CN-120996870-B - Method and system for optimizing rolling processing of point mall order based on big data

CN120996870BCN 120996870 BCN120996870 BCN 120996870BCN-120996870-B

Abstract

The application provides a method and a system for optimizing rolling processing of a point mall order based on big data, wherein the method comprises the steps of adopting a dynamic priority adjustment mechanism for a conflict order list, adjusting order processing priority based on user behavior track data including historical exchange preference to obtain a resource allocation scheme, integrating the order sequence through a queue reordering mechanism, rearranging order based on the resource allocation scheme to obtain a final processing order, preferentially executing the high priority request while maintaining queue stability, extracting supplementary resources from preset inventory through a standby resource set scheduling mechanism if the final processing order has residual conflict to obtain a complete execution plan, wherein the plan covers all order processing paths, monitoring order processing progress based on the complete execution plan, feeding back real-time status through progress tracking and anomaly detection, and performing iterative adjustment to obtain a stably-running order flow.

Inventors

Xie Haige
YE FENGLEI
LI JIE
WU YEJING
YE HUABIN

Assignees

温州城市一卡通服务有限公司

Dates

Publication Date: 20260508
Application Date: 20250902

Claims (7)

1. A method for optimizing rolling processing of a point mall order based on big data, the method comprising: Calculating a demand change parameter by collecting integral business order inflow data and user behavior track data and combining historical load data and a preset load coefficient to determine a peak load level, wherein the peak load level reflects the peak demand of the processing capacity of the system, and the method specifically comprises the following steps: acquiring point business form order inflow data and user behavior track data, wherein the user behavior track data comprises browsing duration records and conversion frequency statistics, extracting average duration values from the browsing duration records, and calculating demand change parameters by combining frequency counts in the conversion frequency statistics, wherein the demand change parameters are obtained by dividing the average duration values by the ratio of the frequency counts; according to the demand change parameters, comparing the demand change parameters with historical load data to determine peak load level; according to the peak period load level, judging that if a demand change parameter exceeds a preset threshold, merging order quantity statistics in order inflow data with the peak period load level to obtain a resource pre-allocation scheme, wherein the resource pre-allocation scheme weights the peak period load level through the order quantity statistics to optimize peak demand reflection, classifies an order queue by adopting a priority ordering algorithm based on the peak period load level to obtain an order sequence for dividing the order into high-priority requests and conventional requests, and when the duty ratio of the high-priority requests in the order sequence exceeds the preset threshold, identifying conflict points with insufficient inventory by comparing the number of the requests in a concurrent access window with real-time inventory multiplied by an inventory safety margin coefficient to obtain a conflict order list, wherein the method specifically comprises the following steps: When the high priority request ratio exceeds a preset threshold, defining a concurrent access window (Concurrent Access Window, CAW), counting all exchange request quantity of the same high priority commodity in any CAW in real time, defining an Inventory safety margin (Inventory SAFETY MARGIN, ISM), wherein the Inventory safety margin is a coefficient S_factor smaller than 1 and represents an Inventory proportion reserved for coping with emergency, determining that resource conflict exists only when the instantaneous request quantity in one concurrent window exceeds the safety capacity of the Inventory, adding all corresponding requested orders into a conflict order list, adopting a preset weighting scoring model for the conflict order list, combining the historical exchange preference characteristics in the user behavior track data, adjusting the order processing priority to obtain a resource allocation scheme, integrating the order sequence through a queue, rearranging the order sequence based on the resource allocation scheme to obtain a final processing sequence, if the final processing sequence has a residual conflict set scheduling mechanism, extracting the corresponding requested orders from the preset standby set, carrying out complete execution and carrying out complete and iteration, and carrying out complete and running, and carrying out complete and stable and running, thereby obtaining a complete and running state.
2. The method for optimizing rolling process of the point mall order based on big data according to claim 1, wherein the calculation of the demand change parameter comprises the step of carrying out weighted statistics on effective browsing duration and high-value exchange events in a preset time window based on time attenuation weight to obtain a quantized value reflecting attention investment of a user.
3. The method for optimizing rolling process of the order of the point mall based on big data according to claim 1, wherein the sorting order queues by using a priority sorting algorithm based on the peak load level, and obtaining a sorted order sequence, the order sequence divides the order into a high priority request and a regular request, the high priority request relates to the point commodity exchanged at high frequency, and the method comprises the following steps: Extracting high-frequency exchange identification features from an order queue by adopting behavior track fusion according to the peak load level evaluation, wherein the extracted high-frequency exchange identification features are obtained by counting exchange frequency counts and layering by combining user requests to obtain a preliminary priority list, and the preliminary priority list marks the high-frequency exchanged point commodities as high-priority requests; For the preliminary priority list, peak demand response data is obtained and combined with a priority ordering mechanism, order queues are subjected to layering processing, the peak demand response data is derived from the peak load level evaluation, a classified order sequence is determined, and the sequence distinguishes high-priority requests from conventional requests; And judging that if the high-priority request duty ratio exceeds a preset threshold value from the classified order sequences, adjusting integral commodity division through resource priority distribution, wherein the integral commodity division is adjusted by adopting a sequence to generate a logic fusion behavior track, and obtaining an optimized sequence response scheme.
4. The method of claim 1, wherein the weighted scoring model further incorporates at least one of user level, order wait time, or commodity scarcity.
5. The method of claim 1, wherein integrating the order sequence by a queue reordering mechanism and rearranging the order based on the resource allocation scheme to obtain a final processing order comprises: Acquiring the order sequence, extracting priority indexes from the resource allocation scheme, and adopting an bubbling sequencing algorithm to perform successive comparison, exchange and integration on the order sequence to obtain a preliminary reordering sequence; Aiming at the preliminary reordering sequence, merging queue stability attributes extracted from the preliminary reordering sequence, judging that if a high priority request exceeds a preset threshold, adjusting the execution position of the high priority request, and determining an intermediate processing sequence; and rearranging the low-priority requests by the intermediate processing sequence and combining sequence integration attributes extracted from the intermediate processing sequence to obtain a final processing sequence.
6. The method for optimizing rolling process of a point mall order based on big data according to claim 1, wherein if there is a remaining conflict in the final processing sequence, extracting supplementary resources from a preset inventory through a standby resource set scheduling mechanism to obtain a complete execution plan covering all order processing paths, comprising: For the final processing sequence, acquiring the residual conflict position, extracting supplementary resources from the preset standby inventory, and determining a conflict resolution path by comparing the conflict position with the matching degree of the inventory resources; Performing resource allocation adjustment execution sequence on conflict positions by adopting resource set management through the conflict resolution path to obtain an intermediate plan under path coverage guarantee; And according to the intermediate plan, the standby resource scheduling is carried out by merging the order processing paths, and if the residual conflict exceeds a preset threshold value, the resource extraction is supplemented again from the preset inventory to obtain the complete execution plan.
7. The system is characterized by comprising a data acquisition and demand calculation module, a demand change parameter calculation module and a peak period load level calculation module, wherein the data acquisition and demand calculation module is used for calculating a demand change parameter by acquiring integral mall order inflow data and user behavior track data, and determining the peak period load level by combining historical load data and a preset load coefficient, and the peak period load level reflects the peak demand of the processing capacity of the system and comprises the following specific steps: acquiring point business form order inflow data and user behavior track data, wherein the user behavior track data comprises browsing duration records and conversion frequency statistics, extracting average duration values from the browsing duration records, and calculating demand change parameters by combining frequency counts in the conversion frequency statistics, wherein the demand change parameters are obtained by dividing the average duration values by the ratio of the frequency counts; according to the demand change parameters, comparing the demand change parameters with historical load data to determine peak load level; The method comprises the steps of determining a peak time load level, determining an order queue, classifying the order queue by a priority ordering algorithm based on the peak time load level, obtaining an order sequence for dividing the order into a high priority request and a conventional request, and obtaining a conflict list by merging the order quantity statistics in order inflow data with the peak time load level to obtain a resource pre-allocation scheme, wherein the resource pre-allocation scheme weights the peak time load level through the order quantity statistics, the order queue is classified by the order classification and queue management module to obtain the order sequence for dividing the order into the high priority request and the conventional request based on the peak time load level, and the resource competition detection module is used for comparing the request quantity in a concurrent access window with the real-time stock quantity multiplied by an inventory safety margin coefficient when the occupation ratio of the high priority request in the order sequence exceeds the preset threshold, wherein the conflict list is obtained specifically comprises the following steps: When the high priority request duty ratio exceeds a preset threshold, defining a concurrent access window (Concurrent Access Window, CAW), the time length of which is deltat, counting all exchange request quantity aiming at the same high priority commodity in any CAW in real time, defining an Inventory safety margin (Inventory SAFETY MARGIN, ISM), wherein the Inventory safety margin is a coefficient S_factor smaller than 1 and represents an Inventory proportion reserved for coping with emergency, the conflict triggering condition is defined as that only when the instantaneous request quantity in one concurrent window exceeds the safety capacity of the Inventory, the conflict is judged to exist, and all orders corresponding to the request are added into a conflict order list, a dynamic priority adjustment module is used for adjusting the priority of order processing aiming at the conflict order list by adopting a preset weighting scoring model and combining with the historical exchange preference characteristics in the user behavior trace data, a queue reordering module is used for integrating the order sequence through a queue mechanism and rearranging the order sequence based on the resource allocation scheme to obtain a final processing sequence, and if the final processing sequence exists, the final processing sequence is judged to exist, and all the corresponding orders are subjected to iteration processing is carried out through the complete and the iteration processing is adjusted, and the complete progress is monitored, and the iteration processing is completed.

Description

Method and system for optimizing rolling processing of point mall order based on big data Technical Field The invention relates to the technical field of information, in particular to a method and a system for optimizing rolling processing of an order of a point mall based on big data. Background The point mall plays a key role in improving user engagement and platform loyalty as an important component of e-commerce and user incentive mechanisms. With the rapid development of digital economy, the point mall needs to process massive order data and meet the diversified exchange demands of users. Especially during peak periods such as sales promotion activities or holidays, the order processing efficiency and the rationality of resource allocation directly influence the user experience and platform operation effect. However, in the conventional order processing manner, when faced with complex and variable user behaviors and dynamic resource requirements, it is often difficult to meet the requirements of modern point shops for efficient and intelligent processing. Currently, the order processing method of the point mall relies on static rules or simple priority scheduling, and is difficult to deal with dynamic changes of user exchange behaviors. For example, existing systems are typically based on fixed order queues and cannot flexibly adjust resource allocation according to user behavior or seasonal consumption trends. This approach tends to result in uneven allocation of resources during peak hours, with some order processing being delayed and other resources being idle. In addition, the existing system lacks real-time monitoring capability to the order processing link, and cannot quickly identify the reason of delay, so that user experience is reduced. This static and split approach limits the efficiency of the operation of the integration mall in complex scenarios. In the point mall order processing, one of the core technical difficulties is how to implement dynamic adjustment of order queues to accommodate fluctuations in user redemption requirements. The complexity of the user behavior trace and consumption pattern makes prediction of order requirements exceptionally difficult. For example, in a large promotional program, users may intensively redeem hot goods, resulting in increased competition for specific resources, which is difficult for existing systems to dynamically allocate processing power according to real-time demand. This lack of dynamic adjustment directly results in inefficient order processing and even system congestion. Another key technical difficulty is the conflict management of order competing resources. When multiple orders compete for limited processing resources, such as limited inventory, during the same time period, the system often cannot efficiently identify and handle such conflicts. Resource competition between orders may cause portions of the orders to be delayed in processing and even fail due to insufficient resources. For example, when a user attempts to redeem a quantity of merchandise, the system may not timely detect competing requests for other orders, resulting in a confusing processing sequence, and the user may eventually face a problem of redemption failure or long waiting time. This lack of conflict management further exacerbates the complexity of order processing. How to dynamically adjust order queues and effectively manage resource competition in a high concurrency scene becomes a key problem for improving the operation efficiency of the integral mall. In actual business, such as during a promotion, a point mall may receive thousands of redemption requests simultaneously, while the inventory and processing power of the hot goods is limited, the system needs to quickly determine which orders may be prioritized and how to avoid processing failures due to resource competition. Therefore, the problems of dynamic adjustment of the order queue and resource conflict management are solved, and the problems become key problems for realizing intelligent and stable operation of the integral mall. Disclosure of Invention In one aspect, the invention provides a method for optimizing rolling processing of an order of a point mall based on big data, which mainly comprises the following steps: The method comprises the steps of collecting integral business city order inflow data and user behavior track data, calculating a demand change parameter, combining historical load data with preset load coefficients, determining a peak period load level, wherein the peak period load level reflects peak demands of system processing capacity, classifying an order queue by adopting a priority ordering algorithm based on the peak period load level to obtain an order sequence for dividing the order into high priority requests and conventional requests, comparing the number of requests in a concurrent access window with real-time stock quantity multiplied by stock safety marginal coefficients