A group unit grouping summary calculation optimization method

Abstract

The application discloses a population unit grouping and summary settlement optimization method, and belongs to the technical field of large-scale high-concurrency operation computing power optimization.The population units in an operation scene are divided into a plurality of operation groups by a preset rule in the application, and the group is taken as the minimum operation unit to complete the data summary settlement in the group and generate the overall operation result of the group;two modes of global unified threshold comparison after multi-group data integration or independent threshold comparison of single-group summary result are supported, and the operation settlement is uniformly completed according to the comparison result.The application is different from the traditional partition and device load allocation mode, takes the group summary batch operation as the core, reduces the total operation times from the root, greatly reduces the high-concurrency operation computing power consumption, can support larger-scale cluster operation under the original hardware condition, blocks the bypass path of technical variants, and is suitable for high-concurrency operation scenes in all fields such as games, simulation, monitoring and data processing.

Description

Technical Field

[0001] This invention relates to the fields of large-scale cluster computing, high-concurrency data processing, and computing power optimization. Specifically, it relates to a group unit grouping and aggregation settlement optimization method, which is applicable to various general high-concurrency processing scenarios that require multi-entity parallel computing, batch data judgment, and group behavior settlement. Background Technology

[0002] In current high-concurrency computing scenarios, the need for large-scale synchronous interaction of group units and batch data settlement is becoming increasingly common. Whether it is simulated combat, cluster monitoring, parallel data processing, or centralized judgment scenarios, synchronous operation of multiple units will bring enormous computing power pressure. Existing industry-standard processing solutions generally adopt a load-sharing model of partitioning, node-sharing, and device-sharing, splitting massive group units into different regions or computing nodes for load distribution.

[0003] Existing technologies have several core flaws: First, traditional load balancing is merely task splitting and load transfer, essentially a means of allocating computing power. Whether splitting by spatial partitioning or by quantity, each group still employs a point-to-point independent computation and judgment model, requiring separate calculation for each action and effect, thus not reducing the total number of computations and failing to conserve computing resources. Second, under the traditional judgment model, each group unit needs to perform threshold comparison with the target object individually, resulting in a large amount of repetitive computation and high logical redundancy. In high-concurrency scenarios, the pressure on computing devices surges, easily leading to processing delays, logical stuttering, and overload anomalies. Third, the partitioning and node-based splitting model can only rely on increasing the number of hardware units to meet high-concurrency demands, leading to continuously increasing hardware maintenance and site rental costs, and failing to achieve cost reduction and efficiency improvement through computational logic optimization. Fourth, the lack of a group-level batch settlement mechanism and the absence of computational logic for data integration within groups and globally unified threshold comparison result in a long-term technical bottleneck in computational efficiency for large-scale high-concurrency scenarios.

[0004] In summary, existing load balancing models only address the load distribution problem and do not optimize the computing architecture, thus failing to reduce the total computational load at its root. To address these technical issues, there is an urgent need to design a novel grouping and aggregation computing scheme that, without upgrading hardware, reduces invalid computations, lowers computing power consumption, and meets the stable operation requirements of large-scale, high-concurrency computing. Summary of the Invention

[0005] Purpose of the invention The purpose of this invention is to overcome the shortcomings of existing technologies and provide a group unit grouping and aggregation settlement optimization method. This method solves the problems of traditional partitioning and diversion only distributing the load without saving computing power, large redundancy in single-point operations, and high-concurrency lag. By using a group aggregation batch operation + global / group threshold comparison architecture, the total amount of computation is greatly reduced and the operating efficiency of computing devices is improved. Technical solution

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A method for optimizing group unit grouping and aggregation settlement includes the following steps: S1: Group division, which divides all group units in the computing scene into several independent computing groups according to any preset rules such as region, type, intensity, attribute, and faction. The number of groups can be fixed or dynamically adjusted according to the number of group units and the load of computing devices. S2: Intra-group integration and settlement. For each independent calculation group, collect the behavioral data, effect data, and status data of all group units in the group, complete the comprehensive summary and settlement within the group, and generate the overall calculation result for each group. This step takes the group as the smallest calculation unit, performs batch merging, deduplication, and unified quantification on the scattered data in the group, eliminates duplicate calculation logic, and reduces the total number of calculations from the source. S3: Threshold comparison execution. Threshold determination is performed based on the overall calculation results of each group, providing two flexible execution methods: The first method is to integrate and merge the overall calculation results of all groups to form unified global data, and then compare it with the unified judgment threshold preset or dynamically generated by the target object; The second method is to take the independent overall calculation result of each group as a unit and compare it with the judgment threshold of the target object one by one; where the target object is the main body of cluster calculation, centralized management unit, global judgment node, etc., and the judgment threshold can be set to a preset fixed value or a dynamically floating value. S4: Output the results. Based on the threshold comparison results, generate the corresponding settlement effect and status judgment in a unified manner to complete large-scale high-concurrency logical operations.

[0007] This invention differs from traditional load balancing methods such as partitioning, node splitting, and traffic distribution. It reduces the total computational load at the computing architecture level, rather than simply splitting tasks. Beneficial effects

[0008] 1. Fundamentally different from traditional load balancing techniques, significantly optimizing computing power. This invention differs from the traditional single-point, individual computation mode, breaking through the conventional approach of load balancing by device or region. Instead, it performs merged and aggregated calculations using groups as units, greatly reducing the number of repetitive and scattered single-point calculations, thus reducing computational pressure at its source. Traditional grouping only distributes tasks without reducing computational load, while this invention directly reduces the total computational scale, achieving true computing power savings.

[0009] 2. Dual-mode threshold comparison provides full coverage, completely blocking loopholes. This invention simultaneously covers two types of threshold calculation logic, supporting both global unified comparison after merging multiple sets of results and independent comparison of single-group summary results. Both modes rely on the core architecture of "intra-group summarization first, optimizing scattered calculations," which is fundamentally different from the traditional single-unit point-to-point independent judgment. Regardless of the comparison variant subsequently used, it falls within the protection scope of this invention.

[0010] 3. Universal applicability with no limitations on the scope of protection. This invention is not limited to a single application scenario or bound to a specific industry. Regardless of whether it is applied to any technical field such as game competition, simulation, cluster monitoring, data processing, or industrial control management, as long as the core logic of group aggregation and batch threshold comparison is executed, it falls within the protection scope of this solution, without the need to adjust the architecture separately for different scenarios.

[0011] 4. Reduced operating costs and strong hardware compatibility. No new computing hardware or equipment upgrades are required. Existing hardware can support larger-scale high-concurrency computing. The freed-up computing power can be used for process optimization, accuracy improvement, and scale expansion, comprehensively improving computing stability and smoothness.

[0012] 5. Stable decision-making logic, adaptable to high-concurrency scenarios. Employing grouped batch processing logic avoids logical conflicts and data deviations inherent in distributed single-point decision-making. This results in more consistent and stable logic settlement under large-scale concurrent scenarios, effectively preventing issues such as stuttering, delays, and settlement anomalies.

[0013] 6. Batch settlement of time slices / frames further improves smoothness. Supports summarization and judgment by fixed time slices or by frame, adapting to real-time calculation and synchronous rendering rhythm for smoother operation. Detailed Implementation

[0014] The present invention will be further described in detail below with reference to specific embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0015] Example 1: Large-scale game battle scenario In massively multiplayer online games (MMORPGs) featuring massive battles involving tens of thousands of players, sieges, and world boss fights, the number of participating units can reach tens of thousands. Traditional single-point calculations are prone to lag. First, all units in the field are divided into several calculation groups based on region, faction, and strength, which can be dynamically adjusted according to the load. Second, data aggregation and settlement within each group are completed using the group as the smallest unit to generate the overall calculation value for each group. Finally, a global unified comparison or independent comparison within each group can be selected to complete the high-concurrency battle settlement in one go.

[0016] This solution can significantly reduce server computing power consumption and stably support large-scale simultaneous combat without upgrading hardware.

[0017] Example 2: Centralized Cluster Target Determination Scenario In scenarios where multiple units collectively challenge a single objective, the number of participating units is large and the computation time is long. The computation groups are divided according to the unit strength and type to complete the summary of the effect and status data within the group and optimize the redundant logic of single-point comparison. The system can choose to compare with a globally unified threshold or compare with groups separately, which greatly reduces the computing power consumption during peak periods and ensures the stable operation of the system.

[0018] Example 3: Grouped Independent Threshold Comparison Scenario In multi-type unit synchronous operation scenarios, the operation is divided into different operation groups according to the behavior type. Each group first summarizes and generates an independent overall value. The group is then directly calculated separately with the target threshold to complete the batch judgment. This reduces the total amount of operation while meeting the differentiated judgment requirements of different types of behavior.

Claims

1. A method for optimizing group unit grouping, aggregation, and settlement, characterized in that, Includes the following steps: S1: Group division, which divides all group units within the computing scenario into several independent computing groups according to a preset division rule; the preset division rule includes any one or more combinations of region division, type division, intensity level division, attribute division, and faction division; S2: Intra-group integrated settlement. For each independent calculation group, collect behavioral data, effect data, and status data of all group units within the group, complete the comprehensive summary settlement within the group, and generate the overall calculation results for each group; use the group as the smallest calculation unit to perform batch merging calculations on various types of data within the group; S3: Threshold comparison execution, threshold determination is performed based on the overall calculation results of each group; There are two execution methods: Method 1: Integrate and merge the overall calculation results of all groups to form unified global data, and compare them with the unified judgment threshold preset or dynamically generated by the target object; Method 2: Take the independent overall calculation result of each group as a unit, and compare it with the judgment threshold of the target object one by one. S4: Output the results. Based on the threshold comparison results, generate the corresponding settlement effect and status judgment in a unified manner to complete large-scale high-concurrency logical operations.

2. The method according to claim 1, characterized in that, In S1, group division can be either fixed or dynamically adaptive, adjusting the number and size of groups in real time based on the number of group units and the load of computing devices within the scene.

3. The method according to claim 1, characterized in that, In S2, the group-wide integrated settlement merges, deduplicatizes, and uniformly quantifies scattered data within the group, eliminating redundant calculation logic and reducing the total number of calculations.

4. The method according to claim 1, characterized in that, The target objects are computing entities such as cluster computing entities, centralized management and control units, and global judgment nodes, and the thresholds are preset fixed values ​​or dynamically floating values.

5. The method according to claim 1, characterized in that, Batch summary settlement and threshold determination are performed based on fixed time slices or frames.

6. The method according to claim 1, characterized in that, It is suitable for various high-concurrency computing scenarios such as large-scale cluster warfare, high-concurrency parallel data processing, batch threshold determination, and group behavior settlement.