Enterprise purchase transaction control method and system based on multi-mode matching and risk pricing

Abstract

The invention discloses an enterprise purchase transaction control method and system based on multi-mode matchmaking and risk pricing, and belongs to the field of supply chain management, man-machine collaborative operation management and artificial intelligence algorithm application. The invention constructs an intelligent system comprising four core modules, namely a demand identification module, a supply matching module, a multi-mode matching and risk pricing module and an execution and performance feedback module. A risk factor weighted evaluation model is established by collecting multi-dimensional data such as real-time inventory, passenger flow prediction, supplier credit and the like, an optimal scheme is automatically screened from six transaction modes such as fixed-price purchase and order-sharing purchase, and dynamic risk pricing is synchronously realized; and through full-process automatic execution and performance data feedback, model iterative optimization is completed. According to the invention, the purchase cost and loss can be significantly reduced, the purchase intelligence level and the supply chain stability are improved, the method is suitable for a multi-industry B2B purchase scene, and automatic control and risk pricing of enterprise purchase transaction are realized through a computer system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of supply chain management, human-computer collaborative management, artificial intelligence algorithm and platform transaction matching, and particularly relates to a business procurement transaction control method and system based on multi-mode matching and risk pricing. BACKGROUND

[0002] With the development of digital economy, B2B procurement as the core link of the supply chain directly affects the operating efficiency of enterprises. At present, although the B2B procurement platform in the market has realized the basic online transaction function, there are still many core pain points in the actual application process, which is difficult to meet the intelligent and efficient procurement needs of enterprises, which is specifically manifested as follows:

[0003] 1) Single and fixed transaction mode: The existing B2B procurement platform mainly supports only the single transaction mode of fixed price procurement, which cannot be flexibly adjusted according to the actual demand characteristics of enterprises (such as procurement quantity, demand urgency), market supply and demand relationship changes, resulting in high procurement cost, especially for small batch procurement enterprises or emergency procurement scenarios, it is difficult to achieve cost optimization.

[0004] 2) Lack of scientific risk identification and price credit correlation mechanism: In the procurement matching process, the platform mainly focuses on the price of the supplier, without fully considering the core risk factors such as the probability of supply interruption, performance history and default cost of the supplier, and without establishing a dynamic correlation model between risk and price, resulting in frequent risks such as supply interruption and unqualified performance in the procurement process, which seriously affects the normal operation of enterprises; At the same time, the lack of consideration of the credit status and cooperation stability of enterprises, it is difficult to realize accurate supply and demand matching.

[0005] 3) The problem of waste of near-expiration consumables and low-turnover inventory is prominent: For the catering, fresh food and other industries, the shelf life of goods is short and the turnover requirement is high. The existing platform lacks a special transaction mechanism for near-expiration goods, and cannot realize the rapid matching and disposal of near-expiration consumables.

[0006] 4) Service resources are difficult to integrate into a unified procurement system: The existing B2B procurement platform mainly focuses on the procurement of physical goods, and excludes service resources such as technician working hours and robot service capabilities from the procurement system, which cannot realize the unified procurement and efficient matching of service resources; A large number of service resources are in idle state, the resource utilization rate is low, and the service procurement needs of enterprises cannot be accurately met.

[0007] 5) Low degree of automation of procurement process, lack of closed-loop execution mechanism: The procurement process of the existing platform still needs a lot of manual intervention, from demand identification, supplier selection, order placement to performance monitoring, which depends on manual experience decision-making, not only low efficiency, but also prone to procurement deviation due to human error.

[0008] In view of the above problems, the present application provides a business procurement transaction control method and system based on multi-mode matching and risk pricing. SUMMARY

[0009] The present application aims to overcome the defects of the existing B2B procurement platform, such as single transaction mode, lack of risk identification, low resource utilization, and insufficient automation, and provides a business procurement transaction control method and system based on multi-mode matching and risk pricing. By constructing four core modules, the present application realizes accurate identification of procurement demand, efficient matching of supply and demand, intelligent selection of transaction mode, dynamic pricing based on risk orientation, and automatic execution and optimization of the whole process, ultimately achieving the goal of reducing procurement cost, improving efficiency, and enhancing supply chain stability.

[0010] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0011] A business procurement transaction control method based on multi-mode matching and risk pricing, comprising the following steps:

[0012] S1, demand identification: collecting real-time inventory data, passenger flow prediction data, historical order trend data, and commodity attribute data, and identifying procurement gap level, demand time sensitivity, and procurement event type by a processor, and outputting standardized procurement demand instructions;

[0013] S2, supply matching: based on the procurement demand instructions, calling credit rating, time performance ability, historical pricing data, geographic matching degree, and supply capacity data in the supplier database, and obtaining a candidate supplier set that meets the demand through a screening algorithm;

[0014] S3, multi-mode matching and risk pricing: constructing a multi-dimensional risk factor weighted evaluation model, comprehensively scoring the candidate supplier set and the selectable transaction mode, and automatically selecting the optimal transaction mode and determining the dynamic optimal price interval according to the scoring results; the multi-dimensional risk factors include cost factor, risk factor, profit factor, time factor, loss factor, and behavior factor; the selectable transaction mode includes fixed price procurement, group purchase, competitive procurement, or pricing auction, near-term price reduction / idle inventory matching, long-term contract price locking, and service capacity procurement;

[0015] S4, execution and performance feedback: according to the optimal transaction mode and price interval, automatically completing ordering, payment, delivery track monitoring, warehouse verification, and abnormal problem correction; collecting whole-process performance data and feeding back to the system database for optimizing the risk factor weighted evaluation model and the transaction mode selection algorithm.

[0016] Further, in step S1, the preset algorithm is a demand prediction algorithm based on a time series prediction model, which includes but is not limited to an LSTM neural network model, specifically including: performing time series analysis on historical order trend data and passenger flow prediction data, combining real-time inventory data and commodity shelf life threshold and safety stock threshold, and calculating procurement gap; according to the ratio of the procurement gap to the demand time window, the gap level is divided into emergency gap, regular gap and standby gap; according to the difference between the commodity turnover period and the demand time window, the time sensitivity level is determined.

[0017] Further, in step S3, the construction process of the multi-dimensional risk factor weighted evaluation model includes:

[0018] S31, determine the specific evaluation index and quantitative standard of each dimension risk factor:

[0019] The cost factor evaluation index includes procurement unit price, logistics transportation cost and order processing cost, and the quantitative standard is the ratio of each cost index to the industry average level;

[0020] The risk factor evaluation index includes supply interruption probability, historical performance deviation rate and default compensation cost, and the quantitative standard is the statistical probability and cost estimation value based on the historical data of the supplier;

[0021] The profit factor evaluation index includes the gross profit contribution of the purchased commodity and the capital occupation period, and the quantitative standard is the ratio of the gross profit contribution to the procurement amount and the ratio of the capital occupation period to the average capital turnover period of the industry;

[0022] The time factor evaluation index is the demand urgency, and the quantitative standard is the difference between the demand time window and the average performance period of the supplier;

[0023] The loss factor evaluation index includes the commodity expiration risk coefficient and the inventory pressure coefficient, and the quantitative standard is the ratio of the remaining shelf life to the total shelf life and the ratio of the current inventory turnover rate to the safety turnover rate;

[0024] The behavior factor evaluation index includes the merchant credit rating and the cooperation stability, and the quantitative standard is the third-party credit rating score and the ratio of the continuous cooperation period to the total cooperation period;

[0025] S32, determine the weight of each dimension risk factor by using the analytic hierarchy process, and correct the weight by using the entropy weight method to obtain the final weighting coefficient;

[0026] S33, calculate the comprehensive score based on the quantitative value of each dimension risk factor and the final weighting coefficient; set a transaction mode selection threshold, and automatically match the optimal transaction mode according to the comprehensive score and the procurement demand characteristics.

[0027] Furthermore, in step S3, the logic for selecting the transaction mode includes:

[0028] When the coefficient of variation of procurement demand is less than the preset threshold, it is determined that the demand is stable, and the fixed price procurement + long-term contract price lock-in mode is given priority.

[0029] When the absolute value of the difference between the total supply capacity of candidate suppliers and the procurement gap exceeds a preset ratio, it is determined to be a supply-demand mismatch, and competitive bidding or fixed-price auction mode is given priority.

[0030] When the risk coefficient of a product nearing its expiration date exceeds a preset threshold, it is determined to be a high risk of inventory expiration, and the automatic matching mode of near-expiration price reduction / idle inventory is selected first.

[0031] When the procurement gap is less than the preset procurement quantity threshold, it is determined that the procurement quantity is small, and the group purchase mode is given priority.

[0032] When the procurement event type is service procurement, the service capacity procurement model should be selected first, matching the idle status and cost level of human labor hours or robot services.

[0033] Furthermore, in step S4, the abnormal problem correction includes: when a delivery delay occurs, automatically triggering the alternative supplier replenishment process; when an inbound inspection fails, automatically initiating a return and exchange application and recording supplier performance abnormal data; the performance data includes order response time, delivery on-time rate, product qualification rate, default rate, and cost deviation value.

[0034] A corporate procurement transaction control system based on multi-mode matching and risk pricing includes a demand identification module, a supply matching module, a multi-mode matching and risk pricing module, and an execution and performance feedback module.

[0035] The demand identification module is used to collect real-time inventory data, customer flow forecast data, historical order trend data and product attribute data, and automatically identify the procurement gap level, demand timeliness sensitivity and procurement event type through a preset algorithm, and output standardized procurement demand instructions.

[0036] The supply matching module is communicatively connected to the demand identification module and is used to retrieve supplier database data based on procurement demand instructions, and obtain a set of candidate suppliers that meet the requirements through a screening algorithm; the supplier database data includes supplier credit rating, timeliness and performance capability, historical quotation data, geographical matching degree and supply capacity data;

[0037] The multi-mode matching and risk pricing module is connected to the demand identification module and the supply matching module respectively. It is the core innovation module of the system and is used to construct a multi-dimensional risk factor weighted evaluation model, comprehensively score the candidate supplier set and the selectable transaction modes, and automatically select the optimal transaction mode and determine the dynamic optimal price range based on the scoring results.

[0038] The execution and fulfillment feedback module is connected to the multi-mode matching and risk pricing module. It is used to automatically complete order placement, payment, delivery trajectory monitoring, warehousing verification and abnormal problem correction according to the optimal transaction mode and price range. It is also used to collect full-process fulfillment data and feed it back to the system database and the multi-mode matching and risk pricing module to realize model iterative optimization.

[0039] Furthermore, the demand identification module includes a data acquisition unit, a data preprocessing unit, and a gap and timeliness identification unit; the data acquisition unit adopts a multi-source data interface, which can access data from enterprise ERP systems, inventory management systems, customer flow statistics systems, and order management systems; the data preprocessing unit is used to perform noise reduction, standardization, and missing value imputation on the acquired data; the gap and timeliness identification unit has a built-in LSTM-based demand forecasting and gap identification algorithm to determine the procurement gap level and timeliness sensitivity level.

[0040] Furthermore, the multi-mode matching and risk pricing module includes a factor construction unit, a weight calculation unit, a comprehensive scoring unit, and a mode matching unit. The factor construction unit is used to define evaluation indicators and quantitative standards for six dimensions: cost, risk, profit, time, loss, and behavior. The weight calculation unit uses a combined weighting method combining the analytic hierarchy process (AHP) and the entropy weighting method to determine the final weighting coefficients of each dimension factor. The comprehensive scoring unit calculates a comprehensive score based on the quantitative values ​​and weighting coefficients. The mode matching unit has built-in trading mode selection logic and thresholds to determine the optimal trading mode and the dynamic optimal price range.

[0041] The beneficial effects of this invention are as follows:

[0042] 1. Cost reduction and efficiency improvement: By matching different scenarios with multiple modes, procurement costs are reduced, near-expiry and inventory waste is reduced, and resource utilization is improved; by coupling risk factors and pricing models, a dynamic and controllable computer pricing mechanism is achieved.

[0043] 2. Enhance intelligence: Relying on LSTM prediction and multi-dimensional risk models to achieve accurate matching, realize fully automated procurement, improve procurement efficiency, and reduce human error.

[0044] 3. Enhance supply chain stability: Incorporate multiple risk factors, establish a dynamic correlation between risk and price, improve performance assurance, and effectively reduce the risk of supply disruption.

[0045] 4. Enhance platform value: The model self-optimizes to form a positive network effect, which can be monetized through multiple means, resulting in significant commercial value.

[0046] 5. Promote business district collaboration: Support cross-industry resource matching, improve the utilization rate of business district resources and enterprise collaboration, and help the development of business district economy. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is an overall flowchart of the present invention;

[0049] Figure 2 This is a diagram illustrating the multi-dimensional risk factor system architecture of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0051] A method for controlling enterprise procurement transactions based on multi-mode matching and risk pricing includes the following steps:

[0052] S1. Demand Identification: Collect real-time inventory data, customer flow forecast data, historical order trend data, and product attribute data (such as product shelf life, unit cost, storage requirements, etc.). Using a preset LSTM-based demand forecasting and gap identification algorithm, perform time-series analysis and feature extraction on the collected data to calculate the procurement gap. Based on the ratio of the procurement gap to the demand time window, classify the gap into emergency gaps (ratio > preset emergency threshold), regular gaps (preset regular threshold ≤ ratio ≤ preset emergency threshold), and reserve gaps (ratio < preset regular threshold). Based on the difference between the product turnover cycle and the demand time window, determine the time sensitivity level (the smaller the difference, the higher the time sensitivity). Finally, output a standardized procurement demand instruction containing the type of procured goods, procurement quantity, gap level, time sensitivity level, and demand time window.

[0053] S2. Supply Matching: Based on the standardized procurement demand instruction output in step S1, core data from the supplier database is retrieved through a data interface. This core data includes supplier credit rating (provided by a third-party credit agency or generated based on historical cooperation data), timeliness and fulfillment capability (historical average fulfillment period, on-time fulfillment rate), historical quotation data (average quotation in the past 3 / 6 months, quotation fluctuation range), geographical matching degree (distance between the supplier and the purchasing company, delivery coverage), and supply capacity (maximum supply volume, supply stability). A screening algorithm based on the analytic hierarchy process is used to quantitatively score the above-mentioned core data of the suppliers, and suppliers with scores higher than the preset qualified threshold are selected to form a candidate supplier set.

[0054] S3. Multi-mode matching and risk pricing: This step is the core innovation step, and specifically includes the following sub-steps:

[0055] S31. Construct a multi-dimensional risk factor system: Identify the risk factors of six core dimensions and their corresponding evaluation indicators and quantitative standards, as detailed below:

[0056] a) Cost Factors: Core evaluation indicators include unit purchase price, logistics and transportation costs, and order processing costs; the quantitative standards are: unit purchase price = (supplier quotation - industry average quotation) / industry average quotation, logistics and transportation cost = logistics cost / purchase amount, and order processing cost = processing cost / purchase amount.

[0057] b) Risk Factors: Core evaluation indicators include the probability of supply disruption, historical performance deviation rate, and cost of breach of contract; the quantitative standards are as follows: the quantitative value of the probability of supply disruption = the number of historical supply disruptions / the total number of historical collaborations, the quantitative value of the performance deviation rate = |actual performance volume - agreed performance volume| / agreed performance volume, and the quantitative value of the cost of breach of contract = the historical average amount of breach of contract compensation / the purchase amount.

[0058] c) Profitability Factors: Core evaluation indicators include gross profit contribution and capital occupation period; the quantitative standards are: gross profit contribution = (expected sales price - purchase price) / purchase price, and capital occupation period = number of days of capital occupation / industry average capital turnover days.

[0059] d) Time factor: The core evaluation indicator is the urgency of demand; the quantitative standard is: Quantitative value of demand urgency = demand time window / average supplier fulfillment cycle;

[0060] e) Loss Factors: Core evaluation indicators include the near-expiry risk coefficient and the inventory pressure coefficient; the quantitative standards are: near-expiry risk coefficient = remaining shelf life of the product / total shelf life of the product, and inventory pressure coefficient = current inventory turnover rate / industry safety stock turnover rate.

[0061] f) Behavioral Factors: Core evaluation indicators include merchant credit rating and cooperation stability; the quantitative standards are: merchant credit rating quantitative value = third-party credit rating score / full score, cooperation stability quantitative value = continuous cooperation period / total cooperation period;

[0062] S32. Determine the weights of each factor: Construct a judgment matrix using the Analytic Hierarchy Process (AHP), and determine the initial weights of each dimension of risk factors through expert scoring; at the same time, use the entropy weight method to correct the initial weights, eliminate the influence of subjective factors, and obtain the final weighting coefficients of each dimension of risk factors; among them, the sum of the weighting coefficients of cost factors, risk factors, and time factors shall not be less than 60%, to ensure the dominant role of core factors.

[0063] S33. Comprehensive Scoring and Pattern Matching: Based on the quantified values ​​of risk factors in each dimension and the final weighted coefficients, a linear weighted summation method is used to calculate the comprehensive score of candidate suppliers and their corresponding transaction patterns. A comprehensive score threshold (e.g., 80 points) is set, and supplier-transaction pattern combinations with comprehensive scores higher than the threshold are selected. Simultaneously, based on procurement demand characteristics (e.g., demand stability, procurement volume, commodity type, etc.), a preset transaction pattern optimization logic is triggered to determine the optimal transaction pattern and its corresponding optimal supplier from the qualified combinations. Based on the comprehensive score results, market supply and demand relationship, and quantified values ​​of risk factors, a regression analysis model is used to determine the dynamic optimal price range.

[0064] The selectable transaction modes include the following six types, which are defined as follows:

[0065] 1) Fixed-price procurement: The supplier provides a fixed price, and the purchasing company completes the procurement at the fixed price. This is suitable for scenarios with stable demand and a balance between supply and demand.

[0066] 2) Group buying (joint purchasing): The platform integrates purchase orders from multiple companies with similar purchasing needs, forms a large-volume purchase order, and then purchases from suppliers. By leveraging economies of scale, costs are shared and unit prices are reduced. This is suitable for small-volume purchasing scenarios.

[0067] 3) Competitive bidding or fixed-price auction mechanism: The purchasing company publishes its procurement needs, and the candidate suppliers submit bids within a preset time. The platform selects the supplier with the best bid based on a comprehensive score; or the purchasing company sets a reserve price, and the suppliers bid, with the highest bidder (for idle / near-expiration resources) or the lowest bidder (for regular procurement) winning the bid. This is suitable for scenarios where supply and demand are mismatched.

[0068] 4) Automatic matching of near-expiry price reductions / idle inventory: For near-expiry products or idle inventory, the platform automatically sets price reduction tiers and matches them with purchasing companies that have corresponding needs, achieving rapid turnover. This is suitable for scenarios with high near-expiry risk and high inventory pressure.

[0069] 5) Long-term contract price-locking model: The purchasing company signs a long-term purchase contract with the supplier, agreeing on a fixed purchase price or a price fluctuation range to ensure stable supply and demand. This model is suitable for scenarios with stable demand and long-term cooperation.

[0070] 6) Service Capability Procurement: Technician working hours, robot service capabilities, etc. are procurable resources. The platform matches service providers with idle resources with purchasing companies that need services. This is suitable for service procurement scenarios.

[0071] The preset transaction mode selection logic is specifically as follows:

[0072] If the coefficient of variation (standard deviation / mean) of the procurement demand is less than the preset threshold (e.g., 0.2), the demand is determined to be stable, and the fixed-price procurement + long-term contract price locking model is given priority to ensure supply stability and price certainty.

[0073] If the absolute value of the difference between the total supply capacity of candidate suppliers and the procurement gap exceeds a preset ratio (e.g., 30%), it is determined to be a supply-demand mismatch, and competitive bidding or a pricing auction mechanism will be given priority to achieve supply-demand balance through market competition.

[0074] If the near-expiry risk coefficient of a product is less than a preset threshold (e.g., 0.3), it is determined that the risk of inventory expiration is high. The near-expiry price reduction / automatic matching mode of idle inventory is selected first, and a tiered price reduction strategy is set to improve turnover efficiency and reduce waste.

[0075] If the procurement gap is less than the preset procurement volume threshold (determined based on the industry average procurement volume), it is judged as a small procurement volume, and the group purchase model is given priority to integrate similar demands and reduce procurement costs.

[0076] If the procurement event type is a service procurement (such as technician working hours, robot services), the service capacity procurement model should be selected first, matching the resource idleness and cost level of the service provider, and giving priority to service providers with high idle rates and costs lower than the industry average.

[0077] S4. Execution and Performance Feedback: Based on the optimal transaction model, optimal supplier, and dynamic optimal price range determined in step S3, perform the following operations:

[0078] S41. Automated Execution: By connecting with the enterprise's financial system and supplier order system through system interfaces, the system automatically completes the order generation, order placement, and payment processes; by using GPS positioning systems and logistics information platform interfaces, the system obtains delivery trajectories in real time, enabling full monitoring of the delivery process; and by automatically triggering the warehousing verification process (physical goods are verified by scanning codes and undergoing quality inspection, while service resources are verified through performance acceptance) upon delivery.

[0079] S42. Anomaly Handling: If a delivery delay occurs (the actual delivery time exceeds the preset proportion of the agreed time), the system will automatically trigger the alternative supplier replenishment process, select the second-best supplier from the candidate supplier set to complete the replenishment; if the warehouse inspection fails (the physical goods are substandard in quality or quantity, or the service resources do not meet the agreed effect), the system will automatically initiate a return or exchange application or a service rework application, and record the supplier's performance anomaly data.

[0080] S43. Data Feedback: Collect full-process performance data, including order response time, on-time delivery rate, product qualification rate / service compliance rate, default rate, cost deviation value (the difference between actual procurement cost and estimated cost), etc.; after standardizing the performance data, feed it back to the system database, and simultaneously synchronize it to the multi-mode matching and risk pricing module to optimize the quantitative standards of risk factors, weighting coefficients and transaction mode selection logic, so as to achieve iterative optimization of the model.

[0081] Secondly, the present invention provides an enterprise procurement transaction control system based on multi-mode matching and risk pricing, used to implement the method described in the first aspect above, including a demand identification module, a supply matching module, a multi-mode matching and risk pricing module, an execution and performance feedback module, and a system database; the modules are connected through bus or network communication to realize real-time data transmission and interaction.

[0082] 1) Demand Identification Module: As the core of the system's demand input, it is used to accurately identify and standardize the output of procurement requirements; it includes a data acquisition unit, a data preprocessing unit, and a gap and timeliness identification unit.

[0083] The data acquisition unit adopts a multi-source data interface design, which can be flexibly connected to enterprise ERP system, inventory management system, customer flow statistics system, order management system and product information management system to realize batch collection and real-time updating of real-time inventory data, customer flow forecast data, historical order trend data and product attribute data;

[0084] The data preprocessing unit is used to clean, denoise, standardize, and impute missing values ​​in the collected multi-source data. Among them, outlier detection algorithms (such as the 3σ criterion) are used to remove outliers from the data, min-max standardization is used to transform data of different dimensions to the same interval, and linear interpolation is used to impute missing data to ensure data quality.

[0085] The gap and timeliness identification unit has a built-in LSTM-based demand forecasting and gap identification algorithm. It uses the LSTM model to perform time-series forecasting on historical order trend data and customer flow forecast data to obtain future demand forecast values; it combines real-time inventory data and commodity safety stock thresholds to calculate the procurement gap; it classifies the gap level according to the ratio of the procurement gap to the demand time window, determines the timeliness sensitivity level according to the difference between the commodity turnover cycle and the demand time window, and finally outputs standardized procurement demand instructions.

[0086] 2) Supply matching module: As the core link in supply and demand matching, it is used to screen out candidate suppliers that meet the requirements from a large number of suppliers; it includes a data retrieval unit and a supplier screening unit;

[0087] The data retrieval unit establishes a communication connection with the supplier database through a RESTful API interface, and accurately retrieves the corresponding core supplier data based on key information such as the type of goods to be purchased, the quantity to be purchased, and the time window of the demand in the standardized procurement demand instruction.

[0088] The supplier screening unit incorporates a screening algorithm based on the analytic hierarchy process (AHP). It uses supplier credit rating, timeliness of performance, historical pricing data, geographical matching degree, and supply capacity as screening indicators to construct a screening indicator system. The weight of each screening indicator is determined through the AHP, and the supplier's indicators are quantitatively scored. Suppliers with scores higher than the preset qualified threshold are selected to form a candidate supplier set, which is then synchronized to the multi-mode matching and risk pricing module.

[0089] 3) Multi-mode matching and risk pricing module: As the core innovative module of the system, it is used to realize multi-dimensional risk assessment, intelligent selection of trading modes and dynamic risk pricing; including factor construction unit, weight calculation unit, comprehensive scoring unit and mode matching unit;

[0090] The factor construction unit is used to define risk factors in six dimensions: cost, risk, profit, time, loss, and behavior. It clarifies the evaluation indicators, data sources, and quantitative standards for each factor, and establishes a standardized risk factor system. The data sources include supplier databases, enterprise operation databases, industry benchmark databases, and third-party credit databases.

[0091] The weight calculation unit adopts a combined weighting method that combines the analytic hierarchy process (AHP) and the entropy weighting method. First, the AHP is used to construct a judgment matrix, and the initial weights of each risk factor are determined by combining expert experience. Then, the entropy weighting method is used to calculate the objective weights of each risk factor, and the final weighting coefficients are obtained by merging them according to a preset ratio (e.g., 60% subjective weight and 40% objective weight) to ensure the scientific and reasonable allocation of weights.

[0092] The comprehensive scoring unit uses a linear weighted summation method to calculate the comprehensive score for each candidate supplier corresponding to different transaction models based on the quantified values ​​of each risk factor and the final weighting coefficient. At the same time, the comprehensive score is corrected by combining market supply and demand data (such as industry supply and demand ratio and price volatility index).

[0093] The pattern matching unit has built-in transaction pattern selection logic and comprehensive scoring threshold. Based on the demand characteristics (demand stability, purchase volume, commodity type, etc.) in the standardized procurement demand instruction, it triggers the corresponding selection logic to determine the optimal transaction pattern and the optimal supplier from the supplier-transaction pattern combination with a comprehensive score higher than the threshold. It also uses a multiple linear regression model with comprehensive score, market supply-demand ratio, and risk factor quantification as independent variables and purchase price as dependent variable to construct a dynamic pricing model and output the optimal price range.

[0094] 4) Execution and Performance Feedback Module: As the core of the system's execution and optimization, it is used to realize the automated execution and performance data feedback loop of the entire procurement process; it includes an order execution unit, a performance monitoring unit, an exception handling unit, and a data feedback unit.

[0095] The order execution unit connects to the enterprise's financial system, supplier order system, and logistics and distribution system through interfaces to automate processes such as order generation, order placement, payment, and delivery scheduling without manual intervention.

[0096] The performance monitoring unit adopts real-time data acquisition technology, obtains the real-time location of delivery vehicles through the GPS positioning interface, obtains delivery node information (such as outbound, en route, and arrival) through the logistics information interface, and obtains inbound verification data through the enterprise's internal warehousing system interface, thereby realizing real-time monitoring of the entire procurement process.

[0097] The anomaly handling unit has preset handling rules and procedures for various anomaly scenarios, including delivery delays, unqualified products, substandard services, and payment failures. When an anomaly event is detected, the corresponding handling procedure is automatically triggered, such as delayed order replacement, returns and exchanges, and fee reductions, and an anomaly handling report is generated.

[0098] The data feedback unit standardizes the entire process of performance data (order response time, on-time delivery rate, product qualification rate, default rate, cost deviation value, etc.) and synchronizes it to the system database. At the same time, it feeds the performance data back to the multi-mode matching and risk pricing module to optimize the quantitative standards, weighting coefficients and transaction mode selection logic of risk factors, so as to realize the closed-loop management of "decision-execution-feedback-optimization".

[0099] 5) System Database: As the core of the system's data storage, it adopts a distributed storage architecture to ensure data security, reliability, and scalability; including a procurement requirements database, supplier database, transaction database, performance database, and model parameter database;

[0100] The procurement demand database is used to store standardized procurement demand instructions and related raw data;

[0101] The supplier database is used to store core data such as suppliers' basic information, credit rating, performance history, quotation data, and supply capacity.

[0102] The transaction database is used to store transaction data such as purchase order information, transaction mode, transaction price, and transaction time;

[0103] The performance database is used to store performance data and exception handling data throughout the entire performance process.

[0104] The model parameter database is used to store model parameters of the multi-mode matching and risk pricing module, including risk factor weights, comprehensive scoring thresholds, and transaction mode selection logic parameters, and supports real-time updates based on feedback data.

[0105] Example 1: Automated Procurement Scenario in the Catering Industry

[0106] Application target: A chain restaurant enterprise that needs to purchase ingredients daily. Its demand is greatly affected by fluctuations in customer traffic, and it has strict requirements for the freshness of ingredients and procurement costs.

[0107] Implementation process: The demand identification module predicts a 30% increase in weekend customer traffic, calculates the procurement gap for vegetables and meat, and outputs standardized procurement demand; the supply matching module selects a set of qualified suppliers; because the procurement volume is less than the industry average, the group buying logic is triggered, integrating the demand of surrounding catering enterprises to form a batch order, matching the supplier with the best comprehensive score and determining the preferential price range; the execution module automatically completes order placement, delivery monitoring, handles returns and exchanges for a small number of damaged ingredients, and feeds the fulfillment data back to the system optimization model.

[0108] 1) Demand Identification: The demand identification module connects to the enterprise's POS system (customer flow data, sales data), inventory management system (real-time food inventory data), and commodity information system (food shelf life, storage requirements, etc.) through interfaces. Based on the LSTM model, it predicts weekend customer flow data and finds that weekend customer flow will increase by 30%, corresponding to a projected increase of 25% in food sales. Combined with real-time inventory data, it calculates that the procurement gap for vegetables is 50kg and the procurement gap for meat is 30kg, with the gap level being a regular gap. Due to the short shelf life of vegetables (2-3 days), the demand time window is 12 hours, and the time sensitivity level is high. Finally, it outputs a standardized procurement demand instruction that includes food type, procurement quantity, gap level, time sensitivity, and demand time window.

[0109] 2) Supply matching: The supply matching module retrieves the core data of 5 vegetable suppliers and 3 meat suppliers from the supplier database based on standardized procurement demand instructions. Among them, the credit rating of all suppliers is A or above, the average fulfillment cycle is 6-8 hours, and the geographical matching degree covers the business district where the store is located. Through the analytic hierarchy process, one vegetable supplier with a historical fulfillment deviation rate of more than 5% is removed, resulting in a candidate supplier set consisting of 4 vegetable suppliers and 3 meat suppliers.

[0110] 3) Multi-mode matching and risk pricing: The multi-mode matching and risk pricing module scores candidate suppliers based on multi-dimensional risk factors. Since the current purchase volume (50kg vegetables and 30kg meat) is less than the industry average (100kg vegetables and 50kg meat), the "small purchase volume, priority for group purchase" optimization logic is triggered. The system integrates the similar food purchase needs (30kg vegetables and 20kg meat) of two other small catering enterprises in the same business district, forming a group purchase order for a total of 80kg vegetables and 50kg meat. Based on the comprehensive score of cost factors (purchase unit price, logistics cost) and risk factors (probability of supply failure, performance deviation rate), the supplier with the highest comprehensive score is selected as the optimal supplier. Through a dynamic pricing model, combined with the scale effect of group purchase, the optimal price range is determined, with the vegetable purchase unit price reduced by 8% and the meat purchase unit price reduced by 5% compared to the market average.

[0111] 4) Execution and Fulfillment Feedback: The execution and fulfillment feedback module automatically generates group purchase orders and completes payment through an interface; it monitors the delivery trajectory in real time, and the supplier completes delivery within the agreed time (within 6 hours); if a small amount of vegetables are found to have minor damage during the warehouse inspection, the system automatically initiates a partial return and exchange application, and the supplier completes replenishment within 2 hours; fulfillment data (delivery on-time rate 95%, product qualification rate 98%, cost deviation -3%) is fed back to the system database to optimize risk factor weights.

[0112] Implementation results: Reduced procurement costs, improved procurement efficiency, and reduced food waste rate.

[0113] Example 2: Automated Handling of Near-Expiry Inventory in the Fresh Food Retail Industry

[0114] Application target: A fresh food retail chain with a large number of near-expiry products, where traditional manual handling is inefficient and wasteful.

[0115] Implementation process: The demand identification module monitors a batch of dairy products nearing their expiration date and detects high risk, triggering an event to handle near-expiration inventory; the supply matching module filters out qualified buyers with relevant purchasing needs; the near-expiration price reduction matching logic is triggered, a tiered price reduction strategy is set, and three optimal buyers are matched; the execution module automatically completes order generation, delivery monitoring, and payment collection, and the fulfillment data is fed back to the system.

[0116] 1) Demand Identification: The demand identification module is integrated into the inventory management system to monitor the shelf life of goods in real time. If a batch of dairy products is found to have only 3 days left of shelf life, the near-expiry risk coefficient is 0.25 (less than the preset threshold of 0.3), and the inventory pressure coefficient is 1.5 (greater than the industry safety stock turnover rate), a near-expiry inventory disposal procurement event is triggered. A standardized procurement demand instruction containing the product type (dairy products), inventory quantity (200 boxes), and near-expiry risk level (high) is output (here the procurement demand is "disposal of near-expiry inventory", i.e., matching the demand side).

[0117] 2) Supply matching: The supply matching module treats near-expiry dairy products as "supplyable resources", retrieves demand data from the supplier database for near-expiry products (such as small convenience stores and catering companies), and selects 10 demanders located in the same city who have dairy product procurement needs and have a good historical credit record to form a candidate demander set.

[0118] 3) Multi-mode matching and risk pricing: The multi-mode matching and risk pricing module triggers the optimal logic of "high near-expiry risk, priority to near-expiry price reduction / automatic matching of idle inventory"; sets a tiered price reduction strategy (60% off the original price, 55% off for purchases exceeding 50 boxes); comprehensively scores based on behavioral factors such as purchase volume, payment cycle, and historical cooperation stability of the demand party, and selects 3 demand parties with larger purchase volume and shorter payment cycle (purchases of 80 boxes, 60 boxes, and 60 boxes respectively) as the optimal demand parties; the final transaction price is determined to be 55% of the original price.

[0119] 4) Execution and Fulfillment Feedback: The execution and fulfillment feedback module automatically generates matching orders for near-expiry goods and notifies 3 demanders; it automatically arranges delivery and monitors the delivery trajectory in real time; after delivery is completed, the demanders complete acceptance confirmation through the system, and the system automatically completes payment collection; fulfillment data (100% on-time delivery rate, 100% acceptance pass rate, and inventory turnover cycle shortened to 2 days) are fed back to the system database.

[0120] Implementation results: All near-expiry dairy products were disposed of, avoiding waste and losses, improving inventory turnover efficiency, and reducing losses.

[0121] Example 3: Procurement Scenario for Manufacturing Service Capabilities

[0122] Application target: A machinery manufacturing company that needs to purchase services such as equipment maintenance and robot inspection. The demand varies significantly between peak and off-peak periods, with a high rate of idle service resources during off-peak periods.

[0123] Implementation process: The demand identification module detects the need for equipment inspection and maintenance during off-peak hours at night, triggering a service procurement event; the supply matching module filters out qualified providers with idle service resources; the service capacity procurement logic is triggered to match service providers with high idle rates, low costs, and excellent qualifications and determine preferential prices; the execution module automatically completes order generation, service quality monitoring, and acceptance payment, and the performance data is fed back to the system.

[0124] 1) Demand Identification: The demand identification module is connected to the enterprise production management system. It detects that during off-peak hours (22:00-6:00 the next day), routine inspection and maintenance of 3 production equipment are required, triggering a service capability procurement event. It outputs a standardized procurement demand instruction that includes the service type (equipment inspection, repair), service time (off-peak hours at night), and service standard.

[0125] 2) Supply matching: The supply matching module retrieves data from the supplier database of service providers with equipment testing and maintenance qualifications, and filters out 5 service providers with idle technician hours and robot testing equipment to form a candidate service provider set; among them, 3 service providers have an idle rate of 80% at night, and 2 have an idle rate of 60%.

[0126] 3) Multi-mode matching and risk pricing: The multi-mode matching and risk pricing module triggers the optimal selection logic of "service procurement, prioritizing service capability procurement mode"; it comprehensively scores service providers based on risk factors such as idle rate, service cost, qualification level, and historical service compliance rate; it selects the service provider with an idle rate of 80%, service cost 15% lower than the industry average, and the highest qualification level as the optimal service provider; and it determines the service price to be 80% of the industry average price.

[0127] 4) Execution and Performance Feedback: The execution and performance feedback module automatically generates service purchase orders, specifying service time and standards; during nighttime services, service quality is monitored in real time through a remote monitoring system; after the service is completed, the system automatically completes acceptance (equipment test data meets standards) and payment; performance data (service compliance rate 100%, cost deviation -15%, resource idle rate reduced by 80%) is fed back to the system database.

[0128] Implementation results: Improved utilization of service resources during off-peak periods, reduced service procurement costs, and improved service efficiency.

[0129] Example 4: Cross-industry business district complementary procurement scenario

[0130] Target audience: A core business district in a city, including businesses of various formats such as catering and fresh food retail, which have complementary needs for resources such as ingredients.

[0131] Implementation process: The demand identification module discovers that a catering company in the business district has surplus fresh vegetables and a fresh food retail company has a vegetable inventory gap, triggering a cross-business complementary procurement event; the supply matching module confirms that both parties are qualified and delivery is convenient, forming a supply and demand matching pair; the near-expiry price reduction matching logic is triggered to determine a reasonable transaction price; the execution module automatically completes order generation, shared logistics delivery and payment, and the performance data is fed back to the system.

[0132] 1) Demand Identification: The demand identification module connects to the inventory management systems of multiple catering enterprises within the business district and discovers that a certain catering enterprise has 20kg of fresh vegetables remaining on the day (not spoiled, with 1 day remaining shelf life); at the same time, it connects to the inventory management system of a fresh food retail enterprise within the business district and discovers that the enterprise has a shortage of 15kg of fresh vegetables in its inventory; this triggers a cross-business complementary procurement event and outputs a standardized procurement demand instruction that includes the supplier (catering enterprise), the demander (fresh food retail enterprise), the product type (fresh vegetables), and the quantity.

[0133] 2) Supply matching: The supply matching module confirms that the quality of the leftover vegetables from catering companies is up to standard (through the quality inspection photos uploaded by the system), the procurement needs of fresh food retail companies are genuine, both parties are companies registered on the platform with good credit ratings, the geographical distance is only 500 meters, and delivery is convenient, thus forming a qualified supply and demand matching pair.

[0134] 3) Multi-mode matching and risk pricing: The multi-mode matching and risk pricing module triggers the optimal logic of "high risk of near expiration + cross-business complementarity, priority to price reduction of near expiration / automatic matching of idle inventory"; considering the short remaining shelf life of vegetables, the transaction price is determined to be 70% of the catering enterprise's procurement cost, which not only ensures that the catering enterprise reduces losses, but also meets the cost needs of fresh food retail enterprises.

[0135] 4) Execution and Performance Feedback: The execution and performance feedback module automatically generates complementary purchase orders and notifies both companies; catering companies complete vegetable delivery through shared logistics within the business district, delivering within 30 minutes; after the fresh food retail companies pass the acceptance inspection, the system automatically completes the payment; performance data (delivery time 30 minutes, acceptance rate 100%, resource waste rate 0, business district resource utilization rate increased by 60%) is fed back to the system database.

[0136] Implementation results: Catering enterprises reduce waste and losses, fresh food retail enterprises reduce procurement costs, business district resource utilization rate is improved, and enterprise collaboration is enhanced.

[0137] In summary, this invention achieves accurate demand identification, efficient supply and demand matching, intelligent selection of transaction models, risk-oriented dynamic pricing, and automated execution and optimization of the entire process in B2B procurement scenarios through the coordinated operation of four core modules: demand identification, supply matching, multi-mode matching and risk pricing, and execution and performance feedback.

[0138] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for controlling enterprise procurement transactions based on multi-mode matching and risk pricing, characterized in that, Includes the following steps: S1. Demand Identification: Collects real-time inventory data, customer flow forecast data, historical order trend data, and product attribute data. The processor executes and identifies the procurement gap level, demand timeliness sensitivity, and procurement event type, and outputs standardized procurement demand instructions. S2. Supply Matching: Based on the procurement demand instruction, the system retrieves credit ratings, timeliness and performance capabilities, historical quotation data, geographical matching degree and supply capacity data from the supplier database, and obtains a set of candidate suppliers that meet the requirements through a screening algorithm. S3. Multi-mode matching and risk pricing: Construct a multi-dimensional risk factor weighted evaluation model to comprehensively score the candidate supplier set and selectable transaction modes, and automatically select the optimal transaction mode and determine the dynamic optimal price range based on the scoring results; the multi-dimensional risk factors include cost factors, risk factors, profit factors, time factors, loss factors, and behavioral factors; the selectable transaction modes include fixed-price procurement, group purchase procurement, competitive bidding or fixed-price auction, near-expiration price reduction / idle inventory matching, long-term contract price locking, and service capability procurement; S4. Execution and Fulfillment Feedback: Based on the optimal transaction model and price range, automatically complete order placement, payment, delivery tracking monitoring, warehousing verification, and correction of abnormal issues; collect full-process fulfillment data and feed it back to the system database to optimize the risk factor weighted evaluation model and transaction model selection algorithm.

2. The method according to claim 1, characterized in that, In step S1, the preset algorithm is a demand forecasting algorithm based on a time-series forecasting model, which specifically includes: performing time-series analysis on historical order trend data and customer flow forecast data, and calculating the procurement gap by combining real-time inventory data with product shelf-life thresholds and safety stock thresholds; classifying the gap into emergency gaps, regular gaps, and reserve gaps based on the ratio of the procurement gap to the demand time window; and determining the time sensitivity level based on the difference between the product turnover cycle and the demand time window.

3. The method according to claim 1, characterized in that, In step S3, the construction process of the multi-dimensional risk factor weighted evaluation model includes: S31. Determine the specific evaluation indicators and quantitative standards for each dimension of risk factors: Cost factor evaluation indicators include purchase unit price, logistics and transportation costs, and order processing costs. The quantitative standard is the ratio of each cost indicator to the industry average level. Risk factor evaluation indicators include the probability of supply disruption, historical performance deviation rate, and cost of breach of contract compensation. The quantitative standard is the statistical probability and cost calculation value based on the supplier's historical data. Profitability factor evaluation indicators include the gross profit contribution of purchased goods and the capital occupation period. The quantitative standards are the ratio of gross profit contribution to purchase amount and the ratio of capital occupation period to industry average capital turnover period. The time factor evaluation index is the urgency of demand, and the quantitative standard is the percentage difference between the demand time window and the supplier's average fulfillment cycle. The evaluation indicators for loss factors include the near-expiry risk coefficient and the inventory pressure coefficient. The quantitative standards are the ratio of the remaining shelf life of the product to the total shelf life and the ratio of the current inventory turnover rate to the safe turnover rate. The behavioral factor evaluation indicators include merchant credit rating and cooperation stability, and the quantitative standards are third-party credit rating score and the ratio of continuous cooperation period to total cooperation period; S32. The weights of risk factors in each dimension are determined by the analytic hierarchy process (AHP), and the weights are corrected by the entropy weight method to obtain the final weighting coefficients. S33. Calculate the comprehensive score based on the quantitative values ​​and final weighting coefficients of risk factors in each dimension; set the threshold for selecting the transaction mode, and automatically match the optimal transaction mode according to the comprehensive score and the characteristics of the procurement needs.

4. The method according to claim 1, characterized in that, In step S3, the logic for selecting the transaction mode includes: When the coefficient of variation of procurement demand is less than the preset threshold, it is determined that the demand is stable, and the fixed price procurement + long-term contract price lock-in mode is given priority. When the absolute value of the difference between the total supply capacity of candidate suppliers and the procurement gap exceeds a preset ratio, it is determined to be a supply-demand mismatch, and competitive bidding or fixed-price auction mode is given priority. When the risk coefficient of a product nearing its expiration date exceeds a preset threshold, it is determined to be a high risk of inventory expiration, and the automatic matching mode of near-expiration price reduction / idle inventory is selected first. When the procurement gap is less than the preset procurement quantity threshold, it is determined that the procurement quantity is small, and the group purchase mode is given priority. When the procurement event type is service procurement, the service capacity procurement model should be selected first, matching the idle status and cost level of human labor hours or robot services.

5. The method according to claim 1, characterized in that, In step S4, the abnormal problem correction includes: when a delivery delay occurs, automatically triggering the alternative supplier replenishment process; when an inbound inspection fails, automatically initiating a return and exchange application and recording supplier performance abnormal data; the performance data includes order response time, delivery on-time rate, product qualification rate, default rate and cost deviation value.

6. A corporate procurement transaction control system based on multi-mode matching and risk pricing, characterized in that, It includes a demand identification module, a supply matching module, a multi-mode matching and risk pricing module, and an execution and performance feedback module; The demand identification module is used to collect real-time inventory data, customer flow forecast data, historical order trend data and product attribute data, and automatically identify the procurement gap level, demand timeliness sensitivity and procurement event type through a preset algorithm, and output standardized procurement demand instructions. The supply matching module is communicatively connected to the demand identification module and is used to retrieve supplier database data based on procurement demand instructions, and obtain a set of candidate suppliers that meet the requirements through a screening algorithm; the supplier database data includes supplier credit rating, timeliness and performance capability, historical quotation data, geographical matching degree and supply capacity data; The multi-mode matching and risk pricing module is connected to the demand identification module and the supply matching module, respectively. It is the core innovative module of the system and is used to construct a multi-dimensional risk factor weighted evaluation model to comprehensively score the candidate supplier set and the selectable transaction modes. Based on the scoring results, it automatically selects the optimal transaction mode and determines the dynamic optimal price range. The price range is calculated based on historical transaction data, market supply and demand parameters and risk factor quantification values ​​through regression models or machine learning models. The execution and fulfillment feedback module is connected to the multi-mode matching and risk pricing module. It is used to automatically complete order placement, payment, delivery trajectory monitoring, warehousing verification and abnormal problem correction according to the optimal transaction mode and price range. It is also used to collect full-process fulfillment data and feed it back to the system database and the multi-mode matching and risk pricing module to realize model iterative optimization.

7. The system according to claim 6, characterized in that, The demand identification module includes a data acquisition unit, a data preprocessing unit, and a gap and timeliness identification unit. The data acquisition unit adopts a multi-source data interface and can access data from enterprise ERP systems, inventory management systems, customer flow statistics systems, and order management systems. The data preprocessing unit is used to denoise, standardize, and impute missing values ​​in the acquired data. The gap and timeliness identification unit has a built-in LSTM-based demand forecasting and gap identification algorithm to determine the procurement gap level and timeliness sensitivity level.

8. The system according to claim 6, characterized in that, The multi-mode matching and risk pricing module includes a factor construction unit, a weight calculation unit, a comprehensive scoring unit, and a mode matching unit. The factor construction unit defines evaluation indicators and quantitative standards for six dimensions: cost, risk, profit, time, loss, and behavior. The weight calculation unit uses a combined weighting method combining the analytic hierarchy process (AHP) and entropy weighting, ensuring that the sum of the weights of cost factors, risk factors, and time factors is not less than a preset ratio, and determines the final weighting coefficients for each dimension factor. The comprehensive scoring unit calculates a comprehensive score based on the quantitative values ​​and weighting coefficients. The mode matching unit incorporates trading mode selection logic and thresholds to determine the optimal trading mode and the dynamic optimal price range.

Images