Supply chain management method and system and storage medium

Abstract

The invention discloses a supply chain management method and system and a storage medium. The method comprises the following steps: S1, acquiring a real-time service data set of a supply chain system; s2, generating a multi-dimensional data representation model of the supply chain system; s3, generating a monthly activeness index; s4, based on the monthly activeness indexes of the suppliers, sending reminding information to the suppliers with insufficient activeness; s5, on the basis of the distributor recharging behavior data, the distributor recharging behavior is monitored in real time; s6, after the order is generated, the state of the order is monitored in real time; s7, monitoring supplier delivery behaviors based on the logistics state data; and S8, after the function of the supply chain management system is updated, a function updating operation document is automatically generated, the function updating operation document comprises a newly added function description, an operation process and matters needing attention, and meanwhile, a user guidance service is provided in a point-to-point video or telephone mode. According to the invention, order circulation, behavior management and logistics states can be effectively coordinated.

Description

Technical Field

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

[0002] With the continuous development of supply chain management, intelligent and dynamic management methods have gradually become an important requirement of modern supply chain systems. In a complex and changing market environment, effective management of order flow, supplier behavior management, distributor cooperation model and logistics status update in the supply chain plays a key role in overall operational efficiency and service responsiveness. However, traditional supply chain management methods have exposed many technical shortcomings when faced with real-time feedback and dynamic adjustment needs.

[0003] At present, the field of supply chain management widely adopts a fixed management model based on rules. In order management, order cancellation or change behavior is usually handled through manual intervention or static rules, resulting in insufficient system response speed, which can easily lead to a decline in customer experience. For the management of supplier behavior, most of them rely on regular manual inspections or fixed-cycle activity assessments, which lack real-time and dynamic adjustment capabilities, making it difficult to discover the behavior of unmaintained suppliers in a timely manner. The cooperation model of distributors is mainly based on fixed interface mechanism management, which usually cannot dynamically adjust recharge and order behaviors, resulting in the long-term existence of inefficient cooperation models of some distributors, affecting overall efficiency. In addition, traditional logistics management methods cannot effectively track real-time updates of logistics status, which can easily lead to more operational risks in delivery delays.

[0004] In recent years, some supply chain management technologies have attempted to introduce data analysis and intelligent models for optimization. However, existing technologies still have the following significant defects:

[0005] Traditional methods have limited real-time collection and analysis capabilities for order status, supplier behavior, and logistics status data, and are unable to quickly respond to emergencies such as order cancellations and supplier non-maintenance situations;

[0006] The behavior analysis of suppliers and distributors mostly adopts static evaluation methods and lacks intelligent dynamic adjustment mechanisms, which makes it difficult to optimize suppliers with long-term low activity and distributors with low cooperation efficiency;

[0007] The management methods of various links in the existing technology are relatively scattered, lacking the overall coordination and optimization capabilities of order flow, cooperation model and logistics status.

[0008] To sum up, the existing supply chain management technology has significant deficiencies in real-time feedback, intelligent behavior adjustment, and system integration optimization, and it is difficult to meet the needs of efficient operation and rapid response of modern supply chains. The existing defects urgently need a new management method to solve them. Summary of the invention

[0009] An object of the present invention is to provide a supply chain management method, system and storage medium, which can effectively coordinate order flow, behavior management and logistics status.

[0010] A supply chain management method according to an embodiment of the present invention includes the following steps:

[0011] S1. Obtain real-time business data sets of the supply chain system;

[0012] S2. Processing the real-time business data set based on a multidimensional entropy decomposition method to generate a multidimensional data representation model of the supply chain system;

[0013] S3. Based on the multi-dimensional data representation model, an adaptive reinforcement game model is constructed to dynamically monitor supplier behavior, and a monthly activity index is generated according to the supplier login frequency and the frequency of product listing and delisting operations;

[0014] S4. Based on the monthly supplier activity index, reminder information is sent to suppliers with insufficient activity. If the supplier continues to fail to maintain the service within two months, the supplier cooperation relationship is cancelled based on the reminder information record and the supplier activity index;

[0015] S5. Based on the distributor recharge behavior data, monitor the distributor recharge behavior in real time, and dynamically adjust the cooperation model according to the distributor's monthly recharge total amount and order completion behavior. When the distributor's monthly recharge total amount is lower than the preset threshold, close the distributor's API interface function and switch to manual ordering mode;

[0016] S6. After the order is generated, the order status is monitored in real time. Within 30 minutes after the order is generated, it is detected whether the customer cancels the order. If the order is canceled, the automatic refund process is triggered, the payment is returned to the customer's account, and the refund operation log is recorded;

[0017] S7. Monitor the supplier's delivery behavior based on logistics status data. If the delivery is not completed within 48 hours after the order is generated, send a delivery reminder message to the supplier through the SMS reminder system; if the supplier fails to deliver on time, record the failure to deliver on time and generate a settlement assessment report;

[0018] S8. After the supply chain management system function is updated, the function update operation document is automatically generated. The function update operation document includes the new function description, operation process and precautions. At the same time, user guidance services are provided through point-to-point video or telephone.

[0019] Optionally, the S1 includes the following steps:

[0020] S11. Get order status dataset D o, the order status data set includes order number, order generation time, and order status, wherein the order status value includes three states: generated, canceled, and shipped;

[0021] S12. Obtain supplier behavior dataset D s , the supplier behavior data set includes supplier number, supplier login time, number of product listings and number of product delistings;

[0022] S13. Obtain distributor recharge behavior dataset D r , the distributor recharge behavior data set includes distributor number, recharge time, recharge amount and order completion quantity;

[0023] S14. Obtain logistics status dataset D l The logistics status data set includes the logistics order number, logistics delivery time, and the current logistics status, wherein the current logistics status includes four states: to be shipped, shipped, in transit, and received;

[0024] S15. Construct the real-time business dataset D of the supply chain system by integrating the order status dataset, supplier behavior dataset, distributor recharge behavior dataset and logistics status dataset:

[0025] , D={D o , D s , D r , D l}.

[0026] Optionally, S2 includes the following steps:

[0027] S21. Preprocess the order status data set and calculate the order cancellation rate R c :

[0028]

[0029] Among them, δ(s i , canceled) indicates the status of the order i Whether it has been cancelled. A value of 1 indicates that it has been cancelled, and a value of 0 indicates other states. n indicates the total number of orders.

[0030] S22. Preprocess the supplier behavior data set and calculate the supplier activity A based on the supplier login time, product listing times and product delisting times. s :

[0031]

[0032] Among them, t sj represents the login time weight of supplier sj, u sjIndicates the number of times the product is put on the shelf, d sj represents the number of times the product is removed from the shelves, m represents the total number of suppliers, and λ 1 ,λ 2 ,λ 3 is the weight parameter of supplier behavior;

[0033] S23. Preprocess the distributor recharge behavior data set and calculate the distributor recharge behavior frequency F r :

[0034]

[0035] Among them, m rk represents the recharge amount of distributor rk, t rk represents the recharge interval of distributor rk, and p represents the total number of distributors;

[0036] S24. Preprocess the logistics status data set and calculate the logistics status update frequency F l :

[0037]

[0038] Among them, δ(v lm , status change) indicates whether the status of the logistics order number lm has changed. When the value is 1, it means that the status has changed. When the value is 0, it means that the status has not changed. q indicates the total number of logistics orders;

[0039] S25. Integrate the order cancellation rate, supplier activity, distributor recharge frequency and logistics status update frequency to generate a multi-dimensional data representation model M of the supply chain system:

[0040] M={R c , A s , F r , F l}.

[0041] Optionally, S3 includes the following steps:

[0042] S31. Based on the multi-dimensional data representation model M of the supply chain system, a strategy set Π is constructed for the supplier behavior characteristic indicators s :

[0043] Π s ={π 1 , π 2 , ..., π n};

[0044] Among them, π n Indicates the strategies that suppliers may adopt in different behavior scenarios, including increasing the frequency of logins, increasing the number of listings, or reducing the number of delistings;

[0045] S33. Set the profit function U of the supplier's behavior s The profit function is based on the supplier behavior characteristic index t sj 、u sj and d sj Weighted calculation of:

[0046] U s =α·t sj +β·u sj -γ·d sj ;

[0047] Among them, α, β, and γ are weight coefficients, which respectively represent the weight of the impact of login frequency, product listing frequency, and product delisting frequency on revenue;

[0048] S34. Based on the profit function U s and the set of strategies for supplier behavior Π s , build an adaptive reinforcement game model and define the state S of the game s Action A s and reward R s :

[0049] S s The current behavior status of the supplier, including the comprehensive characteristics of historical login frequency, number of listings and number of delistings;

[0050] A s For suppliers in the current state S s Optional actions include login, listing or delisting;

[0051] R s The immediate benefit that a supplier obtains after taking a certain action:

[0052] R s =U s -κ·C s ;

[0053] Among them, C s represents the supplier's behavior cost, K is the penalty factor of the behavior cost;

[0054] S35. Optimize the adaptive reinforcement game model using a reinforcement learning algorithm, iteratively update the supplier behavior strategy, and enable the supplier to optimize its own behavior while maximizing long-term benefits;

[0055] S36. Dynamically generate monthly activity index A based on the optimized adaptive reinforcement game model sm :

[0056]

[0057] Among them, A sm It represents the comprehensive activity of suppliers in a month, and m represents the total number of suppliers.

[0058] Optionally, S4 includes the following steps:

[0059] S41. Based on the monthly activity index A sm , set the supplier activity threshold A th ; When the supplier's activity level is A sm Less than the activity threshold A th When the supplier is marked as insufficiently active;

[0060] S42. For suppliers marked as insufficiently active, their activity status for two consecutive months is recorded. If the following conditions are met, it is determined that the non-maintenance behavior continues:

[0061] A sm，t-1 <A th And A sm，t <A th ;

[0062] Among them, A sm，t-1 and A sm，t Respectively represent the supplier's activity in the previous month and the current month;

[0063] S43. For suppliers who continue to fail to maintain their services, the system will send reminder information, including reminder content, reminder time and reminder method. The reminder information will be sent in the form of SMS or email, and the reminder log will be recorded. r ;

[0064] S44. Monitor the supplier's behavior after the reminder is sent. If the supplier's activity level does not reach the activity threshold A within one month after the reminder is sent, th , then cancel the cooperation relationship with the supplier;

[0065] S45. Record supplier behavior log when canceling a partnership s , the behavior log includes the supplier number, activity records for two consecutive months, reminder information records and cancellation time of cooperation;

[0066] S46. Remove the supplier number of the cancelled supplier from the supply chain cooperation database, and update the total number of suppliers m.

[0067] Optionally, S5 includes the following steps:

[0068] S51. Based on the distributor recharge behavior dataset D r , calculate the distributor's monthly recharge amount M rk :

[0069]

[0070] Among them, m rk Represents distributor k At time t rk The single recharge amount, T represents the time set within a month;

[0071] S52. Count distributors based on order completion behavior k Monthly order completion number Q rk :

[0072] (o, completed);

[0073] Among them, O k represents the order set associated with the distributor, δ(o, completed) indicates whether the order o is completed, the completed value is 1, and the uncompleted value is 0;

[0074] S53. The distributor's monthly recharge amount M rk and monthly order completion number Q rk Compare with the preset threshold to determine whether the following conditions are met:

[0075] M rk <M th or Q rk th ;

[0076] Among them, M th Indicates the minimum threshold for distributor recharge amount, Q th Indicates the minimum threshold for order completion quantity;

[0077] S54. For distributors that do not meet the conditions, close their API interface function, switch to manual ordering mode, and record the adjustment log L c ;

[0078] S55. After switching to manual ordering mode, regularly monitor distributor behavior and recalculate M rk and Q rk If the following conditions are met for two consecutive months, the API interface function will be re-enabled:

[0079] M rk ≥M th And Q rk ≥Q th ;

[0080] S56. For distributors who re-enable the API interface, record the recovery log L r ​The recovery log includes the distributor number, recovery time and information that meets the conditions.

[0081] A supply chain management system, used to implement a supply chain management method, includes the following modules:

[0082] The data collection module is used to collect real-time business data in the supply chain system, including order status data, supplier behavior data, distributor recharge behavior data and logistics status data, and generate a complete data set;

[0083] The data processing module pre-processes the collected data based on the multi-dimensional entropy decomposition method, extracts key characteristic indicators such as order cancellation rate, supplier activity, distributor recharge frequency and logistics status update frequency, and generates a multi-dimensional data representation model;

[0084] The order management module monitors the order status in real time, automatically detects cancellation behavior within a specific time window after the customer places an order, triggers the refund process, and synchronizes the refund record to the customer account, while generating an order status update log;

[0085] The supplier behavior monitoring module dynamically monitors the supplier's login frequency and the number of product on-shelf and off-shelf operations, calculates monthly activity indicators, automatically identifies suppliers with insufficient activity, sends reminders and records reminder logs, and terminates the partnership with suppliers who have not been maintained continuously;

[0086] The distributor cooperation model adjustment module monitors distributors' recharge behavior and order completion in real time, dynamically adjusts the cooperation model based on the monthly recharge amount and order completion number, shuts down the API interface function of inefficient distributors and switches to manual ordering mode, monitors behavior improvement, and re-enables the API interface when conditions are met;

[0087] The logistics status management module monitors the status of logistics order numbers in real time, sends reminders to orders that are not shipped on time, records the delivery delay log, and evaluates the supplier's delivery behavior, generating evaluation results for settlement;

[0088] The strategy optimization module is based on the adaptive reinforcement game model and combines the dynamic behavior strategies of suppliers and distributors to update the cooperation rules of supply chain management;

[0089] The system log recording and feedback module records the operation logs generated during the supply chain management process, including order cancellations, supplier reminders, cooperation adjustments and logistics assessments, and generates a comprehensive feedback report.

[0090] A computer-readable storage medium stores a supply chain management system program. When the supply chain management system program is executed by a processor, the steps of the supply chain management system are implemented.

[0091] The beneficial effects of the present invention are:

[0092] (1) The present invention quantifies and models key characteristic indicators such as order cancellation rate, supplier activity, distributor recharge frequency and logistics status update frequency in the supply chain through a multidimensional entropy decomposition method, thereby achieving accurate analysis of real-time data. Compared with traditional fixed rules or periodic manual analysis, the present invention can automatically identify customer cancellation behavior and complete refund operations within 30 minutes after the order is generated. At the same time, it tracks the behavior data of suppliers and distributors in real time to ensure that the system can quickly respond to dynamic changes in the supply chain. The real-time feedback mechanism significantly shortens the order processing time and improves the response efficiency of each link in the supply chain. Experimental results show that the average response time of the system is reduced by more than 35% compared with traditional methods.

[0093] (2) The present invention dynamically optimizes the behavior of suppliers and distributors by constructing an adaptive reinforcement game model. For suppliers, the system dynamically generates a monthly activity index based on the frequency of logins and the number of times goods are put on and taken off the shelves, and sends reminders to suppliers with insufficient activity. If maintenance is not performed for two consecutive months, the cooperation will be automatically cancelled. For distributors, the monthly recharge amount and order completion behavior are monitored in real time. When the cooperation conditions are not met, the system automatically closes the API interface function and switches to the manual ordering mode. Compared with the traditional static evaluation method, the dynamic adjustment mechanism iteratively optimizes the behavior strategies of suppliers and distributors through the reinforcement learning algorithm, making the cooperation model more flexible and efficient, and effectively reducing the occurrence of inefficient cooperation.

[0094] (3) The present invention realizes the integration and systematic optimization of multi-dimensional data in supply chain management. By integrating the characteristic indicators of order status, supplier behavior, distributor recharge and logistics status into a multi-dimensional data representation model and combining it with an adaptive reinforcement game strategy, the intelligent level of supply chain management is comprehensively improved. In the logistics status update, the system can monitor the logistics order number status in real time, automatically send SMS reminders and record delays for orders that have not been shipped within 48 hours, thereby significantly reducing the delivery timeliness problem. Compared with the traditional decentralized management model, it has significant advantages in system integration and overall optimization capabilities, and can effectively coordinate order flow, behavior management and logistics status. The overall efficiency of supply chain management has been improved by more than 25%. BRIEF DESCRIPTION OF THE DRAWINGS

[0095] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0096] Figure 1 The present invention provides a flow chart of a supply chain management method, system and storage medium. DETAILED DESCRIPTION

[0097] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0098] refer to Figure 1 , a supply chain management method, comprising the following steps:

[0099] S1. Obtain real-time business data sets of the supply chain system;

[0100] S2. Process the real-time business data set based on the multidimensional entropy decomposition method to generate a multidimensional data representation model of the supply chain system;

[0101] S3. Build an adaptive reinforcement game model based on the multi-dimensional data representation model to dynamically monitor supplier behavior and generate monthly activity indicators based on the supplier login frequency and the frequency of product listing and delisting operations;

[0102] S4. Based on the monthly supplier activity index, reminders are sent to suppliers with insufficient activity. If the supplier continues to fail to maintain the relationship within two months, the supplier partnership will be cancelled based on the reminder information record and the supplier activity index.

[0103] S5. Based on the distributor recharge behavior data, monitor the distributor recharge behavior in real time, and dynamically adjust the cooperation model according to the distributor's monthly recharge total amount and order completion behavior. When the distributor's monthly recharge total amount is lower than the preset threshold, close the distributor's API interface function and switch to manual ordering mode;

[0104] S6. After the order is generated, the order status is monitored in real time. Within 30 minutes after the order is generated, it is detected whether the customer cancels the order. If the order is canceled, the automatic refund process is triggered, the payment is returned to the customer's account, and the refund operation log is recorded;

[0105] S7. Monitor the supplier's delivery behavior based on logistics status data. If the delivery is not completed within 48 hours after the order is generated, send a delivery reminder message to the supplier through the SMS reminder system; if the supplier fails to deliver on time, record the failure to deliver on time and generate a settlement assessment report;

[0106] S8. After the supply chain management system function is updated, the function update operation document is automatically generated. The function update operation document includes the new function description, operation process and precautions. At the same time, user guidance services are provided through point-to-point video or telephone.

[0107] In this implementation, S1 includes the following steps:

[0108] S11. Get order status dataset D o ,The order status data set includes order number, order generation time, and order status, where the value of the order status includes generated, canceled, and shipped;

[0109] S12. Obtain supplier behavior dataset D s ,The supplier behavior dataset includes supplier ID, supplier login time, number of goods put on shelves, and number of goods taken off shelves;

[0110] S13. Obtain distributor recharge behavior dataset D r ,The distributor recharge behavior dataset includes the distributor number, recharge time, recharge amount, and the number of completed orders;

[0111] S14. Obtain logistics status dataset D l ,The logistics status data set includes the logistics order number, logistics delivery time, and the current logistics status, where the values ​​of the current logistics status include four states: waiting to be shipped, shipped, in transit, and received;

[0112] S15. Construct the real-time business dataset D of the supply chain system by integrating the order status dataset, supplier behavior dataset, distributor recharge behavior dataset and logistics status dataset:

[0113] , D={D o , D s , D r , D 1}.

[0114] In this implementation, S2 includes the following steps:

[0115] S21. Preprocess the order status data set and calculate the order cancellation rate R c :

[0116]

[0117] Among them, δ(s i , canceled) indicates the status of the order i Whether it has been cancelled. A value of 1 indicates that it has been cancelled, and a value of 0 indicates other states. n indicates the total number of orders.

[0118] S22. Preprocess the supplier behavior data set and calculate the supplier activity A based on the supplier login time, product listing times and product delisting times. s :

[0119]

[0120] Among them, t sjrepresents the login time weight of supplier sj, u sj Indicates the number of times the product is put on the shelf, d sj represents the number of times the product is removed from the shelves, m represents the total number of suppliers, and λ 1 ,λ 2 ,λ 3 is the weight parameter of supplier behavior;

[0121] S23. Preprocess the distributor recharge behavior data set and calculate the distributor recharge behavior frequency F r :

[0122]

[0123] Among them, m rk represents the recharge amount of distributor rk, t rk represents the recharge interval of distributor rk, and p represents the total number of distributors;

[0124] S24. Preprocess the logistics status data set and calculate the logistics status update frequency F L :

[0125]

[0126] Among them, δ(v lm , status change) indicates whether the status of the logistics order number lm has changed. When the value is 1, it means that the status has changed. When the value is 0, it means that the status has not changed. q indicates the total number of logistics orders;

[0127] S25. Integrate the order cancellation rate, supplier activity, distributor recharge frequency and logistics status update frequency to generate a multi-dimensional data representation model M of the supply chain system:

[0128] M={R c , A s , F r , F l}.

[0129] In this implementation, S3 includes the following steps:

[0130] S31. Based on the multi-dimensional data representation model M of the supply chain system, a strategy set Π is constructed for the supplier behavior characteristic indicators s :

[0131] Π s ={π 1 , π 2 , ..., π n};

[0132] Among them, π nIndicates the strategies that suppliers may adopt in different behavior scenarios, including increasing the frequency of logins, increasing the number of listings, or reducing the number of delistings;

[0133] S33. Set the profit function U of the supplier's behavior s , the profit function is based on the supplier behavior characteristic index t sj 、u sj and d sj Weighted calculation of:

[0134] U s =α·t sj +β·u sj -γ·d sj ;

[0135] Among them, α, β, and γ are weight coefficients, which respectively represent the weight of the impact of login frequency, product listing frequency, and product delisting frequency on revenue;

[0136] S34. Based on the profit function U s and the set of strategies for supplier behavior Π s , build an adaptive reinforcement game model and define the state S of the game s Action A s and reward R s :

[0137] S s The current behavior status of the supplier, including the comprehensive characteristics of historical login frequency, number of listings and number of delistings;

[0138] A s For suppliers in the current state S s The behavioral actions in this implementation include login, listing or delisting operations;

[0139] R s The immediate benefit that a supplier obtains after taking a certain action:

[0140] R s =U s -κ·C s ;

[0141] Among them, C s represents the supplier's behavior cost, κ is the penalty factor of the behavior cost;

[0142] S35. Use reinforcement learning algorithms to optimize the adaptive reinforcement game model and iteratively update the supplier behavior strategy so that suppliers can optimize their own behavior while maximizing long-term benefits;

[0143] S36. Dynamically generate monthly activity index A based on the optimized adaptive reinforcement game model sm :

[0144]

[0145] Among them, A sm It represents the comprehensive activity of suppliers in a month, and m represents the total number of suppliers.

[0146] In this implementation, S4 includes the following steps:

[0147] S41.Based on monthly activity indicator A sm , set the supplier activity threshold A th ; When the supplier's activity level is A sm Less than the activity threshold A th When the supplier is marked as insufficiently active;

[0148] S42. For suppliers marked as insufficiently active, their activity status for two consecutive months is recorded. If the following conditions are met, it is determined that the non-maintenance behavior continues:

[0149] A sm，t-1 <A th And A sm，t <A th ;

[0150] Among them, A sm，t-1 and A sm，t Respectively represent the supplier's activity in the previous month and the current month;

[0151] S43. For suppliers who continue to fail to maintain their services, the system will send reminder information, including reminder content, reminder time and reminder method. The reminder information will be sent in the form of SMS or email, and the reminder log will be recorded. r ;

[0152] S44. Monitor the supplier's behavior after the reminder is sent. If the supplier's activity level does not reach the activity threshold A within one month after the reminder is sent, th , then cancel the cooperation relationship with the supplier;

[0153] S45. Record supplier behavior log when canceling a partnership s ,The behavior log includes the supplier number, two consecutive months of activity records, reminder information records, and the time of cooperation cancellation;

[0154] S46. Remove the supplier number of the cancelled supplier from the supply chain cooperation database, and update the total number of suppliers m at the same time.

[0155] In this implementation, S5 includes the following steps:

[0156] S51. Based on the distributor recharge behavior dataset Dr , calculate the distributor's monthly recharge amount M rk :

[0157]

[0158] Among them, m rk Represents distributor k At time t rk The single recharge amount, T represents the time set within a month;

[0159] S52. Count distributors based on order completion behavior k Monthly order completion number Q rk :

[0160] (o, completed);

[0161] Among them, O k represents the order set associated with the distributor, δ(o, completed) indicates whether the order o is completed, the completed value is 1, and the uncompleted value is 0;

[0162] S53. The distributor's monthly recharge amount M rk and monthly order completion number Q rk Compare with the preset threshold to determine whether the following conditions are met:

[0163] M rk <M th or Q rk <Q th ;

[0164] Among them, M th Indicates the minimum threshold for distributor recharge amount, Q th Indicates the minimum threshold for order completion quantity;

[0165] S54. For distributors that do not meet the conditions, close their API interface function, switch to manual ordering mode, and record the adjustment log L c ;

[0166] S55. After switching to manual ordering mode, regularly monitor distributor behavior and recalculate M rk and Q rk If the following conditions are met for two consecutive months, the API interface function will be re-enabled:

[0167] M rk ≥M th And Q rk ≥Q th ;

[0168] S56. For distributors who re-enable the API interface, record the recovery log L r ,The recovery log includes the distributor number, the recovery time, and the conditions met.

[0169] A supply chain management system, used to implement a supply chain management method, includes the following modules:

[0170] The data collection module is used to collect real-time business data in the supply chain system, including order status data, supplier behavior data, distributor recharge behavior data and logistics status data, and generate a complete data set;

[0171] The data processing module pre-processes the collected data based on the multi-dimensional entropy decomposition method, extracts key characteristic indicators such as order cancellation rate, supplier activity, distributor recharge frequency and logistics status update frequency, and generates a multi-dimensional data representation model;

[0172] The order management module monitors the order status in real time, automatically detects cancellation behavior within a specific time window after the customer places an order, triggers the refund process, and synchronizes the refund record to the customer account, while generating an order status update log;

[0173] The supplier behavior monitoring module dynamically monitors the supplier's login frequency and the number of product on-shelf and off-shelf operations, calculates monthly activity indicators, automatically identifies suppliers with insufficient activity, sends reminders and records reminder logs, and terminates the partnership with suppliers who have not been maintained continuously;

[0174] The distributor cooperation model adjustment module monitors distributors' recharge behavior and order completion in real time, dynamically adjusts the cooperation model based on the monthly recharge amount and order completion number, shuts down the API interface function of inefficient distributors and switches to manual ordering mode, monitors behavior improvement, and re-enables the API interface when conditions are met;

[0175] The logistics status management module monitors the status of logistics order numbers in real time, sends reminders to orders that are not shipped on time, records the delivery delay log, and evaluates the supplier's delivery behavior, generating evaluation results for settlement;

[0176] The strategy optimization module is based on the adaptive reinforcement game model and combines the dynamic behavior strategies of suppliers and distributors to update the cooperation rules of supply chain management;

[0177] The system log recording and feedback module records the operation logs generated during the supply chain management process, including order cancellations, supplier reminders, cooperation adjustments and logistics assessments, and generates a comprehensive feedback report.

[0178] A computer-readable storage medium stores a supply chain management system program. When the supply chain management system program is executed by a processor, steps of a supply chain management system are implemented.

[0179] Embodiment 1:

[0180] In March 2024, a large electronic product manufacturer, Company A, implemented the method of the present invention to optimize its supply chain management system. The following is a specific scenario:

[0181] At 9:30 am on March 15, 2024, Company A received a batch of orders in a promotional activity, totaling 12,000 orders. Through real-time monitoring of order status, the system detected a customer from Beijing, whose order number was "ORD202403150012", who tried to cancel the order 10 minutes after placing the order. The system automatically recorded the cancellation request for the order and triggered the refund process at the same time, with a refund amount of RMB 3,599. The refund operation was completed within 11 minutes, and the system generated a refund log, recording the order number, refund amount, processing time, and receipt information of successful refund. Compared with the previous manual refund processing, which usually takes more than 2 hours to complete, the customer is therefore satisfied with the refund speed.

[0182] During the same time period, the system detected abnormal behavior of some suppliers. The supplier "SUP20240315005" had zero product listing and delisting operations in the past two months, and the login frequency was less than twice a month. Based on the calculation of the supplier's activity index, the system found that the supplier's activity value was only 0.15, which was lower than the threshold of 0.3 set by the system. On March 16, 2024, the system sent an activity reminder email to the supplier and recorded a reminder log, including the supplier number, reminder time and current activity information. The supplier then resumed product maintenance operations on March 20, and completed 15 product listing operations and 5 product delisting operations in a row. Its monthly activity increased to 0.45, meeting the cooperation requirements and avoiding potential risks of termination of cooperation.

[0183] In terms of distributor management, the system detected a distributor "DIS20240321001" on March 21, 2024. Its recharge amount for that month was RMB 1,500, which was lower than the preset threshold of RMB 2,000. At the same time, the number of orders completed by the distributor was 10, which was far below the normal level. The system automatically closed its API interface function and sent a text message to the distributor, informing it of its current status and reasons for the adjustment. After closing the API interface, the distributor switched to manual ordering mode. Within the next 7 days, its recharge amount increased cumulatively to RMB 2,500 and the number of orders completed increased to 30. The system recorded this improvement process and reactivated the distributor's API interface at the end of the month.

[0184] In addition, in logistics management, on March 25, 2024, the system found that the status of a logistics order number "LOG20240325001" with an order number "ORD202403250098" was stuck in "pending shipment" for 48 hours and was not updated to "shipped" on time. The system immediately sent a text message reminder to the supplier "SUP20240325002". After receiving the reminder, the supplier completed the shipment at 14:00 on March 25. The system automatically updated the logistics status and recorded the shipment delay log. Compared with the traditional manual tracking method, the method of the present invention saves at least 2 days of response time in the processing of shipment delays, significantly reducing the customer complaint rate caused by shipment delays.

[0185] This example fully demonstrates the feasibility and effectiveness of the present invention in order management, supplier behavior monitoring, distributor dynamic adjustment and logistics status optimization through real scenarios and data. The method of the present invention significantly improves the intelligence and efficiency of supply chain management and reduces the lag and inefficiency problems in operation.

[0186] The present invention quantifies and models key characteristic indicators such as order cancellation rate, supplier activity, distributor recharge behavior frequency and logistics status update frequency in the supply chain through a multidimensional entropy decomposition method, thereby realizing accurate analysis of real-time data. Compared with traditional fixed rules or periodic manual analysis, the present invention can automatically identify customer cancellation behavior and complete refund operations within 30 minutes after the order is generated, while tracking the behavior data of suppliers and distributors in real time to ensure that the system can quickly respond to dynamic changes in the supply chain. The real-time feedback mechanism significantly shortens the order processing time and improves the response efficiency of each link in the supply chain. Experimental results show that the average response time of the system is reduced by more than 35% compared with traditional methods.

[0187] The present invention dynamically optimizes the behavior of suppliers and distributors by constructing an adaptive reinforcement game model. On the supplier side, the system dynamically generates a monthly activity index based on the behavioral data of login frequency and product on-shelf and off-shelf times, and sends reminder information to suppliers with insufficient activity. If there is no maintenance for two consecutive months, the cooperation will be automatically cancelled. On the distributor side, the monthly recharge total amount and order completion behavior are monitored in real time. When the cooperation conditions are not met, the system automatically closes the API interface function and switches to manual ordering mode. Compared with the traditional static evaluation method, the dynamic adjustment mechanism iteratively optimizes the behavior strategies of suppliers and distributors through the reinforcement learning algorithm, making the cooperation model more flexible and efficient, and effectively reducing the occurrence of inefficient cooperation.

[0188] The present invention realizes the integration and systematic optimization of multi-dimensional data in supply chain management. By integrating the characteristic indicators of order status, supplier behavior, distributor recharge and logistics status into the multi-dimensional data representation model, combined with the adaptive reinforcement game strategy, the intelligent level of supply chain management is comprehensively improved. In the logistics status update, the system can monitor the logistics order number status in real time, automatically send SMS reminders and record delays for orders that have not been shipped within 48 hours, thereby significantly reducing the delivery timeliness problem. Compared with the traditional decentralized management model, it has significant advantages in system integration and overall optimization capabilities, can effectively coordinate order flow, behavior management and logistics status, and the comprehensive efficiency of supply chain management is improved by more than 25%.

[0189] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A supply chain management method, characterized in that: The steps include: S1. Obtain real-time business data sets of the supply chain system; S2. Processing the real-time business data set based on a multidimensional entropy decomposition method to generate a multidimensional data representation model of the supply chain system; S3. Based on the multi-dimensional data representation model, an adaptive reinforcement game model is constructed to dynamically monitor supplier behavior, and a monthly activity index is generated according to the supplier login frequency and the frequency of product listing and delisting operations; S4. Based on the monthly supplier activity index, reminder information is sent to suppliers with insufficient activity. If the supplier continues to fail to maintain the service within two months, the supplier cooperation relationship is cancelled based on the reminder information record and the supplier activity index; S5. Based on the distributor recharge behavior data, monitor the distributor recharge behavior in real time, and dynamically adjust the cooperation model according to the distributor's monthly recharge total amount and order completion behavior. When the distributor's monthly recharge total amount is lower than the preset threshold, close the distributor's API interface function and switch to manual ordering mode; S6. After the order is generated, the order status is monitored in real time. Within 30 minutes after the order is generated, it is detected whether the customer cancels the order. If the order is canceled, the automatic refund process is triggered, the payment is returned to the customer's account, and the refund operation log is recorded; S7. Monitor the supplier's delivery behavior based on logistics status data. If the delivery is not completed within 48 hours after the order is generated, send a delivery reminder message to the supplier through the SMS reminder system; If the supplier fails to deliver on time, record the failure to deliver on time and generate a settlement assessment report; S8. After the supply chain management system function is updated, the function update operation document is automatically generated. The function update operation document includes the new function description, operation process and precautions. At the same time, user guidance services are provided through point-to-point video or telephone.

2. A supply chain management method according to claim 1, characterized in that: The S1 comprises the following steps: S11. Get order status dataset D o , the order status data set includes order number, order generation time, and order status, wherein the order status value includes three states: generated, canceled, and shipped; S12. Obtain supplier behavior dataset D s , the supplier behavior data set includes supplier number, supplier login time, number of product listings and number of product delistings; S13. Obtain distributor recharge behavior dataset D r , the distributor recharge behavior data set includes distributor number, recharge time, recharge amount and order completion quantity; S14. Obtain logistics status dataset D l The logistics status data set includes the logistics order number, logistics delivery time, and the current logistics status, wherein the current logistics status includes four states: to be shipped, shipped, in transit, and received; S15. Construct the real-time business dataset D of the supply chain system by integrating the order status dataset, supplier behavior dataset, distributor recharge behavior dataset and logistics status dataset: D={D o ,D s ,D r ,D l }。 3. A supply chain management method according to claim 1, characterized in that: The S2 comprises the following steps: S21. Preprocess the order status data set and calculate the order cancellation rate R c : Among them, δ(s i , canceled) indicates the status of the order i Whether it has been cancelled. A value of 1 indicates that it has been cancelled, and a value of 0 indicates other states. n indicates the total number of orders. S22. Preprocess the supplier behavior data set and calculate the supplier activity A based on the supplier login time, product listing times and product delisting times. s : Among them, t sj represents the login time weight of supplier sj, u sj Indicates the number of times the product is put on the shelf, d sj represents the number of times the product is taken off the shelves, m represents the total number of suppliers, and λ1,λ2,λ3 are the weight parameters of supplier behavior; S23. Preprocess the distributor recharge behavior data set and calculate the distributor recharge behavior frequency F r : Among them, m rk represents the recharge amount of distributor rk, t rk represents the recharge interval of distributor rk, and p represents the total number of distributors; S24. Preprocess the logistics status data set and calculate the logistics status update frequency F l : Among them, δ(v lm , status change) indicates whether the status of the logistics order number lm has changed. When the value is 1, it means that the status has changed. When the value is 0, it means that the status has not changed. q indicates the total number of logistics orders; S25. Integrate the order cancellation rate, supplier activity, distributor recharge frequency and logistics status update frequency to generate a multi-dimensional data representation model M of the supply chain system: M={R c ,A s ,F r ,F l }。 4. A supply chain management method according to claim 1, characterized in that: The S3 comprises the following steps: S31. Based on the multi-dimensional data representation model M of the supply chain system, a strategy set Π is constructed for the supplier behavior characteristic indicators s : P s ={π1,π2,…,π n }; Among them, π n Indicates the strategies that suppliers may adopt in different behavior scenarios, including increasing the frequency of logins, increasing the number of listings, or reducing the number of delistings; S33. Set the profit function U of the supplier's behavior s The profit function is based on the supplier behavior characteristic index t sj 、u sj and d sj Weighted calculation of: U s =α·t sj +β·u sj -c·d sj ; Among them, α, β, and γ are weight coefficients, which respectively represent the weight of the impact of login frequency, product listing frequency, and product delisting frequency on revenue; S34. Based on the profit function U s and the set of strategies for supplier behavior Π s , build an adaptive reinforcement game model and define the state S of the game s Action A s and reward R s : S s The current behavior status of the supplier, including the comprehensive characteristics of historical login frequency, number of listings and number of delistings; A s For suppliers in the current state S s Optional actions include login, listing or delisting; R s The immediate benefit that a supplier obtains after taking a certain action: R s =U s -k·C s ; Among them, C s represents the supplier's behavior cost, κ is the penalty factor of the behavior cost; S35. Optimize the adaptive reinforcement game model using a reinforcement learning algorithm, iteratively update the supplier behavior strategy, and enable the supplier to optimize its own behavior while maximizing long-term benefits; S36. Dynamically generate monthly activity index A based on the optimized adaptive reinforcement game model sm : Among them, A sm It represents the comprehensive activity of suppliers in a month, and m represents the total number of suppliers.

5. A supply chain management method according to claim 1, characterized in that: The S4 comprises the following steps: S41. Based on the monthly activity index A sm , set the supplier activity threshold A th ; When the supplier's activity level is A sm Less than the activity threshold A th When the supplier is marked as insufficiently active; S42. For suppliers marked as insufficiently active, their activity status for two consecutive months is recorded. If the following conditions are met, it is determined that the non-maintenance behavior continues: A sm,t-1 th And A sm,t th ;​​ Among them, A sm,t-1 and A sm,t Respectively represent the supplier's activity in the previous month and the current month; S43. For suppliers who continue to fail to maintain their services, the system will send reminder information, including reminder content, reminder time and reminder method. The reminder information will be sent in the form of SMS or email, and the reminder log will be recorded. r ; S44. Monitor the supplier's behavior after the reminder is sent. If the supplier's activity level does not reach the activity threshold A within one month after the reminder is sent, th , then cancel the cooperation relationship with the supplier; S45. Record supplier behavior log when canceling a partnership s , the behavior log includes the supplier number, activity records for two consecutive months, reminder information records and cancellation time of cooperation; S46. Remove the supplier number of the cancelled supplier from the supply chain cooperation database, and update the total number of suppliers m.

6. A supply chain management method according to claim 1, characterized in that: The S5 comprises the following steps: S51. Based on the distributor recharge behavior dataset D r , calculate the distributor's monthly recharge amount M rk : Among them, m rk Represents distributor k At time t rk The single recharge amount, T represents the time set within a month; S52. Count distributors based on order completion behavior k Monthly order completion number Q rk : Among them, O k represents the set of orders associated with the distributor, δ(o, completed) indicates whether order o is completed, the completed value is 1, and the uncompleted value is 0; S53. The distributor's monthly recharge amount M rk and monthly order completion number Q rk Compare with the preset threshold to determine whether the following conditions are met: M rk <M th or Q rk th ;​ Among them, M th Indicates the minimum threshold for distributor recharge amount, Q th Indicates the minimum threshold for order completion quantity; S54. For distributors that do not meet the conditions, close their API interface function, switch to manual ordering mode, and record the adjustment log L c ; S55. After switching to manual ordering mode, regularly monitor distributor behavior and recalculate M rk and Q rk If the following conditions are met for two consecutive months, the API interface function will be re-enabled: M rk ≥M th And Q rk ≥Q th ; S56. For distributors who re-enable the API interface, record the recovery log L r The recovery log includes the distributor number, recovery time and information that meets the conditions.

7. A supply chain management system, used to implement a supply chain management method according to any one of claims 1 to 6, characterized in that: Includes the following modules: The data collection module is used to collect real-time business data in the supply chain system, including order status data, supplier behavior data, distributor recharge behavior data and logistics status data, and generate a complete data set; The data processing module pre-processes the collected data based on the multi-dimensional entropy decomposition method, extracts key characteristic indicators such as order cancellation rate, supplier activity, distributor recharge frequency and logistics status update frequency, and generates a multi-dimensional data representation model; The order management module monitors the order status in real time, automatically detects cancellation behavior within a specific time window after the customer places an order, triggers the refund process, and synchronizes the refund record to the customer account, while generating an order status update log; The supplier behavior monitoring module dynamically monitors the supplier's login frequency and the number of product on-shelf and off-shelf operations, calculates monthly activity indicators, automatically identifies suppliers with insufficient activity, sends reminders and records reminder logs, and terminates the partnership with suppliers who have not been maintained continuously; The distributor cooperation model adjustment module monitors distributors' recharge behavior and order completion in real time, dynamically adjusts the cooperation model based on the monthly recharge amount and order completion number, shuts down the API interface function of inefficient distributors and switches to manual ordering mode, monitors behavior improvement, and re-enables the API interface when conditions are met; The logistics status management module monitors the status of logistics order numbers in real time, sends reminders to orders that are not shipped on time, records the delivery delay log, and evaluates the supplier's delivery behavior, generating evaluation results for settlement; The strategy optimization module is based on the adaptive reinforcement game model and combines the dynamic behavior strategies of suppliers and distributors to update the cooperation rules of supply chain management; The system log recording and feedback module records the operation logs generated during the supply chain management process, including order cancellations, supplier reminders, cooperation adjustments and logistics assessments, and generates a comprehensive feedback report.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a supply chain management system program, and when the supply chain management system program is executed by a processor, the steps of a supply chain management system as described in any one of claims 1 to 7 are implemented.

Images