Packet data scheduling optimization system and method for pay-per-value

Abstract

The application discloses a packet data scheduling optimization system and method for value-based payment, and belongs to the technical field of data processing. The system comprises a data resource evaluation module, a data resource initial allocation module, a data resource scheduling module and a value-based payment processing module. The data resource evaluation module is used for obtaining resource demand of each data group. The data resource initial allocation module is used for obtaining a first data resource set. The data resource scheduling module is used for obtaining an expected decision instruction. If the expected decision instruction is normal processing, then the secondary scheduling data group continues to wait for processing according to a preset rule. Otherwise, the data resource secondary scheduling processing is performed to obtain a second data resource set. The secondary scheduling data group is a residual data group when the data resource initial allocation is performed. The value-based payment processing module is used for obtaining an audit result data after settlement. The application solves the problem that in the prior art, due to the static resource mode, resources are scattered and cut, and when a high complexity task appears, there is a lack of continuous resource bearing, thereby causing low overall data processing efficiency.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a system and method for optimizing packet data scheduling for pay-as-you-go billing. Background Technology

[0002] Existing grouped data scheduling systems trigger data interaction between hospital information systems and medical insurance information systems at fixed times, causing billing data to be uploaded at fixed points in time and triggering subsequent processing. Alternatively, they can be task-driven, where when a user or administrator issues a request, the scheduling engine retrieves the corresponding billing or reimbursement data from the database based on the task's priority and execution rules, and returns the processing result to the interaction end. Or, based on preset matching rules, medical insurance bills and hospital bills are compared one by one. If the rules are met, accounts receivable data is directly generated; otherwise, discrepancy data is generated for subsequent manual intervention.

[0003] For example, the reconciliation system and corresponding computer equipment disclosed in Chinese invention patent CN112508710B include: a scheduling microservice for managing the scheduling of timed tasks of the reconciliation system; an HIS microservice for downloading HIS bills from the hospital information system in response to calls from the scheduling microservice; a medical insurance microservice for parsing the audited medical insurance bills in response to calls from the scheduling microservice, and performing real-time reconciliation of the medical insurance bills and HIS bills according to preset reconciliation data matching rules; in response to successful reconciliation, forming the corresponding data into accounts receivable in finance; and in response to failed reconciliation, forming the corresponding data into discrepancy data.

[0004] For example, Chinese invention patent CN101694712B discloses a billing generation system and method for cluster job scheduling, which includes: a job scheduling monitoring module monitoring and recording the operational indicator data of each job, and simultaneously recording the object attribute information corresponding to the job, and then sending the recorded information to a data processing module. The data processing module receives the information from the job scheduling monitoring module, processes the operational indicator data of each job, and stores it in a database. An interaction module receives a request from the administrator and sends it to an accounting engine. The accounting engine receives the request from the administrator, reads the data that meets the request from the database, analyzes and processes the data according to the request, generates a bill, and then feeds the bill back to the interaction module for presentation to the administrator.

[0005] However, in the process of implementing the inventive technical solution in the embodiments of this application, it was found that the above-mentioned technology has at least the following technical problems: Existing technologies primarily focus on task triggering and bill comparison, such as using timed triggering to achieve fixed-cycle data interaction or rule-driven matching and difference generation of bill data. However, these methods only solve the problem of task initiation and comparison logic, lacking real-time allocation and dynamic optimization of computing and storage resources. In long-term operation, systems generally adopt static, pre-allocated resource allocation strategies, resulting in fragmented and dispersed resources that are difficult to flexibly integrate according to task requirements, leading to low overall data processing efficiency. Summary of the Invention

[0006] To address the technical problem of low overall data processing efficiency caused by the lack of continuous resources when high-complexity tasks occur in existing technologies, embodiments of the present invention provide a grouped data scheduling optimization system and method for value-based payment. The technical solution is as follows: On one hand, a data scheduling optimization system for pay-as-you-go data is provided. This system includes: a data resource assessment module, used to acquire user-uploaded datasets, manage them into groups, obtain data complexity parameters for each data group, analyze the data complexity value of each data group, and thus obtain the resource requirements of each data group. The data complexity value is used to characterize the overall processing difficulty of the data group. On the other hand, a data resource initial allocation module is used to determine the data review order based on the resource requirements analysis of each data group, simultaneously acquire idle resource processing units, and sort them in descending order based on the resource quantity (remaining idle resource quantity) of each idle resource processing unit to obtain the resource processing unit arrangement order, thereby enabling data allocation. The initial allocation of data resources yields a first data resource set. The data resource scheduling module acquires the audit execution parameters of each idle resource processing unit, analyzes them to obtain the expected processing time of each idle resource processing unit, and thus obtains the expected judgment instruction. If the expected judgment instruction is for normal processing, the secondary scheduling data group continues to wait for processing according to preset rules; otherwise, secondary data resource scheduling is performed to obtain a second data resource set. The secondary scheduling data group consists of the remaining data group from the initial data resource allocation. The pay-as-you-go processing module processes the data resource set using pay-as-you-go methods to obtain the settled audit result data. The data resource set includes the first and second data resource sets.

[0007] On the other hand, a grouped data scheduling optimization method for pay-as-you-go pricing is provided. This method includes: acquiring user-uploaded datasets and managing them into groups; acquiring data complexity parameters for each data group and analyzing their data complexity values ​​to obtain resource requirements for each group; the data complexity value characterizes the overall processing difficulty of the data group; based on the resource requirements analysis of each data group, determining the data review order; simultaneously acquiring idle resource processing units and sorting them in descending order based on resource quantity to obtain the resource processing unit arrangement order; and performing initial data resource allocation to obtain a first data resource set; acquiring review execution parameters for each idle resource processing unit and analyzing their expected processing time to obtain the expected judgment instruction; if the expected judgment instruction is for normal processing, then the secondary scheduling data group continues to wait for processing according to preset rules; otherwise, secondary data resource scheduling is performed to obtain a second data resource set, which consists of the remaining data groups from the initial data resource allocation; and performing pay-as-you-go processing on the data resource set to obtain the settled review result data. The data resource set includes the first and second data resource sets. The beneficial effects of the technical solutions provided by the embodiments of the present invention include at least the following: 1. The packet data scheduling optimization system for pay-as-you-go pricing provided by this invention performs complexity analysis on uploaded packet data and calculates the complexity value and resource requirements of the packet data accordingly. This quantifies the processing load of each packet data item and forms a resource requirement list, thereby realizing dynamic resource aggregation and priority-oriented scheduling based on quantified requirements. This ensures that when high-complexity packet data arrives, continuous resource blocks can be aggregated and executed on demand, significantly alleviating the resource fragmentation problem caused by static pre-allocation and improving the overall data processing efficiency. It effectively solves the problem in the prior art where static resource methods cause resources to be scattered and fragmented, resulting in a lack of continuous resources to support high-complexity tasks and thus low overall data processing efficiency.

[0008] 2. This invention determines the grouped data according to resource demand and preset threshold and sorts them by priority. At the same time, it sorts and matches available resource blocks according to resource capacity and continuity, thereby allocating resource blocks to priority grouped data and ordinary grouped data in sequence. This enables critical grouped data to obtain sufficient and continuous resource guarantees first, reducing the waiting and preemption costs of critical tasks.

[0009] 3. By evaluating the execution capabilities of available resource blocks, the expected processing time of the resource blocks is obtained. Then, the expected processing time is compared with the corresponding processing time threshold to obtain the expected judgment instruction. This enables more reasonable execution of secondary scheduling of data resources, thereby reducing the interference of long-tail tasks and improving the overall processing speed of the data group.

[0010] 4. By monitoring the actual occupancy of resource blocks in real time after the initial allocation and marking and refilling the remaining resource blocks after allocation, the system can process very small-scale or newly emerging grouped data, thereby realizing the secondary utilization of resource fragments, improving resource utilization and shortening the processing latency of small-batch urgent tasks. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A schematic diagram of the structure of a packet data scheduling optimization system for value-based billing provided in an embodiment of this application; Figure 2 A flowchart of a packet data scheduling optimization method for value-based billing provided in an embodiment of this application; Figure 3 The upper half of a schematic diagram of data task management for a value-based packet data scheduling optimization system provided in an embodiment of this application; Figure 4 The lower half of the schematic diagram of data task management for a value-based data scheduling optimization system provided in this application embodiment. Detailed Implementation

[0013] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0014] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0015] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0016] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0017] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0018] like Figure 1 The diagram shown is a structural schematic of a data scheduling optimization system for pay-as-you-go billing provided in this application embodiment. The system includes: a data resource assessment module, used to acquire user-uploaded datasets, manage them into groups to obtain data groups, acquire data complexity parameters for each data group, analyze the data complexity value of each data group, and thus obtain the resource requirements of each data group. The data complexity value is used to characterize the overall processing difficulty of the data group. A data resource initial allocation module, used to determine the data review order based on the resource requirements analysis of each data group, synchronously acquire idle resource processing units and sort them in descending order based on resource quantity to obtain the resource processing unit arrangement order, thereby performing initial data resource allocation. The system obtains a first data resource set. A data resource scheduling module acquires the audit execution parameters of each idle resource processing unit, analyzes them to obtain the expected processing time of each idle resource processing unit, and thus obtains the expected judgment instruction. If the expected judgment instruction is for normal processing, the secondary scheduling data group continues to wait for processing according to preset rules (the current order of arrangement); otherwise, secondary data resource scheduling is performed to obtain a second data resource set. The secondary scheduling data group consists of the remaining data group from the initial data resource allocation. A pay-as-you-go processing module processes the data resource set using pay-as-you-go methods to obtain the settled audit result data. The data resource set includes the first data resource set and the second data resource set.

[0019] In this embodiment, the solution addresses the issues of large data volume, complex grouping, and uneven resource allocation in the medical reimbursement system for social security reimbursement, aiming to improve review efficiency and accuracy. The resource processing unit refers to the data processing capacity that can process a specific amount of data. For example, the available memory. By analyzing the data complexity and resource requirements of each data group, these resources can be reasonably allocated to ensure that priority data groups can be processed quickly, while maximizing overall processing efficiency and avoiding resource waste or excessive waiting time.

[0020] It should also be noted that the value-based payment method is specifically a disease-based value payment method, which is a reimbursement method for medical insurance. Specifically, the medical insurance department will set a relatively fixed payment standard based on clinical pathways, disease-related groups (DRGs) or disease-specific values ​​(DIPs). When medical institutions treat patients, regardless of the actual cost, as long as the disease is classified as a certain disease, the medical insurance will settle the payment according to the payment amount corresponding to the preset disease-specific value.

[0021] The resource requirements for each data group are obtained as follows: The data complexity of each data group is analyzed separately. Historical data groups belonging to the same data group type (e.g., all belonging to the lower leg fracture data group) are obtained for the same period in previous years. Historical reference data groups are obtained, and their resource requirements are compared with these historical resource requirements to obtain historical resource requirement differences. A preset resource requirement threshold in the database is obtained and compared with the historical resource requirement differences. Historical reference data groups with historical resource requirement differences less than the threshold are marked as historical control data groups. The average and median resource requirements of the historical control data groups are obtained and averaged to obtain the average processing value of the historical control data groups. This average processing value is used as the resource requirement for that data group, thus obtaining the resource requirements for each data group.

[0022] Furthermore, the data complexity value of each data group is obtained. The specific method is as follows: obtain the data complexity parameters of each data group, which include the total data volume, the number of item categories, the number of field dimensions, and the number of record entries; obtain the preset data complexity comparison benchmark set in the database, and perform comparison and normalization processing with the data complexity parameters to obtain the comparison and normalization processing value; based on the comparison and normalization processing value, introduce the corresponding weighting factor for coupling processing to obtain the data complexity value of each data group; the data complexity comparison benchmark set includes the benchmark value of the total data volume, the benchmark value of the number of item categories, the benchmark value of the number of field dimensions, and the benchmark value of the number of record entries (where the benchmark value is the best historical value stored in the database).

[0023] In this embodiment, as Figure 2 As shown, Figure 2 This is a flowchart illustrating a grouped data scheduling optimization method for value-based payment provided in this application embodiment. The system acquires user-uploaded datasets and groups them to obtain data groups. Then, it acquires the data complexity parameters of each data group and calculates the resource requirements of each group based on these parameters, characterizing the overall processing difficulty of each data group. Simultaneously, the system acquires idle resource processing units and sorts them according to their resource quantity, obtaining the arrangement order of the resource processing units. A data review order is generated based on the resource requirements of each data group. Combined with the arrangement order of the resource processing units, initial allocation of data resources is performed, resulting in a first data resource set. Next, the system analyzes the expected processing time of each idle resource processing unit and generates an expected judgment instruction based on this processing time. If the judgment result is to continue waiting, the secondary scheduling data group remains in a waiting state; if the judgment result is that secondary scheduling is required, secondary data resource scheduling processing is performed, resulting in a second data resource set. Finally, the first and second data resource sets are used together as input for value-based payment processing to generate the settled review result data.

[0024] Total data capacity refers to the overall storage capacity occupied by the data set, used to measure the input / output overhead. The number of project categories refers to the number of business project types involved in the data set. The number of field dimensions refers to the number of fields contained in each record, such as the charge item field, material used field, and expense category field. The more fields, the higher the complexity of parsing and validation. The number of record entries refers to the number of original data records contained in the data set; records can be charge items, reimbursement units, or other business data units. Total data capacity, number of project categories, number of field dimensions, and number of record entries can be obtained by statistically analyzing the uploaded dataset using unified data analysis and management tools, such as SQL query tools.

[0025] The specific method for obtaining the data complexity value for each data set is as follows: ; In the formula, SFi represents the data complexity value of the i-th data group, and i represents the data group number, i=1,2,...,i max i max δ1 represents the total number of data groups, Mi represents the total data capacity of the i-th data group, FM represents the baseline value of the total data capacity, Li represents the number of item categories in the i-th data group, FL represents the baseline value of the number of item categories, Zi represents the number of field dimensions in the i-th data group, FZ represents the baseline value of the number of field dimensions, Ti represents the number of record entries in the i-th data group, FT represents the baseline value of the number of record entries, δ1 represents the weighting factor for the total data capacity, δ2 represents the weighting factor for the number of item categories, δ3 represents the weighting factor for the number of field dimensions, and δ4 represents the weighting factor for the number of record entries.

[0026] The weighting factors for total data capacity, number of project categories, number of field dimensions, and number of record entries can be obtained from the database. For example, the total data capacity weighting factor can be obtained from the database by retrieving the total data capacity dataset of historical data groups belonging to the same data category (specifically, all data within the past three years). The total data capacity of each data group stored in the database is then retrieved, and the difference between this dataset and the total data capacity is calculated to obtain the difference value for each total data capacity. A preset threshold for this difference value is then retrieved from the database and compared with each of the total data capacity differences. The data with a difference value less than the threshold value is assigned the total data capacity weighting factor. The total capacity of historical reference data is used as the basis for obtaining the total capacity of each historical reference data. Then, the weighting factors of each total capacity corresponding to the total capacity of each historical reference data in the database are obtained. The mode and median of each weighting factor are selected and averaged to obtain the processed value of the weighting factor of the total capacity. This processed value of the weighting factor of the total capacity is used as the weighting factor of the total capacity. Other weighting factors, such as the weighting factor of the number of project categories, the weighting factor of the number of field dimensions, and the weighting factor of the number of record entries, are obtained in a similar way to the weighting factor of the total capacity. They can all be obtained by obtaining the mode and median of the historical reference data and averaging them.

[0027] The data complexity values ​​for each data group are obtained by analyzing data complexity parameters, including total data volume, number of item categories, number of field dimensions, and number of record entries. This analysis takes into account the interrelationships between these parameters. For example, total data volume, number of record entries, number of field dimensions, and number of item categories are interconnected and jointly determine the overall complexity of the data group. Generally, the more record entries and the higher the number of field dimensions, the greater the total data volume. Simultaneously, an increase in the number of item categories increases the logical complexity of data processing, as different categories may correspond to different processing rules or calculation processes. An increase in the number of field dimensions not only increases storage requirements but also increases the computational load for processing each record, thus affecting both the total data volume and resource requirements.

[0028] By analyzing the data complexity value of each data group, it is easier to allocate resources based on this data complexity value. The larger the data complexity value, the more resources are required to process the data in that group. Therefore, more data processing resources need to be allocated to process the data better.

[0029] Furthermore, the data review order is obtained through the following method: A preset resource requirement threshold is retrieved from the database. Each data group has its own corresponding resource requirement threshold. Specifically, the resource requirement threshold is the average of the resource requirements of the same data group (e.g., all data related to lower leg fractures) from the same historical period in the database. This averaged value yields the historical resource requirement for the same period. This historical resource requirement is then increased by a preset percentage (e.g., 10%) to obtain the resource requirement threshold (i.e., increasing the historical resource requirement from 100% to 110%). This threshold is then compared with the resource requirements of each data group. If a data group's resource requirement exceeds the threshold, it is marked as a priority data group; otherwise, it is marked as a regular data group. The priority data groups are then sorted in descending order of resource requirement to obtain their ranking. The regular data groups are then arranged according to their data upload order to obtain their ranking. Finally, the rankings of the priority and regular data groups are combined to obtain the data review order. In this data review order, the overall ranking of the priority data groups takes precedence over the overall ranking of the regular data groups.

[0030] In this embodiment, by dividing data groups into priority data groups and ordinary data groups according to resource demand, it is possible to distinguish between data groups with high and low resource consumption during processing, thereby enabling reasonable allocation of processing order during scheduling. If no distinction is made and processing is based solely on upload time order, data groups with high resource demand may be queued at the front, but there may not be enough resource processing units available for subsequent use, forcing them to wait and reducing overall processing efficiency. By sorting priority data groups in descending order of resource demand, it is ensured that data groups with high resource consumption are processed first when suitable processing units are available, avoiding long-term backlogs. At the same time, arranging ordinary data groups in the order of upload ensures that small-scale resource requests are processed quickly, without long delays due to the presence of priority data groups, thereby improving overall scheduling efficiency. This effectively avoids the unreasonable waiting phenomenon caused by relying solely on time order, thus optimizing resource scheduling and accelerating overall processing speed in the expense reimbursement process.

[0031] Furthermore, the initial allocation of data resources is carried out. The specific method is as follows: based on the data review order and the arrangement order of resource processing units, the idle resource processing units are allocated to each data group in sequence according to the preset allocation rules. If a data group is not allocated an idle resource processing unit, the data group is marked as a data group to be allocated. If an idle resource processing unit is not allocated, the idle resource processing unit is marked as a resource processing unit to be allocated.

[0032] In this embodiment, by initially allocating data resources based on the data review order and the resource processing unit arrangement order, the correspondence between resources and data groups can be established in the first instance, thereby ensuring the systematicness and orderliness of the overall allocation process. During the initial allocation process, data groups that have not been allocated to resource processing units are marked to prevent them from being missed and to ensure that they can be rescheduled later; unallocated idle resource processing units are marked to ensure that remaining resources are not idle or wasted. This achieves the goal of completing most of the resource allocation in the initial stage, improving overall processing efficiency, and providing a clear basis for subsequent secondary allocation, ensuring the maximization of resource utilization, and thus realizing the integrity and continuity of the resource scheduling process.

[0033] Furthermore, the preset allocation rules are as follows: Idle resource processing units are allocated based on resource demand. If an idle resource processing unit is allocated to a priority data group, its remaining idle resources are obtained, split, and divided into unallocated resource processing units, which are then used in subsequent allocations. Resources are allocated to each data group based on the data review order. Each unallocated resource processing unit within the resource processing unit arrangement is traversed and allocated accordingly. If, after resource processing unit allocation, the remaining unallocated resource processing unit has less remaining resources than the resource demand of any priority data group, this unallocated resource processing unit is marked as a regular resource processing unit, and the remaining unallocated resource processing units are counted and marked as regular resource processing units. Based on resource demand analysis, regular resource processing units are allocated to each regular data group.

[0034] In this embodiment, idle resource processing units are allocated to priority data groups based on resource demand, ensuring that data groups with high resource demands receive sufficient resource support immediately. This avoids prolonged waiting or task delays due to insufficient resources, thereby improving overall processing efficiency. During the allocation process, remaining idle resources from resource processing units are split and redistributed to subsequent allocations, achieving full utilization of resources and reducing waste. Two different allocation mechanisms are employed: firstly, priority data groups are allocated based on demand, ensuring continuous and stable resource support for large-scale tasks; secondly, ordinary data groups are allocated based on review order, ensuring fairness and timeliness for small-scale tasks. This approach prioritizes critical tasks while maintaining overall system balance and resource utilization, thus achieving accelerated processing and rational resource scheduling during grouped data scheduling.

[0035] Furthermore, the general resource processing units are allocated to each general data group. Specifically, resources are allocated to each general data group based on their order of arrangement. The remaining resources of each general resource processing unit are retrieved and compared with the resource requirements of the general data group. If the remaining resources of a general resource processing unit are greater than or closest to the resource requirements of the general data group, then the general resource processing unit is allocated to that general data group. If a general data group is not allocated a general resource processing unit, then that general data group is marked as an unprocessed general data group. Based on the order of resource processing units, each ordinary resource processing unit is traversed to check if there are any idle resources in each ordinary resource processing unit. If there are, the idle resource quantity is retrieved and compared with the resource demand of the unprocessed ordinary data group. The idle resource is then allocated to each unprocessed ordinary data group. If there are still unprocessed ordinary data groups that have not been allocated to an ordinary resource processing unit, the unprocessed ordinary data group is marked as a secondary scheduling data group and the secondary scheduling data group is pre-queued. If there are no unprocessed ordinary data groups that have not been allocated to an ordinary resource processing unit, the first data resource set is obtained.

[0036] In this embodiment, by matching the order of ordinary data groups with the remaining resources of ordinary resource processing units, ordinary data groups can preferentially obtain resource processing units that are closest to their needs, thereby effectively reducing resource waste and improving allocation accuracy. Ordinary data groups that fail to obtain resource processing units in the initial allocation are marked as unprocessed ordinary data groups, and a secondary scheduling mechanism is further introduced. This mechanism, while traversing ordinary resource processing units, calls upon their idle resources in real time and compares them with the needs of unprocessed ordinary data groups, maximizing the utilization of remaining scattered resources in the system. If there are still ordinary data groups that cannot be allocated, they are marked as secondary scheduling data groups, and their queuing positions are dynamically adjusted according to the expected working time of the resource processing units: when the remaining working time of a resource processing unit exceeds a threshold, the secondary scheduling data group will no longer wait under that unit, but will be readjusted to other available resource processing units for queuing. This avoids unreasonable waiting caused by ordinary data groups being occupied by resource processing units for extended periods. If there are no unprocessed ordinary data groups that have not been assigned to ordinary resource processing units, it means that the ordinary data group can be processed normally by the resource processing unit. After processing, the first data resource set can be obtained. It should be noted that the data in the first data resource set, such as details of reimbursable expenses, is used for subsequent medical insurance reimbursement through value-based payment analysis.

[0037] Furthermore, the expected processing time of each idle resource processing unit is obtained. Specifically, the following methods are used: 1) Obtain the audit execution parameters for each idle resource processing unit, including instruction-level parallelism, instruction cache bandwidth, non-blocking memory request concurrency, and task context switching latency; 2) Obtain the preset audit execution standard set from the database and perform proportional normalization and inverse normalization processing with the audit execution parameters to obtain the normalized processing ratio and inverse normalization ratio; 3) Based on the normalized processing ratio and inverse normalization ratio, introduce corresponding weighting factors for coupling processing to obtain the comprehensive performance value of each idle resource processing unit; 4) Based on the data complexity value of each data group and the database processing... Row matching is performed to obtain the expected processing time for each data group; the comprehensive performance value of each idle resource processing unit is matched with the database to obtain the processing time correction coefficient for each idle resource processing unit; multiplicative coupling analysis is performed based on the expected processing time of each data group and the processing time correction coefficient of each idle resource processing unit to obtain the expected processing time of each idle resource processing unit; the comprehensive performance value of each idle resource processing unit is used to represent the data processing execution capability of the idle resource processing unit; the audit execution standard set includes instruction-level parallelism standard value, instruction cache bandwidth standard value, non-blocking memory request concurrency standard value, and task context switching latency standard value.

[0038] In this embodiment, instruction-level parallelism, instruction cache bandwidth, non-blocking memory request concurrency, and task context switching latency can all be uniformly obtained by reading the scheduling log of the resource processing unit itself.

[0039] The comprehensive performance value of each idle resource processing unit is obtained using the following method: ; In the formula, XNj represents the overall performance value of the j-th idle resource processing unit, and j represents the number of the idle resource processing unit, j=1,2,...,j max j max XBj represents the total number of idle resource processing units, NB represents the instruction-level parallelism of the j-th idle resource processing unit, XKj represents the instruction cache bandwidth of the j-th idle resource processing unit, NK represents the instruction cache bandwidth standard value, XMj represents the non-blocking memory request concurrency of the j-th idle resource processing unit, NM represents the non-blocking memory request concurrency standard value, XYj represents the task context switching latency of the j-th idle resource processing unit, NY represents the task context switching latency standard value, σ1 represents the instruction-level parallelism weighting factor, σ2 represents the instruction cache bandwidth weighting factor, σ3 represents the non-blocking memory request concurrency weighting factor, and σ4 represents the task context switching latency weighting factor.

[0040] Instruction-level parallelism weighting factors, instruction cache bandwidth weighting factors, non-blocking memory request concurrency weighting factors, and task context switching latency weighting factors can be obtained from a database. For example, the instruction-level parallelism weighting factor can be obtained by retrieving the instruction-level parallelism values ​​stored in the database, performing difference processing on each instruction-level parallelism value, obtaining a preset instruction-level parallelism difference threshold in the database, and comparing it with each instruction-level parallelism difference value. Instruction-level parallelism values ​​with a difference value less than the instruction-level parallelism difference threshold are used as historical reference instruction-level parallelism values. This process yields the historical reference values. For instruction-level parallelism, obtain the instruction-level parallelism weighting factors corresponding to each historical reference instruction-level parallelism in the database. Select the mode and median of each instruction-level parallelism weighting factor, and then average them to obtain the processed value of the instruction-level parallelism weighting factor. This processed value of the instruction-level parallelism weighting factor is used as the instruction-level parallelism weighting factor. Other weighting factors, such as instruction cache bandwidth weighting factor, non-blocking memory request concurrency weighting factor, and task context switching latency weighting factor, are obtained in a similar way to the instruction-level parallelism weighting factor. They can all be obtained by obtaining the mode and median of historical reference data and averaging them.

[0041] The comprehensive performance value of each idle resource processing unit is obtained by analyzing the execution parameters, including instruction-level parallelism, instruction cache bandwidth, non-blocking memory request concurrency, and task context switching latency. This takes into account the interrelationships between these parameters. For example, higher instruction-level parallelism increases the number of instructions that can be processed simultaneously, but if the instruction cache bandwidth is insufficient, the parallelism cannot be fully utilized, leading to potential instruction execution stalls. Higher non-blocking memory request concurrency increases the number of memory access requests that can be initiated simultaneously, but due to limitations in instruction cache bandwidth and context switching efficiency, memory waits or cache conflicts may occur, thereby prolonging task execution latency. Longer task context switching latency increases system overhead during multi-task or multi-threaded execution, affecting the effective utilization of instruction parallelism and reducing the throughput efficiency of non-blocking memory requests.

[0042] The expected processing time for each data group is obtained by matching the data complexity value of each data group with the database. The specific method is as follows: each data group is analyzed one by one, the historical data complexity value corresponding to a certain data group stored in the database is obtained, and the difference is processed with the data complexity value of the data group to obtain the difference of the historical data complexity value. The preset difference threshold in the database is obtained and compared with the difference of the historical data complexity value. The historical data complexity value corresponding to the difference threshold is taken as the historical reference data complexity value. The expected processing time of each data group corresponding to the historical reference data complexity value is marked as the expected processing time of each historical data group. The mode and median of the expected processing time of each historical data group are respectively processed by mean processing to obtain the historical reference average value. The historical reference average value is taken as the expected processing time of the data group. Thus, the expected processing time of each data group is obtained.

[0043] The processing time correction coefficient for each idle resource processing unit is obtained by matching the comprehensive performance value of each idle resource processing unit with the database. The specific method is as follows: Each idle resource processing unit is analyzed one by one, and the historical comprehensive performance value corresponding to a certain idle resource processing unit stored in the database is obtained. The difference between the historical comprehensive performance value and the historical comprehensive performance value is obtained. The preset difference threshold in the database is obtained and compared with the difference between the historical comprehensive performance values. The historical comprehensive performance values ​​with differences less than the difference threshold are used as historical reference comprehensive performance values. The processing time correction coefficient of each idle resource processing unit corresponding to each historical reference comprehensive performance value is used as the processing time correction coefficient of each historical idle resource processing unit. The mode and median of the processing time correction coefficients of each historical idle resource processing unit are taken and then the mean is processed to obtain the historical reference average value. The historical reference average value is used as the processing time correction coefficient of the idle resource processing unit. Thus, the processing time correction coefficient of each idle resource processing unit is obtained.

[0044] The expected processing time for each idle resource processing unit is obtained by multiplying the expected processing time of each data group by the processing time correction of the corresponding idle resource processing unit that processes that data group. It should be noted that each resource processing unit processes only one data group at a time; therefore, the processing time correction coefficient of the resource processing unit is the same as the processing time correction coefficient for processing that data group.

[0045] By analyzing the audit execution parameters of each idle resource processing unit and combining normalization processing with weighting factor coupling, the comprehensive performance value of each idle resource processing unit is obtained, thereby calculating the expected processing time of each idle resource processing unit and achieving accurate assessment of resource processing capacity. Analyzing the expected processing time can identify resource processing units that may exhibit a long-tail effect—resources that take longer to process complex or large amounts of data—preventing subsequent data groups from being queued to these resource units, thus preventing slow processing speeds from causing a decline in overall scheduling efficiency. Simultaneously, based on dynamic matching and adjustment of expected processing times, data groups requiring rapid processing can be rationally allocated to idle resource processing units with stronger processing capabilities or shorter processing times, thereby optimizing resource utilization, shortening the overall data processing wait time, and improving the processing efficiency of the entire value-based payment data audit and medical insurance reimbursement process. This ensures that data groups with high complexity and large resource demands receive priority access to suitable resources, effectively avoiding system bottlenecks, achieving balanced and dynamic optimization of resource scheduling, improving data processing throughput and overall execution efficiency, and ensuring the efficiency, stability, and controllability of subsequent settlement and reimbursement processes.

[0046] Furthermore, the expected judgment instruction is obtained through the following method: each data group is matched with the database to obtain the processing time threshold for each data group. Each data group has its corresponding historical processing time for the same period in previous years. The processing time of the same data group (such as the lower leg fracture group) in the same period of previous years is averaged to obtain the historical average processing time. This average processing time is then increased by a preset percentage (such as 10%) to obtain the processing time threshold for each data group (that is, increasing the original 100% historical average processing time to 110%). Based on the comparison between the expected processing time of each idle resource processing unit and the preset processing time threshold, if the expected processing time of an idle resource processing unit is less than the processing time threshold, the expected judgment instruction for that idle resource processing unit is to reschedule the data group to continue waiting for processing; otherwise, the expected judgment instruction for that idle resource processing unit is to perform secondary data resource scheduling processing.

[0047] In this embodiment, by comparing the processing time threshold of each data group with the expected processing time of the idle resource processing unit, an expected judgment instruction is obtained, thereby achieving dynamic control of the data group scheduling order. If the expected processing time of an idle resource processing unit is less than the processing time threshold, the corresponding data group is marked as a secondary scheduling data group to continue waiting for processing, ensuring that high-complexity data groups that are prioritized can obtain matching resources in a timely manner, improving overall processing efficiency. Conversely, if the expected processing time exceeds the threshold, secondary scheduling of data resources is immediately performed, and the secondary scheduling data is redistributed to other suitable resource units to queue for processing, thereby avoiding resource waste and excessive processing delays. This effectively balances resource utilization and processing efficiency, ensuring that the system can achieve efficient and reasonable scheduling under different data complexity and resource load conditions, and improving the stability and real-time performance of value-based payment data processing and medical insurance reimbursement processes.

[0048] Furthermore, secondary scheduling of data resources is performed. Specifically, the scheduling resource processing units are arranged according to their adjacency. These units include idle ordinary resource processing units and ordinary resource processing units currently performing data audits. Based on the resource demand analysis of the secondary scheduling data groups, adjacent scheduling resource processing units are selected (resource processing units that possess joint scheduling and collaborative execution capabilities at the computing resource level; for example, if a large-scale data group requires 20 units of resources to complete processing, and a single CPU core, i.e., a single resource processing unit, only has 8 units of remaining resources, it cannot complete the task independently. If these two cores share the same L3 cache and can quickly exchange data through high-speed interconnect, they can be used for joint scheduling). Considered as "adjacent scheduling resource processing units," during scheduling, two cores are joined together, equivalent to a joint execution unit with a resource capacity of 16, thus enabling them to jointly handle the processing of larger data groups. If the remaining resource capacity of a certain scheduling resource processing unit exceeds the resource demand of the secondary scheduling data group, then that scheduling resource processing unit is marked as a candidate scheduling resource processing unit. If the sum of the remaining resources of some adjacent scheduling resource processing units exceeds the resource demand of the secondary scheduling data group, then that group of adjacent scheduling resource processing units is marked as a candidate scheduling resource processing unit. This results in various candidate scheduling resource processing units, where the same scheduling resource processing unit can form different combinations with adjacent scheduling resource processing units.

[0049] In this embodiment, by analyzing the resource requirements of the secondary scheduling data group and arranging and combining adjacent scheduling resource processing units, alternative scheduling resource processing units that can meet the requirements are dynamically selected, thereby achieving efficient utilization of remaining resources. This avoids waiting and delays caused by the inability of a single resource processing unit's idle resources to meet the data group's needs. Furthermore, by combining adjacent resource processing units to form a set of available resources, the processing capacity for highly complex or large-scale data groups is improved. Secondary scheduling of data resources optimizes the overall resource allocation structure, enhances the system's flexibility and efficiency in processing data groups of varying complexity, reduces resource idleness and processing bottlenecks, and ensures that the data review process for value-based payment can achieve fast and efficient processing while guaranteeing fair allocation, thus supporting the data processing needs of medical insurance reimbursement or other critical businesses.

[0050] It should be noted that the same scheduling resource processing unit can form different combinations with adjacent scheduling resource processing units. For example, if there is a scheduling resource processing unit A that is adjacent to scheduling resource processing units B and C respectively, but B and C are not adjacent, then scheduling resource processing unit A can be combined with B, or combined with C, or combined with both B and C at the same time.

[0051] like Figure 3 and Figure 4 As shown, Figure 3 The upper half of the data task management diagram of the packet data scheduling optimization system for value-based payment provided in the embodiments of this application is shown. Figure 4 The lower half of the schematic diagram of data task management for a value-based data scheduling optimization system provided in this application embodiment shows the actual operation status of each resource processing unit, including the current running progress, remaining runtime, the specific data groups waiting in order under that resource processing unit, and whether the CPU utilization and memory utilization are normal.

[0052] In summary, this embodiment analyzes the complexity of uploaded grouped data and calculates the complexity value and resource requirements accordingly. This quantifies the processing load of each group of data and forms a resource requirement list, thereby achieving dynamic resource aggregation and priority-oriented scheduling based on quantified requirements. This ensures that when high-complexity grouped data arrives, continuous resource blocks can be aggregated and executed on demand, significantly alleviating the resource fragmentation problem caused by static pre-allocation and improving overall data processing efficiency. It effectively solves the problem in existing technologies where static resource allocation leads to fragmented resources, resulting in a lack of continuous resources to support high-complexity tasks and thus low overall data processing efficiency.

[0053] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0054] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0056] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0057] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0058] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A system for packet data scheduling optimization oriented towards pay-per- value, characterized by, include: The data resource assessment module is used to acquire user-uploaded datasets, group them to obtain data groups, acquire data complexity parameters for each data group, analyze the data complexity value of each data group, and thus obtain the resource requirements of each data group. The data complexity value is used to characterize the overall processing difficulty of the data group. The data resource initial allocation module is used to analyze the resource demand of each data group, obtain the data review order, synchronously obtain idle resource processing units and sort them in descending order based on the resource quantity of the idle resource processing units to obtain the resource processing unit arrangement order, thereby performing the initial allocation of data resources and obtaining the first data resource set. The data resource scheduling module is used to obtain the audit execution parameters of each idle resource processing unit, analyze and obtain the expected processing time of each idle resource processing unit, thereby obtaining the expected judgment instruction. If the expected judgment instruction is normal processing, the secondary scheduling data group is counted and continues to wait for processing according to the preset rules. Otherwise, the data resource secondary scheduling processing is performed to obtain the second data resource set. The secondary scheduling data group is the remaining data group when the data resources are initially allocated. The method for obtaining the expected processing time of each idle resource processing unit is as follows: Obtain the audit execution parameters of each idle resource processing unit, including instruction-level parallelism, instruction cache bandwidth, non-blocking memory request concurrency, and task context switching latency; Obtain the preset audit execution standard set from the database, and perform proportional normalization and inverse proportional normalization processing on the audit execution parameters to obtain the normalization processing ratio and the inverse normalization processing ratio. Based on the normalization processing ratio and the inverse normalization processing ratio, introduce the corresponding weighting factor for coupling processing to obtain the comprehensive performance value of each idle resource processing unit. Based on the data complexity values ​​of each data group, the expected processing time for each data group is obtained by matching them with the database. The processing time correction coefficient of each idle resource processing unit is obtained by matching the comprehensive performance value of each idle resource processing unit with the database. Multiplicative coupling analysis is performed based on the expected processing time of each data group and the processing time correction coefficient of each idle resource processing unit to obtain the expected processing time of each idle resource processing unit. The overall performance value of each idle resource processing unit is used to represent the data processing and execution capability of the idle resource processing unit; The audit execution standard set includes standard values ​​for instruction-level parallelism, instruction cache bandwidth, non-blocking memory request concurrency, and task context switching latency. The pay-as-you-go processing module is used to process the data resource set using pay-as-you-go methods to obtain the audit result data after settlement. The data resource set includes a first data resource set and a second data resource set.

2. The value-based packet data scheduling optimization system of claim 1, wherein: The specific method for obtaining the data complexity value of each data group is as follows: Obtain the data complexity parameters for each data group, including the total data volume, number of project categories, number of field dimensions, and number of record entries; Obtain the preset data complexity comparison benchmark set in the database, and perform comparison and normalization processing with the data complexity parameters to obtain the comparison and normalization processing value. Based on the comparison and normalization processing value, introduce the corresponding weighting factor for coupling processing to obtain the data complexity value of each data group. The data complexity comparison benchmark set includes benchmark values ​​for total data volume, number of project categories, number of field dimensions, and number of record entries.

3. The value-based packet data scheduling optimization system of claim 1, wherein: The specific method for obtaining the data review order is as follows: Get the preset resource demand threshold in the database and compare it with the resource demand of each data group. If the resource demand of a certain data group is above the resource demand threshold, mark the data group as a priority data group; otherwise, mark it as a normal data group. Sort the priority data groups in descending order of resource demand to obtain the order of priority data groups; Arrange the ordinary data groups according to the order in which the data was uploaded, and obtain the arrangement order of each ordinary data group; The data review order is obtained by combining the order of priority data groups with the order of ordinary data groups. In the data review order, the overall order of priority data groups takes precedence over the overall order of ordinary data groups.

4. The value-based packet data scheduling optimization system of claim 1, wherein: The specific method for initial allocation of data resources is as follows: Based on the data review order and the resource processing unit arrangement order, idle resource processing units are sequentially allocated to each data group according to the preset allocation rules. If a data group is not allocated an idle resource processing unit, it is marked as a data group to be allocated. If an idle resource processing unit is not allocated, it is marked as a resource processing unit to be allocated.

5. The value-based packet data scheduling optimization system of claim 4, wherein: The preset allocation rule is as follows: Based on resource demand, idle resource processing units are allocated. If an idle resource processing unit is allocated to a priority data group, the remaining idle resources of that idle resource processing unit are obtained, split, and then the resource processing units to be allocated are obtained and put into subsequent allocation. Based on the data review order, resources are allocated to each data group. The resource processing units to be allocated are traversed in the order of resource processing units and then allocated. If, after the resource processing units are allocated, there is a remaining resource processing unit whose remaining resource quantity is less than the resource requirement of any priority data group, then the remaining resource processing unit is marked as a normal resource processing unit, and the remaining resource processing units to be allocated are counted and marked as normal resource processing units. Based on resource demand analysis, ordinary resource processing units are allocated to each ordinary data group.

6. The value-based packet data scheduling optimization system of claim 5, wherein: The method for allocating ordinary resource processing units to each ordinary data group is as follows: Resources are allocated to each ordinary data group based on the order of their arrangement. The remaining resources of each ordinary resource processing unit are retrieved and compared with the resource requirements of the ordinary data group. If the remaining resources of an ordinary resource processing unit are greater than or closest to the resource requirements of the ordinary data group, then the ordinary resource processing unit is allocated to that ordinary data group. If a regular data group is not assigned a regular resource processing unit, then the regular data group is marked as an unprocessed regular data group. Based on the order of resource processing units, each ordinary resource processing unit is traversed to check if there are any idle resources in each ordinary resource processing unit. If there are, the idle resource quantity is retrieved and compared with the resource demand of the unprocessed ordinary data group. The idle resource is then allocated to each unprocessed ordinary data group. If there are still unprocessed ordinary data groups that have not been allocated to an ordinary resource processing unit, the unprocessed ordinary data group is marked as a secondary scheduling data group and the secondary scheduling data group is pre-queued. If there are no unprocessed ordinary data groups that have not been allocated to an ordinary resource processing unit, the first data resource set is obtained.

7. The value-based packet data scheduling optimization system of claim 1, wherein: The specific method for obtaining the expected determination instruction is as follows: ​ Each data group is matched with the database to obtain the processing time threshold for each data group; Based on the comparison between the expected processing time of each idle resource processing unit and the preset processing time threshold, if the expected processing time of an idle resource processing unit is less than the processing time threshold, the expected determination instruction of the idle resource processing unit is to continue waiting for the data group to be processed in a secondary scheduling manner; otherwise, the expected determination instruction of the idle resource processing unit is to perform secondary scheduling processing of the data resources.

8. The value-based packet data scheduling optimization system of claim 1, wherein: The specific method for performing secondary data resource scheduling is as follows: ​ The scheduling resource processing units are arranged according to their adjacent relationships. The scheduling resource processing units include idle general resource processing units and general resource processing units that are performing data auditing. Based on the resource demand analysis of the secondary scheduling data group, each adjacent scheduling resource processing unit is selected. If the remaining resource amount of a certain scheduling resource processing unit is greater than the resource demand of the secondary scheduling data group, then the scheduling resource processing unit is marked as a candidate scheduling resource processing unit. If the sum of the remaining resources of some adjacent scheduling resource processing units is greater than the resource demand of the secondary scheduling data group, then the group of adjacent scheduling resource processing units is marked as a candidate scheduling resource processing unit. Thus, each candidate scheduling resource processing unit is obtained. The same scheduling resource processing unit can form different combinations with adjacent scheduling resource processing units.

9. The method for the value-based packet data scheduling optimization system of any one of claims 1-8, wherein, Includes the following steps: The system acquires user-uploaded datasets and manages them into groups. It then obtains the data complexity parameters for each data group, analyzes them to obtain the data complexity value, and thus calculates the resource requirements for each data group. The data complexity value is used to characterize the overall processing difficulty of the data group. Based on the resource demand analysis of each data group, the data review order is obtained. Idle resource processing units are simultaneously acquired and sorted in descending order based on resource quantity to obtain the resource processing unit arrangement order. Thus, the initial allocation of data resources is carried out to obtain the first data resource set. The audit execution parameters of each idle resource processing unit are obtained, and the expected processing time of each idle resource processing unit is analyzed to obtain the expected judgment instruction. If the expected judgment instruction is normal processing, the secondary scheduling data group is counted and continues to wait for processing according to the preset rules. Otherwise, the secondary scheduling of data resources is performed to obtain the second data resource set. The secondary scheduling data group is the remaining data group when the data resources are initially allocated. The data resource set is processed on a pay-as-you-go basis to obtain the audit result data after settlement. The data resource set includes a first data resource set and a second data resource set.

Images