chain synergy.

[0031] In step S108, the key influencing factors affecting the synergy index data are determined by analyzing the synergy index data through a decision tree algorithm, the purchase cycle duration data including the planned purchase time and the actual arrival time in the historical purchase data are obtained, the Pearson correlation coefficient and the significance level of the key influencing factors and the purchase cycle duration data are calculated, and the final purchase decision data is generated according to the key influencing factors and the corresponding Pearson correlation coefficient and significance level.

[0032] According to the synergy index data, a feature matrix including order response speed and matching accuracy fields is constructed. A decision tree algorithm is applied to analyze the feature matrix, and key influencing factors with an importance score greater than 0.1 are extracted. Purchase cycle duration data is extracted from historical purchase data. The Pearson correlation coefficient and the significance level of each key influencing factor and the purchase cycle duration data are calculated to form a factor-coefficient mapping table with the key influencing factors, the Pearson correlation coefficient and the significance level as the core fields, wherein the significance level is calculated by hypothesis testing on the Pearson correlation coefficient. According to the correlation coefficient and the significance level in the factor-coefficient mapping table, the final purchase decision data including the factor optimization direction and the decision threshold is generated.

[0033] In one embodiment, when constructing the feature matrix, the order response speed and matching accuracy and other collaborative index data need to be extracted from the platform operation log. The feature matrix takes orders as rows and fields as response speed, accuracy and other related variables such as order quantity and material type to form a structured data table for subsequent analysis. When training the decision tree, the model needs to be provided with label data. In this embodiment, the label is defined as "procurement collaboration influence degree" and is calculated by expert scoring combined with historical procurement data. For example, the procurement collaboration influence degree of an order = 0.4 x (response speed / historical average response speed) + 0.3 x (matching accuracy / 100) + 0.3 x (order quantity / historical maximum order quantity). The feature matrix (response speed, accuracy, etc.) and label data are input into the decision tree model, and the model is trained by minimizing the mean square error (MSE). For example, when analyzing the feature matrix using the decision tree algorithm, the system inputs the matrix into the model and calculates the importance score of each field. The importance score is calculated by the decrease in node impurity: each time the model evaluates the proportion of impurity reduction (such as the decrease in MSE) brought about by the feature split, and the total contribution of the feature in all splits is the importance score. For example, the importance score of response speed is 0.35, indicating that the feature contributes 35% to reducing the prediction error of the model. Assuming the analysis result shows that the importance score of response speed is 0.35, the matching accuracy is 0.25, the order quantity is 0.15, and the material type is 0.08. Filtering fields with a score greater than 0.1, response speed, matching accuracy and order quantity are obtained as key influencing factors. These factors reflect the core drivers of supply chain efficiency. Orders with fast response speed usually shorten the delivery cycle, high accuracy reduces error costs, and order quantity affects procurement priority. In one possible implementation, the procurement cycle length data is extracted from historical procurement data. The system selects all orders in June 2025 and records the time from ordering to delivery. For example, the procurement cycle of order B6-001 is 5 days, and the procurement cycle of order B6-002 is 7 days. The average cycle of 100 orders is 6 days. These data provide a basis for subsequent correlation analysis and ensure that the analysis results are close to actual business. Specifically, when calculating the Pearson correlation coefficient between the key influencing factors and the procurement cycle length, the system analyzes the response speed, matching accuracy and order quantity with the procurement cycle respectively. Assuming that the correlation coefficient between response speed and procurement cycle is -0.65, indicating that the faster the response speed, the shorter the cycle; the matching accuracy is 0.40, indicating that the higher the accuracy, the cycle is slightly shortened; the order quantity is 0.20, which has a weak impact. The significance level is determined by statistical test. Assuming that the significance level of response speed is 0.01, the significance level of matching accuracy is 0.05, and the significance level of order quantity is 0.10. The factor-coefficient mapping table is generated, including fields: key factor, correlation coefficient, and significance level.For example, response speed (-0.65, 0.01), matching accuracy (0.40, 0.05), order quantity (0.20, 0.10). The table intuitively shows the influence strength and reliability of each factor on the procurement cycle. For example, when generating procurement decision data according to the factor-coefficient mapping table, the system sets the optimization direction and decision threshold based on the correlation coefficient and significance level. The correlation coefficient of response speed is -0.65 and the significance is high, the optimization direction is to prefer suppliers with response speed less than 0.5 hours, and the decision threshold is 0.5 hours. The correlation coefficient of matching accuracy is 0.40, the optimization direction is to ensure that the accuracy is higher than 90%, and the threshold is 90%. The order quantity has a weak influence, and the threshold is set to no more than 1000 pieces per order to avoid excessive concentration of resources. The final decision data is output in table form, including factors, optimization direction and threshold, guiding order bidding. For example, the system adjusts the strategy to preferentially bid orders to supplier C with fast response speed, records the adjustment time 2025-07-02-09:00, and ensures the improvement of supply chain efficiency. It can be understood that when applying the decision tree algorithm to analyze the feature matrix, the system can retrieve the historical review data of relevant field experts in the group expert library, for example, invite procurement field experts to verify the correlation analysis result of "order response speed" and "procurement cycle", and the experts can adjust the feature importance score threshold based on industry experience (such as adjusting "importance score greater than 0.1" to 0.15). When calculating the Pearson correlation coefficient, the expert can set the industry standard of the significance level (such as p<0.03) to ensure that the analysis result conforms to the actual group procurement.

[0034] In step S109, configuration parameters are extracted from the final procurement decision data, and the order procurement strategy and resource scheduling rules of the platform are iteratively updated according to the configuration parameters to obtain full-chain running state data, wherein the configuration parameters refer to parameters including supplier priority weight, response time index weight, maximum carrying order quantity of the supplier, and emergency order resource quota proportion.

[0035] Configuration parameters are extracted from the final procurement decision data, and the configuration parameters refer to parameters including supplier priority weight, response time index weight, maximum carrying order quantity of the supplier, and emergency order resource quota proportion. The order allocation algorithm built-in the platform is updated according to the configuration parameters. The platform running data after updating the order allocation algorithm is obtained, and full-chain running state data including strategy version, update time and index change is generated.

[0036] For example, in a supply chain management platform, extracting configuration parameters from final procurement decision data is a key step in optimizing order procurement. Configuration parameters include supplier priority weight, response time indicator weight, maximum supplier order capacity, and emergency order resource quota ratio. These parameters directly affect the efficiency of the order allocation algorithm. Supplier priority weight reflects the importance of the supplier in procurement, based on historical performance such as response speed and accuracy. For example, supplier A has a weight of 0.8 due to fast response speed and high accuracy, while supplier B has a weight of 0.5. Response time indicator weight is used to balance speed and cost, assuming that the platform prioritizes speed, with a weight of 0.6. The maximum supplier order capacity limits the concentration of orders from a single supplier, such as a maximum of 800 orders for supplier A. The emergency order resource quota ratio reserves resources for critical orders, set at 20% to handle unexpected demand. In one possible implementation, when updating the order procurement algorithm, the platform adjusts the logic based on the configuration parameters. For example, the algorithm prioritizes suppliers with a weight higher than 0.7, while ensuring a response time of less than 0.5 hours. The system sets the order limit for supplier A to 800 orders, and any excess is bid to other eligible suppliers. Critical orders are prioritized for fast-responding suppliers according to the 20% quota. This update ensures that the algorithm aligns with business needs and improves procurement efficiency. For example, when obtaining platform running data after updating, the system records the strategy version, update time, and changes in indicators. Full-chain running state data reflects the effectiveness of the algorithm. For example, strategy version V2.0 was updated on 2025-07-03-10:00, response speed decreased from 0.4 hours to 0.3 hours, and matching accuracy increased from 92% to 95%. Data also includes order completion rate, delayed order ratio, and other indicators, showing the improvement in supply chain efficiency. These data provide the basis for subsequent optimization, ensuring continuous improvement in decision-making. It can be understood that the design of the above configuration parameters and algorithm updates closely integrates key factors from historical procurement data, such as response speed and accuracy, ensuring that parameter settings are scientific and reasonable. Comprehensive recording of running state data provides real-time monitoring and feedback mechanisms for the platform, helping to dynamically adjust strategies to adapt to market demand changes.

[0037] In step S110, the final procurement result is sent to the winning supplier in the form of a communication notice, and the final procurement result is combined with the internally preset fulfillment system to complete subsequent contract signing and fulfillment tracking, and the tracking results form supplier big data for supplier portrait evaluation.

[0038] Exemplarily, the final procurement result is sent to the winning supplier in the form of a transaction notice, and after the procurement transaction result is synchronized to the internally preset performance system, a contract signing process is automatically triggered, and the performance process (such as delivery progress and quality acceptance) is tracked throughout. The platform supports contract performance evaluation functions, collects data such as on-time delivery rate and after-sales service quality, and quantitatively evaluates suppliers. Evaluation data is used to build a supplier profile (such as performance capability score and risk level label), form a supplier big data, and provide data support for subsequent supplier selection and procurement strategy optimization.

[0039] The above only lists some preferred embodiments of the present application, but the present application is not limited thereto, and many improvements and changes can be made. Any improvement and change made on the basis of the basic principles of the present application shall be considered to fall within the scope of protection of the present application.

Claims

1. A procurement management method based on an intelligent supply chain management platform, characterized in that: The method includes: The process involves acquiring material demand data submitted by various departments, standardizing the timestamps and quantities of this data, and then categorizing and organizing the standardized data according to department name and material type. This results in standardized data containing auxiliary classification dimensions, time dimensions, and quantity dimensions. The auxiliary classification dimensions include department and material type. From this standardized data, the single-order purchase quantity for each department is extracted, and the fluctuation range of the purchase quantity is calculated. Abnormal purchase orders are removed from the standardized data based on preset fluctuation thresholds to obtain real-time demand data. An initial procurement plan is generated based on the real-time demand data and preset procurement management rules. The purchase quantities in the initial procurement plan are then grouped according to preset grouping rules. The data is grouped to obtain a grouped procurement plan table. Based on the order characteristic data in the grouped procurement plan table, combined with the access identifiers and blacklist identifiers in the pre-set supplier qualification database, the matching degree between the order characteristic data and the supplier characteristic data is calculated. Each procurement order is publicly invited to submit quotations to a pre-set number of suppliers with the highest matching degree, possessing access identifiers and not blacklisted. The order status is determined based on the confirmation receipts from the suppliers, and the order status is tracked to obtain a tracking log. Order processing data and supplier response time are extracted from the tracking log. A pre-trained ARIMA model is used to calculate the predicted supplier response time, obtaining the response time prediction data. Here, the supplier response time refers to the time from order placement to delivery. The time interval between a supplier initiating a quote invitation and the supplier sending a confirmation reply is considered. Response time prediction data includes supplier response times predicted for future orders. This prediction data is compared to a preset response time threshold. If the predicted response time exceeds the threshold, the order procurement process is optimized by adjusting order priorities and resource allocation to determine the optimized order procurement strategy. The optimized order procurement strategy is then executed, and platform operation data is retrieved from the platform operation logs. Collaboration indicator data, including order response speed and matching accuracy, is calculated based on this data. Finally, a decision tree algorithm is used to analyze the collaboration indicator data. The key influencing factors affecting the collaborative performance index data are identified. Historical procurement data, including planned procurement time and actual delivery time, is obtained. The Pearson correlation coefficient and significance level between the key influencing factors and the procurement cycle duration data are calculated. Based on the key influencing factors and their corresponding Pearson correlation coefficients and significance levels, final procurement decision data is generated. Configuration parameters are extracted from the final procurement decision data. The platform's order procurement strategy and resource scheduling rules are iteratively updated based on these configuration parameters to obtain full-chain operational status data. The configuration parameters include supplier priority weight, response time indicator weight, maximum order capacity of suppliers, and emergency order resource quota ratio.

2. The method according to claim 1, characterized in that, The process involves acquiring material demand data submitted by various departments, standardizing the timestamps and quantities of the material demand data, and then classifying and organizing the standardized material demand data according to department name and material type to form standardized data containing auxiliary classification dimensions, time dimensions, and quantity dimensions. The auxiliary classification dimensions include department and material type, including: By acquiring material demand data submitted by various departments through a real-time data acquisition system, preliminary data including timestamps, demand quantities, departments, and material types are obtained. The timestamps and demand quantities in the preliminary data are standardized according to preset standardization rules to generate demand records with time identifiers; The demand records with time stamps were categorized and organized according to department name and material type to obtain standardized data that includes auxiliary classification dimensions, time dimensions, and quantity dimensions.

3. The method according to claim 1, characterized in that, The process involves extracting the single purchase quantity of individual orders from each department from standardized data, calculating the fluctuation range of the purchase quantity, and removing abnormal purchase orders from the standardized data based on a preset fluctuation threshold to obtain real-time demand data, including: Extract the single purchase quantity of each department's individual orders from the standardized data, and calculate the standard deviation of each department based on the sequence of multiple single purchase quantities of each department, which serves as the fluctuation range value. The single purchase quantity of multiple individual orders of each department includes the single purchase quantity of the current order and the single purchase quantity of historical orders. If the fluctuation range exceeds the preset fluctuation threshold, the current purchase order will be marked as an abnormal purchase order, and the abnormal purchase order will be removed to obtain real-time demand data.

4. The method according to claim 1, characterized in that, The process involves generating an initial procurement plan table based on real-time demand data and preset procurement management rules, and then grouping the procurement data in the initial procurement plan table according to preset grouping rules to obtain a grouped procurement plan table, including: The material type, demand quantity, demand department, and timestamp fields in the real-time demand data are matched with preset procurement management rules to generate an initial procurement plan table. The preset management rules include framework procurement ratio allocation rules and supplier capacity limit rules. The initial procurement plan table includes the name, specifications, suggested procurement quantity, minimum / maximum procurement quantity limit, demand department, and planned procurement time of the procured materials. The suggested procurement quantity is calculated based on the demand quantity, preset safety stock threshold, and supplier capacity. According to the preset grouping rules, the procurement data in the initial procurement plan table is aggregated or split to form multiple procurement groups. The grouping rules include classification by demanding department, classification by material type, classification by supplier suitability category, and classification by procurement urgency. For each procurement group, data standardization processing is performed to generate a group procurement plan table that includes group number, material category, set of demanding departments, summary of procurement quantity, planned procurement time interval and corresponding supplier range. The standardization processing includes unifying the procurement unit, merging similar materials, and marking the group leader or the supplier to be contacted.

5. The method according to claim 1, characterized in that, The process involves calculating the matching degree between order feature data and supplier feature data based on order feature data in the grouped procurement plan table, combined with the access identifiers and blacklist identifiers in the preset supplier qualification database. For each procurement order, a predetermined number of suppliers with the highest matching degree, possessing the access identifier and not being on the blacklist are publicly invited to submit quotations. The order status is determined based on the confirmation receipts from the suppliers, and the order status is tracked to obtain a tracking log, including: Purchase orders are extracted from the grouped procurement plan table to generate order feature data, which includes material type, purchase quantity, and order urgency. Based on the access identifier and blacklist identifier in the supplier qualification database, a supplier feature dataset is constructed. The access identifier is determined according to the supplier's supply capacity and on-time delivery rate, and the blacklist is determined according to historical default records. The matching degree between order feature data and supplier feature data is calculated by weighted Euclidean distance algorithm. Combined with the supplier's supply capacity, each purchase order is publicly invited to quote from a preset number of suppliers with the highest matching degree, access identifier and no blacklist identifier. The order status is marked based on the confirmation receipt from the supplier. The time when the order was sent to the supplier for a quotation invitation, the supplier's response time, and status change information are recorded. A tracking log containing basic order information, supplier information, and status trajectory is generated. The confirmation receipt includes one of three status information: confirmed, rejected, and pending.

6. The method according to claim 1, characterized in that, The process involves extracting order processing data and supplier response times from tracking logs, and using a pre-trained ARIMA model to calculate predicted supplier response times, resulting in predicted response time data. Supplier response time refers to the time interval from when an order sends a quote request to the supplier to when the supplier sends a confirmation receipt. The predicted response time data includes supplier response times predicted for future orders, including: Based on the time when the quotation invitation was sent to the supplier, the confirmation receipt time, and the corresponding supplier information, the actual response time is calculated and abnormal orders are removed. Combined with the supplier's historical response time, comprehensive response time data including timestamp, current response time, and historical response time is generated. Data preprocessing methods were used to clean the comprehensive response time data, remove outliers and fill in missing values ​​to obtain standardized time series data. Standardized time series data is input into a pre-trained ARIMA model to generate response time prediction data containing prediction time points, supplier identifiers, and predicted response time values. The ARIMA model is pre-trained using historical time series data consisting of the supplier's historical response times.

7. The method according to claim 1, characterized in that, The step of comparing the predicted response time data with a preset response time threshold, and if it exceeds the preset response time threshold, then optimizing the order procurement behavior by adjusting order priority and resource allocation, and determining the optimized order procurement strategy, includes: The predicted response time data of each supplier is compared with the preset response time threshold. If the response time is greater than the preset response time threshold, the supplier is marked as exceeding the threshold. For orders where suppliers exceeding the standards successfully bid, the task priority will be increased according to the urgency level; Reduce the quota of bidable orders for suppliers that exceed the limit, and increase the quota for high-quality suppliers; In the order procurement process, the weighting of indicators that improve on-time delivery rate is determined, the matching degree between orders and suppliers is recalculated, and an optimized order procurement strategy is determined.

8. The method according to claim 1, characterized in that, The process involves executing order procurement tasks using an optimized order procurement strategy, obtaining platform operation data from platform operation logs, and calculating collaborative performance metrics based on this data. These collaborative performance metrics include order response speed and matching accuracy, and include: The platform executes order procurement tasks through an optimized order procurement strategy, sends quotation invitations to suppliers for each order, and records order procurement data in real time to obtain platform operation logs. The platform operation logs include order number, time of sending quotation invitations to suppliers, reply time, and actual successful bidders. Obtain operational data from the platform's operational logs, and for orders that have received confirmation receipts normally, calculate the average interval between the time it takes to send a quotation invitation to the supplier and the time it takes to receive the receipt as the order response speed; Compare the suppliers who actually won bids in the platform's operation logs with the target suppliers preset in the strategy, and calculate the percentage of orders that match each other as the matching accuracy rate. Generate collaborative metrics data that include response speed and matching accuracy.

9. The method according to claim 1, characterized in that, The process involves analyzing the synergy indicator data using a decision tree algorithm to identify key influencing factors. This includes acquiring historical procurement data containing planned and actual delivery times, calculating the Pearson correlation coefficient and significance level between the key influencing factors and the procurement cycle duration data, and generating final procurement decision data based on the key influencing factors and their corresponding Pearson correlation coefficients and significance levels. Construct a feature matrix that includes fields for order response speed and matching accuracy based on collaborative indicator data; The feature matrix was analyzed using a decision tree algorithm to extract key influencing factors with an importance score greater than 0.1; Extract procurement cycle duration data from historical procurement data; Calculate the Pearson correlation coefficient and significance level of each key influencing factor and the procurement cycle duration data, and form a factor-coefficient mapping table with key influencing factors, Pearson correlation coefficient and significance level as the core fields. The significance level is calculated by hypothesis testing on the Pearson correlation coefficient. Based on the correlation coefficients and significance levels in the factor-coefficient mapping table, final procurement decision data containing factor optimization directions and decision thresholds is generated.

10. The method according to claim 1, characterized in that, The process involves extracting configuration parameters from the final procurement decision data, iteratively updating the platform's order procurement strategy and resource scheduling rules based on these parameters, and obtaining full-chain operational status data. The configuration parameters include, but are not limited to, parameters such as supplier priority weight, response time indicator weight, maximum supplier order capacity, and emergency order resource quota ratio. The final procurement results are sent to the winning supplier in the form of a notice of award. The final procurement results are combined with the internally pre-set performance system to complete the subsequent contract signing and performance tracking. The tracking results form supplier big data for supplier profile evaluation.

Images