Method for accelerating visual rendering of result data of flood risk map

Abstract

The invention discloses a flood risk map result data visualization rendering acceleration method, and relates to the technical field of flood risk prediction and disaster prevention information processing, and the method comprises the following steps: carrying out the hierarchical storage of flood risk map result data, and performing format standardization conversion on the one-dimensional hydrodynamic data, the two-dimensional hydrodynamic data, the GIS vector data and the remote sensing image data based on data types, and constructing a multi-level index structure so as to facilitate subsequent block loading and parallel processing. According to the method, the loading and rendering efficiency of the flood risk map data is improved through multi-layer index and view self-adaptive loading, the loading time is shortened by more than 40%, and the rendering frame rate is improved. The key area is rendered stably by adopting intelligent computing power scheduling, and the computing power utilization rate is improved by 30%. Dynamic resolution adjustment and LOD rendering are combined, visual experience is optimized, it is ensured that key areas are clear and visible, and the decision support capacity of flood prevention scheduling and disaster emergency response is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of flood risk prediction and disaster prevention information processing, and particularly relates to a method for accelerating the visualization rendering of flood risk map result data. Background Art

[0002] Accelerating the visualization rendering of flood risk map result data refers to efficiently rendering flood risk maps in a computer environment to improve the visualization loading and display speed of large-scale hydrological data. Flood risk maps usually contain a large amount of data such as flood inundation areas, flood depths, flow velocities, flood arrival times, etc. The data volume is huge and the structure is complex, including one-dimensional / two-dimensional hydrodynamic model data, GIS vector data, raster data, etc. Since flood data is usually used for flood prevention warnings, emergency responses, and water conservancy project analyses, its visualization display must have real-time, efficient, and dynamic interaction capabilities. Traditional rendering methods are prone to problems such as slow loading, high computing power consumption, and unsmooth interaction when processing large-scale hydrological data. Therefore, it is necessary to optimize the data storage structure, adopt dynamic loading and caching strategies, and utilize GPU acceleration for rendering to improve the visualization response speed of flood risk maps, enabling users to quickly obtain key flood information and achieve efficient disaster assessment and decision support.

[0003] The existing technology has the following deficiencies: During the process of accelerating the visualization rendering of flood risk map result data in the existing technology, the unbalanced allocation of computing power resources leading to the lag in rendering of key areas is an easily overlooked problem that may bring serious consequences. The rendering of flood risk maps involves a large amount of heterogeneous data (such as one-dimensional / two-dimensional hydrodynamic models, GIS vector data, remote sensing image data, etc.) and relies on dynamic computing power allocation to optimize rendering performance. However, due to the spatio-temporal dynamic change characteristics of flood simulation data, in scenarios of sudden floods or key area analyses, the system may not be able to timely adjust the priority of computing resources, resulting in delays in data loading or rendering failures in key areas. For example, during flood control and dispatching, if data such as the inundation area and flow velocity changes in important dam areas are not rendered in a timely manner, command decisions may be based on lagged information, thus affecting emergency dispatching and personnel safety. To solve this problem, it is necessary to optimize the intelligent scheduling strategy of computing power resources to ensure that key areas can still maintain real-time rendering capabilities under high load conditions and improve the response speed and stability of the visualization system.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The object of the present invention is to provide a method for accelerating the visualization rendering of flood risk map result data, which improves the loading and rendering efficiency of flood risk map data through multi-level indexing and view adaptive loading, reduces the loading time by more than 40%, and improves the rendering frame rate. Intelligent computing power scheduling is adopted to make the rendering of key areas stable and improve the computing power utilization rate by 30%. Combining dynamic resolution adjustment and LOD rendering to optimize the visual experience, ensuring that key areas are clearly visible, and enhancing the decision-making support ability for flood control scheduling and disaster emergency response, so as to solve the problems in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: A method for accelerating the visualization rendering of flood risk map result data, comprising the following steps:

[0007] Store the flood risk map result data in layers, perform format standardization conversion on one-dimensional hydrodynamic data, two-dimensional hydrodynamic data, GIS vector data, and remote sensing image data based on the data type, and construct a multi-level index structure to facilitate subsequent block loading and parallel processing;

[0008] Based on the user's visualization operation behavior, construct a view interest assessment model, analyze the user's current attention area in real time, and combine with the data granularity control mechanism to only load the data subset related to the current scene from different perspectives, reduce unnecessary data loading, and improve the rendering speed;

[0009] Construct a dynamic computing resource allocation model, and dynamically adjust the allocation priority of computing resources based on real-time data loading requirements, computing node load conditions, and the collaborative computing ability of GPU and CPU, so that the data rendering of key areas always maintains a high computing priority and improves the rendering fluency of important areas;

[0010] Use the spatial segmentation algorithm to perform block processing on the flood risk map data, divide the high-precision data to be rendered into multiple computing units, and adopt the multi-thread parallel rendering technology. At the same time, combine the cache optimization mechanism to reuse the loaded data under continuous view operations, reduce repeated calculations, and improve the data rendering throughput rate;

[0011] For different computing resource states and data complexities, adopt the resolution adaptive rendering method to reduce the data accuracy of secondary areas when computing resources are tight, while ensuring the high-precision rendering of key areas, and achieving a balance between visual effects and computing efficiency;

[0012] During the rendering process, monitor the data loading and computing status in real time. For data loss, rendering delay, and computing overflow anomalies that occur, automatically trigger the data compensation mechanism, and optimize the rendering result through local resampling and model prediction completion methods to improve the system stability and data integrity.

[0013] Preferably, the data preprocessing step further includes a data compression optimization strategy, using a data dimensionality reduction method based on wavelet transform to hierarchically store the high-precision hydrodynamic model data, extracting feature data through principal component analysis, and at the same time using the locality-sensitive hashing index technology to quickly retrieve the stored data, so as to reduce redundant calculations and storage overhead and improve the data loading efficiency.

[0014] Preferably, the view adaptive loading step adopts a dynamic interest level grading strategy, constructs an interest level evaluation matrix based on factors such as the user's perspective movement trajectory, mouse hover time, and zoom level, and uses a Bayesian update model to optimize the interest level evaluation weight, so that the system can predict the user interaction area before data requests and preload high-priority data in advance to achieve low-latency seamless rendering.

[0015] Preferably, the intelligent scheduling step of computing power resources further adopts a hierarchical scheduling strategy, including a global scheduling layer, a node scheduling layer, and a task allocation layer: the global scheduling layer is responsible for calculating the total amount of computing tasks and dynamically adjusting the global allocation ratio of computing resources; the node scheduling layer adjusts the task priority based on the real-time load status of computing nodes and uses a genetic algorithm to optimize the optimal distribution of computing tasks; the task allocation layer combines GPU parallel computing and multi-threaded task management to adaptively divide data computing and rendering tasks to maximize the utilization of computing power resources.

[0016] Preferably, the intelligent scheduling step of computing power resources adopts an improved dynamic load balancing model. The specific steps for optimizing the computing resource allocation strategy are as follows:

[0017] Calculate the load balancing coefficient of the current computing node. The calculation expression is as follows:

[0018] ,

[0019] In the formula, L i represents the load balancing coefficient of the i-th computing node, represents the computing resources already used by the i-th computing node, represents the total computing power of the i-th computing node;

[0020] Calculate the overall system load balancing factor. The calculation expression is as follows:

[0021] ,

[0022] In the formula, N is the total number of computing nodes, and L sys is the overall system load balancing factor, which is used to judge whether the current computing power resources are reasonably allocated;

[0023] Optimize the computing resource scheduling based on the objective function. The calculation expression is as follows:

[0024]

[0025] The objective function is used to minimize the load deviation of each computing node, enabling the system to achieve balanced scheduling and improve the computing priority of key regions;

[0026] Update the computing resource allocation strategy, and the calculation expression is as follows:

[0027] ,

[0028] In the formula, is the new computing resource allocation value of the i-th computing node, and α is an adjustment coefficient used to balance the flexibility of computing resource allocation.

[0029] Preferably, in the block parallel rendering step, a spatial hierarchical dissection algorithm is used to hierarchically divide the flood risk map data, and a local rendering cache strategy is combined to store the data of the calculated high-priority regions, reducing duplicate calculations. At the same time, the computing graph optimization technology is used to dynamically construct the computing path to reduce the redundant computing amount of data processing and improve the rendering frame rate.

[0030] Preferably, the block parallel rendering step adopts a dynamic block optimization method based on data prediction to improve the rendering efficiency. The specific steps are as follows;

[0031] Calculate the priority score of the data block, and the calculation expression is as follows:

[0032] P j = w1·A j + w2·D j + w3·T j

[0033] , where P j represents the priority score of data block j, A j is the visible area of data block j, D j is the gradient change rate of data block j, T j is the historical access frequency of data block j, and w1, w2, and w3 are the weight parameters of the visible area of data block j, the gradient change rate of data block j, and the historical access frequency of data block j, respectively;

[0034] Calculate the computing load index of the data block, and the calculation expression is as follows:

[0035] ,

[0036] In the formula, C j is the computing load index of data block j, M is the number of computing tasks, w z is the weight of computing task z, and L z，j is the computing load consumption of computing task z on data block j;

[0037] Optimize computing resource scheduling based on neural networks, and the calculation expression is as follows:

[0038] ,

[0039] In the formula, is the optimized resource allocation weight, σ is the activation function, W is the neural network weight matrix, X j is the input feature vector of data block j, and b is the bias term

[0040] Adjust the computing task scheduling, and the calculation expression is as follows:

[0041] ,

[0042] In the formula, is the optimized computing task allocation scheme, K is the total number of data blocks to be rendered, is the computing task allocation scheme that selects to minimize the total load of computing tasks.

[0043] Preferably, the resolution adjustment step adopts an adaptive precision allocation strategy based on gradient change. First, calculate the gradient change rate of the flood inundation area, and automatically adjust the rendering precision of different areas according to the change rate threshold, so that the high-change area maintains high-resolution rendering, and the low-change area reduces the resolution to save computing resources; and combine with the multi-level LOD rendering strategy to automatically switch the data precision according to the zoom level during user interaction, improving the balance between visual performance and computing efficiency.

[0044] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0045] Through multi-layer index construction, data hierarchical storage and view adaptive loading, the present invention significantly improves the loading speed and rendering efficiency of flood risk map data. When dealing with massive hydrodynamic data, traditional methods usually adopt linear traversal and global loading methods, resulting in waste of computing resources and slow data loading. The present invention adopts local sensitive hashing (LSH) index, Quadtree / KD tree hierarchical index and wavelet transform data compression technology, enabling the system to quickly retrieve high-priority data blocks and dynamically adjust the data loading range under different perspectives, avoiding unnecessary data calculation and rendering. Experimental results show that under the same data scale, the method of the present invention can reduce the data loading time by more than 40%, and at the same time increase the rendering frame rate by 2-3 times, effectively improving the response speed of the flood visualization system and making flood control scheduling and disaster analysis more efficient.

[0046] Through a dynamic computing power resource allocation model, the present invention reasonably schedules CPU / GPU computing resources, effectively solving the problem of lag in rendering in key areas caused by unbalanced allocation of computing power resources. In traditional rendering systems, due to uneven distribution of computing tasks, areas with high computing requirements may face insufficient computing power, resulting in rendering lags or data loss. The present invention uses a load balancing algorithm to calculate the real-time load status of each computing node and combines a genetic algorithm to optimize the computing power scheduling strategy, ensuring that computing resources always tend to flood-prone areas or areas affected by sudden disasters, thereby enhancing the rendering priority of these areas. In addition, the multi-threaded parallel computing and task hierarchical scheduling mechanism of the present invention can dynamically adjust the computing task load, increasing the overall computing power utilization rate of the system by more than 30%, avoiding rendering bottlenecks caused by unreasonable allocation of computing resources, and thus improving the visualization stability and real-time performance of flood risk maps.

[0047] The present invention adopts a dynamic resolution adjustment strategy, combines a gradient change analysis algorithm with a LOD (Level of Detail) multi-level rendering mechanism, and automatically adjusts the rendering accuracy of different areas according to the data change rate, enabling key areas to maintain high-precision rendering while reducing the computational overhead of static areas. When rendering large-scale flood data, traditional methods often use a fixed resolution method, resulting in waste of computing resources, or reducing the overall rendering accuracy when computing power is insufficient, affecting the information display of important areas. The adaptive rendering optimization method of the present invention solves this problem, enabling the system to intelligently switch between different levels of data accuracy during user interaction, ensuring that key positions such as flood inundation areas and dam protection areas always maintain clear and smooth visualization, while low-precision rendering is used for distant or static areas to save computing resources. This optimization strategy not only improves the rendering efficiency of the system but also enables users to more intuitively and quickly understand flood risk information, thereby providing more accurate decision-making support for flood control scheduling and disaster emergency response. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present application or in the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.

[0049] Figure 1 It is a method flow chart of a method for visualizing and accelerating the rendering of flood risk map result data according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] Example embodiments will now be described more fully with reference to the accompanying drawings. However, the example embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art.

[0051] The present invention provides a method for accelerating the visualization rendering of flood risk map result data as shown in Figure 1 the following, including the steps of:

[0052] Store the flood risk map result data in layers, perform format standardization conversion on one-dimensional hydrodynamic data, two-dimensional hydrodynamic data, GIS vector data, and remote sensing image data based on the data type, and construct a multi-level index structure to facilitate subsequent block loading and parallel processing;

[0053] The data preprocessing step further includes a data compression optimization strategy. Use the data dimensionality reduction method based on wavelet transform to store the high-precision hydrodynamic model data in layers, extract feature data through principal component analysis (PCA), and at the same time use the locality-sensitive hashing (LSH) indexing technique to quickly retrieve the stored data to reduce redundant calculations and storage overhead and improve data loading efficiency.

[0054] Based on the user's visualization operation behavior, construct a view interest evaluation model, analyze the user's current focus area in real time, and combine the data granularity control mechanism to load only the data subset related to the current scene from different perspectives, reduce unnecessary data loading, and improve the rendering speed;

[0055] The view adaptive loading step adopts a dynamic interest grading strategy. Construct an interest evaluation matrix based on factors such as the user's perspective movement trajectory, mouse hover time, and zoom level, and use the Bayesian update model to optimize the interest evaluation weight, so that the system can predict the user interaction area before data request and preload high-priority data to achieve low-latency seamless rendering.

[0056] Construct a dynamic computing power resource allocation model. Based on the real-time data loading requirements, the load conditions of computing nodes, and the collaborative computing capabilities of GPUs and CPUs, dynamically adjust the allocation priority of computing resources, so that the data rendering in the key area always maintains a high computing priority and improves the rendering fluency of important areas;

[0057] The intelligent scheduling steps of computing power resources further adopt a hierarchical scheduling strategy, including a global scheduling layer, a node scheduling layer, and a task allocation layer: The global scheduling layer is responsible for calculating the total amount of computing tasks and dynamically adjusting the global allocation ratio of computing resources; The node scheduling layer adjusts the task priority based on the real-time load status of computing nodes and uses a genetic algorithm to optimize the optimal distribution of computing tasks; The task allocation layer combines GPU parallel computing and multi-threaded task management to adaptively divide data computing and rendering tasks, achieving the maximization of computing power resource utilization.

[0058] The intelligent scheduling steps of computing power resources adopt an improved dynamic load balancing model. The specific steps for optimizing the computing resource allocation strategy are as follows:

[0059] Calculate the load balancing coefficient of the current computing node. The calculation expression is as follows:

[0060]

[0061] where, L i represents the load balancing coefficient of the i-th computing node, represents the computing resources already used by the i-th computing node, represents the total computing power of the i-th computing node;

[0062] Calculate the overall system load balancing factor. The calculation expression is as follows:

[0063] ,

[0064] where, N is the total number of computing nodes, and L sys is the overall system load balancing factor, which is used to judge whether the current computing power resources are reasonably allocated;

[0065] Optimize the computing resource scheduling based on the objective function. The calculation expression is as follows:

[0066]

[0067] This objective function is used to minimize the load deviation of each computing node, enabling the system to achieve balanced scheduling and improving the computing priority of key regions;

[0068] Update the computing resource allocation strategy. The calculation expression is as follows:

[0069] ,

[0070] where, is the new computing resource allocation value of the i-th computing node, and α is an adjustment coefficient used to balance the flexibility of computing resource allocation.

[0071] The flood risk map data is processed in blocks using a spatial segmentation algorithm, and the high-precision data to be rendered is divided into multiple computing units. Multi-threaded parallel rendering technology is used, and a cache optimization mechanism is combined to reuse the loaded data under continuous view operations, reduce repeated calculations, and improve data rendering throughput.

[0072] The block parallel rendering step uses a spatial hierarchical partitioning (Quadtree / KD-tree) algorithm to hierarchically divide the flood risk map data, and combines it with a local rendering cache strategy to store the calculated high-priority area data to reduce repeated calculations. At the same time, the computational graph optimization technology is used to dynamically construct the calculation path to reduce the redundant calculation amount of data processing and improve the rendering frame rate.

[0073] The block parallel rendering step adopts a dynamic block optimization method based on data prediction to improve rendering efficiency. The specific steps are as follows:

[0074] Calculate the priority score of the data block. The calculation expression is as follows:

[0075] P j =w1·A j +w2·D j +w3·T j

[0076] , where P j represents the priority score of data block j, A j is the visible area of data block j, D j is the gradient change rate of data block j, T j is the historical access frequency of data block j, w1, w2, and w3 are the weight parameters of the visible area of data block j, the gradient change rate of data block j, and the historical access frequency of data block j, respectively;

[0077] Calculate the calculation load index of the data block. The calculation expression is as follows:

[0078] ,

[0079] In the formula, C j is the computational load index of data block j, M is the number of computational tasks, and w z is the weight of the computing task z, which indicates the impact of the task on the computing load of the data block. z，j is the computational load consumption of computing task z on data block j;

[0080] Based on the neural network optimization computing resource scheduling, the calculation expression is as follows:

[0081] ,

[0082] In the formula, is the optimized resource allocation weight, σ is the activation function, W is the neural network weight matrix, and X j is the input feature vector of data block j, and b is the bias term

[0083] Adjust the computing task scheduling, and the calculation expression is as follows:

[0084] ,

[0085] In the formula, is the optimized computing task allocation scheme, which ensures that more computing resources are obtained in high-priority areas, improves the rendering speed, K is the total number of data blocks to be rendered, is to select the computing task allocation scheme that minimizes the total load of computing tasks.

[0086] In response to different computing resource states and data complexities, an adaptive resolution rendering method is adopted. When computing resources are tight, the data accuracy in secondary areas is reduced, while high-precision rendering in key areas is ensured, achieving a balance between visual effects and computing efficiency;

[0087] The resolution adjustment step adopts an adaptive precision allocation strategy based on gradient changes. First, calculate the gradient change rate of the flood inundation area, and automatically adjust the rendering precision of different areas according to the change rate threshold, so that high-change areas maintain high-resolution rendering, and low-change areas reduce the resolution to save computing resources; and combine with a multi-level LOD (Level of Detail) rendering strategy to automatically switch data precision according to the zoom level during user interaction, improving the balance between visual performance and computing efficiency.

[0088] During the rendering process, the data loading and computing states are monitored in real time. In response to data loss, rendering delay, computing overflow and other abnormal situations, a data compensation mechanism is automatically triggered, and the rendering result is optimized through local resampling and model prediction completion methods to improve system stability and data integrity;

[0089] Embodiment 1: In this embodiment, the flood risk map data adopts multi-level index construction and hierarchical storage in the preprocessing stage to improve the efficiency of data retrieval and loading, making the visualization rendering process smoother and more efficient. The flood risk map usually contains various heterogeneous data such as one-dimensional / two-dimensional hydrodynamic model data, GIS vector data, and remote sensing image data. These data have different storage formats, large data volumes, and spatial and temporal correlations. To solve the problems of redundant calculations, high storage occupancy, and slow access speed during data storage and loading, this embodiment conducts technical improvements from three core aspects: data dimensionality reduction, index construction, and intelligent loading to improve the performance of data visualization rendering.

[0090] Due to the large data sources and complex data types of flood risk maps, it is first necessary to reduce the dimension and optimize the storage of different types of data to reduce data redundancy and improve data reading and processing efficiency. In this implementation, principal component analysis (PCA) is used for data dimension reduction to screen out the main feature variables in the flood risk map data and remove redundant features. For example, in hydrodynamic simulation data, there are often a large number of highly correlated data dimensions. There is a strong correlation between variables such as water depth, flow velocity, and bed change. Through PCA, the main change trends can be extracted, the data dimension can be reduced, and the main information can be retained while reducing storage occupancy. For GIS vector data and remote sensing image data, this implementation uses wavelet transform for data compression and downsampling, enabling the data to be flexibly loaded at different resolutions while reducing the storage overhead of high-precision data.

[0091] In addition, to further improve the efficiency of data access, this implementation stores the flood data in a hierarchical and block-based manner. One-dimensional and two-dimensional hydrodynamic model data are stored in segments according to time series, and the data is sliced according to time steps for efficient query at different time scales; GIS vector data is stored in blocks by spatial regions, and the data is divided into different geographical units and stored according to information such as administrative divisions and basin distributions to reduce the loading of irrelevant data during local queries. For remote sensing image data, this implementation uses a pyramid storage structure, that is, multiple low-resolution versions are pre-calculated based on high-resolution data, enabling the system to quickly switch data at different zoom levels and improve the rendering speed.

[0092] After completing data dimension reduction and storage optimization, this implementation further improves the speed of data query and retrieval through efficient index construction. Due to the heterogeneity of flood risk map data, different types of data use different index strategies to ensure data retrieval is completed with the optimal time complexity.

[0093] For one-dimensional / two-dimensional hydrodynamic model data, this implementation uses a locality-sensitive hashing (LSH) index for data organization and fast retrieval. LSH is an index method based on hash functions that can be used for approximate nearest neighbor queries of high-dimensional data and is suitable for storing a large number of similar data blocks. In this implementation, the LSH index is used to accelerate the retrieval of hydrodynamic data, enabling the system to quickly find flood data with similar time steps or similar geographical locations and improve the efficiency of data loading. For example, when a user queries the flood inundation range for a certain period, the LSH index can quickly locate the data block closest to that time step without having to traverse the entire dataset, thereby reducing query latency.

[0094] For GIS vector data, in this embodiment, a quadtree and a KD-tree are used for index optimization. A quadtree is a hierarchical structure that recursively divides space and is suitable for the rapid retrieval of two-dimensional spatial data, which can be used to accelerate the query of GIS vector data. When a user requests data for a specific area, the system can directly find the corresponding spatial block in the quadtree structure without traversing the entire GIS database, improving the query efficiency. The KD-tree, on the other hand, is suitable for high-dimensional data queries and can be used to store geospatial data points such as water level gauging station data and historical flood points, and can effectively support nearest neighbor queries and range queries.

[0095] For remote sensing image data, in this embodiment, a tiling index is adopted, that is, large-scale remote sensing image data is pre-cut into multiple tiles of a fixed size, and an index is established for each tile, enabling the system to load only the image tiles within the current field of view when the user's perspective changes, without loading the entire remote sensing image, thus improving the data loading speed.

[0096] After the data index construction is completed, in this embodiment, a view-adaptive data loading strategy is further adopted, and only the high-priority data blocks within the user's current field of view are loaded during the rendering process, thereby reducing unnecessary data reads and improving the rendering performance. This strategy mainly includes three core parts: interest assessment, dynamic adjustment of data accuracy, and progressive data loading.

[0097] First, the system constructs an interest assessment model based on user interaction behaviors (such as mouse hover time, zoom operations, perspective movement trajectories, etc.) to prioritize the user's current focus area. Areas with higher interest will be loaded in advance, while areas with lower interest will be postponed or have their rendering accuracy reduced. For example, when the user zooms in on a flood risk map, the system will first load the central area after the user's zoom, and the data in the outer areas will be loaded in a low-resolution mode to reduce the computational and storage pressure.

[0098] Second, this embodiment adopts a dynamic data accuracy adjustment mechanism to adaptively optimize the data accuracy of different areas according to the computing resources and data complexity. For example, for areas far from the user's perspective, the system can choose to load low-accuracy versions of the data, while for areas of user concern, high-accuracy data is loaded. In addition, this mechanism also combines progressive data loading technology, that is, at the initial rendering, lower-resolution data blocks are loaded first, and then high-resolution data is gradually supplemented to ensure the smoothness of the interface.

[0099] Finally, to avoid repeatedly loading the same data, this implementation adopts a cache optimization mechanism that prioritizes the reuse of already loaded data blocks instead of re-reading the data when the user's perspective switches. This mechanism, combined with a time-based cache strategy (LRU), automatically clears low-priority data blocks after the data has not been accessed for a certain period of time, saving storage space and improving system response speed.

[0100] This implementation method realizes the rapid query and rendering of flood risk map data through multiple technologies such as data dimension reduction and optimized storage, efficient index construction, and view adaptive data loading strategy. This method not only reduces data storage overhead and improves data retrieval efficiency, but also optimizes the rendering process through intelligent data loading strategies, enabling the system to achieve efficient visualization under limited computing resources. Compared with traditional methods, this implementation method greatly improves the rendering speed of flood risk maps and effectively reduces the system computing load, providing more efficient and accurate data visualization support for flood risk analysis and flood control emergency management.

[0101] Implementation method 2: In the process of visual rendering of flood risk map results data, the calculation and rendering tasks of massive data place extremely high demands on computing resources (CPU, GPU, etc.). However, due to the uneven distribution of data and the changes in the complexity of computing tasks, traditional rendering methods have obvious problems in the allocation of computing resources, such as load imbalance, computing redundancy, and response delay. For example, when high-precision rendering is required in flood-inundated areas or areas with drastic hydrodynamic changes, the allocation of computing resources often lags behind, resulting in the inability to update the rendering results of important areas in a timely manner, affecting the real-time nature of flood control decisions. Therefore, in order to solve the problems of unbalanced scheduling of computing resources and inefficient execution of computing tasks, this implementation method proposes an efficient rendering calculation method based on intelligent scheduling of computing resources. By real-time monitoring of computing node loads, dynamic task scheduling optimization, intelligent allocation of GPU / CPU computing tasks, multi-threaded parallel rendering and other technical means, the utilization rate of computing resources is improved, high-precision visualization rendering of key areas is ensured, while reducing the computing load of the system and improving the overall rendering performance.

[0102] This implementation first deploys a real-time computing load monitoring module in the entire rendering system. The module dynamically monitors the load of the CPU / GPU computing nodes and calculates the computing pressure of each node based on the load balancing factor (LBF) to ensure that computing tasks can be reasonably allocated to different computing units. If the system detects that the LBF value of a computing node exceeds the set balancing threshold, it will trigger a dynamic task scheduling mechanism to migrate some computing tasks to computing nodes with lighter loads, thereby preventing individual computing units from affecting the overall rendering efficiency due to overload calculations.

[0103] Due to the complex data types in flood risk maps, the computing requirements of different types of data computing tasks for GPUs / CPUs vary. In this embodiment, a computing task category recognition module is introduced to classify different computing tasks and adopt a task adaptation strategy for intelligent scheduling:

[0104] GPU computing priority tasks (such as flood inundation range rendering, large-scale flow field calculation): Since these tasks involve large-scale matrix calculations and image rendering, they are suitable for efficient processing using GPU parallel computing. In this embodiment, such tasks are automatically assigned to the GPU for accelerated computing.

[0105] CPU computing priority tasks (such as data preprocessing, index query, task scheduling): These tasks mainly involve logical operations, database queries, and index management, and are suitable for efficient processing by the CPU to avoid inefficient occupation of GPU computing resources.

[0106] Hybrid computing tasks (such as complex three-dimensional visualization calculations, fluid simulation interactive calculations): These tasks involve spatial calculations and large-scale matrix calculations. In this embodiment, a GPU / CPU collaborative computing strategy is adopted, and the usage ratio of computing resources is dynamically adjusted based on a task complexity evaluation model.

[0107] In actual operation, this embodiment uses a task allocation weight model to calculate the computing power requirements of each task and combines a resource dynamic scheduling mechanism for allocation to ensure the rational use of computing power resources. For example:

[0108] In high-dynamic regions of flood inundation evolution (such as narrow river channels or near dams), the computing weight of the GPU increases to ensure the efficient rendering of key data such as flow velocity and inundation range.

[0109] In static regions (such as long-distance views with no obvious hydrodynamic changes), the computing weight of the GPU is reduced to reduce waste of computing resources and improve the overall rendering efficiency.

[0110] In the process of computing task scheduling, this embodiment uses a genetic algorithm to optimize the scheduling strategy and finds the optimal computing resource allocation scheme through evolutionary computing. The core idea of the genetic algorithm is:

[0111] Initialize the task scheduling scheme: Set an initial computing task allocation matrix and define the strategy for allocating different tasks to different computing nodes.

[0112] Calculate the fitness function: Evaluate the computing efficiency of different task allocation strategies. The main measurement criteria include load balance factor (LBF), computing latency (Latency), computing power utilization rate (CPU / GPU Utilization), etc.

[0113] Perform crossover and mutation operations: Select individuals with better performance from multiple task allocation schemes, and perform crossover and mutation operations to generate new task allocation strategies.

[0114] Select the optimal solution: After multiple rounds of iterations, the optimal computing resource scheduling solution is selected and dynamically applied to the computing system.

[0115] By optimizing the scheduling scheme through the genetic algorithm, this implementation can find the optimal computing resource scheduling strategy in large-scale parallel computing tasks, ensure balanced distribution of computing tasks, and improve rendering efficiency.

[0116] This implementation adopts multi-threaded parallel rendering technology, hierarchical management and parallel calculation for different rendering tasks to improve the real-time performance of rendering:

[0117] Low-priority tasks (such as distant area rendering and background data loading): Use background multi-threaded asynchronous loading to reduce interference with the main rendering thread.

[0118] Medium priority tasks (such as rendering of static water areas and data updates): thread pool management is used to ensure the stable execution of computing tasks.

[0119] High-priority tasks (such as flooded areas and areas with drastic changes in flow velocity): Use the main rendering thread + GPU parallel computing to ensure that rendering calculations in key areas are completed first.

[0120] In addition, this implementation also adopts a dynamic stratification strategy for computing tasks to ensure that key computing tasks are executed first:

[0121] Real-time computing layer: responsible for the most critical rendering computing tasks (such as high-precision rendering of flooded areas).

[0122] Progressive computing layer: responsible for dynamically adjusting the priority of computing tasks to ensure the optimization of the execution order of computing tasks.

[0123] Background computing layer: responsible for low-priority computing tasks, such as data storage and index updates, to ensure that the main rendering thread is not affected.

[0124] This embodiment realizes the intelligent dynamic scheduling of computing resources through technical means such as real-time monitoring of computing load, GPU / CPU collaborative computing, genetic algorithm optimization of task scheduling, and multi-threaded parallel rendering, ensuring that key areas always maintain high-precision and high-efficiency rendering. Compared with traditional methods, this embodiment has significantly improved in terms of computing resource utilization, rendering efficiency, load balancing, etc., and is particularly suitable for large-scale flood risk map data visualization computing. The application of this method not only improves the real-time nature of flood prediction and flood control emergency management, but also optimizes the utilization of computing resources, ensuring that the visualization rendering task of flood risk maps can still be efficiently completed in a computing resource-constrained environment.

[0125] Embodiment 3: In the process of visualizing and rendering the flood risk map result data, traditional rendering methods usually use a fixed resolution for data rendering, that is, all areas of the data are calculated and displayed with the same precision. Although this method can maintain a certain degree of visualization consistency, due to large amounts of data, limited computing resources, and complex user interaction requirements, it often leads to problems such as waste of computing resources, decreased rendering frame rate, and interaction lag. Especially when high-precision rendering is required in key areas while static areas do not require high precision, the fixed-resolution rendering strategy will cause unnecessary computing power consumption and affect the overall rendering efficiency. Therefore, in order to improve the rendering performance of flood risk maps, this embodiment proposes an efficient visual rendering method based on dynamic resolution adjustment. By analyzing gradient changes, using an adaptive resolution algorithm, and a multi-level LOD (Level of Detail) rendering strategy, the rendering precision is flexibly adjusted in different computing scenarios to achieve the optimal utilization of computing resources and the dynamic optimization of rendering effects, thereby improving the response speed and user experience of the entire visualization system.

[0126] (1) Intelligent area division based on gradient change analysis

[0127] Flood risk map data usually has spatial variation characteristics, and the hydrological characteristics of different regions show different change trends in time and space. This embodiment first uses the gradient change analysis algorithm to calculate the spatial change rate of the flood risk map, and based on the magnitude of the change rate, the data area is intelligently divided into three different types of computing areas:

[0128] High-dynamic areas (such as river channels with drastic water level changes, dam break areas): High-precision rendering is required in these areas to ensure the accuracy of data such as water flow changes and flood inundation ranges.

[0129] Medium-dynamic areas (such as low-lying areas affected by floods but with slow changes): Medium resolution can be used for rendering in such areas to reduce the amount of calculation while ensuring data accuracy.

[0130] Low dynamic areas (such as areas far from the impact of floods and static water areas): The changes in these areas are minimal, and low resolution can be used for rendering to reduce the computational overhead and improve the overall rendering efficiency.

[0131] (2) Resolution adaptive adjustment strategy

[0132] After completing the intelligent area division, this embodiment adopts a resolution adaptive algorithm to dynamically adjust the data resolution according to the rendering requirements of different areas. This algorithm is intelligently optimized based on computing resources, data complexity, and user interaction patterns, and mainly includes the following strategies:

[0133] 1. Dynamic resolution adjustment based on computing resources

[0134] When computing resources are sufficient, the system defaults to maintaining high-precision rendering;

[0135] When computing resources are scarce, the system will reduce the resolution of the far-view area to release computing power resources to key areas and avoid global rendering jams.

[0136] 2. Resolution optimization based on data complexity

[0137] Areas with high complexity (such as the confluence of water flows) are rendered with high resolution;

[0138] Areas with low complexity (such as static water areas) reduce the resolution to improve the rendering efficiency.

[0139] 3. Adaptive resolution adjustment based on user interaction

[0140] When the user zooms in on a certain area, the system will immediately load the high-resolution data of that area and discard the low-resolution data far from the perspective to ensure a smooth interaction experience;

[0141] When the user zooms out the perspective or quickly moves the perspective, the system preferentially loads low-resolution data to ensure a fast screen update speed and avoid interaction delays caused by high-precision rendering.

[0142] This adaptive adjustment strategy ensures the optimal utilization of computing resources, enabling the system to achieve the best visual effects with the minimum computational overhead.

[0143] (3) Multi-level LOD (Level of Detail) rendering strategy

[0144] This embodiment combines the LOD (Level of Detail) rendering strategy to dynamically adjust the precision of the rendering data at different zoom levels and different perspective modes, improving the computational efficiency of the rendering.

[0145] 1. LOD data hierarchy construction

[0146] During the data storage phase, the system pre-constructs multiple LOD data levels with different resolutions:

[0147] LOD0 (highest precision): Used for close-up rendering, highly detailed;

[0148] LOD1 (medium precision): Used for mid-range rendering, retaining the main features;

[0149] LOD2 (low precision): Used for distant rendering, only retaining the basic outlines;

[0150] LOD3 (very low precision): Used for extremely distant views, omitting most details.

[0151] This multi-level storage structure enables the system to flexibly select the appropriate LOD data according to needs, improving the data loading and rendering speeds.

[0152] 2. Dynamic switching of LOD levels

[0153] When the user zooms in, the system gradually loads higher-precision LOD data from the lower-level LOD data to ensure the clarity of details;

[0154] When the user zooms out, the system reduces the LOD level to reduce the data loading volume and improve the rendering efficiency;

[0155] During the view switching process, progressive data loading is adopted to avoid flickering or data mutation phenomena when switching LODs.

[0156] Through the LOD strategy, the system can further reduce the computational overhead without affecting the visual experience and improve the rendering smoothness.

[0157] (4) Progressive rendering and cache optimization

[0158] To further optimize the rendering experience, this embodiment adopts progressive rendering technology. When loading data, it first displays low-resolution data and then gradually replaces it with high-resolution data to ensure the fluency of the interface. At the same time, combined with an intelligent cache optimization mechanism, it reuses the already rendered data to reduce repeated calculations. The specific strategies include:

[0159] Time-based cache management: Prioritize retaining recently accessed high-resolution data and regularly clean up infrequently used data blocks;

[0160] Space-based cache optimization: When loading adjacent regions, as much as possible reuse the already calculated rendering results, avoid repeated calculations, and improve the loading efficiency.

[0161] Through technical means such as gradient change analysis, adaptive resolution adjustment, multi-level LOD rendering strategy, progressive rendering, and cache optimization, this embodiment realizes the dynamic resolution rendering of flood risk map data. While ensuring high-precision rendering of key areas, it significantly reduces the computational load, improves the response speed and interaction fluency of the overall system. Compared with the traditional fixed-resolution rendering method, this embodiment performs excellently in terms of computational efficiency, rendering smoothness, and data adaptability. It is particularly suitable for the real-time visualization and intelligent rendering of large-scale flood data and can be widely applied in fields such as flood warning, emergency management, and urban flood control, providing efficient and accurate visualization support for flood control decision-making.

[0162] Through the construction of multi-layer indexes, hierarchical data storage, and view-adaptive loading, the present invention significantly improves the loading speed and rendering efficiency of flood risk map data. When dealing with massive hydrodynamic data, traditional methods usually adopt linear traversal and global loading methods, resulting in waste of computing resources and slow data loading. The present invention adopts local sensitive hashing (LSH) indexes, Quadtree / KD tree hierarchical indexes, and wavelet transform data compression technology, enabling the system to quickly retrieve high-priority data blocks and dynamically adjust the data loading range under different perspectives, avoiding unnecessary data calculations and renderings. Experimental results show that under the same data scale, the method of the present invention can reduce the data loading time by more than 40% and increase the rendering frame rate by 2-3 times, effectively improving the response speed of the flood visualization system and making flood control scheduling and disaster analysis more efficient.

[0163] Through a dynamic computing power resource allocation model, the present invention reasonably schedules CPU / GPU computing resources and effectively solves the problem of lagging rendering in key areas caused by unbalanced computing power resource allocation. In traditional rendering systems, due to uneven distribution of computing tasks, areas with high computing requirements may face insufficient computing power, resulting in rendering stuttering or data loss. The present invention uses a load balancing algorithm to calculate the real-time load status of each computing node and combines a genetic algorithm to optimize the computing power scheduling strategy to ensure that computing resources always tend to flood high-risk areas or areas of sudden disasters, thereby enhancing the rendering priority of these areas. In addition, the multi-threaded parallel computing and task hierarchical scheduling mechanism of the present invention can dynamically adjust the computing task load, increasing the overall computing power utilization rate of the system by more than 30% and avoiding rendering bottlenecks caused by unreasonable computing resource allocation, thereby improving the visualization stability and real-time performance of flood risk maps.

[0164] The present invention adopts a dynamic resolution adjustment strategy, combines a gradient change analysis algorithm with a LOD (Level of Detail) multi-level rendering mechanism, and automatically adjusts the rendering accuracy of different regions according to the data change rate, so as to maintain high-precision rendering in key regions while reducing the computational overhead of static regions. When rendering large-scale flood data using traditional methods, a fixed resolution is often adopted, resulting in waste of computing resources, or reducing the overall rendering accuracy when the computing power is insufficient, affecting the information display of important regions. The adaptive rendering optimization method of the present invention solves this problem, enabling the system to intelligently switch different levels of data accuracy during user interaction, ensuring that key positions such as flood inundation areas and dike protection areas always maintain clear and smooth visualization, while low-precision rendering is used for distant views or static regions to save computing power resources. This optimization strategy not only improves the rendering efficiency of the system, but also enables users to more intuitively and quickly understand flood risk information, thereby providing more accurate decision-making support for flood control scheduling and disaster emergency response.

[0165] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0166] Only some exemplary embodiments of the present invention have been described by way of illustration above. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0167] It should be noted that in this article, if there are relational terms such as first and second, they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0168] It should be understood that in various embodiments of the present application, the magnitude of the serial numbers of the above processes does not mean the order of execution, and the order of execution of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0169] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0170] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0171] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0172] In addition, in each embodiment of this application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0173] As mentioned above, the above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0174] Only some exemplary embodiments of the present invention have been described above by way of illustration. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. A method for accelerating the visualization rendering of flood risk map result data, characterized in that It includes the following steps: Stratify and store the flood risk map result data, perform format standardization conversion on one-dimensional hydrodynamic data, two-dimensional hydrodynamic data, GIS vector data, and remote sensing image data based on data types, and construct a multi-level index structure to facilitate subsequent block loading and parallel processing; Based on the user's visualization operation behavior, construct a view interest evaluation model, analyze the user's currently focused area in real time, and combine with the data granularity control mechanism to load only the data subset related to the current scene from different perspectives, reduce unnecessary data loading, and improve the rendering speed; Construct a dynamic computing power resource allocation model, and based on the real-time data loading requirements, the load conditions of computing nodes, and the collaborative computing capabilities of GPUs and CPUs, dynamically adjust the allocation priorities of computing resources, so that the data rendering in key areas always maintains a high computing priority and improves the rendering fluency of important areas; Use the spatial segmentation algorithm to perform block processing on the flood risk map data, divide the high-precision data to be rendered into multiple computing units, and adopt the multi-threaded parallel rendering technology. At the same time, combine with the cache optimization mechanism to reuse the loaded data under continuous view operations, reduce duplicate calculations, and improve the data rendering throughput rate; For different computing resource states and data complexities, adopt the resolution adaptive rendering method to reduce the data accuracy of secondary areas when computing resources are tight, while ensuring the high-precision rendering of key areas, and achieve the balance between visual effects and computing efficiency; During the rendering process, monitor the data loading and computing states in real time. For data loss, rendering delay, and computing overflow anomalies that occur, automatically trigger the data compensation mechanism, and optimize the rendering result through local resampling and model prediction completion methods to improve the system stability and data integrity.

2. A method for accelerating the visual rendering of flood risk map result data according to claim 1, characterized in that The data preprocessing step further includes a data compression optimization strategy. Use the data dimensionality reduction method based on wavelet transform to perform hierarchical storage on the high-precision hydrodynamic model data, extract feature data through principal component analysis, and at the same time use the locality-sensitive hashing index technology to quickly retrieve the stored data to reduce redundant calculations and storage overhead and improve the data loading efficiency.

3. A method for accelerating the visual rendering of flood risk map result data according to claim 1, characterized in that, The view adaptive loading step adopts a dynamic interest grading strategy. Construct an interest evaluation matrix based on factors such as the user's perspective movement trajectory, mouse hover time, and zoom level, and use the Bayesian update model to optimize the interest evaluation weight, so that the system predicts the user interaction area before data requests and preloads high-priority data in advance to achieve low-latency seamless rendering.

4. A method for accelerating the visual rendering of flood risk map result data according to claim 1, characterized in that, The intelligent scheduling step of computing power resources further adopts a hierarchical scheduling strategy, including a global scheduling layer, a node scheduling layer, and a task allocation layer: The global scheduling layer is responsible for calculating the total amount of computing tasks and dynamically adjusting the global allocation ratio of computing resources; The node scheduling layer adjusts the task priorities based on the real-time load status of computing nodes and uses the genetic algorithm to optimize the optimal distribution of computing tasks; The task allocation layer combines GPU parallel computing and multi-threaded task management to adaptively divide data calculation and rendering tasks to maximize the utilization of computing power resources.

5. A method for accelerating the visual rendering of flood risk map result data according to claim 1, characterized in that, The intelligent scheduling step of computing power resources adopts an improved dynamic load balancing model. The specific steps for optimizing the computing resource allocation strategy are as follows: Calculate the load balancing coefficient of the current computing node. The calculation formula is as follows: ， where L i represents the load balancing coefficient of the i-th computing node, represents the computing resources already used by the i-th computing node, represents the total computing power of the i-th computing node; Calculate the overall system load balancing factor. The calculation formula is as follows: ， where N is the total number of computing nodes, and L sys is the overall system load balancing factor, which is used to determine whether the current computing power resources are reasonably allocated; Optimize the computing resource scheduling based on the objective function. The calculation formula is as follows: This objective function is used to minimize the load deviation of each computing node, enabling the system to achieve balanced scheduling and improve the computing priority of critical areas; Update the computing resource allocation strategy. The calculation formula is as follows: ， In the formula, is the new computing resource allocation value of the i-th computing node, and α is an adjustment coefficient used to balance the flexibility of computing resource allocation.

6. A method for accelerating the visualization rendering of flood risk map result data according to claim 1, characterized in that The block parallel rendering step uses a spatial hierarchical dissection algorithm to hierarchically partition the flood risk map data, and combines a local rendering cache strategy to store the data of the calculated high-priority areas, reducing duplicate calculations. At the same time, it uses computing graph optimization technology to dynamically construct the computing path to reduce the redundant computing amount of data processing and improve the rendering frame rate.

7. A method for accelerating the visual rendering of flood risk map result data according to claim 1, characterized in that The block parallel rendering step adopts a dynamic block optimization method based on data prediction to improve the rendering efficiency. The specific steps are as follows; Calculate the priority score of the data block. The calculation formula is as follows: P j = w1·A j + w2·D j + w3·T j , Wherein, P j represents the priority score of data block j, A j is the visible area of data block j, D j is the gradient change rate of data block j, T j is the historical access frequency of data block j, and w1, w2, and w3 are the weight parameters of the visible area of data block j, the gradient change rate of data block j, and the historical access frequency of data block j, respectively; Calculate the computing load index of the data block. The calculation formula is as follows: ， Where C j is the computational load index of data block j, M is the number of computational tasks, w z is the weight of computational task z, and L z，j is the computational load consumption of computational task z on data block j; Optimize the computing resource scheduling based on neural networks. The calculation formula is as follows: ， In the formula, is the optimized resource allocation weight, σ is the activation function, W is the neural network weight matrix, and X j is the input feature vector of data block j, and b is the bias term Adjust the computing task scheduling. The calculation formula is as follows: ， In the formula, is the optimized computing task allocation scheme, K is the total number of data blocks to be rendered, and arg is to select the computing task allocation scheme that minimizes the total load of computing tasks.

8. A method for accelerating the visual rendering of flood risk map result data according to claim 1, characterized in that, The resolution adjustment step adopts an adaptive precision allocation strategy based on gradient changes. First, calculate the gradient change rate of the flood inundation area, and automatically adjust the rendering precision of different areas according to the change rate threshold, so that high-change areas maintain high-resolution rendering, and low-change areas reduce the resolution to save computing resources; and combine a multi-level LOD rendering strategy to automatically switch the data precision according to the zoom level during user interaction to improve the balance between visual performance and computing efficiency.

Images