Virtual game interaction identification analysis system
By using an adaptive observation sampling and multi-scale preprocessing module and a fractional-order dynamic fusion analysis module, the problem of high concurrency and real-time performance discrepancy in high-frequency interactive data processing in virtual games was solved. This enabled real-time processing of critical data and dynamic adjustment of high-frequency non-critical data, ensuring the real-time performance and accuracy of the game.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-06
- Publication Date
- 2026-03-27
AI Technical Summary
Existing technologies struggle to balance high concurrency and real-time performance when processing high-frequency interactive data in virtual games, leading to the loss of critical instantaneous data and high computational costs, thus failing to support real-time anti-cheating detection or dynamic difficulty adjustment.
An adaptive observation sampling and multi-scale preprocessing module is adopted. The sampling parameter drift is limited by the fractional time-varying gain adaptive sampling law and projection operator. The key data and high-frequency non-key data are divided. A fractional dynamic fusion analysis module is designed. The analysis strategy is dynamically adjusted by the fractional trapezoidal sliding surface. The key data is processed in real time and the high-frequency non-key data is processed in batches. The response speed and noise suppression are optimized by combining the time-varying scaling term.
It ensured the integrity of key data and the stability of game parameters, reduced processing latency, met the real-time requirements of anti-cheating, improved the accuracy of anti-cheating, and guaranteed the reliability of the system through proactive monitoring and analysis.
Smart Images
Figure CN121731772A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing and analysis technology, and specifically to a virtual game interaction recognition and analysis system. Background Technology
[0002] In the current field of interactive data analysis for virtual games (including large-scale online games, VR / AR games, metaverse spaces, etc.), although the technology is relatively mature, there are still many unique pain points in handling ultra-large-scale, high-frequency interactions and deep semantic understanding.
[0003] Existing game interaction data has an extremely high frequency, such as displacement data, skill casting data, and item exchange data. Traditional batch processing solutions cause delays in analysis results and cannot support real-time anti-cheat detection or dynamic difficulty adjustment. On the other hand, pure real-time stream processing has extremely high server bandwidth and computing costs when facing millions of players online at the same time. It is difficult to strike a balance between full high-frequency sampling and computing costs, often resulting in the loss of critical instantaneous data. There is a gap between high concurrency and real-time performance. Summary of the Invention
[0004] The purpose of this invention is to provide a virtual game interaction recognition and analysis system to solve the technical problem of the gap between high concurrency and real-time performance in existing technical solutions.
[0005] The objective of this invention can be achieved through the following technical solutions: A virtual game interaction recognition and analysis system, comprising: Adaptive observation sampling and multi-scale preprocessing module: Divide game interaction data into key data and high-frequency non-key data, design fractional time-varying gain adaptive sampling law, dynamically adjust the sampling rate for high-frequency non-key data, and design projection operator to limit sampling parameter drift and prevent loss of key instantaneous data. Fractional-order dynamic fusion analysis module: By dynamically adjusting the analysis strategy through a fractional-order trapezoidal sliding surface, it performs real-time stream processing on key data and batch processing on high-frequency non-key data. It balances the initial response speed and steady-state noise suppression by designing a time-varying scaling term, and monitors and analyzes the application effect of the sliding surface of the fusion scaling term from different dimensions, providing dynamic prompts.
[0006] Furthermore, when classifying and prioritizing game interaction data, the data is divided into key data and high-frequency non-key data based on the degree of impact of the data on the core functions of the game. Key data corresponds to high priority, and high-frequency non-key data corresponds to low priority.
[0007] Furthermore, a fractional-order time-varying gain adaptive sampling law is designed to balance sampling accuracy in real time. The relevant expression is: ;in, The real-time sampling rate for displacement data; The basic sampling rate for displacement; The gain coefficient controls the sensitivity of the sampling rate to data fluctuations. It is a fractional differential operator with a range of values (0, 1); This refers to displacement fluctuation error.
[0008] Furthermore, a projection operator is designed to limit the sampling rate range, preventing the loss of critical instantaneous displacement due to excessive downsampling of high-frequency non-critical data, or the waste of resources due to excessive upsampling. The relevant expression is as follows: Where proj() is the projection operator; This represents the minimum displacement sampling rate. This represents the maximum displacement sampling rate. Set up for key data .
[0009] Furthermore, the collected key data and high-frequency non-key data are preprocessed. Based on the preprocessed data, a fractional step-shaped sliding surface is constructed, involving the following expression: ;in, Output value for the sliding surface; For fractional-order trend fusion operators; These are multi-scale weighted coefficients, corresponding to key data. High-frequency non-critical data correspond ;k is the index of the categorical data, k is 1, 2; The error is the k-th type of data; m is the order of the sliding surface. This is a sign function that distinguishes the direction of the error.
[0010] Furthermore, when When the threshold is reached, real-time stream processing and analysis are performed on key data; when When the threshold is ≤, perform batch analysis on high-frequency non-critical data.
[0011] Furthermore, a time-varying scaling term is injected into the sliding surface calculation to optimize the initial response speed and steady-state noise suppression. The relevant expression is as follows: ;in, Output values of the sliding surface for injecting time-varying scaling terms; ;in, This is a time-varying scaling factor; This is the initial scaling factor; The attenuation coefficient; is a power parameter; e is a constant.
[0012] Furthermore, when conducting regulatory analysis on the sliding surface application effect of the fusion scaling item from different dimensions, the output reliability value and output stability value after the sliding surface application are calculated and analyzed together. If the output reliability value and output stability value both meet the corresponding output reliability requirements and output stability requirements, then the sliding surface application effect of the fusion scaling item is determined to be normal. Conversely, if the application effect of the sliding surface of the fusion scaling item is abnormal, a warning will be issued.
[0013] Furthermore, the expression for calculating the reliable output value is: Where SK is the output reliability value; These are the captured value, misjudged value, and missed detection value after the application of the sliding surface; These are the capture threshold, false positive threshold, and missed detection threshold, respectively; max() indicates that the maximum value is obtained. The expression for calculating the output stable value is: Where SW is the stable output value; N is the number of times the sliding surface output value is sampled; Let be the sliding surface output value of the i-th sample; i is the sampling index, i=1,2,3,…,N; This represents the average value of the output values from the sliding surface. To output a stable standard value.
[0014] Furthermore, different dimensions include the output reliability dimension and the output stability dimension of the sliding surface application.
[0015] Compared to existing solutions, the beneficial effects achieved by this invention are: This invention defines data priorities by classification, which can provide a basis for subsequent sampling strategies; it applies fractional sampling laws to dynamically adjust the sampling rate for low-priority, high-frequency, non-critical data, which can effectively reduce bandwidth consumption and achieve dynamic response; and it constrains the sampling rate range by using projection operators to prevent parameter drift from causing the loss of critical data or waste of resources. These three elements form a closed loop, ensuring the integrity of critical data and the stability of game parameters.
[0016] This invention, through the coordinated operation of different steps, enables dynamic processing of critical data and high-frequency non-critical data, effectively reducing processing latency and meeting the real-time requirements of anti-fraud. By fusing fractional-order differentials with historical trends and designing time-varying scaling terms to suppress noise, it effectively improves anti-fraud accuracy. Furthermore, by monitoring and analyzing the sliding surface application effect of the fused scaling terms from different dimensions, it achieves proactive monitoring and dynamic prompts regarding the sliding surface application effect, further ensuring the reliability of the technical solution. Finally, through the coordinated operation of sliding surface dynamic decision-making, fusion analysis, scaling term optimization, and proactive monitoring analysis, it effectively solves the problem of the gap between high concurrency and real-time performance. Attached Figure Description
[0017] The invention will now be further described with reference to the accompanying drawings.
[0018] Figure 1 This is a flowchart illustrating the operation of a virtual game interaction recognition and analysis system according to the present invention. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] like Figure 1 As shown, this invention is a virtual game interaction recognition and analysis system, including an adaptive observation sampling and multi-scale preprocessing module and a fractional-order dynamic fusion analysis module; specifically: The adaptive observation sampling and multi-scale preprocessing module divides game interaction data into critical data and high-frequency non-critical data. It designs a fractional-order time-varying gain adaptive sampling law to dynamically adjust the sampling rate for high-frequency non-critical data, and designs a projection operator to limit sampling parameter drift and prevent the loss of critical instantaneous data. Specific steps include: When classifying and prioritizing game interaction data, the data is divided into key data and high-frequency non-key data based on the degree of impact of the data on the core functions of the game, such as anti-cheat detection and dynamic difficulty adjustment. Key data corresponds to high priority, and high-frequency non-key data corresponds to low priority. Key data includes skill casting and item exchange; skill casting specifically includes skill ID, casting time, and target player ID; item exchange specifically includes item ID, IDs of both parties involved in the exchange, and exchange time; this type of data directly affects anti-cheat measures and dynamic difficulty, requiring 100% sampling; for example, the rapid skill casting and high-frequency item exchange of cheats correspond to the intensity of teamwork. High-frequency, non-critical data includes displacement; displacement specifically includes player coordinates; this type of data is high-frequency, but a single data point has little impact on core functions, and only the trend needs to be statistically analyzed, so adaptive dynamic sampling is used. For high-frequency, non-critical data, a fractional-order time-varying gain adaptive sampling law is designed to balance sampling accuracy in real time. The relevant expression is as follows: ;in, The real-time sampling rate for displacement data; The basic sampling rate for displacement; This is the gain coefficient, which controls the sensitivity of the sampling rate to data fluctuations; the default value is 0.5. It is a fractional differential operator with a value range of (0, 1), and a default value of 0.6, which reflects the weight of the impact of historical data fluctuations on the current sampling rate; It is of fractional order; For displacement fluctuation error, , The displacement coordinate value of the player at time t refers to the horizontal coordinate in the game scene, such as the x-axis coordinate, which represents the player's current position. The player's displacement coordinates at time t-1, which is the sampling time before time t, serve as the reference point for calculating the current displacement change. This represents the instantaneous change in the player's displacement between time t and time t-1. This represents the average value of the displacement change over the past 5 seconds. It needs to be explained that when the displacement changes abruptly, such as instantaneous teleportation caused by external cheats, Increase the fractional differential operator to amplify the error and control the real-time sampling rate. Increase to the maximum value; when the displacement is stable, for example... Reduce the real-time sampling rate to a minimum. Furthermore, in the embodiments of the present invention, the default values of the parameters in the expressions are all determined and illustrated based on the simulation test results data from the previous period. In actual applications, they can be customized and adjusted according to the application requirements and specifications of the actual application scenario. Moreover, the data calculations performed by the expressions in this application are all standardized before the calculations are performed, including but not limited to extracting the numerical values of the calculation data and performing normalization processing. These are all existing conventional technical means, and the specific implementation steps are not described here.
[0021] Furthermore, the design of the projection operator limits the sampling rate range to prevent excessive downsampling of high-frequency non-critical data, which could result in the loss of critical instantaneous displacements, such as jumping motions, or excessive upsampling, which could waste resources. The relevant expressions are as follows: ;where proj() is the projection operator, which forces the sampling rate to remain within a reasonable range; The minimum displacement sampling rate is set to ensure that displacement trends are not missed; The maximum displacement sampling rate limits bandwidth usage. It is important to note that for key data, settings should be configured. To ensure no data loss; In this embodiment of the invention, defining data priorities by category can provide a basis for subsequent sampling strategies; applying fractional sampling laws to dynamically adjust the sampling rate for low-priority, high-frequency, non-critical data can effectively reduce bandwidth usage and achieve dynamic response; constraining the sampling rate range by projection operators prevents parameter drift from causing the loss of critical data or waste of resources. These three elements form a closed loop, ensuring the integrity of critical data and the stability of game parameters.
[0022] Fractional-order dynamic fusion analysis module: This module dynamically adjusts the analysis strategy using a fractional-order trapezoidal sliding surface. It performs real-time stream processing on critical data and batch processing on high-frequency, non-critical data. It balances initial response speed with steady-state noise suppression by designing a time-varying scaling term, and monitors and dynamically prompts the sliding surface application effect of the fusion scaling term from different dimensions. Specific steps include: The key data and high-frequency non-key data collected are preprocessed. The preprocessing includes, but is not limited to, noise reduction and normalization, which are existing conventional technical solutions. The specific implementation steps are not described here. Based on the preprocessed data, a fractional step-shaped sliding surface is constructed as the core decision indicator for dynamically adjusting the analysis strategy. The relevant expression is as follows: ;in, The output value of the sliding surface reflects the degree of data anomaly and the processing priority; A fractional-order trend fusion operator is used to integrate historical data with current trends, thereby improving decision robustness. These are multi-scale weighted coefficients, corresponding to key data. High-frequency non-critical data correspond ;k is the index of the categorical data, k is 1, 2; For the k-th type of data error, Corresponding to key data errors, , For skill release frequency, The skill release frequency threshold can be determined based on the median of all skill release frequencies obtained from previous simulation tests. Corresponding to high-frequency non-critical data errors, the specific values are related to displacement fluctuation errors. Same; m is the order of the sliding surface, which controls the degree of nonlinear weighting of the error; The sign function distinguishes the direction of the error; It should be noted that by capturing dynamic changes in data through fractional derivatives, and by ensuring that key data guides decision-making through weighting coefficients, the larger the sliding surface output, the more significant the data anomalies, which should be prioritized for real-time processing.
[0023] Based on the output value of the sliding surface The output is dynamically allocated to computing resources to perform a fusion analysis of real-time stream processing and batch processing. when When the threshold is reached, real-time stream processing and analysis of key data are performed, including anti-cheating detection and dynamic difficulty adjustment; the threshold here can be determined based on the development and design requirements of the virtual game, or based on the simulation test data in the early stage, and the specific value is not limited; During anti-cheat detection, the skill casting interval is compared in real time to see if it is less than the minimum allowed interval. If it is, cheating is determined not to exist; otherwise, cheating is determined to exist. When making dynamic difficulty adjustments, the player's skill hit rate is statistically analyzed in real time. If the player hits five times in a row, the difficulty of the defense rules is increased; otherwise, the existing difficulty of the defense rules is maintained. when When the threshold is less than or equal to, high-frequency non-critical data is batch-processed and analyzed: displacement samples are packaged every 5 seconds, and the displacement trajectory is analyzed to see if it conforms to the map terrain. If it conforms to the map terrain, the existing obstacle refresh position is maintained; if it does not conform to the map terrain, the existing obstacle refresh position is adjusted. Among these, analyzing whether the displacement trajectory conforms to the map terrain is an existing conventional technical solution, and the specific implementation steps are not described here. Furthermore, a time-varying scaling term is injected into the sliding surface calculation to optimize the initial response speed and steady-state noise suppression. The relevant expression is as follows: ;in, Output values of the sliding surface for injecting time-varying scaling terms; ;in, The time-varying scaling factor is used to adjust the sensitivity of the sliding surface; This is the initial scaling factor, with a default value of 1.2, to ensure a fast response in the initial stage; This is the attenuation coefficient, with a default value of 0.3, which controls the attenuation rate of the scaling factor. This is a power parameter, with a default value of 0.6, which adjusts the degree of non-linearity in the decay; e is a constant. It should be noted that combining the scaling term with the sliding surface in the initial stage... Large scale allows for rapid detection of critical data anomalies; later stages Reducing the size of the displacement sample can suppress minor fluctuations and balance real-time performance with stability.
[0024] Furthermore, by constructing a sliding surface, performing fusion analysis based on the sliding surface, and optimizing performance with time-varying scaling terms, the three are related and progressive. The sliding surface is the basis for resource allocation, fusion analysis is the goal, and time-varying scaling terms are the optimization means, which together achieve a dynamic balance between real-time processing of critical data and batch processing of non-critical data.
[0025] When conducting regulatory analysis on the sliding surface application effect of the fusion scaling item from different dimensions, the output reliability value and output stability value after the sliding surface application are calculated and analyzed together. If the output reliability value and output stability value both meet the corresponding output reliability requirements and output stability requirements, the sliding surface application effect of the fusion scaling item is judged to be normal. Conversely, if the application effect of the sliding surface of the fusion scaling item is abnormal, a warning will be issued. It should be noted that the different dimensions in the embodiments of the present invention include the output reliability dimension and the output stability dimension of the sliding surface application. These dimensions are processed based on the data obtained from the supervision, and the processed output reliability value and output stability value are analyzed. This enables diversified proactive supervision analysis of the sliding surface application effect of the fusion scaling item. Compared with the supervision analysis of a single supervision indicator, the embodiments of the present invention can effectively improve the sliding surface application effect and proactive supervision effect of the fusion scaling item. The expression for calculating the reliable output value is as follows: Where SK is the output reliability value; These are the captured value, misjudged value, and missed detection value after the application of the sliding surface. , These represent the number of correctly detected anomalies and the number of actual anomalies, respectively. , These represent the number of normal values that were mistakenly identified as abnormal, and the total number of times the data was detected as abnormal, respectively. , These represent the number of undetected anomalies; These are the capture threshold, false positive threshold, and missed detection threshold, respectively; max() indicates that the maximum value is obtained. It should be explained that the output reliability value is used to digitally process and integrate the captured data, misjudged data, and missed detection data of the regulator to digitally represent the output reliability state of the sliding surface application; It should be noted that the number of correctly detected anomalies is specifically the number of times when the sliding surface output value exceeds the preset anomaly threshold and there is actually abnormal game behavior; for example: if a player uses a skill with no cooldown, it is a real anomaly, and the sliding surface output value reaches 60 points, exceeding the preset anomaly threshold of 50 points, which is judged as an anomaly and counted as 1 correct detection. The actual number of anomalies refers to the total number of real abnormal behaviors recorded by the game server, such as skills with no cooldown, passing through walls, and cross-map item swapping. The number of normal occurrences that are misjudged as abnormal is specifically defined as the number of times the output value of the sliding surface exceeds the preset abnormal threshold, but is actually the number of times the player's normal behavior occurs; for example, if a player's skill casting frequency temporarily increases due to network fluctuations, the sliding surface may misjudge it as a skill with no cooldown, which is counted as 1 misjudgment. The total number of times an anomaly was detected, specifically the total number of times the sliding surface was judged as an anomaly, including both correct detections and false positives; The number of undetected anomalies refers to the number of times that abnormal game behavior actually exists, but the sliding surface output value does not reach the preset anomaly threshold; for example, if a player passes through a wall but the displacement sampling rate is too low, the sliding surface output value is only 30 points, which is lower than the threshold of 50, and it is not detected, which is counted as 1 missed detection. In addition, the preset abnormal thresholds corresponding to the output values of the sliding surface, as well as the capture threshold, misjudgment threshold, and missed detection threshold, can all be determined based on the development and design requirements of the virtual game, or based on the simulation test data in the early stage. The specific values are not limited. The expression for calculating the output stable value is: Where SW is the stable output value; N is the number of times the sliding surface output value is sampled; Let be the sliding surface output value of the i-th sample; i is the sampling index, i=1,2,3,…,N; This represents the average value of the output values from the sliding surface. To output a stable standard value, the specific value can be the median of the standard deviation of all sliding surface output values from previous simulation tests; When both the output reliability value and the output stability value meet the corresponding output reliability requirements and output stability requirements, the specific output reliability requirements are as follows: The specific output stability requirements are as follows: ; In addition, the output reliability requirements and output stability requirements can also be customized according to the application requirements and specifications of the actual application scenario. The specific range is not limited here.
[0026] In this embodiment of the invention, through the coordinated operation of the above steps, dynamic processing of critical data and high-frequency non-critical data can be achieved, effectively reducing processing latency and meeting the real-time requirements of anti-fraud. By fusing fractional-order differentials with historical trends and designing time-varying scaling terms to suppress noise, the accuracy of anti-fraud can be effectively improved. By monitoring and analyzing the sliding surface application effect of the fused scaling terms from different dimensions, proactive monitoring and dynamic prompts for the sliding surface application effect are achieved, further ensuring the reliability of the technical solution. Through the coordinated operation of sliding surface dynamic decision-making, fusion analysis, scaling term optimization, and proactive monitoring analysis, the problem of the gap between high concurrency and real-time performance can be effectively solved.
[0027] In the several embodiments provided by this invention, it should be understood that the disclosed system can be implemented in other ways. For example, the embodiments of the invention described above are merely illustrative; for example, the division of modules is only a logical functional division, and there may be other division methods in actual implementation.
[0028] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0029] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated module can be implemented in hardware or in the form of hardware plus software functional modules.
[0030] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the essential characteristics of the present invention.
[0031] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. A virtual game interaction recognition and analysis system, characterized in that, include: Adaptive observation sampling and multi-scale preprocessing module: Divide game interaction data into key data and high-frequency non-key data, design fractional time-varying gain adaptive sampling law, dynamically adjust the sampling rate for high-frequency non-key data, and design projection operator to limit sampling parameter drift and prevent loss of key instantaneous data. Fractional-order dynamic fusion analysis module: By dynamically adjusting the analysis strategy through a fractional-order trapezoidal sliding surface, it performs real-time stream processing on key data and batch processing on high-frequency non-key data. It balances the initial response speed and steady-state noise suppression by designing a time-varying scaling term, and monitors and analyzes the application effect of the sliding surface of the fusion scaling term from different dimensions, providing dynamic prompts.
2. The virtual game interaction recognition and analysis system according to claim 1, characterized in that, When classifying and prioritizing game interaction data, the data is divided into key data and high-frequency non-key data according to the degree of impact of the data on the core functions of the game. Key data corresponds to high priority, and high-frequency non-key data corresponds to low priority.
3. The virtual game interaction recognition and analysis system according to claim 1, characterized in that, The design of a fractional-order time-varying gain adaptive sampling law to balance sampling accuracy in real time involves the following expressions: ;in, The real-time sampling rate for displacement data; The basic sampling rate for displacement; The gain coefficient controls the sensitivity of the sampling rate to data fluctuations. It is a fractional differential operator with a range of values (0, 1); This refers to displacement fluctuation error.
4. The virtual game interaction recognition and analysis system according to claim 3, characterized in that, The design of the projection operator limits the sampling rate range to prevent excessive downsampling of high-frequency non-critical data from losing critical instantaneous displacements, or excessive upsampling from wasting resources. The relevant expression is: ； Where proj() is the projection operator; This represents the minimum displacement sampling rate. This represents the maximum displacement sampling rate. Set up for key data .
5. The virtual game interaction recognition and analysis system according to claim 4, characterized in that, The collected key data and high-frequency non-key data are preprocessed. Based on the preprocessed data, a fractional step-shaped sliding surface is constructed, involving the following expressions: ;in, Output value for the sliding surface; For fractional-order trend fusion operators; These are multi-scale weighted coefficients, corresponding to key data. High-frequency non-critical data correspond ;k is the index of the categorical data, k is 1, 2; The error is the k-th type of data; m is the order of the sliding surface. This is a sign function that distinguishes the direction of the error.
6. The virtual game interaction recognition and analysis system according to claim 5, characterized in that, when When the threshold is reached, real-time stream processing and analysis are performed on key data; when When the threshold is ≤, perform batch analysis on high-frequency non-critical data.
7. The virtual game interaction recognition and analysis system according to claim 6, characterized in that, A time-varying scaling term is injected into the sliding surface calculation to optimize the initial response speed and steady-state noise suppression. The relevant expression is as follows: ;in, Output values of the sliding surface for injecting time-varying scaling terms; ;in, This is a time-varying scaling factor; This is the initial scaling factor; The attenuation coefficient; is a power parameter; e is a constant.
8. The virtual game interaction recognition and analysis system according to claim 7, characterized in that, When conducting regulatory analysis on the sliding surface application effect of the fusion scaling item from different dimensions, the output reliability value and output stability value after the sliding surface application are calculated and analyzed together. If the output reliability value and output stability value both meet the corresponding output reliability requirements and output stability requirements, the sliding surface application effect of the fusion scaling item is judged to be normal. Conversely, if the application effect of the sliding surface of the fusion scaling item is abnormal, a warning will be issued.
9. A virtual game interaction recognition and analysis system according to claim 8, characterized in that, The expression for calculating the output reliability value is: Where SK is the output reliability value; These are the captured value, misjudged value, and missed detection value after the application of the sliding surface; These are the capture threshold, false positive threshold, and missed detection threshold, respectively; max() indicates that the maximum value is obtained. The expression for calculating the output stable value is: Where SW is the stable output value; N is the number of times the sliding surface output value is sampled; Let be the sliding surface output value of the i-th sample; i is the sampling index, i=1,2,3,…,N; This represents the average value of the output values from the sliding surface. To output a stable standard value.
10. A virtual game interaction recognition and analysis system according to claim 1, characterized in that, Different dimensions, including the output reliability dimension and output stability dimension of sliding surface application.