Coastal town storm surge disaster intelligent prevention and control system and method

Abstract

The invention belongs to the technical field of storm surge disaster early warning, and relates to a coastal town storm surge disaster intelligent prevention and control system and method.The system comprises an internet-of-things sensing layer, a data resource platform and an application supporting platform, the application supporting platform comprises a forecasting module, and the forecasting module is based on a multi-layer nested storm surge model; through nesting calculation of an open sea grid and a near-shore grid, the dam overtopping water flow and the water level and range of a submerged area are predicted; the early warning module is used for dynamically analyzing hydrology and water quality data in combination with a risk threshold value, generating an early warning signal and visually issuing the early warning signal through a G IS engine; the rehearsal module is used for carrying out real-time spatial model rendering on a prediction result of the prediction module by adopting an OpenGLES underlying graphic API and a GPU rendering pipeline optimization technology so as to realize restoration of dynamic change of a physical world; according to the system, the precision and the real-time performance are considered, a forecasting, early warning, rehearsing and pre-planning full-linkage module is constructed, and the decision response speed is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of storm surge disaster early warning, and specifically relates to an intelligent prevention and control system and method for storm surge disasters in coastal towns. Background Art

[0002] Fujian is located in the southeastern coastal area of ​​my country. It has alternating land and sea, and a changeable climate. It is often attacked by various marine disasters such as typhoons and storm surges, and suffers huge economic and social losses every year. As pointed out in the "2020 Fujian Province Marine Disaster Bulletin", the direct economic losses caused by marine storm surge disasters in Fujian Province in 2020 reached 124 million yuan.

[0003] Therefore, early warning of storm surge disasters is extremely necessary. Coastal towns are often affected by marine disasters such as typhoons and storm surges. Thanks to the rapid progress of computer technology, numerical simulation has become an important means of storm surge research. However, existing prevention and control technologies have the following key problems:

[0004] Limitations of Traditional Storm Surge Models Although numerical simulation techniques (such as ADCIRC and FVCOM) can simulate the nonlinear coupling effects of storm surge and astronomical tide, their application in large-scale and small-scale scenarios still faces severe challenges:

[0005] Resolution and efficiency conflict: Traditional models often use a single grid. Covering multiple scales, from offshore (kilometer-level) to nearshore (meter-level), requires extremely high computing resources. For example, while using a meter-level grid nearshore can improve flooding prediction accuracy, it leads to an exponential increase in computational effort, making it difficult to meet real-time warning requirements.

[0006] Inadequate model coupling: Existing models still rely on simplified assumptions for coupled simulations of waves, tides, and currents, resulting in insufficient accuracy in complex hydrodynamic processes, such as wave breaking and sudden topographic changes. While non-hydrostatic models (such as SWASH) can directly solve the Navier-Stokes equations, they lack GPU acceleration, limiting computational efficiency.

[0007] Currently, the marine industry still relies on traditional two-dimensional digital maps for analysis and simulation, and the acquired data is presented solely through charts. This lack of intuitive scene display results in a lengthy process for assessing storm surge hazards in coastal towns, requiring a certain amount of experience to quickly make decisions. When coastal towns face storm surge hazards, risk assessment still relies on manual judgment combined with equipment monitoring. This approach lacks real-time integration with rehearsal modules, making it impossible to dynamically display the impact of the inundation process on specific areas. This approach suffers from the drawbacks of being inaccurate and inadequate. Summary of the Invention

[0008] In order to solve the problem that the existing storm surge model prioritizes computational efficiency, resulting in large prediction accuracy errors, delayed decision-making of the disaster prevention and control system, and reliance on manual experience and decentralized module design, resulting in low response efficiency, the present invention provides a coastal urban storm surge disaster intelligent prevention and control system and method, which takes into account both accuracy and real-time performance, constructs a fully linked module of forecasting, warning, rehearsal, and plan, and improves the decision-making response speed.

[0009] The technical solutions of the present invention are as follows:

[0010] An intelligent storm surge disaster prevention and control system for coastal towns includes the following hierarchical architecture:

[0011] IoT perception layer, data resource platform, and application support platform;

[0012] The IoT perception layer is used to collect multi-source disaster monitoring data in real time, including:

[0013] Hydrological and water quality data sensors are deployed in seawalls and river areas to collect hydrological and water quality data in real time;

[0014] Drones and cameras are used to obtain image data of surface elevation, inundation extent, and engineering facility status;

[0015] The IoT communication module is used to transmit the collected multi-source disaster monitoring data and realize the real-time feedback of the monitoring data;

[0016] The data resource platform is used for standardized processing and efficient management of multi-source disaster monitoring data, including:

[0017] Data access module: connected to the Internet of Things communication module, used to receive and integrate multi-source disaster monitoring data and complete multi-source disaster monitoring data fusion;

[0018] Data processing module: used to pre-process the collected multi-source disaster monitoring data;

[0019] Data storage module: used to store pre-processed multi-source disaster monitoring data and support dynamic data compression and rapid retrieval;

[0020] Application support platform for disaster simulation and decision support, including:

[0021] Forecast module: The forecast module is connected to the data storage module for data exchange, and based on a multi-level nested storm surge model, the module uses offshore grids and nearshore grids for nested calculations to predict the amount of water flowing over the embankment, the water level and range of the submerged area;

[0022] Early warning module: Dynamically analyzes hydrological and water quality data based on risk thresholds, generates early warning signals, and publishes them visually through the GIS engine;

[0023] Preview module: Using OpenGLES underlying graphics API and GPU rendering pipeline optimization technology, the prediction module's prediction results are rendered in real-time spatial models to achieve dynamic changes in the physical world.

[0024] Plan module: Generate disaster prevention strategies based on the rendering results of the rehearsal module, and iteratively optimize operation scheduling and storm surge air defense plans.

[0025] Furthermore, the hydrological and water quality data sensors include water level sensors, rainfall sensors, flow rate sensors and pollution monitoring sensors.

[0026] Furthermore, the IoT communication module supports 4G / 5G and LoRa transmission.

[0027] Furthermore, the data processing module specifically uses Python's pandas library or R language data processing package to clean the ocean hydrological and geological data, including removing outliers and filling in missing values.

[0028] Furthermore, the data storage module is specifically: the data storage module realizes dynamic data compression and fast retrieval based on a hierarchical storage structure and a B-tree index mechanism.

[0029] Furthermore, the OpenGLES underlying graphics API and GPU rendering pipeline optimization technology are specifically as follows:

[0030] Convert the model data format in the forecast module into 3D Tiles format to achieve spatial data coordinate mapping;

[0031] For different model data in the multi-level nested storm surge model, accurate mapping of spatial data coordinates is achieved through OpenGL ES scaling matrix, viewport matrix and translation matrix;

[0032] LOD technology is used to generate multi-level detail models, and ETC1 texture compression technology, S3TC texture compression technology and vertex data compression algorithm are combined to achieve dynamic compression of model data;

[0033] Instanced rendering: batch draw repeated geometry through glDrawArraysInstanced to reduce the number of GPU draw calls;

[0034] Shader optimization: simplify the fragment shader and vertex shader logic, and use the discard keyword to eliminate invalid pixel calculations;

[0035] Rendering pipeline optimization: using depth testing technology and occlusion culling technology to reduce GPU thread consumption;

[0036] In terms of simulation algorithm optimization, the flood simulation algorithm is optimized to reduce computational complexity. By improving the structure and logic of the algorithm, adopting a more efficient fluid dynamics model or using spatial hashing technology to accelerate collision detection, the amount of computation can be reduced and the accuracy of the simulation can be improved.

[0037] Furthermore, the spatial model rendering includes dynamically simulating flooding, rainfall effects, and atmospheric temperature effects, and the spatial model rendering is implemented through the UE5 engine.

[0038] Furthermore, the system builds a three-dimensional visualization interface based on the UE5 engine, and uses Java as the background development language to integrate hydrological models and Internet of Things technologies to achieve B / S end response within seconds.

[0039] Furthermore, the hydrological model includes a hydrodynamic model, a shock wave model and a water depth integral model.

[0040] A method for intelligently preventing and controlling storm surge disasters in coastal towns, the method being as follows:

[0041] Step 1: Multi-source disaster monitoring data collection and access;

[0042] The IoT sensing layer acquires real-time water level, rainfall, flow rate, and pollution monitoring data, and uses drones and cameras to collect surface elevation, inundation range, and engineering facility image data. The water level, rainfall, flow rate, pollution monitoring data, surface elevation, inundation range, and engineering facility image data are multi-source disaster monitoring data.

[0043] Transmit multi-source disaster monitoring data back to the data resource platform via 4G / 5G or LoRa communication protocols;

[0044] Step 2: Data standardization processing and storage;

[0045] Clean and remove outliers from multi-source disaster monitoring data; based on a hierarchical storage structure and B-tree indexing mechanism, store the processed data in a comprehensive database, supporting dynamic retrieval and rapid access;

[0046] Step 3: Multi-level nested storm surge model prediction;

[0047] A nested computational model of offshore and coastal grids was constructed, and an improved non-hydrostatic wave equation was used to solve tidal changes. The conjugate gradient method was used to accelerate iterative calculations, combined with multi-GPU parallel training to eliminate communication errors, and the overbank flow, flooding area level, and extent were predicted.

[0048] Step 4: Dynamic rendering and preview of disaster scenes;

[0049] Convert standardized data into 3D Tiles format and configure the rendering pipeline through the OpenGLES underlying graphics API; use LOD technology to generate multi-level detail models, combine instanced rendering and depth testing to reduce GPU thread consumption; use the UE5 engine to simulate flooding, rainfall, and atmospheric diffusion effects in real time;

[0050] Step 5: Warning issuance and emergency plan generation;

[0051] Dynamically generate early warning signals based on sensor data and risk thresholds, and publish them to the B / S end through GIS engine visualization; generate personnel evacuation routes and engineering scheduling plans (or disaster prevention strategies) based on pre-rehearsal results;

[0052] Step 6: System response and decision support;

[0053] Integrate the hydrological model and IoT data interface through the Java backend to achieve B / S end response within seconds;

[0054] Based on the three-dimensional visualization interface, the typhoon path and flood evolution process are dynamically displayed to provide a decision-making basis for flood prevention and control.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] (1) High-precision and high-efficiency storm surge prediction capabilities;

[0057] Meter-level resolution simulation: Through multiple nested grid technology, the coupling calculation of large offshore grids (kilometer level) and fine nearshore grids (meter level) is realized, which significantly improves the prediction accuracy of water level and range of the flooded area, and the prediction error is small.

[0058] GPU parallel acceleration: Multi-GPU parallel training and conjugate gradient method optimization are used to improve model calculation speed, allowing large-scale calculations to be completed on ordinary PC devices, reducing dependence on supercomputers.

[0059] (2) Real-time dynamic rendering of ultra-large-scale data

[0060] Fast loading and realistic restoration: Based on the OpenGLES underlying graphics API and GPU rendering pipeline optimization, it achieves fast data loading and dynamically simulates flooding, rainfall particles and atmospheric diffusion effects, with a high degree of restoration of the physical world.

[0061] Dynamic resource management: Through paging storage, texture compression (ETC1 / S3TC) and vertex data quantization technology, GPU memory usage is reduced to support long-term stable operation of complex scenes.

[0062] (3) Full-chain collaborative decision support;

[0063] The forecast, warning, rehearsal, and contingency plan modules are seamlessly linked: the forecast module provides short-term water situation forecasts, the warning module dynamically publishes risk thresholds, the rehearsal module renders disaster scenarios in real time, and the contingency plan module generates optimized scheduling plans, significantly shortening the overall decision-making response time.

[0064] Deep integration of multi-source data: Integrate IoT sensors, drone images, historical databases, and GIS data through the data resource platform to eliminate data silos and improve early warning accuracy.

[0065] (4) Significant economic and social benefits;

[0066] Reduce disaster losses: Through accurate predictions and real-time rehearsals, assist government departments in deploying disaster prevention measures in advance;

[0067] Improve emergency response capabilities: The system supports three-dimensional visualization interaction between B / S terminals. Decision makers can use the UE5 engine to visually view the flooding evolution path and formulate scientific evacuation plans.

[0068] (5) Technological innovation and industry leadership;

[0069] For the first time, the non-hydrostatic model is combined with GPU accelerated rendering technology to solve the problem of lack of digital twin applications in the field of marine disasters and promote the transformation of domestic disaster prevention and control from "experience-driven" to "data-driven".

[0070] It integrates cutting-edge technologies such as ocean models, GIS engines, and the Internet of Things to provide standardized solutions for flood prevention, tide control, and typhoon resistance in coastal cities, and has the applicability to other coastal areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 Provide a technical roadmap for the system;

[0072] Figure 2 A roadmap for OpenGL ES low-level image rendering. DETAILED DESCRIPTION

[0073] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0074] See also Figure 1 , a smart storm surge disaster prevention and control system for coastal cities, including the following hierarchical architecture:

[0075] IoT perception layer, data resource platform, and application support platform;

[0076] The IoT perception layer is used to collect multi-source disaster monitoring data in real time, including:

[0077] Hydrological and water quality data sensors are deployed in seawalls and river areas to collect hydrological and water quality data in real time;

[0078] Drones and cameras are used to obtain image data of surface elevation, inundation extent, and engineering facility status;

[0079] The IoT communication module is used to transmit the collected multi-source disaster monitoring data and realize the real-time feedback of the monitoring data;

[0080] The data resource platform is used for standardized processing and efficient management of multi-source disaster monitoring data, including:

[0081] Data access module: connected to the Internet of Things communication module, used to receive and integrate multi-source disaster monitoring data and complete multi-source disaster monitoring data fusion;

[0082] Data processing module: used to pre-process the collected multi-source disaster monitoring data;

[0083] Data storage module: used to store pre-processed multi-source disaster monitoring data and support dynamic data compression and rapid retrieval;

[0084] Application support platform for disaster simulation and decision support, including:

[0085] Forecast module: The forecast module is connected to the data storage module for data exchange, and based on a multi-level nested storm surge model, the module uses offshore grids and nearshore grids for nested calculations to predict the amount of water flowing over the embankment, the water level and range of the submerged area;

[0086] Early warning module: Dynamically analyzes hydrological and water quality data based on risk thresholds, generates early warning signals, and publishes them visually through the GIS engine;

[0087] Preview module: Using OpenGLES underlying graphics API and GPU rendering pipeline optimization technology, the prediction module's prediction results are rendered in real-time spatial models to achieve dynamic changes consistent with the physical world;

[0088] Plan module: Generate disaster resistance strategies based on the rendering results of the rehearsal module, and iteratively optimize operation scheduling and storm surge air defense plans.

[0089] In the above forecast module, the offshore grid and the nearshore grid are nested for calculation to predict the amount of water flowing over the bank, the water level and range of the inundated area, specifically:

[0090] Large offshore grid: covers an area of ​​50-100 kilometers, with a grid resolution of 1-5 kilometers. It uses static pressure approximation to solve two-dimensional shallow water equations and quickly calculate tidal wave propagation.

[0091] Nearshore fine grid: Focusing on the area within 5 km of the coastline, with a grid resolution of 10-50 meters, it uses an improved non-hydrostatic model (such as NHWAVE or SWASH variants) to directly solve the three-dimensional Navier-Stokes equations to capture fine processes such as wave breaking and turbulent mixing.

[0092] Dynamic boundary coupling: Water level and velocity data are transferred through adaptive grid transition layers (such as sponge layer boundary conditions) to avoid numerical oscillation. The output of the offshore grid is used as the initial and boundary conditions of the coastal grid to achieve two-way data synchronization;

[0093] The multiple nested storm model is a meteorological model technology used to simulate and predict complex storm systems. It is mainly used for the research and forecast of severe convective weather systems such as hurricanes and typhoons.

[0094] The multi-nested storm model simulates storms by using grids of different resolutions in different regions:

[0095] 1. Outer coarse grid: covers a large area and provides a large-scale environmental field

[0096] 2. Middle-level medium-resolution grid: Focus on the storm development area

[0097] 3. Inner high-resolution grid: Accurately simulate the storm core structure

[0098] Main features: 1. Variable resolution: different grid spacing is used in different areas;

[0099] 2. Bidirectional nesting: There is bidirectional data exchange between each layer of grid;

[0100] 3. Dynamic Adaptation: The nesting area position can be adjusted as the storm moves;

[0101] 4. Computational efficiency: It saves more computing resources than full-domain high resolution;

[0102] Common models: WRF (Weather Research and Forecasting Model) supports multiple nested configurations and is widely used in typhoon forecasting;

[0103] HWRF (Hurricane Weather Research and Forecasting Model) is designed specifically for hurricane forecasting and uses a moving nested grid technique;

[0104] The GFDL hurricane model, developed by the National Oceanic and Atmospheric Administration, uses a multi-nested grid system.

[0105] This modeling technology significantly improves the simulation accuracy of complex storm systems and is an important tool in modern numerical weather forecasting.

[0106] The above technology is prior art and will not be described in detail here.

[0107] In one embodiment of the present invention, the hydrological and water quality data sensors include a water level sensor, a rainfall sensor, a flow rate sensor, and a pollution monitoring sensor.

[0108] In one embodiment of the present invention, the Internet of Things communication module supports 4G / 5G and LoRa transmission.

[0109] In one embodiment of the present invention, the data processing module specifically uses Python's pandas library or R language data processing package to clean the marine hydrological and geological data, including removing outliers and filling missing values ​​to ensure the continuity and consistency of the data.

[0110] In one embodiment of the present invention, the data storage module is specifically: the data storage module realizes dynamic data compression and fast retrieval based on a hierarchical storage structure and a B-tree index mechanism;

[0111] A hierarchical storage structure is used to organize data into different layers or levels. Each layer of data has its own specific compression format and storage strategy. An efficient B-tree indexing mechanism is established to quickly locate the required data blocks, thereby accelerating data loading.

[0112] After completing data preprocessing, the data is compressed and stored. An efficient data compression algorithm can not only significantly reduce the storage space of the data, thereby reducing storage costs, but also effectively reduce the bandwidth pressure during data transmission and improve data transmission efficiency. A variety of advanced compression technologies are used. For texture data, compression algorithms based on wavelet transform or discrete cosine transform are used. They can well preserve image details while significantly reducing the amount of data. For vertex data, algorithms such as quantization or predictive coding are used. The algorithms can remove redundant information in the data and achieve efficient compression. In terms of storage, the design of a reasonable file structure and indexing mechanism is the key to ensuring efficient data loading and access. A hierarchical storage structure is used to organize data according to different layers or levels. The data at each layer has its own specific compression format and storage strategy. Establishing an efficient B-tree indexing mechanism can quickly locate the required data blocks, thereby speeding up data loading. Data may need to be transmitted and shared between multiple systems or platforms.

[0113] The compression algorithm and file format have good compatibility and scalability. This not only avoids format conversion and data loss during data transmission, but also facilitates subsequent data analysis and visualization.

[0114] See also Figure 2 In one embodiment of the present invention, the OpenGLES underlying graphics API and GPU rendering pipeline optimization technology are specifically:

[0115] Convert the model data format in the forecast module into 3D Tiles format to achieve spatial data coordinate mapping;

[0116] For different model data in the multi-level nested storm surge model, accurate mapping of spatial data coordinates is achieved through OpenGL ES scaling matrix, viewport matrix and translation matrix;

[0117] Specifically, the data format is converted according to the specific requirements of OpenGL ES. Different specific methods are used for different model data to convert the model into the standard 3dtiles structured description data format, so that the model can be loaded more efficiently. Secondly, for the scaling and offset operations of the data, the OpenGL ES scaling matrix, viewport matrix, clipping space transformation, and translation matrix are used. The position and size of the data in the rendering space are precisely controlled to ensure that the data can be correctly mapped to the corresponding position. At the same time, according to the actual needs of the rendering scene, the data needs to be scaled and offset to ensure that the data can be correctly mapped to the corresponding position in the rendering space.

[0118] In terms of GPU memory management, since the amount of data processed can reach over 100GB, this paper uses LOD technology to generate multi-level detail models. This technology combines ETC1 texture compression, S3TC texture compression, and vertex data compression algorithms to achieve dynamic compression of model data. By combining data block processing with LOD technology, large-scale spatial data can be effectively processed, achieving efficient and high-quality rendering. This allows for the restoration of 100GB+ of data and the dynamic changes in the physical world.

[0119] Dynamically allocate and release GPU memory based on scene requirements. Use OpenGL functions such as glGenBuffers, glBindBuffer, glBufferData, and glDeleteBuffers to create, bind, update, and destroy resources.

[0120] Instanced rendering: batch draw repeated geometry through glDrawArraysInstanced to reduce the number of GPU draw calls;

[0121] Shader optimization: Simplify the fragment shader and vertex shader logic, and use the discard keyword to eliminate invalid pixel calculations; that is, by removing or simplifying unnecessary calculation steps and conditional judgments in the shader, the execution efficiency of the shader can be improved.

[0122] Rendering Pipeline Optimization: Utilize depth testing and occlusion culling techniques to reduce GPU thread consumption; focus on reducing draw calls and utilizing depth testing and occlusion culling techniques. By merging draw calls with the same material and geometry attributes, the GPU drawing load can be reduced. Optimizing the render queue, sorting objects by material or depth, and using glMultiDrawArrays or enabling options such as GL_DEPTH_TEST and GL_CULL_FACE, we can preemptively cull pixels occluded by other objects, avoiding unnecessary shading calculations and further improving rendering efficiency.

[0123] In terms of simulation algorithm optimization, the flood simulation algorithm is optimized to reduce computational complexity. By improving the structure and logic of the algorithm, adopting a more efficient fluid dynamics model or using spatial hashing technology to accelerate collision detection, the amount of computation can be reduced and the accuracy of the simulation can be improved.

[0124] In flood simulation, the fluid dynamics model is the core. The following is an example of optimization in the embodiment:

[0125] 1. Shallow water equation optimization:

[0126] Local one-dimensional method (LOD): decomposes the two-dimensional problem into two one-dimensional problems and solves them along the x and y directions respectively, thereby reducing the computational complexity.

[0127] Explicit and implicit hybrid format: an explicit format (such as MacCormack format) is used for the advection term, and an implicit format is used for the friction term and the pressure term, thereby relaxing the time step restriction.

[0128] Adaptive Mesh Refinement (AMR): Automatically refine the mesh in flood fronts and complex terrain areas, and coarsen the mesh in other areas to reduce the amount of calculation.

[0129] 2. Fluid flow simulation using the Lattice Boltzmann Method, a mesoscopic computational fluid dynamics method that lies between microscopic molecular dynamics models and macroscopic continuum models. LBM describes the macroscopic behavior of fluids by simulating the distribution and evolution of fluid particles. It has the flexibility to handle complex boundary conditions and multi-physics coupling problems, and is easy to parallelize and program.

[0130] Advantages: Excellent parallelism, suitable for GPU acceleration;

[0131] Implementation optimization: Use the D2Q9 or D3Q19 model, combined with the non-equilibrium extrapolation format to handle complex boundaries.

[0132] Use spatial hashing technology to accelerate collision detection. In fluid simulation, collision detection mainly occurs between fluid particles and terrain / buildings (such as SPH method) and the calculation of interaction forces between particles (neighbor search).

[0133] Spatial hashing is a technology that divides space into grids (hash buckets) and maps spatial positions to grid indexes through hash functions, thereby quickly locating objects.

[0134] The application of spatial hashing in flood simulation is as follows:

[0135] SPH fluid simulation: Neighbor search is the main computational bottleneck, and spatial hashing can accelerate force calculation.

[0136] Flood and building collision: voxelize building boundaries and store them in a spatial hash table to quickly detect fluid particle collisions.

[0137] Multi-GPU parallelism: Different grid regions are assigned to different GPUs based on spatial hash partitioning.

[0138] The principle of simulation algorithm optimization is briefly introduced, and the details are not repeated here. This application is known to those skilled in the art and is a prior art.

[0139] In one embodiment of the present invention, the spatial model rendering includes dynamically simulating flooding, rainfall effects, and atmospheric temperature effects, and the spatial model rendering is implemented by the UE5 engine.

[0140] The integration of UE5 provides the system with excellent three-dimensional visual expression, which can simulate the topographic characteristics, hydrological conditions and various effects that storm surges may cause in coastal towns in a highly realistic and detailed manner, helping to improve the accuracy of risk assessment and significantly enhance the risk awareness of decision makers and the public.

[0141] In one embodiment of the present invention, the system constructs a three-dimensional visualization interface based on the UE5 engine, and uses Java as the background development language to integrate hydrological models and Internet of Things technologies to achieve a B / S end response in seconds.

[0142] Java is used as the backend development language to ensure the stable and efficient operation of the data processing process. The system integrates professional hydrological models to quickly analyze and calculate the collected meteorological, oceanographic and topographic data, providing accurate forecasts of storm surge events. At the same time, the system integrates Internet of Things technology and relies on sensor networks deployed in key areas to monitor key parameters such as water levels and rainfall in real time to enhance the immediacy and accuracy of early warning signals. At the level of simulation rehearsals and plan preparation, the system provides scientific evidence for the formulation of disaster prevention and mitigation strategies by simulating the scope and severity of disasters under different storm surge scenarios. Decision makers can scientifically and rationally customize emergency response measures and personnel evacuation plans based on the simulation results and in-depth analysis reports generated by the system. In this process, the application of digital twin technology makes the simulation process more realistic, thereby improving the practical effectiveness and implementation of the plan.

[0143] In one embodiment of the present invention, the hydrological model includes a hydrodynamic model, a shock wave model, and a water depth integral model;

[0144] A hydrodynamic model is a mathematical model used to describe the relationship between the forces acting on and the motion of water. Based on the fundamental equations of fluid mechanics, a mathematical model is established to numerically simulate the dynamic processes of flowing water. The MIKE 21 hydrodynamic model employed in this embodiment is suitable for large-scale tidal and storm surge simulations.

[0145] The shock wave model (using the HLLC model) is a high-resolution approximate Riemann solver for solving hyperbolic conservation law equations. It is particularly suitable for the numerical simulation of shock waves, contact discontinuities, and rarefaction waves. Especially in the nearshore area of ​​the wave breaking zone, the rapid changes in water level during storm surges are similar to shock wave phenomena. Through shock wave models such as HLLC, contact discontinuities and rarefaction waves are captured at high resolution, accurately simulating the steep gradient changes of the tidal wave front.

[0146] The water depth integral model (using the LISFLOOD-FP model) is the first choice for urban waterlogging. It can intuitively display the depth and range of inundation. The model has an advantage in computational efficiency (supporting million-level grids) in storm surges. By simplifying three-dimensional flow into two-dimensional calculations through vertical integration, it achieves a balance between efficiency and accuracy in storm surge disaster simulation.

[0147] The above models are all existing technical models, and those skilled in the art will know how to use them.

[0148] A method for intelligently preventing and controlling storm surge disasters in coastal towns, the method being as follows:

[0149] Step 1: Multi-source disaster monitoring data collection and access;

[0150] The IoT sensing layer acquires real-time water level, rainfall, flow rate, and pollution monitoring data, and uses drones and cameras to collect surface elevation, inundation range, and engineering facility image data. The water level, rainfall, flow rate, pollution monitoring data, surface elevation, inundation range, and engineering facility image data are multi-source disaster monitoring data.

[0151] Transmit multi-source disaster monitoring data back to the data resource platform via 4G / 5G or LoRa communication protocols;

[0152] Step 2: Data standardization processing and storage;

[0153] Clean and remove outliers from multi-source disaster monitoring data; based on a hierarchical storage structure and B-tree indexing mechanism, store the processed data in a comprehensive database, supporting dynamic retrieval and rapid access;

[0154] Step 3: Multi-level nested storm surge model prediction;

[0155] A nested computational model of offshore and coastal grids was constructed, and an improved non-hydrostatic wave equation was used to solve tidal changes. The conjugate gradient method was used to accelerate iterative calculations, combined with multi-GPU parallel training to eliminate communication errors, and the overbank flow, flooding area level, and extent were predicted.

[0156] Step 4: Dynamic rendering and preview of disaster scenes;

[0157] Convert standardized data into 3D Tiles format and configure the rendering pipeline through the OpenGLES underlying graphics API; use LOD technology to generate multi-level detail models, combine instanced rendering and depth testing to reduce GPU thread consumption; use the UE5 engine to simulate flooding, rainfall, and atmospheric diffusion effects in real time;

[0158] Step 5: Warning issuance and emergency plan generation;

[0159] Dynamically generate early warning signals based on sensor data and risk thresholds, and publish them to the B / S end through GIS engine visualization; generate personnel evacuation routes and engineering scheduling plans (or disaster prevention strategies) based on pre-rehearsal results;

[0160] The generation of personnel evacuation routes and engineering scheduling plans (or disaster prevention strategies) based on the rehearsal results is specifically achieved by those skilled in the art observing the results of the rehearsal simulation and customizing the disaster prevention strategies based on experience.

[0161] Step 6: System response and decision support;

[0162] Integrate the hydrological model and IoT data interface through the Java backend to achieve B / S end response within seconds;

[0163] Based on the three-dimensional visualization interface, the typhoon path and flood evolution process are dynamically displayed to provide a decision-making basis for flood prevention and control.

[0164] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention's description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A smart storm surge disaster prevention and control system for coastal towns, characterized by , including the following hierarchical architecture: IoT perception layer, data resource platform, and application support platform; The IoT perception layer is used to collect multi-source disaster monitoring data in real time, including: Hydrological and water quality data sensors are deployed in seawalls and river areas to collect hydrological and water quality data in real time; Drones and cameras are used to obtain image data of surface elevation, inundation extent, and engineering facility status; Internet of Things communication module, used to transmit collected multi-source disaster monitoring data, and realize real-time monitoring data return; The data resource platform is used for standardized processing and efficient management of multi-source disaster monitoring data, including: Data access module: connected to the Internet of Things communication module, used to receive and integrate multi-source disaster monitoring data, Complete the fusion of multi-source disaster monitoring data; Data processing module: used to pre-process the collected multi-source disaster monitoring data; Data storage module: used to store pre-processed multi-source disaster monitoring data and support data dynamic compression Zoom and quick search; Application support platform for disaster simulation and decision support, including: Forecast module: The forecast module is connected with the data storage module to interact with data. The sub-nested storm surge model uses offshore grids and nearshore grids for nested calculations to predict the amount of water flowing over the embankment. Water level and extent of the inundated area; Early warning module: Dynamically analyze hydrological and water quality data based on risk thresholds, generate early warning signals and send them to GIS Engine visualization release; Preview module: Using OpenGLES underlying graphics API and GPU rendering pipeline optimization technology, the prediction module's prediction results are rendered in real-time spatial models to achieve dynamic changes in the physical world. Contingency plan module: Generates disaster prevention strategies based on the rendering results of the preview module.

2. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The hydrological and water quality data sensors include water level sensors, rainfall sensors, flow rate sensors and pollution monitoring sensors.

3. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The IoT communication module supports 4G / 5G and LoRa transmission.

4. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The data processing module specifically uses Python's pandas library or R language's data processing package to clean the ocean hydrological and geological data, including removing outliers and filling in missing values.

5. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The data storage module is specifically: the data storage module is based on a hierarchical storage structure and a B-tree index mechanism to achieve dynamic data compression and fast retrieval.

6. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The OpenGLES underlying graphics API and GPU rendering pipeline optimization technology are specifically: Convert the model data format in the forecast module into 3D Tiles format to achieve spatial data coordinate mapping; For different model data in the multi-level nested storm surge model, accurate mapping of spatial data coordinates is achieved through OpenGL ES scaling matrix, viewport matrix and translation matrix; LOD technology is used to generate multi-level detail models, and ETC1 texture compression technology, S3TC texture compression technology and vertex data compression algorithm are combined to achieve dynamic compression of model data; Instanced rendering: batch draw repeated geometry through glDrawArraysInstanced to reduce the number of GPU draw calls; Shader optimization: simplify the fragment shader and vertex shader logic, and use the discard keyword to eliminate invalid pixel calculations; Rendering pipeline optimization: using depth testing technology and occlusion culling technology to reduce GPU thread consumption; In terms of simulation algorithm optimization, the flood simulation algorithm is optimized to reduce computational complexity. By improving the structure and logic of the algorithm, adopting a more efficient fluid dynamics model or using spatial hashing technology to accelerate collision detection, the amount of computation can be reduced and the accuracy of the simulation can be improved.

7. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The spatial model rendering includes realizing dynamic simulation of flooding, rainfall effect and atmospheric temperature effect, and the spatial model rendering is realized by UE5 engine.

8. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The system builds a three-dimensional visualization interface based on the UE5 engine, and uses Java as the background development language to integrate hydrological models and Internet of Things technologies to achieve B / S end response in seconds.

9. The intelligent storm surge disaster prevention and control system for coastal towns according to claim 1 is characterized in that: The hydrological model includes a hydrodynamic model, a shock wave model and a water depth integral model.

10. A smart method for preventing and controlling storm surge disasters in coastal towns, characterized in that: The method is as follows: Step 1: Multi-source disaster monitoring data collection and access; The IoT sensing layer acquires real-time water level, rainfall, flow rate, and pollution monitoring data, and uses drones and cameras to collect surface elevation, inundation range, and engineering facility image data. These data, which are multi-source disaster monitoring data, are transmitted back to the data resource platform via 4G / 5G or LoRa communication protocols. Step 2: Data standardization processing and storage; Clean and remove outliers from multi-source disaster monitoring data; Based on the hierarchical storage structure and B-tree index mechanism, the processed data is classified and stored in a comprehensive database, supporting dynamic retrieval and fast call; Step 3: Multi-level nested storm surge model prediction; A nested computational model of offshore and coastal grids was constructed, and an improved non-hydrostatic wave equation was used to solve tidal changes. The conjugate gradient method was used to accelerate iterative calculations, combined with multi-GPU parallel training to eliminate communication errors, and the overbank flow, flooding area level, and extent were predicted. Step 4: Dynamic rendering and preview of disaster scenes; Convert standardized data into 3D Tiles format and configure the rendering pipeline through the OpenGLES underlying graphics API; use LOD technology to generate multi-level detail models, combine instanced rendering and depth testing to reduce GPU thread consumption; use the UE5 engine to simulate flooding, rainfall, and atmospheric diffusion effects in real time; Step 5: Warning issuance and emergency plan generation; Dynamically generate early warning signals based on sensor data and risk thresholds, and publish them to the B / S end through GIS engine visualization; generate personnel evacuation routes and engineering scheduling plans (or disaster prevention strategies) based on pre-rehearsal results; Step 6: System response and decision support; Integrate the hydrological model and IoT data interface through the Java backend to achieve B / S end response within seconds; Based on the three-dimensional visualization interface, the typhoon path and flood evolution process are dynamically displayed to provide a decision-making basis for flood prevention and control.

Images