Full-process automatic processing system and method for online orders

Abstract

The invention discloses an online order full-process automatic processing system and method. The system comprises an order state analysis module, a cross-platform collaborative adaptation module, a resource scheduling planning module, an environment load evaluation module and a task scheduling optimization module. According to the invention, by analyzing the trend of the key time sequence change of the order life cycle, the circulation bottleneck can be identified and the time section exceeding the threshold value can be generated, so that the task sequence is matched with the system load, the scheduling coordination and the response precision are enhanced, the execution capability is dynamically compared to form the optimal distribution instruction, and the cooperation efficiency and the transmission continuity are improved. Based on a utilization rate change identification trend, resource distribution is balanced, scheduling flexibility and controllability are enhanced, resource and energy consumption cooperative adjustment is realized through environment load evaluation, operation stability and environment adaptability are improved, high efficiency and continuous optimization of an order flow are realized, and processing capability and operation toughness of a system under high concurrency are enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of business process management, in particular to an online order full-process automatic processing system and method. BACKGROUND

[0002] The technical field of business process management belongs to the technical category of the integration of information processing and enterprise management, mainly involving the modeling, execution, monitoring and optimization of business processes through information means. The core matters of this technical field include the modeling and definition of business processes, the automatic driving of process engines, cross-system data interaction, the decision support of rule engines, and the real-time monitoring of process execution status. Business process management aims to achieve business operation standardization, process visualization and resource optimal allocation through process automation and collaboration, and is an important part of modern enterprise information systems, widely used in complex business processes that require multi-link collaboration in industries such as manufacturing, finance, e-commerce, and government affairs.

[0003] Among them, the traditional online order full-process processing system refers to a system that manages orders from creation, review, payment, distribution, delivery to after-sales in e-commerce and enterprise online transaction platforms through manual operation or semi-automatic methods. This type of system relies on manual switching between different systems, such as confirming order information in the order management system, checking payment status in the payment system, executing inventory locking or warehouse release instructions in the inventory system, and entering delivery information in the logistics system. Manual judgment is used to execute exception handling and rule branch control. Its main technical means include database-based order record query, system data synchronization through interface, and partial automation task flow through fixed rule scripts, but the overall process still has strong manual dependence and cross-system coordination lag problems.

[0004] Traditional online order full-process processing relies on manual switching between multiple systems, and manual judgment during processing causes uncertain response time. When the number of business links increases or transaction peaks occur, data transmission between systems is prone to lag or omission, fixed rule scripts are difficult to handle complex scenario exception branches, resulting in lack of flexibility and dynamic adjustment capability in process scheduling, cross-system data synchronization relies on periodic interface calls, which is prone to task accumulation under network fluctuations or sudden data volume increases, frequent manual intervention increases the probability of errors and processing delays, ultimately causing poor connection between order processing links, unbalanced system resource allocation, and significant impact on overall operation efficiency and user experience. SUMMARY

[0005] The present application provides an online order full-process automatic processing system and method to solve the problems existing in the prior art.

[0006] Technical solution: This invention discloses an automated online order processing system, the system comprising: The order status analysis module, based on the lifecycle characteristics of orders, analyzes the changing trends of order creation time, payment status data, and inventory lock time sequence to extract abrupt change nodes of order flow bottlenecks and generate time segments exceeding thresholds. The cross-platform collaborative adaptation module matches the cross-platform task flow parameters and operation data of the corresponding mutation node based on the time segment exceeding the threshold, and constructs the load linkage interval. It calculates the state difference normalization index of the flow point in the load linkage interval, thereby constructing the offset intensity index group. Based on the offset intensity index group, the task flow path optimization instruction group is obtained. The resource scheduling and planning module calls the task flow path optimization instruction group, extracts the changes in resource utilization before and after instruction execution, identifies the concentration of resource allocation under a unit task in combination with the order flow trend, extracts the fluctuation segment to form the resource allocation path feature segment, and calculates the aggregation trend value of the resource allocation path feature segment. The environmental load assessment module collects environmental load data during order processing for characteristic sections of the resource allocation path, analyzes the overlap duration of pressure changes and resource occupancy surges, and outputs the results of resource and environmental synergy analysis for characteristic sections of the resource allocation path.

[0007] Furthermore, the order status parsing module includes: The time series extraction submodule extracts the time series of order creation time based on the life cycle characteristics of orders, calculates the difference in the number of orders at adjacent time points, extracts the peak-to-trough distance of the difference in the number of orders within each period, and filters out the segments where the peak-to-trough distance exceeds the threshold to obtain the time period fluctuation range. The payment status analysis submodule calls the time period fluctuation range, identifies the corresponding payment status data, analyzes the absolute change range of payment in adjacent time periods, compares it with the set payment status threshold, takes the latest time point in the time period corresponding to the absolute change range of payment exceeding the payment status threshold as the mutation node, establishes a time interval to cover the mutation node, and takes this time interval as the payment status mutation range. The status priority marking submodule identifies the corresponding inventory lock delay and order characteristic data based on the payment status change range. Combining the payment status and creation time fluctuations, it calculates the status complexity offset value of the change node, sets an offset threshold, marks the change node that exceeds the offset threshold, and backtracks a preset time window based on the change node that exceeds the offset threshold to obtain the time segment exceeding the threshold. The priority of the time segment exceeding the threshold is marked as high priority.

[0008] Furthermore, the cross-platform collaborative adaptation module includes: The data flow matching submodule samples at preset intervals based on time segments exceeding thresholds, extracts system task flow parameters and operation data from adjacent sampling points, calculates the rate of change of task flow parameters and operation data from adjacent sampling points, and if the rate of change of task flow parameters and operation data exceeds a preset rate of change threshold, the corresponding sampling point is designated as a load mutation node; if the time period consisting of multiple consecutive load mutation nodes overlaps with the timestamp of order creation or the difference is less than a preset time threshold, the time period consisting of these multiple consecutive load mutation nodes is designated as a load linkage interval. The flow deviation judgment submodule calls the load linkage interval and takes the load change node in the load linkage interval as the flow point; calculates the state difference normalization index G for each flow point; if G is greater than or equal to the preset offset boundary, then G is taken as the deviation intensity of the corresponding flow point; if the deviation intensity is greater than or equal to the preset benchmark judgment value, the corresponding flow point is marked as deviation, and an offset intensity index group is established based on the position of the flow point marked as deviation. The task flow instruction extraction submodule, based on the offset intensity index group, filters the position of the corresponding time point in the task instruction set, extracts the corresponding task instructions and task instruction values, sorts the task instructions according to the time sequence, removes duplicate task instructions, and obtains the task flow path optimization instruction group.

[0009] Furthermore, the resource scheduling and planning module includes: The resource utilization extraction submodule extracts resource utilization data before and after the execution of task instructions based on the task flow path optimization instruction group; obtains resource utilization data in multiple cycles before and after the execution of task instructions, identifies resource boundary conditions, and obtains the resource utilization difference value corresponding to the task instructions based on the resource boundary conditions. The resource allocation path identification submodule calls the resource utilization difference value. If the resource utilization difference value is greater than a preset significance threshold, it is determined that the resource status has changed substantially. Then, it identifies the resource allocation path change trajectory in the unit task path, calculates the resource quantity fluctuation corresponding to the change in the unit task path trajectory, compares the path decrease with the resource quantity fluctuation, and calculates the path fluctuation coupling degree index. It filters out the path fluctuation coupling degree index that is greater than or equal to the preset index, and filters out the segments where the path decrease and the resource quantity fluctuation change direction are consistent and the deviation of their change rate is within a preset range. These segments are defined as the resource allocation path feature segment. The allocation response interval identification submodule samples the characteristic segment of the resource allocation path and calculates the aggregate trend value Q of the characteristic segment of the resource allocation path using the following formula: ; in, The first representing the path displacement The change value of each sampling point This represents the average value of the path displacement change at the sampling points. Representing the The time interval between sampling points Represents the total time interval. It is a weighting factor, and N represents the total number of sampling points.

[0010] Furthermore, the environmental load assessment module includes: The path trend extraction submodule extracts node task records and periodic path data from the resource allocation path feature segments. The periodic path data refers to data related to resource usage or path flow within a specific business cycle. The specific business cycle is a segment corresponding to the resource allocation path feature segments and having specific attributes, which are attributes manually set according to requirements. The node task records and periodic path data are normalized, and the normalized values ​​are weighted and summed to obtain the comprehensive fluctuation trend value for the specific business cycle. For a single business cycle, the aggregated trend values ​​corresponding to multiple resource allocation path feature segments within the single business cycle are summed to obtain a global aggregated trend value; the overall resource scheduling health of the system is evaluated based on the global aggregated trend value. The resource occupancy surge identification submodule identifies resource occupancy surge periods within characteristic segments of the resource allocation path using the following method: Based on the characteristic segments of the resource allocation path, and according to the environmental load data in the order processing process, resource occupancy and pressure records within the same period are selected, the data change magnitude is compared, and time nodes with magnitudes greater than the resource occupancy jump threshold are filtered to obtain the resource occupancy jump period. The environmental load quantification submodule identifies the intersection duration of the path amplitude change period and the resource occupancy surge period within the resource occupancy surge period. Based on the degree of impact of the intersection duration on the system, it performs preliminary weighting processing on the intersection duration. Then, it normalizes the intersection duration, the frequency of path amplitude change, the merged amplitude value, and the comprehensive fluctuation trend value of a specific business cycle after the preliminary weighting processing, and performs weighted fusion on the normalized values. The fusion result serves as the resource and environmental collaborative processing benefit analysis result for the resource allocation path characteristic segment.

[0011] Furthermore, the system also includes a task scheduling optimization module: The task scheduling optimization module, based on the results of the resource and environment collaborative processing benefit analysis, filters task scheduling instructions affected by environmental interference, classifies platform switching time and adjustment trigger frequency, identifies time periods when load fluctuations exceed preset thresholds, and obtains the frequency of order processing task load interference. The frequency of order processing task load interference includes system switching frequency, number of adjustment command triggers, and number of times the jump time period exceeds the limit.

[0012] Furthermore, the task scheduling optimization module includes: The instruction filtering submodule, based on the results of the resource and environment collaborative processing benefit analysis, identifies the platform operation mode and platform status of resource allocation path characteristic segments whose scores are lower than the preset analysis threshold, compares the disturbance duration with the effective duration of task scheduling instructions, and obtains the set of task scheduling instructions affected by environmental interference. The business classification submodule calls the set of task scheduling instructions affected by environmental interference, extracts the business unit switching time and adjustment trigger frequency corresponding to the task scheduling instructions, arranges them in order of switching time, classifies the adjustment type by frequency, and statistically analyzes the frequency band corresponding to the adjustment type and the platform operation mode within the frequency band to obtain the platform adjustment frequency band distribution value. The interference identification submodule adjusts the frequency band distribution value according to the platform, collects the load jump amplitude and duration within the frequency band, determines whether the jump amplitude threshold is exceeded, thereby identifying the frequency band interference intensity, sorting the interference intensity of all frequency bands, determining the frequency band and jump time period of the interference intensity, and obtaining the frequency of order processing task load interference impact.

[0013] A method for automating the entire online order processing process includes the following steps: S1: Based on the lifecycle characteristics of orders, by analyzing the changing trends of order creation time, payment status data, and inventory lock-in time sequence, extract the abrupt change nodes of order flow bottlenecks and generate time segments exceeding the threshold. S2: Based on the time interval exceeding the threshold, match the cross-platform task flow parameters and operation data of the corresponding mutation node, construct the load linkage interval, calculate the state difference normalization index of the flow point in the load linkage interval, construct the offset intensity index group, and obtain the task flow path optimization instruction group based on the offset intensity index group. S3: Call the task flow path optimization instruction group, extract the resource utilization rate changes before and after instruction execution, identify the resource allocation concentration under unit task in combination with order flow trend, extract the fluctuation segment to form the resource allocation path feature segment, and calculate the aggregation trend value of the resource allocation path feature segment. S4: For characteristic segments of the resource allocation path, collect environmental load data during the order processing process, analyze the overlap duration of pressure changes and resource occupation surge periods, and output the resource and environmental collaborative processing benefit analysis results for characteristic segments of the resource allocation path.

[0014] A computer device includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the aforementioned method for fully automated online order processing.

[0015] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned method for fully automated online order processing.

[0016] Beneficial effects: In this invention, by analyzing the trend of key time sequence changes in the order lifecycle, bottlenecks can be accurately identified and priority marker values ​​can be generated, so that the task processing order matches the system load, enhancing the real-time coordination and response accuracy of task scheduling. The task flow path forms the optimal allocation instruction after dynamically comparing the platform's execution capabilities, improving the efficiency of multi-platform collaboration and the continuity of task transmission. During resource allocation, trend identification is performed based on utilization rate changes, making resource distribution more balanced and enhancing the flexibility and controllability of resource scheduling. Environmental load assessment enables the system to achieve coordinated adjustment of resources and energy consumption when pressure changes, improving operational stability and environmental adaptability. Ultimately, it achieves high efficiency and continuous optimization of the entire order execution process, strengthening the system's processing capacity and overall operational resilience under high concurrency conditions. Attached Figure Description

[0017] Figure 1 This is a system flowchart of the present invention;

[0018] Figure 2 This is a flowchart of the order status parsing module in this invention;

[0019] Figure 3 This is a flowchart of the cross-platform collaborative adaptation module in this invention;

[0020] Figure 4 This is a flowchart of the resource scheduling and planning module in this invention;

[0021] Figure 5 This is a flowchart of the environmental load assessment module in this invention;

[0022] Figure 6 This is a flowchart of the task scheduling optimization module in this invention. Detailed Implementation

[0023] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0025] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0026] Please see Figure 1 This invention provides a technical solution: an automated online order processing system comprising:

[0027] The order status parsing module analyzes the changes in order creation time, payment status, and inventory lock-in time sequence based on the lifecycle characteristics of the order, extracts the abrupt change nodes of the order flow bottleneck, generates time segments exceeding the threshold, and generates order status priority marker values.

[0028] The cross-platform collaborative adaptation module matches the cross-platform task flow parameters and operation strategies of the corresponding mutation node based on the order status priority mark value. It compares the platform load level in the marked segment with the current execution capacity, filters the flow instructions corresponding to the abnormal flow points, and obtains the task flow path optimization instruction group.

[0029] The resource scheduling and planning module optimizes instruction groups based on task flow paths, extracts changes in resource utilization before and after instruction execution, identifies the concentration of resource allocation under a unit task by combining order flow trends, extracts fluctuation segments to form resource allocation path feature segments, and obtains resource scheduling optimization trends.

[0030] The environmental load assessment module invokes resource scheduling optimization trends, collects environmental load data during order processing, analyzes the overlap duration of pressure changes and resource consumption surges, and outputs the results of resource and environmental collaborative processing benefit analysis.

[0031] Based on the results of the resource and environment collaborative processing benefit analysis, the task scheduling optimization module filters task scheduling instructions affected by environmental interference, classifies platform switching time and adjustment trigger frequency, identifies the period of load jump exceeding the limit, and obtains the frequency of load interference affecting order processing tasks.

[0032] Order status priority markers include creation time fluctuation range, payment status change magnitude, and inventory lock delay characteristics. Task flow path optimization instruction groups include flow deviation, load status anomaly value, and execution capability offset rate. Resource scheduling optimization trends include resource utilization rate change range, resource allocation type, and allocation response magnitude. Resource and environment collaborative processing benefit analysis results include resource occupation fluctuation duration, pressure surge overlap period, and disturbance duration. Order processing task load interference impact frequency includes system switching frequency, adjustment instruction trigger count, and jump period exceeding limit count.

[0033] Please see Figure 2 The order status parsing module includes:

[0034] The time series extraction submodule extracts the time series of order creation time based on the life cycle characteristics of orders, calculates the difference in the number of orders at adjacent time points, extracts the peak-to-trough distance within each period, and filters out segments where the peak-to-trough distance exceeds a threshold to obtain the time period fluctuation range value.

[0035] Based on the analysis of historical data using order lifecycle characteristics, time-series data of order creation time is extracted. For example, from September 23rd to September 29th on an e-commerce platform, the number of orders created per hour is recorded, forming a time-series data of 168 sampling points. Based on the time series, the absolute difference in the number of orders created between adjacent sampling points is calculated. For example, if there are 10,000 orders in the first hour and 10,500 orders in the second hour, the time difference is 500 orders. The difference between adjacent sampling points is continuously calculated to obtain an order difference sequence. The order difference sequence is then processed by a moving average with a window size of 3 to smooth the data and identify periodic fluctuation patterns. The smoothed time difference sequence is observed to identify the fluctuation cycle of order creation volume, such as a daily cycle of 24 hours and a weekly cycle of 7 days. The peaks and troughs of order creation volume in each cycle are extracted. A peak is defined as an order creation volume exceeding the weekly peak. The local maximum point is defined as 15% of the period average, and the trough is defined as the local minimum point below 15% of the period average. The peak-trough distance is calculated as the time from the time of the peak to the time of the first trough. For example, in a daily cycle, there is a peak at 10:00 and a trough at 15:00, with a peak-trough distance of 5 hours. The current period's peak-trough distance is compared with the historical average distance to calculate the period difference value. For example, the historical average distance is 4.5 hours, and the current observation is 5 hours, with a period difference of 0.5 hours. The period difference value is compared with a preset period difference threshold, which is derived from the analysis of order data over the past year to capture abnormal changes in order creation patterns. Through historical data statistical analysis, the threshold is set to 0.8 hours. The setting process is as follows: collect daily period peak-trough distance data of historical order creation volume, calculate the daily distance change rate, and calculate the standard deviation of all daily distance change rates. Set the threshold to twice the standard deviation, for example, the standard deviation. The time period is set to 0.4 hours, and the threshold is set to 0.8 hours. The segment that deviates from the regular pattern is identified, and the segment with a period difference of more than 0.8 hours is selected to obtain the time period fluctuation range value.

[0036] The payment status analysis submodule calls the time period fluctuation range value, identifies the corresponding payment status data, analyzes the absolute change range of payment in adjacent time periods, compares it with the set payment status threshold, locates all abrupt change nodes, and obtains the range of payment status change amplitude.

[0037] The system calls upon a time-period fluctuation range, for example, creating a time-period fluctuation range from 10:00 to 12:00 on September 24th. It identifies payment status data within this range, including the real-time status and timestamps of different payment stages of orders. For example, if 15,000 payment status data entries are identified within the range: 10,000 orders awaiting payment, 3,000 orders in progress, 1,500 successfully paid orders, and 500 failed payment orders, the system analyzes the absolute change in payment status within adjacent time periods. The 2-hour fluctuation range is divided into 10-minute time periods for analysis, calculating the absolute change in the number of successfully paid orders within each 10-minute time period. For example, from 10:00 to 10:10 on September 24th, there were 500 successfully paid orders; from 10:10 to 10:20, there were 800 successfully paid orders; the absolute change in payment status within adjacent time periods was 300 orders. This calculation continues to obtain the fluctuation data. The system tracks the absolute change in payment over all adjacent 10-minute time intervals within a given timeframe. Each absolute change is compared to a preset payment status threshold, determined based on the platform's daily payment capabilities and historical data analysis. This process identifies fluctuations in payment status. Using statistical analysis of daily fluctuations in payment success rates over the past six months, a threshold of 10% of the order quantity is set. For example, if the average number of orders within a 10-minute timeframe is 1000, the threshold is 100 orders. When the absolute change in payment over adjacent time intervals exceeds the 100-order threshold, that point is identified as a mutation node. For instance, if the absolute change in payment between 10:10 and 10:20 is 300 orders, exceeding the 100-order threshold, then 10:20 is marked as a mutation node. All mutation nodes are then located. Subsequently, the system merges the identified, temporally continuous, or adjacent mutation nodes to form one or more consecutive time intervals. For example, if mutation nodes are identified at 10:20, 10:30, and 10:40, the system will merge them into a continuous mutation interval, such as 10:10-10:40 (this interval covers all analysis periods that generated mutation nodes). The final output "Payment Status Mutation Range" refers to one or more such sets of specific time ranges that identify drastic fluctuations in payment status.

[0038] The status priority marking submodule identifies the corresponding inventory lock delay and order characteristic data based on the range of payment status change, extracts the delay and characteristic data, combines the payment status and creation time fluctuations, calculates the status complexity offset value of the change node, sets the offset threshold, marks the time segment exceeding the threshold, and obtains the order status priority marking value of the time segment exceeding the threshold.

[0039] Based on the range of payment status abrupt changes, such as between 10:00 and 12:00 on September 24th, multiple payment status abrupt change nodes are located. For each node, inventory lock delay and order characteristic data are identified. Inventory lock delay data refers to the time interval between order creation and inventory lock, for example, a product inventory lock delay of 5 seconds. Order characteristic data includes product type, order amount, and buyer membership level. For example, during periods of high delay, order characteristics are high-value physical goods with high buyer membership levels. Inventory lock delay data and order characteristic data are extracted; for example, the inventory lock delay time array [3s, 5s] is extracted. [8s, 4s] and the order characteristic array (high-value physical goods, membership level 5), combined with payment status (payment success rate decreased by 15%) and creation time fluctuation (creation volume increased sharply by 20% within 10 minutes), calculate the state complexity offset value of the mutation node. The state complexity offset value represents the degree of deviation of the order processing status from the expected status. The calculation process is as follows: the inventory lock delay time is normalized from 0 to 1, for example, a 5-second delay is normalized to 0.5. Secondly, the order characteristics are quantified, for example, "high-value physical goods" is quantified to 0.8, "membership level 5" is quantified to 0.9, and the payment status fluctuation is also considered. The quantification is set at 0.7. Payment status fluctuation refers to the specific magnitude of changes in payment status. Creation time fluctuation is quantified at 0.9. The state complexity offset value of the mutation node is calculated using a weighted average method. The calculation result is the sum of the products of each weight and its corresponding normalized or quantified value. The weights are set based on historical data and expert experience to reflect the degree of influence of different factors on complexity. The setting process is as follows: review historical failures and performance bottlenecks, evaluate each factor, and use the analytic hierarchy process (AHP) to calculate the weights. For example, the weight of inventory lock delay is set at 0.4; the weight of order characteristics is set at 0.25; and the weight of payment status fluctuation is set at 0. 2. The time fluctuation weight is set to 0.15, and the sum of the weights is 1. Sample data is used for calculation, resulting in an offset value of 0.675. The calculated state complexity offset value is compared with a preset offset threshold. The offset threshold is set based on the average historical state complexity offset value plus two standard deviations during operation. This identifies time segments with high-risk states. For example, through historical data analysis, the average state complexity offset value is 0.3, the standard deviation is 0.15, and the offset threshold is set to 0.6. When the offset value of 0.675 exceeds the threshold of 0.6, the sudden change node (e.g., time point 10:20) is marked as a high-risk node. Subsequently, the system will use this high-risk node as the endpoint to backtrack forward by a preset time window (e.g., 20 minutes) or back to the previous stable state node, thereby determining a complete time segment (e.g., September 24th, 10:00-10:20).This identified time segment will be marked as an over-threshold time segment. For example, September 24th, 10:00-10:20 will be marked as an over-threshold time segment, resulting in an order status priority label value. If the offset value exceeds this threshold (e.g., the offset value calculated in the scheme is 0.675, exceeding the threshold of 0.6), then the corresponding time segment will be marked as "high priority" as an over-threshold time segment to identify high-risk states that require priority attention and processing. Conversely, if the offset value does not exceed the threshold, the time segment will be considered "low priority" or a normal state.

[0040] Please see Figure 3 The cross-platform collaboration and adaptation module includes:

[0041] The data flow matching submodule extracts the system task flow parameters and operation data of the corresponding mutation node based on the order status priority mark value (i.e., the time segment exceeding the threshold), calculates the load fluctuation amplitude according to the sampling interval, and aligns the order creation time and load fluctuation amplitude with the same time to obtain the load linkage interval group.

[0042] Based on the order status priority flag, for example, the time segment from 10:00 to 10:20 on September 24th is marked as high priority. Task flow parameters and operation data for the corresponding mutation nodes within this segment are extracted. Task flow parameters include the number of cross-interface calls, average response time, and error rate. A task refers to a specific business operation or system interaction step in the entire order processing chain. Operation data includes interface call frequency, database operation concurrency, and fluctuation range. For example, within this segment, it is extracted that A called interface B 1000 times (the interface refers to the technical contract for communication between systems or services, which is an API (Application Programming Interface)), with an average response time of 150 milliseconds and an error rate of 5%. B's database operation concurrency is 500. The load fluctuation range is calculated according to a preset sampling interval. (The above describes extracting task flow parameters and operation data for mutation nodes; that is, the rate of change of task flow parameters and operation data should be between adjacent mutation nodes. Why is sampling set?) The system will use this high-risk node as the endpoint and backtrack for a preset time window (e.g., 20 minutes) or back to the previous stable node to determine a complete time segment (e.g., 10:00-10:20 on September 24). This means that there may only be one mutation node in this segment. The calculation of this segment is based on the rate of change or absolute difference of task flow parameters and operation data between adjacent mutation nodes, or sampling within the high-priority time segment. For example, the average response time changes from 120 milliseconds to 180 milliseconds within a certain minute, with a fluctuation range of 60 milliseconds, and the database operation concurrency changes from 400 to 600, with a fluctuation range of 200. Combined with preset threshold rules (e.g., average response time fluctuation exceeds 50ms, error rate increases by more than 3%, call frequency change rate exceeds 20%), when multiple parameters meet the mutation conditions at the same time point or adjacent time points, it is defined as a load mutation node. If multiple parameters change synchronously within a continuous time period, and these changes overlap or are close to the order creation time in terms of timestamp, then that time period is determined to be a "load linkage interval".

[0043] The flow deviation judgment submodule calls the load linkage interval group to extract the platform state sequence, calculates the flow change and state difference at each flow point in the platform state sequence, normalizes the state differences, and compares the normalized values ​​with the preset offset boundary using the following formula:

[0044] ;

[0045] Calculate the deviation intensity of each transfer point, extract abnormal transfer points with deviation intensity greater than the benchmark judgment value, and establish an offset intensity index group.

[0046] in, The deviation intensity represents the flow point. The real-time value representing the flow point i. This represents a preset reference value. The real-time value representing the flow point j. Represents the average value of the state sequence. The total number of state sequences. The weighting factor represents the deviation from the flow point i.

[0047] The call load linkage interval is defined as a "transfer point" as a sudden change node identified within the load linkage interval (i.e., a point where key performance parameters such as average response time, call frequency, or concurrency change abruptly at a certain timestamp). Each transfer point corresponds to a real-time status value of a key business platform (such as order service or inventory service), which serves as the real-time value of the transfer point. For example, if a mutation point occurs at 10:05 on September 24th, and the average response time of platform A is 180ms, then this value will be used as the transition point. Load fluctuation amplitude is used to help identify abrupt change nodes, rather than as a direct state value. For example, it includes a list of average response time fluctuation amplitudes and database operation concurrency fluctuation amplitudes at different timestamps to extract the platform state sequence. For instance, it extracts the average response time data of platform A (order service) per minute over the past hour to form the platform state sequence. ,in This represents the average real-time response time of platform A at transfer point i, for example... Compare the flow changes (i.e., the load fluctuation amplitude in the above load linkage interval group, such as the average response time fluctuation amplitude of 60 milliseconds) with the platform state differences (i.e., the platform state sequence). Compared with the preset reference value (differences), calculate the state difference normalization index. This indicator measures the degree of deviation of the current flow state from the ideal state. Considering its ability to combine real-time state values ​​with historical state distributions, and the introduction of weighting factors to highlight the deviation intensity of key flow points—"key flow points" referring to sampled data points representing a specific business process in a time series—it helps to more accurately identify and locate potential anomalies or bottlenecks. The specific formula used is: ,in, The deviation intensity represents the degree of deviation between the actual state and the expected reference value at a specific circulation stage. The larger the value, the more serious the deviation. Representing the The real-time value of the flow status of each flow point, such as the average response time of platform A at the current moment, is obtained by directly collecting data from the monitoring logs. For example, the average response time of platform A at a certain moment is 180 milliseconds. This represents a preset reference value, such as the expected average response time for platform A. It is set based on the platform's average performance metrics and service level agreement (SLA) requirements during its historical stable operation. For example, if platform A's SLA requires an average response time of less than 120 milliseconds, then this value is set accordingly. ; This represents the average response time series of platform A over the past n sampling points (e.g., the past 60 minutes). The data is collected minute by minute by a real-time monitoring system. For example, the response time series of platform A over the past 5 minutes is... ; The average value of the state sequence is calculated as follows: For example, for the sequence above, ; This represents the total number of state sequences. For example, if one data point is collected every minute over the past 60 minutes, then... ; Representatives and circulation points The deviation-based weighting factor reflects the importance or criticality of different stages in the overall order processing flow. It is set based on business process analysis and risk assessment, with higher weights for critical business stages (such as payment and inventory) and lower weights for general business stages. For example, the payment stage after order creation is of higher importance. In order to perform calculation examples, we set... , Platform state sequence for ,but , , Substitute the above values ​​into the formula to perform the calculation:

[0048] First, calculate the square root term in the denominator: ;

[0049] so, ;

[0050] Then, calculate the absolute value term in the numerator: ;

[0051] Finally, calculate : ;

[0052] The normalized state difference normalization index The data is compared with a preset offset boundary, which is set based on the upper bound of the 99% confidence interval of the normal fluctuation range in historical data. This boundary aims to distinguish between normal fluctuations and abnormal deviations. For example, by comparing historical... Statistical analysis was performed on the values, and the offset boundary was set to 2.5. When the value exceeds the preset offset boundary, G is used as the deviation intensity of the corresponding flow point, for example... Extract the flow location index where the deviation intensity is greater than the benchmark judgment value, which is based on historical data during periods of significant performance issues or business interruptions. The value is set as an average, for example, by analyzing abnormal events that caused business losses in the past year, and statistically analyzing the events at the time of occurrence. The values ​​are calculated, and their average value is taken as the benchmark judgment value, which is set to 2.8. This means that A value higher than 2.8 indicates that a significant deviation is occurring or about to occur, due to the calculated... If the deviation exceeds the benchmark value of 2.8, the transfer point is marked as a deviation. For example, the payment transfer process at 10:05 on September 24 is marked as a deviation. An offset strength index group is established. The result shows that the status of the transfer point deviates far from the preset reference value and shows significant abnormality compared with historical fluctuations, which requires further attention.

[0053] The task flow instruction extraction submodule uses the offset intensity index group to filter the position of the corresponding time point in the task instruction set, extracts the instruction value, sorts it by time sequence, removes duplicate instructions, and obtains the task flow path optimization instruction group.

[0054] Based on the offset intensity index group, for example, the index group marks the deviation in the payment flow process at 10:05 on September 24th. The position of the time point in the corresponding task instruction set is filtered. The task instruction set is a list of instructions to optimize or adjust the order processing flow. For example, the instruction set at this time point contains instructions such as "increase the number of concurrent connections of the payment gateway", "switch to the backup payment channel", and "reduce the priority of some non-core business". The instruction values ​​are extracted. For example, "increase the number of concurrent connections of the payment gateway" corresponds to "increase by 20%" and "switch to the backup payment channel" corresponds to "activate channel B". The extracted instructions are sorted by time sequence. For example, the 10:05 instruction is executed first, followed by the 10:06 instruction. Duplicate instructions are removed. For example, the "increase the number of concurrent connections of the payment gateway" instruction is triggered at consecutive sampling points. Only the first triggered instruction is retained to obtain the task flow path optimization instruction group.

[0055] Please see Figure 4 The resource scheduling and planning module includes:

[0056] The resource utilization extraction submodule optimizes the instruction group based on the task flow path, extracts resource utilization data before and after the execution of task instructions, identifies resource boundary conditions, and obtains the resource utilization difference value.

[0057] Optimize instruction groups based on task flow paths. For example, if an instruction group contains instructions like "increase the number of concurrent connections to the payment gateway by 20%" and "switch to backup payment channel B", extract resource utilization data within the call loop before and after the task instruction execution. For instance, for the instruction "increase the number of concurrent connections to the payment gateway by 20%", extract the CPU utilization, memory utilization, and network bandwidth utilization of the payment gateway node before execution (e.g., 10:04 AM on September 24th) and after execution (e.g., 10:06 AM on September 24th). Obtain data from multiple cycles before and after instruction execution to identify resource boundary conditions. Boundary conditions refer to the maximum resource capacity that the system can withstand under normal operating conditions. Maximum or minimum resource utilization. For example, the CPU utilization limit for a payment gateway node is 85%, the memory utilization limit is 90%, and the network bandwidth utilization limit is 70%. When the resource utilization exceeds the limit, the performance degrades. The resource utilization before and after instruction execution is normalized and compared with the corresponding resource boundary conditions. The difference in resource utilization is calculated as follows: Resource utilization difference = (Resource utilization before instruction execution - Resource utilization after instruction execution) / Utilization limit of the resource. For example, if the CPU utilization is 80% before adjustment and 75% after adjustment, and the CPU limit is 85%, the resource utilization difference is obtained.

[0058] The resource allocation path identification submodule calls the resource utilization difference value to identify the resource allocation path change trajectory in the unit task path, judges the resource quantity trend corresponding to the change in the unit task path trajectory, compares the path reduction and resource quantity fluctuation in adjacent time periods, calculates the path fluctuation coupling index, filters synchronous fluctuation segments, and obtains the resource allocation path feature segment.

[0059] The system calls upon resource utilization difference values, such as a 5% decrease in payment gateway CPU utilization. If this value meets a preset significance threshold, a substantial change in resource status is confirmed. This allows for the identification of resource allocation path changes within a single task path. A single task path refers to the sequence of resource nodes traversed by the order processing flow and their resource consumption. For example, a payment task, from receiving a request to completing processing, sequentially passes through resource nodes such as the load balancer, application server, payment gateway, and database. The resource allocation path change trajectory refers to the changes in the resource nodes used by the single task, the resource allocation on each node, and the task flow direction before and after the task instruction adjustment. For example, after executing the "Switch to backup payment channel B" instruction, the payment task is routed to backup payment gateway B, changing its resource allocation path. The system then judges the resource quantity trend corresponding to the change in the single task path trajectory. For example, after switching to backup payment channel B, the CPU utilization of the primary payment gateway decreases, while the CPU utilization of backup payment gateway B increases, indicating a shift of resources from one gateway to another, forming a resource quantity change trend. The path decrease and resource quantity fluctuation are compared within adjacent time periods. For example, within a certain adjacent time period... Within the context of a 20% decrease in the load on the main payment gateway task path and a 15% decrease in CPU utilization, a path fluctuation coupling index is calculated (in this embodiment, this index is set by analyzing historical adjustment data and employing statistical methods, such as calculating the correlation coefficient between the change in task path load and the corresponding change in resource utilization (e.g., the Pearson correlation coefficient), thus deriving a value that measures the strength of the correlation between the two). This index measures the degree of correlation between changes in task path and changes in resource quantity. For example, through historical adjustment data analysis, setting the index value to 0.7 indicates a high degree of coupling between path changes and resource changes. Based on this high degree of coupling, it is determined that the resource changes in the current time period are directly caused by path adjustments, thereby filtering synchronous fluctuation segments. For example, segments where the path reduction ("path reduction" refers to the decrease in "load" or "traffic" of a unit task on a specific path) is in the same direction as the change in resource quantity fluctuation and the deviation of their change rates (i.e., |path reduction% - resource fluctuation%|) is within ±10% are filtered out. The set of synchronous fluctuation segments that meet the coupling and amplitude conditions is defined as the resource allocation path characteristic segment.

[0060] The allocation response interval identification submodule analyzes the time and path displacement trends based on the resource allocation path feature segments, determines the consistency of direction and amplitude changes, extracts and aggregates the sampling point data within the segment, using the following formula:

[0061] ;

[0062] The aggregation trend value of the fluctuation segments within the characteristic segment is also the resource scheduling optimization trend.

[0063] in, The aggregation trend value represents the fluctuation segments within the characteristic region. The first representing the path displacement The change value of each sampling point This represents the average value of the path displacement change at the sampling points. Representing the The time interval between sampling points Represents the total time interval. It is a weighting factor, and N represents the total number of sampling points.

[0064] Based on the characteristic segments of the resource allocation path, for example, during a certain time period, the path of payment tasks switches from the primary payment gateway to the backup payment gateway B, accompanied by a decrease in the resource utilization of the primary gateway and an increase in the resource utilization of the backup gateway. Analyze the time and path displacement trends. The path displacement trend refers to the direction and speed of change of the resource allocation path over time. For example, the analysis reveals that during a specific time period, payment tasks gradually shift from the primary gateway to the backup gateway, forming a path displacement trend from high load to low load. Determine the consistency of direction and the characteristics of magnitude change. Directional consistency refers to whether the direction of path displacement is consistent with the expected goal (e.g., load balancing), while magnitude change characteristics are... This refers to the scale and speed of path displacement. For example, it determines whether the direction of path displacement is consistent with the goal of transferring load from overloaded nodes, and whether the magnitude of the transfer is within a controllable range. The sampled data within the characteristic segment is aggregated. Instead of re-filtering fluctuation segments, the segment with continuous resource transfer characteristics (i.e., fluctuation segments) identified in the previous step is quantified. For example, if a path continuously transfers resources for 10 minutes, and this segment was identified as a characteristic segment in the previous step, this step aggregates the segments to form a trend observation over a longer time scale. For example, multiple 10-minute transfer segments are aggregated into a 30-minute path optimization cycle, using the formula: Calculate the aggregation trend value of fluctuation segments within the characteristic region. This formula aims to quantify the fluctuation and aggregation trends of resource allocation paths over a period of time. Considering its ability to reflect the degree to which path displacement deviates from the average value, and incorporating time weights, it helps identify persistent and significant path change patterns, providing a quantitative basis for resource scheduling optimization. The aggregation trend value represents the fluctuation segments within the characteristic section, indicating the overall fluctuation and aggregation intensity of the resource allocation path displacement within a specific time period. The larger the value, the more intense and continuous the path fluctuation. Representing the The change in path displacement at each sampling point, for example, within a 1-minute sampling interval, is calculated by taking the change in the CPU utilization of the main payment gateway as the change in path displacement. Assuming the change occurs at the 1st sampling point... At each sampling point, CPU utilization dropped from 70% to 65%. ; The average value of the path displacement change at each sampling point is calculated using all sampling points. The arithmetic mean, for example, within a 30-minute aggregation period, if all sample points... Average ,but ; Representing the The time interval between sampling points, for example, at a sampling frequency of 1 minute. ; This represents the total time interval, that is, the total duration of the fluctuation segment. For example, the total length of the aggregation period is 30 minutes. It is a weighting factor used to adjust the impact of time intervals on the convergent trend value. Its setting is based on historical data analysis and sensitivity to short-term fluctuations. If it is more sensitive to recent fluctuations, then... Larger values, conversely Smaller values, for example, to emphasize recent or ongoing path displacement changes, will Set to 1.5; This represents the total number of sampling points. For example, within a total time interval of 30 minutes, with a sampling interval of 1 minute, then... To perform the calculation example, a fluctuation segment is set to last for 5 minutes, with a sampling interval of 1 minute, i.e. , Minutes, path displacement change at each sampling point They are respectively ,but , minute, .

[0065] Substitute the above values ​​into the formula to calculate: ;

[0066] The resource scheduling optimization trend was obtained. The result shows that the aggregation trend value of the fluctuation segment within the characteristic section is 0.5364, which quantifies the fluctuation intensity of the path displacement. This value will be used in subsequent modules as an important basis for assessing environmental load and guiding task scheduling. It helps the system identify the overall effect of resource scheduling and the direction that needs further optimization.

[0067] Please see Figure 5 The environmental load assessment module includes:

[0068] The path trend extraction submodule calls the resource scheduling optimization trend, extracts node task records and periodic path data, identifies the fluctuation amplitude and frequency within a single period, and obtains the fluctuation trend value of the periodic path.

[0069] The resource scheduling optimization trend is invoked. After obtaining the synchronous fragment data for resource scheduling optimization, node task records and periodic path data are extracted. Here, "node" refers to the specific task node identified within the aggregated synchronous fragment; "node task record" refers to the historical data of tasks running on the computing node, including task start time, end time, and resource consumption. For example, extracting task records for order processing on the product service node over the past few days. Periodic path data refers to the statistical pattern of resource usage or flow paths followed by tasks within a specific business cycle (this cycle corresponds to the time window of the synchronous fragment; i.e., if the current synchronous fragment occurs between 10:05 and 10:35 on September 24th, then the specific business cycle refers to the daily peak period of 10:00-11:00 in historical statistics). For example, identifying that order processing tasks are concentrated between 9:00 and 11:00 daily... The main payment gateway node consumes significant CPU resources. Analysis focuses on stability within a specific business cycle covered by the synchronization segment, identifying the fluctuation amplitude and frequency of key indicators within that cycle. Fluctuation amplitude refers to the difference between the lowest and highest points of key indicators in task records or periodic path data within that specific cycle (e.g., 9 AM to 11 AM). For example, if the CPU utilization of the main payment gateway node fluctuates from 20% to 80% between 9 AM and 11 AM on a trading day, the fluctuation amplitude within that business cycle is 60%. Fluctuation frequency refers to the number of times key indicators exceed a preset threshold or change direction within that specific cycle. For example, the CPU utilization exceeds the 70% threshold 5 times during this period.

[0070] To quantify this stability and identify a trend, the volatility trend value of the cycle path is calculated as follows: First, multiple indicators, such as volatility amplitude and volatility frequency, are normalized (e.g., mapped to the 0-1 range); then, based on the degree of influence of each indicator on system stability, weights are assigned using expert experience or machine learning models (e.g., volatility amplitude is weighted at 0.6, and volatility frequency at 0.4); finally, a weighted summation is performed to obtain the comprehensive volatility trend value of the business cycle path. For example, after normalization, the volatility trend value calculated in the above example is 0.72.

[0071] By calculating and recording this fluctuation trend value over multiple consecutive business cycles (e.g., the 9 AM to 11 AM synchronization period over several consecutive days), a time series can be formed, revealing whether the stability of resource usage under this periodic path is improving or deteriorating. The aggregated trend value of 0.5364 mentioned above is a global indicator derived by linearly weighting and summing the fluctuation trend values ​​of multiple key periodic paths (e.g., the 9-11 AM payment gateway path, the 2-4 PM inventory service path, etc.) according to their weights in relation to business importance (e.g., global aggregated trend value = payment gateway path trend value × 0.6 + inventory service path trend value × 0.4). This is used to macroscopically assess the overall system's resource scheduling health and drive higher-level optimization decisions, thus closely linking the macroscopic aggregated trend value with the microscopic periodic path analysis based on synchronization segments.

[0072] The resource usage surge identification submodule is based on the fluctuation trend value of the periodic path. According to the environmental load data in the order processing process, it selects the resource usage and system pressure records within the same period, compares the data change amplitude on a daily basis, and filters the time nodes with amplitude greater than the resource usage surge threshold to obtain the resource usage surge period.

[0073] Based on the fluctuation trend value of the periodic path, for example, the CPU utilization of the main payment gateway node fluctuates by 60% within a certain daily period, with a fluctuation frequency of 10 times. According to the environmental load data in the order processing process, the environmental load data includes factors such as the number of concurrent users, order submission rate, and interface response latency. For example, when analyzing the daily periodic fluctuation trend, it is related to the current moment when the number of concurrent users reaches its peak and the order submission rate increases. Resource usage and pressure records within the same period are selected. Resource usage records include CPU utilization, memory usage, and network bandwidth. Pressure records include the number of database connections, message queue length, and transaction processing latency. For example, within the same daily period, the CPU utilization record of the main payment gateway node and the database connection record are selected, and the data change magnitude is compared on a daily basis. For example, comparing today's peak CPU utilization of the main payment gateway node with the peak of the same period yesterday, today's peak is 80%, yesterday's is 60%, and the change magnitude is 20%. Time nodes with a change magnitude greater than the resource usage jump threshold are filtered. The threshold is set according to the resource utilization change and business capacity under the operating status to identify a significant increase in resource usage. The setting process is: collect the maximum daily change magnitude data of resource utilization over the past six months and calculate the average daily change magnitude. and standard deviation The threshold is set to the average value plus three times the standard deviation. For example, if the average daily change is 5% and the standard deviation is 2%, the threshold is set to 11%. When the daily change in CPU utilization exceeds 11%, it is considered a jump in resource usage. For example, if the change in CPU utilization is 20%, exceeding the 11% threshold, this time point is marked as a resource usage jump point, and the period of resource usage jump is obtained.

[0074] The environmental load quantification submodule identifies the intersection duration of path fluctuations and resource usage surges based on the resource usage surge period, performs weighted processing, and normalizes the usage duration of specific resources under differentiated cycles by referring to the frequency of fluctuations, the combined fluctuation value, and the cycle fluctuation, and outputs the results of resource and environmental collaborative processing benefit analysis. Differentiated cycles refer to time periods with different business characteristics and load patterns, such as business peak periods (e.g., 10:00-12:00 AM), business trough periods, or a complete 24-hour daily cycle. These cycles are treated differently because of their different system performance.

[0075] Based on the specific segment and its trend value output by the previous resource scheduling and planning module as the evaluation background, the system identifies the intersection duration of the resource usage surge period within that segment (e.g., during 10:00-11:00 on September 25th, the main payment gateway experiences a surge in CPU utilization; "resource usage surge" refers to the identified period of increased resource usage, such as the "duration" during which CPU utilization jumps from a normal level to a high level and remains there) and the identified path change period ("path change" refers to the time period during which the task flow path changes; for example, during the surge period, the payment task flow path switches from the main payment gateway to the backup payment gateway for 15 minutes). The system then performs a preliminary weighting process based on the degree of impact. The weighting factor is set according to the degree of impact of the intersection on overall performance and user experience; for example, if the user experience decreases during the intersection period, the weighting factor is increased. The factor is set to 1.5 otherwise to 1.0; it also combines the frequency of fluctuations (the number of times the path changes within a specific time period, for example, 3 path switches within one hour), the combined fluctuation value (the sum of the changes in resource utilization of key nodes on multiple paths, for example, the sum of the decrease in CPU utilization of the main payment gateway and the increase in CPU utilization of the backup gateway, where the key nodes are manually set) and the periodic fluctuation (specifically, the comprehensive fluctuation trend value of a specific business cycle calculated in the previous steps); in order to eliminate the differences in the dimensions and orders of magnitude of different indicators under differentiated cycles, the four core quantitative indicators, "intersection duration", "frequency of fluctuations", "combined fluctuation value" and "periodic fluctuation", after preliminary weight adjustment, are uniformly normalized. Then, the normalized indicator results are weighted and fused. "Indicator" refers to all input items that are weighted and fused when calculating the final result, and the output is the resource and environment collaborative processing benefit analysis result of the specific segment.

[0076] Please see Figure 6 The task scheduling optimization module includes:

[0077] The instruction filtering submodule filters and matches task scheduling instructions based on the results of the resource and environment collaborative processing benefit analysis, identifies the operating mode and platform status, compares the disturbance duration with the effective duration of the task scheduling instructions, and removes task scheduling instructions whose effective duration is greater than the disturbance duration, thus obtaining a set of task scheduling instructions affected by environmental interference.

[0078] Based on the results of the resource and environment collaborative processing benefit analysis, for example, the analysis results show that the resource and environment collaborative processing benefit score was 0.75 during the period from 10:00 to 11:00 on September 25th. This score serves as a benchmark input for evaluating the current system operation status. When the score is lower than a preset threshold (e.g., 0.8), it indicates that the current resource and environment collaboration is in a suboptimal state, triggering the need to filter subsequent (i.e., tasks to be executed after the current benefit score is generated) task scheduling instructions. Task scheduling instructions are operation instructions that adjust resource allocation and task priorities, such as "reduce the priority of non-core tasks" and "expand payment gateway instances". Combining the system status reflected by the score, the platform time period and platform status are identified. The platform time period refers to the operating mode within a specific time period, such as peak period and off-peak period, while the platform status refers to the health status. For example, "high load" and "normal operation". For example, if the current period (which corresponds to the aforementioned aggregated synchronization segment) is identified as the peak business period and the platform is in a high load operation state, the disturbance period and the instruction period are compared. The disturbance period refers to the time period affected by environmental interference. For example, a surge in environmental load leads to continuous high load operation for 30 minutes. The instruction effective period refers to the time period during which the task scheduling instruction is effective or lasts. For example, the expansion instruction effective period is 10 minutes. The comparison is made to determine whether the instruction can play a role within the disturbance period. Instructions with low matching degree are eliminated. For example, if the disturbance period is 5 minutes and the instruction effective period is 30 minutes, the instruction matching degree is low (e.g., the disturbance only lasts for 5 minutes, but the instruction takes 30 minutes to take effect). In the short term, the instruction cannot be used to effectively respond to sudden problems, so the matching degree is low. When the instruction cycle is less than or close to the disturbance cycle, it means that the instruction can intervene in time and alleviate the system pressure during the duration of the disturbance. At this time, the matching degree is high, and it belongs to the high-efficiency instruction that can be executed first. It cannot respond to short-term disturbances and obtains a set of task scheduling instructions affected by environmental interference. This set consists of effective instructions with a matching degree that have an instruction cycle less than the disturbance cycle.

[0079] The business classification submodule calls the set of task scheduling instructions affected by environmental interference, extracts the switching time and adjustment trigger frequency of the corresponding business unit of the task scheduling, arranges them in order of switching time, classifies the adjustment type by frequency, statistically analyzes the frequency band corresponding to the adjustment type and the platform operation mode within the frequency band, and obtains the platform adjustment frequency band distribution value.

[0080] The system invokes a set of task scheduling instructions affected by environmental interference. For example, the instruction set might include instructions like "expand payment gateway instance" or "switch to backup payment channel." It extracts the corresponding business unit switching time and adjustment trigger frequency. The business unit switching time refers to the point in time when the execution of the instruction causes a change in the business processing flow. For example, the "switch to backup payment channel" instruction switched at 10:00:30 on September 25th. The adjustment trigger frequency refers to the number of times a business unit or resource is adjusted by instructions within a specific time period. For example, the "payment gateway" was adjusted 5 times in the past hour. The instructions are arranged in order of switching time. For example, the 10:00:30 instruction is scheduled first, followed by the 10:00:45 instruction. The business units are categorized by frequency range based on the adjustment trigger frequency. For instance, an adjustment frequency of 0-2 times / hour is categorized as "low-frequency adjustment", 3-5 times / hour as "medium-frequency adjustment", and more than 5 times / hour as "high-frequency adjustment". The correspondence between frequency bands (frequency bands refer to the aforementioned low, medium, and high-frequency adjustment ranges) and operating cycle segments is statistically analyzed. For example, when the "payment gateway" business unit is found to be in the high-frequency adjustment range, it corresponds to the peak operating cycle segment, and the platform's adjustment frequency band distribution value is obtained.

[0081] The interference identification submodule adjusts the frequency band distribution value according to the platform, collects the load jump amplitude and duration within the frequency band, determines whether the jump amplitude threshold is exceeded, identifies the frequency band interference intensity, sorts the interference intensity of all frequency bands, determines the frequency band and jump time period of the interference intensity, and obtains the frequency of order processing task load interference impact.

[0082] Based on the platform's frequency band distribution values, for example, in high-frequency adjustment bands, the payment gateway business unit has a higher adjustment trigger frequency. The system collects the load fluctuation amplitude and duration within the frequency band. The load fluctuation amplitude refers to the instantaneous change in load indicators within the adjustment band. For example, in a high-frequency adjustment band, the payment gateway CPU utilization jumps from 30% to 80%, an amplitude of 50%. The duration refers to the time from the start of the load fluctuation to its return to normal. For example, if the CPU fluctuation lasts for 10 seconds, the system determines whether the fluctuation amplitude threshold is exceeded. The threshold is set based on the statistical distribution of load fluctuations and capacity during operation, identifying load fluctuations that threaten stability. The setting process involves collecting CPU utilization fluctuation amplitude data per minute over the past year and calculating the average of all fluctuation amplitudes. and standard deviation The threshold is set to the average value plus 2.5 times the standard deviation. For example, if the average jump amplitude is 5% and the standard deviation is 5%, then the threshold is set to 17.5%. When the CPU utilization jump amplitude exceeds 17.5%, it is considered interference that requires attention. For example, if the CPU jump amplitude is 50%, exceeding the 17.5% threshold, the frequency band interference intensity is identified. The frequency band interference intensity is a quantified index that combines the jump amplitude and the duration. For example, the interference intensity is quantified as the jump amplitude multiplied by... The interference intensity of frequency bands is sorted. For example, all identified frequency band interference intensity values ​​are sorted from high to low to determine the interference intensity frequency bands and jump periods. For example, the sorting results show that the "high frequency adjustment band" has the highest interference intensity during the period from 10:00:00 to 10:00:10 on September 25. Based on the determined key interference intensity frequency bands and jump periods, the number of times the order processing task was actually subjected to load jump impacts exceeding the threshold during this period is further counted to obtain the frequency of load interference impact on the order processing task.

[0083] The method for automating the entire online order processing process is based on the aforementioned automated online order processing system and includes the following steps:

[0084] S1: Based on the lifecycle characteristics of orders, analyze the changes in creation time and payment status, extract the peak mutation of inventory lock delay and the time period of order characteristic changes, identify the state switching point where payment status and delay changes are synchronized, mark the task level, and generate order status priority tag values.

[0085] S2: Based on the order status priority marker value, extract the flow change data and load change magnitude, compare the flow change and load fluctuation magnitude, analyze the correspondence between flow change and execution capability response, and obtain the task flow path optimization instruction group.

[0086] S3: Based on the task flow path optimization instruction group, extract the changes in resource utilization value and resource allocation path gradient before and after the response, analyze the decrease in resource allocation path and the increase in resource quantity of nodes, identify path decay paths and compare them with the original loop data, filter the segments with path decline characteristics, and generate resource scheduling optimization trends.

[0087] S4: Based on the resource scheduling optimization trend, identify the environmental load data within the corresponding time period, analyze the overlapping periods of resource occupancy surge and path decay path, extract the associated time periods, and obtain the results of resource and environmental collaborative processing benefit analysis.

[0088] S5: Based on the results of the resource and environment collaborative processing benefit analysis, extract the scheduling control strategy records within the time period, analyze the time interval and adjustment trigger frequency of business switching, filter high-frequency adjustment strategy fragments, and obtain the frequency of order processing task load interference.

[0089] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

[0090] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

Claims

1. A fully automated online order processing system, characterized in that, The system includes: The order status analysis module, based on the lifecycle characteristics of orders, analyzes the changing trends of order creation time, payment status data, and inventory lock time sequence to extract abrupt change nodes of order flow bottlenecks and generate time segments exceeding thresholds. The cross-platform collaborative adaptation module matches the cross-platform task flow parameters and operation data of the corresponding mutation node based on the time segment exceeding the threshold, and constructs the load linkage interval. It calculates the state difference normalization index of the flow point in the load linkage interval, thereby constructing the offset intensity index group. Based on the offset intensity index group, the task flow path optimization instruction group is obtained. The resource scheduling and planning module calls the task flow path optimization instruction group, extracts the changes in resource utilization before and after instruction execution, identifies the concentration of resource allocation under a unit task in combination with the order flow trend, extracts the fluctuation segment to form the resource allocation path feature segment, and calculates the aggregation trend value of the resource allocation path feature segment. The environmental load assessment module collects environmental load data during order processing for characteristic sections of the resource allocation path, analyzes the overlap duration of pressure changes and resource occupancy surges, and outputs the results of resource and environmental synergy analysis for characteristic sections of the resource allocation path.

2. The fully automated online order processing system according to claim 1, characterized in that, The order status parsing module includes: The time series extraction submodule extracts the time series of order creation time based on the life cycle characteristics of orders, calculates the difference in the number of orders at adjacent time points, extracts the peak-to-trough distance of the difference in the number of orders within each period, and filters out the segments where the peak-to-trough distance exceeds the threshold to obtain the time period fluctuation range. The payment status analysis submodule calls the time period fluctuation range, identifies the corresponding payment status data, analyzes the absolute change range of payment in adjacent time periods, compares it with the set payment status threshold, takes the latest time point in the time period corresponding to the absolute change range of payment exceeding the payment status threshold as the mutation node, establishes a time interval to cover the mutation node, and takes this time interval as the payment status mutation range. The status priority marking submodule identifies the corresponding inventory lock delay and order characteristic data based on the payment status change range. Combining the payment status and creation time fluctuations, it calculates the status complexity offset value of the change node, sets an offset threshold, marks the change node that exceeds the offset threshold, and backtracks a preset time window based on the change node that exceeds the offset threshold to obtain the time segment exceeding the threshold. The priority of the time segment exceeding the threshold is marked as high priority.

3. The fully automated online order processing system according to claim 1, characterized in that, The cross-platform collaborative adaptation module includes: The data flow matching submodule samples at preset intervals based on time segments exceeding thresholds, extracts system task flow parameters and operation data from adjacent sampling points, calculates the rate of change of task flow parameters and operation data from adjacent sampling points, and if the rate of change of task flow parameters and operation data exceeds a preset rate of change threshold, the corresponding sampling point is designated as a load mutation node; if the time period consisting of multiple consecutive load mutation nodes overlaps with the timestamp of order creation or the difference is less than a preset time threshold, the time period consisting of these multiple consecutive load mutation nodes is designated as a load linkage interval. The flow deviation judgment submodule calls the load linkage interval and takes the load change node in the load linkage interval as the flow point; calculates the state difference normalization index G for each flow point; if G is greater than or equal to the preset offset boundary, then G is taken as the deviation intensity of the corresponding flow point; if the deviation intensity is greater than or equal to the preset benchmark judgment value, the corresponding flow point is marked as deviation, and an offset intensity index group is established based on the position of the flow point marked as deviation. The task flow instruction extraction submodule, based on the offset intensity index group, filters the position of the corresponding time point in the task instruction set, extracts the corresponding task instructions and task instruction values, sorts the task instructions according to the time sequence, removes duplicate task instructions, and obtains the task flow path optimization instruction group.

4. The fully automated online order processing system according to claim 3, characterized in that, The resource scheduling and planning module includes: The resource utilization extraction submodule extracts resource utilization data before and after the execution of task instructions based on the task flow path optimization instruction group; obtains resource utilization data in multiple cycles before and after the execution of task instructions, identifies resource boundary conditions, and obtains the resource utilization difference value corresponding to the task instructions based on the resource boundary conditions. The resource allocation path identification submodule calls the resource utilization difference value. If the resource utilization difference value is greater than a preset significance threshold, it is determined that the resource status has changed substantially. Then, it identifies the resource allocation path change trajectory in the unit task path, calculates the resource quantity fluctuation corresponding to the change in the unit task path trajectory, compares the path decrease with the resource quantity fluctuation, and calculates the path fluctuation coupling degree index. It filters out the path fluctuation coupling degree index that is greater than or equal to the preset index, and filters out the segments where the path decrease and the resource quantity fluctuation change direction are consistent and the deviation of their change rate is within a preset range. These segments are defined as the resource allocation path feature segment. The allocation response interval identification submodule samples the characteristic segment of the resource allocation path and calculates the aggregate trend value Q of the characteristic segment of the resource allocation path using the following formula: ； in, The first representing the path displacement The change value of each sampling point This represents the average value of the path displacement change at the sampling points. Representing the The time interval between sampling points Represents the total time interval. It is a weighting factor, and N represents the total number of sampling points.

5. The fully automated online order processing system according to claim 4, characterized in that, The environmental load assessment module includes: The path trend extraction submodule extracts node task records and periodic path data from the resource allocation path feature segments. The periodic path data refers to data related to resource usage or path flow within a specific business cycle. The specific business cycle is a segment corresponding to the resource allocation path feature segments and having specific attributes, which are attributes manually set according to requirements. The node task records and periodic path data are normalized, and the normalized values ​​are weighted and summed to obtain the comprehensive fluctuation trend value for the specific business cycle. For a single business cycle, the aggregated trend values ​​corresponding to multiple resource allocation path feature segments within the single business cycle are summed to obtain a global aggregated trend value; the overall resource scheduling health of the system is evaluated based on the global aggregated trend value. The resource occupancy surge identification submodule identifies resource occupancy surge periods within characteristic segments of the resource allocation path using the following method: Based on the characteristic segments of the resource allocation path, and according to the environmental load data in the order processing process, resource occupancy and pressure records within the same period are selected, the data change magnitude is compared, and time nodes with magnitudes greater than the resource occupancy jump threshold are filtered to obtain the resource occupancy jump period. The environmental load quantification submodule identifies the intersection duration of the path amplitude change period and the resource occupancy surge period within the resource occupancy surge period. Based on the degree of impact of the intersection duration on the system, it performs preliminary weighting processing on the intersection duration. Then, it normalizes the intersection duration, the frequency of path amplitude change, the merged amplitude value, and the comprehensive fluctuation trend value of a specific business cycle after the preliminary weighting processing, and performs weighted fusion on the normalized values. The fusion result serves as the resource and environmental collaborative processing benefit analysis result for the resource allocation path characteristic segment.

6. The fully automated online order processing system according to claim 5, characterized in that, The system also includes a task scheduling optimization module: The task scheduling optimization module, based on the results of the resource and environment collaborative processing benefit analysis, filters task scheduling instructions affected by environmental interference, classifies platform switching time and adjustment trigger frequency, identifies time periods when load fluctuations exceed preset thresholds, and obtains the frequency of order processing task load interference. The frequency of order processing task load interference includes system switching frequency, number of adjustment command triggers, and number of times the jump time period exceeds the limit.

7. The fully automated online order processing system according to claim 6, characterized in that, The task scheduling optimization module includes: The instruction filtering submodule, based on the results of the resource and environment collaborative processing benefit analysis, identifies the platform operation mode and platform status of resource allocation path characteristic segments whose scores are lower than the preset analysis threshold, compares the disturbance duration with the effective duration of task scheduling instructions, and obtains the set of task scheduling instructions affected by environmental interference. The business classification submodule calls the set of task scheduling instructions affected by environmental interference, extracts the business unit switching time and adjustment trigger frequency corresponding to the task scheduling instructions, arranges them in order of switching time, classifies the adjustment type by frequency, and statistically analyzes the frequency band corresponding to the adjustment type and the platform operation mode within the frequency band to obtain the platform adjustment frequency band distribution value. The interference identification submodule adjusts the frequency band distribution value according to the platform, collects the load jump amplitude and duration within the frequency band, determines whether the jump amplitude threshold is exceeded, thereby identifying the frequency band interference intensity, sorting the interference intensity of all frequency bands, determining the frequency band and jump time period of the interference intensity, and obtaining the frequency of order processing task load interference impact.

8. A method for automating the entire online order processing process, characterized in that, The method is used to implement the online order full-process automated processing system according to any one of claims 1-5, and includes the following steps: S1: Based on the lifecycle characteristics of orders, by analyzing the changing trends of order creation time, payment status data, and inventory lock-in time sequence, extract the abrupt change nodes of order flow bottlenecks and generate time segments exceeding the threshold. S2: Based on the time interval exceeding the threshold, match the cross-platform task flow parameters and operation data of the corresponding mutation node, construct the load linkage interval, calculate the state difference normalization index of the flow point in the load linkage interval, construct the offset intensity index group, and obtain the task flow path optimization instruction group based on the offset intensity index group. S3: Call the task flow path optimization instruction group, extract the resource utilization rate changes before and after instruction execution, identify the resource allocation concentration under unit task in combination with order flow trend, extract the fluctuation segment to form the resource allocation path feature segment, and calculate the aggregation trend value of the resource allocation path feature segment. S4: For characteristic segments of the resource allocation path, collect environmental load data during the order processing process, analyze the overlap duration of pressure changes and resource occupation surge periods, and output the resource and environmental collaborative processing benefit analysis results for characteristic segments of the resource allocation path.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor executes the computer program, it implements the online order full-process automated processing method as described in claim 8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the online order full-process automated processing method as described in claim 8.

Images