A three-dimensional rendering optimization method based on user behavior trend analysis

Abstract

A 3D rendering optimization method based on user behavior trend analysis includes: S1, collecting user interaction behavior data within a 3D scene to construct a user interaction time-series dataset; S2, performing time series analysis on the user interaction time-series dataset to identify the frequency and pattern of user interaction behavior, and analyzing and predicting user behavior trends in real time based on the identification results; S3, deriving the user's future operation path within the 3D scene based on the predicted user behavior trends, and identifying frequently accessed 3D model elements based on the future operation path. This invention provides a 3D rendering optimization method based on user behavior trend analysis, solving the technical problems in existing 3D rendering technologies that cannot fully utilize user behavior data and cannot optimize rendering based on user behavior trends, resulting in low rendering efficiency, resource waste, poor human-computer interaction experience, and ultimately affecting the display effect of 3D models and the actual application of related business systems.

Description

Technical Field

[0001] This invention relates to the field of 3D rendering technology, and in particular to a 3D rendering optimization method based on user behavior trend analysis. Background Technology

[0002] In many fields such as industrial production, scene visualization, and intelligent interaction, the demand for real-time rendering and smooth interaction of 3D scenes is increasing. How to maximize the performance of 3D rendering with limited system resources has become a research focus in this field.

[0003] To improve 3D rendering performance, various optimization schemes have been proposed in existing technologies. For example, some schemes perform passive rendering optimization based on the static technical parameters of 3D models. By using occlusion query technology to identify and remove complex models that are completely occluded and replace them with proxy geometry, and combining it with Level of Detail (LOD) technology to adjust the rendering resolution according to the distance between the model and the camera, the overall rendering smoothness is improved while ensuring a certain level of rendering quality. Other schemes rely on machine learning platforms to predict the optimal rendering cycle based on historical rendering parameter data. They then dynamically allocate rendering resources and select rendering methods according to the rendering cycle to optimize rendering efficiency. These schemes are often applied to specific 3D scenes with repetitive model elements, such as communication equipment rooms. In addition, in interactive scenarios such as augmented reality, existing technologies also construct dynamically updated digital twin models and use lightweight algorithms to optimize model data. Combined with indoor positioning technology, they achieve the overlay of virtual and real scenes and support interaction methods such as gestures and voice, easily adjusting the information display format according to user behavior.

[0004] However, existing 3D rendering optimization methods still have significant technical shortcomings, making it difficult to meet the dual demands of current 3D scenes for efficient rendering and a high-quality interactive experience. Firstly, existing optimization schemes do not fully collect and utilize actual user interaction data in 3D scenes, relying solely on static or inherent data such as model geometric features, camera positions, and historical rendering parameters as optimization criteria. This fails to capture the frequency, patterns, and trends of user behavior during 3D scene operations, resulting in a disconnect between rendering optimization and actual user needs. Secondly, existing technologies are all passive rendering optimizations, only able to adjust rendering parameters and allocate resources based on preset technical rules or historical data. The inability to predict users' future operation paths and resource needs based on their behavioral trends, and the inability to pre-configure resources for frequently accessed 3D model elements, results in slow loading speeds and poor display quality for frequently accessed elements, while infrequently accessed elements consume a large amount of system resources, causing serious resource waste and low overall rendering efficiency. Thirdly, due to the lack of analysis and adaptation to user behavior trends, the 3D scene cannot be dynamically adjusted according to users' operating habits, resulting in poor continuity and adaptability of the human-computer interaction process, poor user interaction experience, and thus seriously affecting the display effect of 3D models, limiting the practical application value of 3D rendering technology in various business systems.

[0005] In summary, how to fully utilize user interaction data in 3D scenes, analyze user behavior trends to achieve proactive and intelligent optimization of 3D rendering, rationally allocate rendering resources, and improve human-computer interaction experience while enhancing rendering efficiency and display quality has become a pressing technical problem to be solved in this field. Summary of the Invention

[0006] This invention provides a 3D rendering optimization method based on user behavior trend analysis, which solves the technical problems in existing 3D rendering technologies that cannot fully utilize user behavior data and cannot optimize rendering based on user behavior trends, resulting in low rendering efficiency, resource waste, poor human-computer interaction experience, and ultimately affecting the display effect of 3D models and the actual application of related business systems.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a 3D rendering optimization method based on user behavior trend analysis, comprising the following steps: S1. Collect user interaction behavior data in the 3D scene and construct a user interaction time series dataset; S2. Perform time series analysis on the user interaction time series dataset to identify the frequency and pattern of user interaction behavior, and analyze and predict user behavior trends in real time based on the identification results; S3. Based on the predicted user behavior trends, deduce the user's future operation path in the three-dimensional scene, and identify the three-dimensional model elements that the user frequently accesses according to the future operation path. S4. Perform targeted adjustments to the rendering parameters for the identified frequently accessed 3D model elements, and generate a list of adapted rendering parameters. S5. Import the rendering parameter list into the 3D rendering engine to complete the rendering configuration, and at the same time establish a rendering resource allocation strategy that matches the rendering parameters to complete the optimization of 3D rendering.

[0008] Preferably, the user interaction behavior data in step S1 includes the three-dimensional spatial coordinate data corresponding to the user's zoom, rotate, and click operations in the three-dimensional scene, as well as the timestamp data corresponding to each operation.

[0009] Data processing involves applying time series analysis to the collected data to identify patterns and frequencies of user behavior. This includes data cleaning, removing invalid and outlier values, and extracting meaningful features from consecutive timestamps, such as the time interval and duration of the behavior. Statistical analysis or machine learning techniques (such as cluster analysis and sequence pattern mining) are then used to identify common behavioral patterns.

[0010] The data is simplified by initializing a queue, inserting the root node of the entire scene (the whole scene) into the queue. Nodes are then extracted sequentially from the queue, and each node is evaluated to determine if further subdivision is needed. An evaluation function is applied to each node; if the node's level of detail exceeds a predetermined threshold, it is split into four child nodes; otherwise, it is marked as a leaf node. All leaf nodes are rendered, representing the final visible detail in the model.

[0011] Preferably, in step S2, an exponential smoothing time series analysis model is used to process the user interaction time series dataset, output the prediction results of user behavior trends in real time, and iteratively update the behavior pattern recognition results and behavior trend prediction results based on the newly collected interactive behavior data in real time.

[0012] Collect the coordinates and time points of the user's zoom, rotation, and click actions in the 3D scene on the web, using the following formula: ; The interaction data combination is obtained, where It is a coordinate. This refers to a specific time point. Time series analysis is performed on data combination D using the following formula: ; Calculate each time point Activity density, generating a behavior frequency model ;in, It is the attenuation parameter. It's a point in time. It is a point in time. The density of interactive activities. Using a behavior frequency model. To perform cluster analysis, define the formula: ; Identify patterns in user behavior ;in, It is a point in time. behavioral patterns It is a set of behavioral patterns. Based on the identified user behavior patterns The influence of each pattern is calculated using the following formula: ; Generate user behavior trend analysis results ;in, It is a behavioral pattern. It is the weight of the pattern. This is the result of user behavior trend analysis.

[0013] Preferably, the rendering parameters in step S4 include the rendering resolution and texture quality parameters of the 3D model elements; the directional adjustment specifically refers to improving the rendering resolution and texture quality of frequently accessed 3D model elements.

[0014] The frequently accessed 3D model elements are identified from the operation path prediction results (FP) using the following formula: ; Calculate the access priority of multiple elements to obtain an access priority list; where, This represents the weight of the i-th element. Indicates the distance from the current viewpoint. It is an attenuation factor used to adjust for the effect of distance. This represents the calculated access priority result. Based on the access priority list, the formula is: ; Set a dynamic value Select the elements whose rendering parameters need to be adjusted; among them, This represents the adjustment coefficient, set to 0.5 to filter out high-priority elements. This represents the maximum value in the access priority list. The threshold is calculated based on the dynamic threshold. The formula used is: ; Select those above the threshold The elements are used to generate the set of elements that need to be adjusted. ;in, This represents the set of elements whose rendering parameters need adjustment, including objects for resolution and texture quality enhancement. The set of elements requiring adjustment... To adjust the parameters, use the following formula: ; Generate a list of rendering parameter adjustments Where, adjust represents the adjustment function. and These represent the adjustment range for resolution and texture quality, respectively. This represents a list of rendering parameter adjustments.

[0015] Preferably, in step S3, when deriving the user's future operation path, the frequency of user interaction behavior, behavior pattern, and behavior trend prediction results are used as the core basis, and the spatial topological relationship of the three-dimensional scene is combined to determine the access order and access range of model elements for the user's future operation.

[0016] Preferably, in step S3, when identifying frequently accessed 3D model elements, the expected access frequency and expected dwell time of each 3D model element are calculated based on future operation paths. The access popularity score of each model element is obtained through a quantitative scoring model. A scoring threshold is set, and model elements with access popularity scores greater than or equal to the scoring threshold are identified as frequently accessed 3D model elements.

[0017] Input the user behavior trend analysis results into the processing unit, analyze the data to identify the transition probability of behavior patterns, use the transition probability to predict the user's future operation path and resource requirements in the 3D scene, and obtain the future operation path prediction results.

[0018] Node evaluation function design The node evaluation function is the core of the LOD algorithm; it determines whether to perform finer-grained segmentation on the current node. It assesses the roughness of the terrain by analyzing parameters such as elevation changes and slope, thus determining the LOD level. Nodes that do not require further subdivision are marked as leaf nodes, and their corresponding simplified models are used during rendering.

[0019] Obtaining future operation path prediction results Input user behavior trend analysis results In the data processing unit, the preliminary transition probability matrix is ​​calculated using the formula: ; Calculate the transition probabilities to obtain the preliminary transition probability matrix. ;in, Representing state arrive Number of observations This represents the total number of states. (This refers to the initial transition probability matrix.) Normalization is performed to ensure that the sum of probabilities for each row is 1, using the formula: ; Adjust the probability distribution to generate a normalized transition probability matrix. ;in, Represents the initial transition probability. This represents the normalized transition probability. Using the normalized transition probability matrix... The steady-state distribution is calculated using an iterative method: ; Until convergence, determine the long-term access probability of each state to obtain the steady-state distribution. ;in, Representing the The steady-state distribution of the next iteration. Using the steady-state distribution... and transition probability matrix Predicting future operational paths, calculation formula: ; The model predicts the user's future operation path and resource requirements in a 3D scene, obtaining the future operation path prediction result FP; where... Represents a steady-state distribution. This represents the normalized transition probability matrix.

[0020] Preferably, in step S5, when importing the rendering parameter list into the 3D rendering engine to complete the configuration, the adjusted rendering parameters are first associated and matched with the unique identifiers of the corresponding 3D model elements to generate a structured rendering parameter list, and then the structured rendering parameter list is imported into the 3D rendering engine to complete the batch update and configuration of the engine rendering parameters.

[0021] Preferably, the rendering resource allocation strategy in step S5 is as follows: prioritize allocating rendering computing power and video memory resources to frequently accessed 3D model elements, and proportionally reduce the rendering resource occupancy ratio of infrequently accessed 3D model elements to achieve dynamic optimization allocation of rendering resources.

[0022] Configure the 3D rendering engine using the application rendering parameter adjustment list, optimize the loading speed and display quality of frequently accessed elements, and establish an optimized resource allocation strategy.

[0023] Graphical Context Creation and Management Initialize the GraphicsContext, creating a separate GraphicsContext for each important view or frequently accessed element for dedicated resource management and optimization. Optimize memory usage for texture and geometry data by employing compression techniques and multi-level caching strategies to reduce memory consumption and improve access speed.

[0024] Optimize resource allocation strategy From the current configuration of the 3D rendering engine, modify the rendering parameters of each element according to the rendering parameter adjustment list, using the formula: ; in The adjusted configuration is calculated using the adjustment factor. ;in, Represents the weighting coefficient. Represents the element's original rendering parameters. Represents the adjustment value. This indicates the adjusted configuration result. Based on the adjusted configuration. The formula used is: ; The average value of the configuration items is used as the benchmark for resource allocation. Generate resource allocation benchmarks; among which, Represents the weighting coefficient. For the result of the i-th configuration, This represents the total number of configuration items. Use resource allocation benchmarks. Perform resource allocation filtering using the formula: ; Select a value greater than or equal to the baseline. The configuration options generate an optimized resource allocation strategy. ;in, This indicates the optimized resource allocation strategy selected, including all strategies that are at or above the baseline. Configuration items.

[0025] Preferably, it also includes a dynamic optimization closed-loop step: real-time collection of new user interaction behavior data, iterative updates of user behavior trend prediction results, identification results of frequently accessed 3D model elements, rendering parameter list and rendering resource allocation strategy, to achieve real-time dynamic optimization of 3D rendering and ensure the continuity of user interaction experience.

[0026] Preferably, the method is applied to one or more of the following scenarios: digital twin industrial scenarios, 3D display scenarios for power energy equipment, 3D visualization scenarios for communication equipment rooms, and general digital 3D interactive scenarios.

[0027] This invention provides a 3D rendering optimization method based on user behavior trend analysis. Addressing the core pain points of existing 3D rendering optimization technologies—namely, their inability to fully utilize user behavior data, the disconnect between rendering optimization and user needs, low rendering efficiency, significant resource waste, and poor human-computer interaction—this invention achieves a technological breakthrough from passive parameter adaptation to proactive demand prediction through deep integration of user behavior trend analysis and 3D rendering optimization. It offers the following beneficial effects: 1. Achieve precise analysis and trend prediction of user behavior, fundamentally solving the problem of the disconnect between rendering optimization and user needs. This patent collects the coordinates and time data of user zooming, rotating, and clicking operations in a 3D scene, and uses time series analysis to identify the frequency and patterns of user behavior. It can perform real-time and dynamic analysis and prediction of user behavior trends, allowing the rendering adjustments of the 3D scene to accurately match the actual operation needs of users. This completely breaks through the limitations of existing technologies that rely solely on static model parameters and historical rendering data for optimization, fully ensuring the continuity and satisfaction of the user interaction process.

[0028] 2. Achieve precise pre-configuration of rendering resources, significantly improving rendering efficiency and display quality while significantly reducing resource waste. This patent deduces the user's future operation path based on user behavior trend prediction results, accurately identifies frequently accessed 3D model elements, and specifically improves their core rendering parameters such as resolution and texture quality. Through pre-configured resource strategies, it ensures fast loading and excellent display effects for frequently accessed elements. At the same time, it establishes an optimized resource allocation strategy, prioritizing the allocation of system rendering resources to frequently accessed elements and proportionally reducing the resource occupation of low-frequency, non-accessed elements. Without relying on hardware performance upgrades or 3D engine capability upgrades, it maximizes rendering efficiency, effectively reduces system resource waste, and significantly improves the overall system performance.

[0029] 3. Achieve dynamic closed-loop optimization of rendering configuration, comprehensively enhancing human-computer interaction experience and business application value. This patent enables targeted configuration of the 3D rendering engine through an adjusted rendering parameter list. Simultaneously, based on real-time user-generated new interaction data, it dynamically iterates user behavior trend prediction results, high-frequency access element identification results, rendering parameters, and resource allocation strategies, forming a complete dynamic optimization closed loop. This solves the problems of interaction lag and poor adaptability caused by static and passive optimization in existing technologies, significantly improving the display effect of 3D models and the human-computer interaction experience. It effectively expands the practical application value of 3D rendering technology in various business systems such as digital twins, industrial equipment visualization, power energy scenarios, and 3D management of communication equipment rooms.

[0030] 4. The solution is highly versatile and applicable to a wide range of scenarios. The rendering optimization method of this patent is not limited to specific 3D scenes, but can be adapted to various application scenarios such as digital twin 3D scenes, general digital 3D interactive scenes, industrial equipment model display, and 3D visualization of communication equipment rooms. It has strong practical application value and industry adaptability. Attached Figure Description

[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0032] like Figure 1 As shown, a 3D rendering optimization method based on user behavior trend analysis includes the following steps: S1. Collect user interaction behavior data in the 3D scene and construct a user interaction time series dataset; S2. Perform time series analysis on the user interaction time series dataset to identify the frequency and pattern of user interaction behavior, and analyze and predict user behavior trends in real time based on the identification results; S3. Based on the predicted user behavior trends, deduce the user's future operation path in the three-dimensional scene, and identify the three-dimensional model elements that the user frequently accesses according to the future operation path. S4. Perform targeted adjustments to the rendering parameters for the identified frequently accessed 3D model elements, and generate a list of adapted rendering parameters. S5. Import the rendering parameter list into the 3D rendering engine to complete the rendering configuration, and at the same time establish a rendering resource allocation strategy that matches the rendering parameters to complete the optimization of 3D rendering.

[0033] This approach fundamentally solves the core problems of existing 3D rendering technologies failing to fully utilize user behavior data and the disconnect between rendering optimization and user needs. It achieves a shift from passive rendering optimization driven by static parameters to proactive rendering optimization driven by user behavior trends. Through a complete technical chain, it completes the entire process from interactive data collection, trend analysis, path prediction to rendering optimization, effectively improving 3D rendering efficiency, reducing system resource waste, significantly improving the human-computer interaction experience, ensuring the display effect of 3D models, and providing a complete and implementable overall solution for rendering optimization of various 3D scenes.

[0034] Preferably, the user interaction behavior data in step S1 includes the three-dimensional spatial coordinate data corresponding to the user's zoom, rotate, and click operations in the three-dimensional scene, as well as the timestamp data corresponding to each operation.

[0035] It can provide accurate, reliable and non-redundant core data sources for subsequent time series analysis, avoid the computing power overhead caused by invalid data collection, adapt to the real-time requirements of 3D rendering scenarios, and ensure that the collected data can fully reflect the user's core operational behaviors in the 3D scene. This lays a solid data foundation for accurately identifying user behavior frequency, patterns and predicting behavior trends, and ensures the accuracy of subsequent behavior analysis work.

[0036] Data processing involves applying time series analysis to the collected data to identify patterns and frequencies of user behavior. This includes data cleaning, removing invalid and outlier values, and extracting meaningful features from consecutive timestamps, such as the time interval and duration of the behavior. Statistical analysis or machine learning techniques (such as cluster analysis and sequence pattern mining) are then used to identify common behavioral patterns.

[0037] The data is simplified by initializing a queue, inserting the root node of the entire scene (the whole scene) into the queue. Nodes are then extracted sequentially from the queue, and each node is evaluated to determine if further subdivision is needed. An evaluation function is applied to each node; if the node's level of detail exceeds a predetermined threshold, it is split into four child nodes; otherwise, it is marked as a leaf node. All leaf nodes are rendered, representing the final visible detail in the model.

[0038] Preferably, in step S2, an exponential smoothing time series analysis model is used to process the user interaction time series dataset, output the prediction results of user behavior trends in real time, and iteratively update the behavior pattern recognition results and behavior trend prediction results based on the newly collected interactive behavior data in real time.

[0039] It can achieve real-time and efficient analysis and prediction of user behavior trends. The algorithm model has low computing power and fast response speed, which perfectly adapts to the low latency requirements of 3D rendering. The dynamic iterative update mechanism allows the behavior trend prediction results to keep up with the real-time changes in user operations, continuously refresh the behavior pattern recognition and trend prediction conclusions, ensure the timeliness and accuracy of trend analysis, and provide a reliable basis for accurate inference of the user's future operation path.

[0040] Collect the coordinates and time points of the user's zoom, rotation, and click actions in the 3D scene on the web, using the following formula: ; The interaction data combination is obtained, where It is a coordinate. This refers to a specific time point. Time series analysis is performed on data combination D using the following formula: ; Calculate each time point Activity density, generating a behavior frequency model ;in, It is the attenuation parameter. It's a point in time. It is a point in time. The density of interactive activities. Using a behavior frequency model. To perform cluster analysis, define the formula: ; Identify patterns in user behavior ;in, It is a point in time. behavioral patterns It is a set of behavioral patterns. Based on the identified user behavior patterns The influence of each pattern is calculated using the following formula: ; Generate user behavior trend analysis results ;in, It is a behavioral pattern. It is the weight of the pattern. This is the result of user behavior trend analysis.

[0041] Preferably, the rendering parameters in step S4 include the rendering resolution and texture quality parameters of the 3D model elements; the directional adjustment specifically refers to improving the rendering resolution and texture quality of frequently accessed 3D model elements.

[0042] This technology allows users to achieve superior display effects for the model elements they are most interested in, ensuring fast loading and clear display of frequently accessed elements. Unlike existing technologies that adjust parameters indiscriminately, it avoids the ineffective investment of resources in low-frequency accessed elements. While ensuring the overall smoothness of 3D scene rendering, it accurately meets users' visual experience needs for core model elements and enhances users' interactive experience in 3D scenes.

[0043] The frequently accessed 3D model elements are identified from the operation path prediction results (FP) using the following formula: ; Calculate the access priority of multiple elements to obtain an access priority list; where, This represents the weight of the i-th element. Indicates the distance from the current viewpoint. It is an attenuation factor used to adjust for the effect of distance. This represents the calculated access priority result. Based on the access priority list, the formula is: ; Set a dynamic value Select the elements whose rendering parameters need to be adjusted; among them, This represents the adjustment coefficient, set to 0.5 to filter out high-priority elements. This represents the maximum value in the access priority list. The threshold is calculated based on the dynamic threshold. The formula used is: ; Select those above the threshold The elements are used to generate the set of elements that need to be adjusted. ;in, This represents the set of elements whose rendering parameters need adjustment, including objects for resolution and texture quality enhancement. The set of elements requiring adjustment... To adjust the parameters, use the following formula: ; Generate a list of rendering parameter adjustments Where, adjust represents the adjustment function. and These represent the adjustment range for resolution and texture quality, respectively. This represents a list of rendering parameter adjustments.

[0044] Preferably, in step S3, when deriving the user's future operation path, the frequency of user interaction behavior, behavior pattern, and behavior trend prediction results are used as the core basis, and the spatial topological relationship of the three-dimensional scene is combined to determine the access order and access range of model elements for the user's future operation.

[0045] This makes the predicted operation path more closely match the actual spatial structure of the 3D scene, greatly improving the accuracy of path prediction. It effectively avoids optimization deviations caused by predictions that are out of touch with the scene structure, ensuring that the derived user operation path is highly matched with the actual operation requirements. This makes the identification of frequently accessed model elements more targeted and provides a guarantee for the accurate implementation of subsequent rendering optimizations.

[0046] Preferably, in step S3, when identifying frequently accessed 3D model elements, the expected access frequency and expected dwell time of each 3D model element are calculated based on future operation paths. The access popularity score of each model element is obtained through a quantitative scoring model. A scoring threshold is set, and model elements with access popularity scores greater than or equal to the scoring threshold are identified as frequently accessed 3D model elements.

[0047] The selection criteria for high-frequency elements have been quantified and standardized, completely avoiding the identification bias caused by subjective judgment. This ensures the objectivity and consistency of the selection of high-frequency access elements in different scenarios, allowing the identified model elements to truly reflect the user's core access needs. It also ensures that subsequent rendering parameter adjustments and resource allocation can accurately target the real core model elements, thereby improving the actual implementation effect of rendering optimization measures.

[0048] Input the user behavior trend analysis results into the processing unit, analyze the data to identify the transition probability of behavior patterns, use the transition probability to predict the user's future operation path and resource requirements in the 3D scene, and obtain the future operation path prediction results.

[0049] Node evaluation function design The node evaluation function is the core of the LOD algorithm; it determines whether to perform finer-grained segmentation on the current node. It assesses the roughness of the terrain by analyzing parameters such as elevation changes and slope, thus determining the LOD level. Nodes that do not require further subdivision are marked as leaf nodes, and their corresponding simplified models are used during rendering.

[0050] Obtaining future operation path prediction results Input user behavior trend analysis results In the data processing unit, the preliminary transition probability matrix is ​​calculated using the formula: ; Calculate the transition probabilities to obtain the preliminary transition probability matrix. ;in, Representing state arrive Number of observations This represents the total number of states. (This refers to the initial transition probability matrix.) Normalization is performed to ensure that the sum of probabilities for each row is 1, using the formula: ; Adjust the probability distribution to generate a normalized transition probability matrix. ;in, Represents the initial transition probability. This represents the normalized transition probability. Using the normalized transition probability matrix... The steady-state distribution is calculated using an iterative method: ; Until convergence, determine the long-term access probability of each state to obtain the steady-state distribution. ;in, Representing the The steady-state distribution of the next iteration. Using the steady-state distribution... and transition probability matrix Predicting future operational paths, calculation formula: ; The model predicts the user's future operation path and resource requirements in a 3D scene, obtaining the future operation path prediction result FP; where... Represents a steady-state distribution. This represents the normalized transition probability matrix.

[0051] Preferably, in step S5, when importing the rendering parameter list into the 3D rendering engine to complete the configuration, the adjusted rendering parameters are first associated and matched with the unique identifiers of the corresponding 3D model elements to generate a structured rendering parameter list, and then the structured rendering parameter list is imported into the 3D rendering engine to complete the batch update and configuration of the engine rendering parameters.

[0052] It achieves precise association and matching between the adjusted rendering parameters and the corresponding 3D model elements, effectively avoiding the rendering optimization failure problem caused by mismatch between parameters and model elements. The generation of the structured rendering parameter list enables the 3D rendering engine to efficiently complete batch parameter updates and configurations, improves the execution efficiency of rendering parameter configuration, and ensures that the adjusted rendering parameters can be implemented and take effect in the rendering engine quickly and accurately, so that the rendering optimization strategy can be transformed into the actual rendering effect in a timely manner.

[0053] Preferably, the rendering resource allocation strategy in step S5 is as follows: prioritize allocating rendering computing power and video memory resources to frequently accessed 3D model elements, and proportionally reduce the rendering resource occupancy ratio of infrequently accessed 3D model elements to achieve dynamic optimization allocation of rendering resources.

[0054] A clear rendering resource allocation strategy enables dynamic and refined allocation of system rendering resources. It prioritizes the allocation of limited rendering computing power and video memory resources to frequently accessed 3D model elements, while proportionally reducing the resource consumption of infrequently accessed elements. This fundamentally solves the problem of resource waste caused by unreasonable resource allocation in existing technologies, maximizes the utilization rate of system rendering resources, and achieves the optimal solution for rendering performance without relying on hardware performance improvements, ensuring the rendering efficiency and display quality of frequently accessed elements.

[0055] Configure the 3D rendering engine using the application rendering parameter adjustment list, optimize the loading speed and display quality of frequently accessed elements, and establish an optimized resource allocation strategy.

[0056] Graphical Context Creation and Management Initialize the GraphicsContext, creating a separate GraphicsContext for each important view or frequently accessed element for dedicated resource management and optimization. Optimize memory usage for texture and geometry data by employing compression techniques and multi-level caching strategies to reduce memory consumption and improve access speed.

[0057] Optimize resource allocation strategy From the current configuration of the 3D rendering engine, modify the rendering parameters of each element according to the rendering parameter adjustment list, using the formula: ; in The adjusted configuration is calculated using the adjustment factor. ;in, Represents the weighting coefficient. Represents the element's original rendering parameters. Represents the adjustment value. This indicates the adjusted configuration result. Based on the adjusted configuration. The formula used is: ; The average value of the configuration items is used as the benchmark for resource allocation. Generate resource allocation benchmarks; among which, Represents the weighting coefficient. For the result of the i-th configuration, This represents the total number of configuration items. Use resource allocation benchmarks. Perform resource allocation filtering using the formula: ; Select a value greater than or equal to the baseline. The configuration options generate an optimized resource allocation strategy. ;in, This indicates the optimized resource allocation strategy selected, including all strategies that are at or above the baseline. Configuration items.

[0058] Preferably, it also includes a dynamic optimization closed-loop step: real-time collection of new user interaction behavior data, iterative updates of user behavior trend prediction results, identification results of frequently accessed 3D model elements, rendering parameter list and rendering resource allocation strategy, to achieve real-time dynamic optimization of 3D rendering and ensure the continuity of user interaction experience.

[0059] By collecting new user interaction data in real time, continuously updating behavior trend prediction results, high-frequency element identification results, rendering parameters and resource allocation strategies, rendering optimization measures can be dynamically adjusted to keep up with changes in user operation behavior. This effectively avoids the problem of decreased interactive experience caused by static optimization mode, ensures the continuity and smoothness of user operation in 3D scene, and ensures that 3D rendering effect always matches the user's real-time needs.

[0060] Preferably, the method is applied to one or more of the following scenarios: digital twin industrial scenarios, 3D display scenarios for power energy equipment, 3D visualization scenarios for communication equipment rooms, and general digital 3D interactive scenarios. It overcomes the application limitations of existing 3D rendering optimization methods, which are often confined to specific scenarios. This user behavior trend analysis-based 3D rendering optimization method can be adapted to various scenarios such as digital twin industrial scenarios, 3D display scenarios for power energy equipment, 3D visualization scenarios for communication equipment rooms, and general digital 3D interactive scenarios. This significantly improves the universality and industry adaptability of the technical solution, expands the boundaries of its practical application, and enhances its commercial application value. It provides efficient and feasible optimization solutions for 3D scene rendering in different industries.

[0061] This invention abandons the passive 3D rendering optimization logic based on static model parameters in existing technologies. It pioneers an active rendering optimization system driven by user behavior trend analysis. By collecting core user interaction data in 3D scenes and using an exponential smoothing time series analysis model, it achieves accurate identification of user behavior frequency and patterns and real-time dynamic prediction of behavior trends. Combining the spatial topology of 3D scene, it derives the user's future operation path and quantifies and identifies frequently accessed 3D model elements. It then optimizes rendering parameters for high-frequency elements and establishes a dynamic allocation strategy for appropriate rendering resources. At the same time, it constructs a full-process dynamic optimization closed loop based on newly added user interaction data, realizing the transformation from technology parameter-driven to user demand-driven 3D rendering optimization. This allows rendering optimization to accurately match the actual operation needs of users, completely solving the problems of existing technologies that cannot fully utilize user behavior data, have low rendering efficiency, waste resources, and have poor human-computer interaction experience.

[0062] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A 3D rendering optimization method based on user behavior trend analysis, characterized in that, Includes the following steps: S1. Collect user interaction behavior data in the 3D scene and construct a user interaction time series dataset; S2. Perform time series analysis on the user interaction time series dataset to identify the frequency and pattern of user interaction behavior, and analyze and predict user behavior trends in real time based on the identification results; S3. Based on the predicted user behavior trends, deduce the user's future operation path in the three-dimensional scene, and identify the three-dimensional model elements that the user frequently accesses according to the future operation path. S4. Perform targeted adjustments to the rendering parameters for the identified frequently accessed 3D model elements, and generate a list of adapted rendering parameters. S5. Import the rendering parameter list into the 3D rendering engine to complete the rendering configuration, and at the same time establish a rendering resource allocation strategy that matches the rendering parameters to complete the optimization of 3D rendering.

2. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, The user interaction behavior data in step S1 includes the three-dimensional spatial coordinate data corresponding to the user's zoom, rotate, and click operations in the three-dimensional scene, as well as the timestamp data corresponding to each operation.

3. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, In step S2, an exponential smoothing time series analysis model is used to process the user interaction time series dataset, output the prediction results of user behavior trends in real time, and iteratively update the behavior pattern recognition results and behavior trend prediction results based on the newly collected interactive behavior data in real time.

4. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, The rendering parameters in step S4 include the rendering resolution and texture quality parameters of the 3D model elements; the directional adjustment specifically refers to improving the rendering resolution and texture quality of frequently accessed 3D model elements.

5. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, In step S3, when deriving the user's future operation path, the frequency of user interaction behavior, behavior pattern, and behavior trend prediction results are used as the core basis. Combined with the spatial topological relationship of the three-dimensional scene, the access order and access range of model elements for the user's future operation are determined.

6. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, In step S3, when identifying frequently accessed 3D model elements, the expected access frequency and expected dwell time of each 3D model element are calculated based on the future operation path. The access popularity score of each model element is obtained through a quantitative scoring model. A scoring threshold is set, and model elements with access popularity scores greater than or equal to the scoring threshold are identified as frequently accessed 3D model elements.

7. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, In step S5, when importing the rendering parameter list into the 3D rendering engine to complete the configuration, the adjusted rendering parameters are first associated and matched with the unique identifiers of the corresponding 3D model elements to generate a structured rendering parameter list. Then, the structured rendering parameter list is imported into the 3D rendering engine to complete the batch update and configuration of the engine's rendering parameters.

8. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, The rendering resource allocation strategy described in step S5 is as follows: prioritize allocating rendering computing power and video memory resources to frequently accessed 3D model elements, and proportionally reduce the proportion of rendering resources occupied by non-frequently accessed 3D model elements to achieve dynamic optimization allocation of rendering resources.

9. The 3D rendering optimization method based on user behavior trend analysis according to claim 1, characterized in that, It also includes dynamic optimization closed-loop steps: real-time collection of new user interaction behavior data, iterative updates of user behavior trend prediction results, identification results of frequently accessed 3D model elements, rendering parameter list and rendering resource allocation strategy, to achieve real-time dynamic optimization of 3D rendering and ensure the continuity of user interaction experience.

10. The 3D rendering optimization method based on user behavior trend analysis according to any one of claims 1-9, characterized in that, The method can be applied to one or more of the following scenarios: digital twin industrial scenarios, 3D display scenarios for power energy equipment, 3D visualization scenarios for communication equipment rooms, and general digital 3D interactive scenarios.

Images