Flood high-precision simulation method, system, device and storage medium

Abstract

The application discloses a flood high-precision simulation method, system and device and a storage medium, and comprises the following steps: acquiring measured rainfall data and high-resolution terrain data of a target area, and determining first hydrodynamic parameters of the target area according to the high-resolution terrain data; using a hydrodynamic model to perform coarse-scale grid flood simulation on the target area according to the first hydrodynamic parameters, and obtaining low-resolution flood simulation data of the target area by taking the measured rainfall data as a boundary condition; and inputting the low-resolution flood simulation data and the high-resolution terrain data into a pre-trained super-resolution generative adversarial network to obtain high-resolution flood simulation data of the target area. The application can improve the efficiency of flood simulation while ensuring the simulation accuracy, and can be widely applied in the field of artificial intelligence technology.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and in particular to a high-precision flood simulation method, system, device and storage medium. BACKGROUND

[0002] Cities are densely populated and rich in social and economic wealth. Once a flood disaster occurs, it will cause more serious casualties and economic property losses. In order to ensure the sustainability of society and economy, it is necessary to properly manage the urban flood problem. Accurate and rapid simulation of flood disasters can help accurately identify urban flood risks and scientifically guide flood disaster prevention and emergency management measures, which are crucial to reducing the casualties and economic losses caused by flood disasters. It is urgent to carry out research on urban flood simulation technology to achieve high-precision and rapid simulation of flood inundation risks.

[0003] Accurate and rapid simulation of flood disasters can help accurately identify urban flood risks and scientifically implement flood disaster prevention and emergency management measures. However, in existing flood simulation methods, the water power model has low calculation efficiency and cannot meet the needs of rapid flood warning; the machine learning model lacks physical mechanism constraints and has poor reliability in simulating complex flood processes. Therefore, how to organically integrate mechanism models and machine learning algorithms to compensate for the inherent defects of the two types of algorithms and significantly improve the calculation efficiency while ensuring accuracy is a problem that needs to be solved. SUMMARY

[0004] The present application aims to at least partially solve one of the problems of the prior art.

[0005] To this end, one object of the present application is to provide a high-precision flood simulation method that improves the efficiency of flood simulation while ensuring simulation accuracy.

[0006] Another object of the present application is to provide a high-precision flood simulation system.

[0007] In order to achieve the above technical purpose, the technical solution adopted by the embodiments of the present application comprises:

[0008] In a first aspect, the embodiments of the present application provide a high-precision flood simulation method, comprising the following steps:

[0009] Obtaining measured rainfall data and high-resolution topographic data of a target area, and determining first hydrodynamic parameters of the target area according to the high-resolution topographic data;

[0010] Using a hydrodynamic model to perform coarse-scale grid flood simulation on the target area according to the first hydrodynamic parameters, and obtaining low-resolution flood simulation data of the target area, taking the measured rainfall data as boundary conditions;

[0011] inputting the low-resolution flood simulation data and the high-resolution terrain data into a pre-trained super-resolution generative adversarial network to obtain high-resolution flood simulation data of the target region.

[0012] Further, in an embodiment of the present application, the super-resolution generative adversarial network is trained by the following steps:

[0013] obtaining high-resolution terrain samples of a plurality of test regions, and determining second hydrodynamic parameters of the corresponding test regions according to the high-resolution terrain samples;

[0014] randomly generating sample rainfall data, using the sample rainfall data as boundary conditions, and using a hydrodynamic model to perform coarse-scale grid flood simulation and fine-scale grid flood simulation on the test regions according to the second hydrodynamic parameters, to obtain low-resolution flood simulation results and high-resolution flood simulation results of each test region;

[0015] training a pre-constructed generative adversarial network according to the low-resolution flood simulation results, the high-resolution terrain samples, and the high-resolution flood simulation results to obtain the trained super-resolution generative adversarial network.

[0016] Further, in an embodiment of the present application, the training of the pre-constructed generative adversarial network according to the low-resolution flood simulation results, the high-resolution terrain samples, and the high-resolution flood simulation results specifically includes:

[0017] inputting the low-resolution flood simulation results and the high-resolution terrain samples into a generator of the generative adversarial network to obtain high-resolution flood prediction results;

[0018] inputting the high-resolution flood prediction results and the high-resolution flood simulation results into a discriminator of the generative adversarial network to obtain a discrimination result;

[0019] determining a first loss value of the generator according to the discrimination result, and determining a second loss value of the discriminator according to the high-resolution flood prediction results and the high-resolution flood simulation results;

[0020] updating parameters of the generator according to the first loss value, and updating parameters of the discriminator according to the second loss value.

[0021] Further, in an embodiment of the present application, the generator comprises a shallow feature extraction module, a deep feature extraction module and a feature fusion module, the shallow feature extraction module is configured to extract shallow features of the input data, the deep feature extraction module is configured to extract deep features according to the shallow features, and the feature fusion module is configured to perform feature fusion on the deep features.

[0022] Further, in an embodiment of the present application, the inputting of the low-resolution flood simulation result and the high-resolution terrain sample into the generator of the generative adversarial network to obtain a high-resolution flood prediction result comprises:

[0023] performing interpolation processing on the low-resolution flood simulation result to obtain an interpolated flood simulation result with the same resolution as the high-resolution terrain sample;

[0024] extracting shallow features of the interpolated flood simulation result and the high-resolution terrain sample respectively by the shallow feature extraction module to obtain flood simulation shallow features and terrain sample shallow features;

[0025] extracting deep features of the flood simulation shallow features and the terrain sample shallow features respectively by the deep feature extraction module to obtain flood simulation deep features and terrain sample deep features;

[0026] performing feature fusion on the flood simulation deep features and the terrain sample deep features by the feature fusion module to obtain spatial detail residual data;

[0027] injecting the spatial detail residual data into the interpolated flood simulation result to obtain the high-resolution flood prediction result.

[0028] Further, in an embodiment of the present application, the hydrodynamic model is:

[0029]

[0030] wherein t represents time, U is a vector composed of conservative variables, E and G are fluxes in x and y directions respectively, S is a source term, h represents water level, u and v represent flow velocities in x and y directions averaged along the water level, q e is a mass source term per unit area, g represents gravitational acceleration, S 0x and S 0y are bottom slope terms in x and y directions respectively, S fx and S fy are friction slopes in x and y directions respectively, and n represents Manning roughness coefficient.

[0031] Further, in one embodiment of the present application, the measured rainfall data is used as a boundary condition to perform a coarse-grid flood simulation of the target area according to the first hydrodynamic parameters using a hydrodynamic model, to obtain low-resolution flood simulation data of the target area, which specifically includes:

[0032] dividing the target area into a plurality of coarse-grid cells;

[0033] using the measured rainfall data as a boundary condition to calculate water level data and flow velocity data of each coarse-grid cell according to the first hydrodynamic parameters using a hydrodynamic model;

[0034] generating the low-resolution flood simulation data according to the water level data and the flow velocity data of each coarse-grid cell.

[0035] In a second aspect, an embodiment of the present application provides a high-precision flood simulation system, which includes:

[0036] a data acquisition module configured to acquire measured rainfall data and high-resolution terrain data of a target area, and determine first hydrodynamic parameters of the target area according to the high-resolution terrain data;

[0037] a low-resolution flood simulation module configured to use the measured rainfall data as a boundary condition to perform a coarse-grid flood simulation of the target area according to the first hydrodynamic parameters using a hydrodynamic model, to obtain low-resolution flood simulation data of the target area;

[0038] a high-resolution flood reconstruction module configured to input the low-resolution flood simulation data and the high-resolution terrain data into a pre-trained super-resolution generative adversarial network, to obtain high-resolution flood simulation data of the target area.

[0039] In a third aspect, an embodiment of the present application provides a high-precision flood simulation device, which includes:

[0040] at least one processor;

[0041] at least one memory configured to store at least one program;

[0042] when the at least one program is executed by the at least one processor, the at least one processor is caused to implement the high-precision flood simulation method described above.

[0043] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores a program executable by a processor, and the program executable by the processor is used to execute the high-precision flood simulation method described above when executed by the processor.

[0044] The advantages and beneficial effects of the present application will be partially given in the following description, partially will become obvious from the following description, or will be understood by the practice of the present application.

[0045] The embodiment of the present application obtains the measured rainfall data and high-resolution terrain data of the target area, determines the first water dynamic parameter of the target area according to the high-resolution terrain data, uses the water dynamic model to perform coarse-scale grid flood simulation on the target area according to the first water dynamic parameter, taking the measured rainfall data as the boundary condition, obtains the low-resolution flood simulation data of the target area, inputs the low-resolution flood simulation data and the high-resolution terrain data into the pre-trained super-resolution generative adversarial network, and obtains the high-resolution flood simulation data of the target area. The embodiment of the present application uses the water dynamic model to perform coarse-scale grid flood simulation on the target area to obtain the low-resolution flood simulation data, and inputs the low-resolution flood simulation data and the high-resolution terrain data into the super-resolution generative adversarial network for super-resolution reconstruction to obtain the high-resolution flood simulation data. Compared with directly using the water dynamic model to perform fine-scale network flood simulation, the calculation efficiency is improved, so that the simulation accuracy is ensured while the efficiency of flood simulation is improved. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following introduces the drawings needed to be used in the embodiments of the present application. It should be understood that the drawings introduced in the following are only for facilitating the clear description of part of the embodiments in the technical solutions of the present application, and other drawings can also be obtained by those skilled in the art without paying creative labor on the premise of the drawings.

[0047] Figure 1 The step flow chart of the high-precision flood simulation method provided by the embodiment of the present application is shown in the figure.

[0048] Figure 2 The training process schematic diagram of the super-resolution generative adversarial network provided by the embodiment of the present application is shown in the figure.

[0049] Figure 3 The schematic diagram of the generator generating high-resolution flood prediction results provided by the embodiment of the present application is shown in the figure.

[0050] Figure 4 The structural block diagram of the high-precision flood simulation system provided by the embodiment of the present application is shown in the figure.

[0051] Figure 5 The structural block diagram of the high-precision flood simulation device provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0052] Embodiments of the present application are described below in detail, examples of which are shown in the drawings, wherein the same or similar notations represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present application, and cannot be understood as a limitation of the present application. For the step numbers in the following embodiments, they are only set for the convenience of setting out the description, and the order between the steps is not limited in any way, and the execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0053] In the description of the present application, the meaning of multiple is two or more, and if the first, the second is described, it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or implicitly indicating the sequence of indicated technical features. In addition, unless otherwise defined, all technical and scientific terms used herein have the same meaning as generally understood by those skilled in the art.

[0054] At present, the common flood disaster simulation methods in China mainly include mechanism models represented by hydrology and hydrodynamic models and machine learning models based on various AI algorithms. The mechanism model uses numerical algorithm to solve the control equation, and the calculation result has the best reliability, but has problems such as complex solution and low calculation efficiency. The machine learning model is mainly based on data driving, does not need to solve complex control equation, and has higher calculation efficiency; but the machine learning model has defects such as large training difficulty, cannot guarantee the conservation of physical quantity of the result, and poor simulation fitting ability for complex process. In view of the limitations of the mechanism model and the machine learning model, it is urgent to explore a fusion way, that is, to retain the accuracy of the mechanism model and the high efficiency and flexibility of the machine learning model, and to develop a new generation of simulation technology which can simulate complex flood process and ensure fast and accurate simulation results. The related research results can provide core technical support for the construction of the urban flood rapid early warning and forecasting system.

[0055] The embodiment of the present application trains a super-resolution neural network based on the flood simulation results of the hydrodynamic model under the general rainfall condition, and realizes the accurate and rapid simulation of the flood process under the millimeter and sub-millimeter fine spatial resolution by fusing the hydrodynamic model and the super-resolution generative adversarial network for flood rapid simulation.

[0056] With reference to Figure 1 The embodiment of the present application provides a high-precision flood simulation method, which specifically includes the following steps:

[0057] S101, obtaining measured rainfall data and high-resolution terrain data of a target area, and determining first hydrodynamic parameters of the target area according to the high-resolution terrain data;

[0058] S102, using the water dynamic model to simulate the flood of the target region according to the first water dynamic parameter, and obtaining low-resolution flood simulation data of the target region by taking the measured rainfall data as a boundary condition;

[0059] S103, inputting the low-resolution flood simulation data and the high-resolution terrain data into the pre-trained super-resolution generative adversarial network to obtain high-resolution flood simulation data of the target region.

[0060] Specifically, the embodiment of the present application uses a water dynamic model to simulate the flood of the target region to obtain low-resolution flood simulation data, and inputs the low-resolution flood simulation data and high-resolution terrain data into a super-resolution generative adversarial network for super-resolution reconstruction to obtain high-resolution flood simulation data, which improves the calculation efficiency compared with directly using the water dynamic model to simulate the flood of the target region, thereby improving the efficiency of the flood simulation while ensuring the simulation accuracy.

[0061] The super-resolution generative adversarial network of the embodiment of the present application takes high-resolution terrain and low-resolution flood simulation results such as water level and flow rate as input, and takes high-resolution flood simulation results as output, and the training process of the super-resolution generative adversarial network of the embodiment of the present application is described below.

[0062] Further as an optional implementation, the super-resolution generative adversarial network is trained by the following steps:

[0063] S201, obtaining high-resolution terrain samples of a plurality of test regions, and determining second water dynamic parameters of the corresponding test regions according to the high-resolution terrain samples;

[0064] S202, randomly generating sample rainfall data, using the water dynamic model to simulate the flood of the test regions according to the second water dynamic parameters, and obtaining low-resolution flood simulation results and high-resolution flood simulation results of the test regions by taking the sample rainfall data as a boundary condition;

[0065] S203, training the pre-constructed generative adversarial network according to the low-resolution flood simulation results, the high-resolution terrain samples and the high-resolution flood simulation results, and obtaining the trained super-resolution generative adversarial network.

[0066] Specifically, sample rainfall data of different durations, rain types and return periods are generated as boundary conditions to drive the water dynamic model to perform flood simulation at a grid spatial resolution of 0.5 to 8 m, to obtain low-resolution flood simulation results and high-resolution flood simulation results; and an ultra-resolution generative adversarial network is trained based on the obtained low-resolution flood simulation results, high-resolution flood simulation results and high-resolution terrain samples. After the training is completed, a typical field measured rainfall is selected, a water dynamic model is used to simulate at a coarse grid resolution, and the trained ultra-resolution generative adversarial network is used to perform ultra-resolution to a meter-level or sub-meter-level spatial scale.

[0067] Further as an optional implementation, the pre-constructed generative adversarial network is trained according to the low-resolution flood simulation results, the high-resolution terrain samples and the high-resolution flood simulation results, and specifically includes:

[0068] S2031, inputting the low-resolution flood simulation results and the high-resolution terrain samples into a generator of the generative adversarial network to obtain high-resolution flood prediction results;

[0069] S2032, inputting the high-resolution flood prediction results and the high-resolution flood simulation results into a discriminator of the generative adversarial network to obtain a discrimination result;

[0070] S2033, determining a first loss value of the generator according to the discrimination result, and determining a second loss value of the discriminator according to the high-resolution flood prediction results and the high-resolution flood simulation results;

[0071] S2034, updating parameters of the generator according to the first loss value, and updating parameters of the discriminator according to the second loss value.

[0072] Specifically, as Figure 2 shown is a training process diagram of an ultra-resolution generative adversarial network provided by an embodiment of the application, which mainly includes a generator and a discriminator. The generator extracts deep features of terrain and low-resolution simulation results, generates high-resolution flood prediction results, and the discriminator evaluates the generated high-resolution flood prediction results. In a cyclic process, the generator and the discriminator are mutually opposed, when the generated high-resolution flood prediction results cannot be effectively distinguished from high-resolution flood simulation results output by the water dynamic model by the discriminator, it means that reliable high-resolution flood prediction results are generated, and a trained ultra-resolution generative adversarial network is obtained.

[0073] Further as an optional implementation, the generator includes a shallow feature extraction module, a deep feature extraction module and a feature fusion module, the shallow feature extraction module is used to extract shallow features of input data, the deep feature extraction module is used to extract deep features according to the shallow features, and the feature fusion module is used to perform feature fusion on multiple deep features.

[0074] Specifically, the generator comprises three modules of shallow feature extraction, deep feature extraction and feature fusion, which are used to extract deep features from the terrain data and the low-resolution simulation result and fuse to generate the high-resolution simulation result. The discriminator is based on a convolutional neural network architecture, which is used to evaluate the difference between the generated simulation result and the real data, so as to realize the high-resolution reconstruction of the hydrodynamic simulation result.

[0075] Further as an optional implementation, the low-resolution flood simulation result and the high-resolution terrain sample are input into the generator of the generative adversarial network to obtain the high-resolution flood prediction result, which specifically comprises:

[0076] S20311, interpolating the low-resolution flood simulation result to obtain an interpolated flood simulation result with the same resolution as the high-resolution terrain sample;

[0077] S20312, respectively extracting features of the interpolated flood simulation result and the high-resolution terrain sample through a shallow feature extraction module to obtain shallow flood simulation features and terrain sample shallow features;

[0078] S20313, respectively extracting features of the shallow flood simulation features and the terrain sample shallow features through a deep feature extraction module to obtain deep flood simulation features and terrain sample deep features;

[0079] S20314, fusing the deep flood simulation features and the terrain sample deep features through a feature fusion module to obtain spatial detail residual data;

[0080] S20315, injecting the spatial detail residual data into the interpolated flood simulation result to obtain the high-resolution flood prediction result.

[0081] Specifically, as shown in Figure 3 the schematic diagram of the generator provided by the embodiment of the present application for generating the high-resolution flood prediction result. Before shallow feature extraction, the low-resolution simulation result is first interpolated to ensure the same resolution as the terrain data. The shallow feature extraction module obtains the shallow features of the terrain and the interpolated hydrodynamic simulation result based on 3x3 convolution. The deep feature extraction module is composed of a residual Swin Transformmer module (RSTB), a convolution layer (Conv) and a jump connection layer (Patch Embedding), which are used to extract global and local deep features of the terrain and the simulation result respectively. The feature fusion module is used to fuse the deep features of the terrain and the simulation result, and generate spatial detail residuals. The spatial detail residuals are fused with the interpolated low-resolution simulation result, and the high-resolution flood prediction result is generated.

[0082] As a further optional implementation, the water power model is:

[0083]

[0084] wherein t represents time, U is a vector composed of conservative variables, E and G are fluxes in x and y directions respectively, S is a source term, h represents water level, u and v represent water velocities in x and y directions respectively, q e is a mass source term per unit area, g represents gravitational acceleration, S 0x and S 0y are bottom slope terms in x and y directions respectively, S fx and S fy are friction slopes in x and y directions respectively, and n represents Manning roughness coefficient.

[0085] As a further optional implementation, the measured rainfall data is used as a boundary condition, and a water power model is used to perform coarse-scale grid flood simulation on a target area according to first water power parameters to obtain low-resolution flood simulation data of the target area, which specifically includes:

[0086] S1021, dividing the target area into a plurality of coarse-scale grids;

[0087] S1022, using the measured rainfall data as a boundary condition, and using a water power model to calculate water level data and flow velocity data of each coarse-scale grid according to the first water power parameters;

[0088] S1023, generating low-resolution flood simulation data according to the water level data and flow velocity data of each coarse-scale grid.

[0089] The method steps of the embodiments of the present application are described above. It can be recognized that the embodiments of the present application use a water power model to perform coarse-scale grid flood simulation on a target area to obtain low-resolution flood simulation data, and input the low-resolution flood simulation data and high-resolution terrain data into a super-resolution generative adversarial network for super-resolution reconstruction to obtain high-resolution flood simulation data. Compared with directly using a water power model to perform fine-scale grid flood simulation, the calculation efficiency is improved, thereby ensuring the simulation accuracy while improving the efficiency of flood simulation.

[0090] With reference to Figure 4 , the embodiments of the present application provide a high-precision flood simulation system, which comprises:

[0091] a data acquisition module configured to acquire measured rainfall data and high-resolution terrain data of a target area, and determine first water power parameters of the target area according to the high-resolution terrain data;

[0092] a low-resolution flood simulation module configured to use a hydrodynamic model to perform a coarse-scale grid flood simulation on the target region according to the first hydrodynamic parameters, to obtain low-resolution flood simulation data of the target region, using the measured rainfall data as boundary conditions;

[0093] a high-resolution flood reconstruction module configured to input the low-resolution flood simulation data and high-resolution terrain data into a pre-trained super-resolution generative adversarial network to obtain high-resolution flood simulation data of the target region.

[0094] The contents in the method embodiments are applicable to the system embodiments, the system embodiments achieve the same functions as the method embodiments, and achieve the same beneficial effects as the method embodiments.

[0095] Referring to Figure 5 The embodiment of the present application provides a high-precision flood simulation device, which comprises:

[0096] at least one processor;

[0097] at least one memory configured to store at least one program;

[0098] When the at least one program is executed by the at least one processor, the at least one processor implements the high-precision flood simulation method.

[0099] The contents in the method embodiments are applicable to the device embodiments, the device embodiments achieve the same functions as the method embodiments, and achieve the same beneficial effects as the method embodiments.

[0100] The embodiment of the present application further provides a computer readable storage medium, wherein the computer readable storage medium stores a program executable by a processor, and the program executable by the processor is used to execute the high-precision flood simulation method.

[0101] The computer readable storage medium of the embodiment of the present application can execute the high-precision flood simulation method provided by the method embodiment of the present application, and can execute the steps of any combination of the method embodiments, and has the corresponding functions and beneficial effects of the method.

[0102] The embodiment of the present application further discloses a computer program product or a computer program, which comprises computer instructions stored in a computer readable storage medium. A processor of a computer device can read the computer instructions from the computer readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method shown in the figure. Figure 1

[0103] ​In alternative embodiments, the functions / operations in the flow diagrams can occur in sequences other than those depicted. For example, two operations shown in succession can in fact be executed substantially concurrently or the operations can sometimes be executed in the reverse order depending upon the functionality / operations involved. Such variations are contemplated to be within the scope of the present application. Embodiments presented and described in the flow diagrams are examples only and are used to provide an enabling teaching for the present application. The processes disclosed are not limited to the order or specific blocks described. Alternative embodiments are contemplated, in which the order of the blocks is changed and where some blocks are performed in parallel rather than sequentially.

[0104] Moreover, while the present application has been described in the context of functional modules, it is to be understood that one or more of the functions and / or features described above can be integrated within a single physical device and / or software module, or one or more functions and / or features can be implemented in separate physical devices or software modules. It will also be appreciated that detailed discussion of the actual implementation of each module is not necessary for an enabling understanding of the application. Rather, the actual implementation is most readily derived from the description of the functionality of the various functional modules, in conjunction with the understanding of the properties, functions and interrelationships of the various functional modules presented in the context of the device disclosed herein. Therefore, the scope of the application is best understood from the appended claims, in conjunction with the full description and examples provided. It is to be understood that the specific concepts presented are merely illustrative of the application and are not intended to limit the scope of the application as defined by the claims. The scope of the application is defined by the claims and the full extent of equivalents to which such claims are entitled.

[0105] If the above functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or part of the prior art or the part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for making a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the methods in the embodiments of the present application. The aforementioned storage medium includes various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, etc.

[0106] The logic and / or steps represented in the flow diagrams or otherwise described herein, for example, can be embodied in non-transitory computer- readable media, executed by an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions, or in conjunction with which the instructions can be executed. In the context of this specification, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. The computer-readable medium can be, for example but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, device, or propagation medium.

[0107] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can also be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example, via optical scanning of the paper or other medium, then compiled, interpreted, or otherwise processed in a suitable manner, if necessary, and stored in a computer memory.

[0108] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the above embodiments, various steps or methods can be implemented in software or firmware that is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any of the following technologies, known in the art, or combinations thereof, can be used: a discrete logic circuit having logic gates for implementing logic functions upon data signals, an application specific integrated circuit having appropriate combinational logic gates, a programmable gate array (PGA), a field programmable gate array (FPGA), and / or the like.

[0109] In the above description of the present specification, reference is made to the description of terms such as "one embodiment / one example", "another embodiment / another example", or "certain embodiments / certain examples" and the like, which means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. The illustrative appearances of the above-mentioned terms in the description are not necessarily referred to the same embodiment or example throughout the specification. Moreover, the particular features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0110] While the embodiments of the application have been shown and described, it is to be understood that the embodiments described are merely exemplary and are not to be construed as limiting the scope of the application. The scope of the application is defined by the appended claims and their equivalents.

[0111] The above is the specific description of the preferred embodiment of the application, but the application is not limited to the above-mentioned embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the application, and these equivalent modifications or replacements are all included in the scope defined by the claims of the application.

Claims

1. A high-precision simulation method for flood, characterized in that, The method comprises the following steps: obtaining measured rainfall data and high-resolution terrain data of a target area, and determining first hydrodynamic parameters of the target area according to the high-resolution terrain data; using a hydrodynamic model to perform coarse-grid flood simulation on the target area according to the first hydrodynamic parameters, and obtaining low-resolution flood simulation data of the target area, wherein the measured rainfall data is used as a boundary condition; inputting the low-resolution flood simulation data and the high-resolution terrain data into a pre-trained super-resolution generative adversarial network to obtain high-resolution flood simulation data of the target area; the super-resolution generative adversarial network is obtained by the following steps of training: obtaining high-resolution terrain samples of a plurality of test areas, and determining second hydrodynamic parameters of the corresponding test areas according to the high-resolution terrain samples; randomly generating sample rainfall data, using a hydrodynamic model to perform coarse-grid flood simulation and fine-grid flood simulation on the test areas according to the second hydrodynamic parameters, respectively, and obtaining low-resolution flood simulation results and high-resolution flood simulation results of each test area, wherein the sample rainfall data is used as a boundary condition; training a pre-constructed generative adversarial network according to the low-resolution flood simulation results, the high-resolution terrain samples, and the high-resolution flood simulation results to obtain the trained super-resolution generative adversarial network; the training of the pre-constructed generative adversarial network according to the low-resolution flood simulation results, the high-resolution terrain samples, and the high-resolution flood simulation results comprises: inputting the low-resolution flood simulation results and the high-resolution terrain samples into a generator of the generative adversarial network to obtain high-resolution flood prediction results; inputting the high-resolution flood prediction results and the high-resolution flood simulation results into a discriminator of the generative adversarial network to obtain a discrimination result; determining a first loss value of the generator according to the discrimination result, and determining a second loss value of the discriminator according to the high-resolution flood prediction results and the high-resolution flood simulation results; updating parameters of the generator according to the first loss value, and updating parameters of the discriminator according to the second loss value; wherein the generator comprises a shallow feature extraction module, a deep feature extraction module, and a feature fusion module, the shallow feature extraction module is used to extract shallow features of input data, the deep feature extraction module is used to extract deep features according to the shallow features, and the feature fusion module is used to fuse a plurality of deep features; the inputting of the low-resolution flood simulation results and the high-resolution terrain samples into the generator of the generative adversarial network to obtain the high-resolution flood prediction results comprises: performing interpolation processing on the low-resolution flood simulation results to obtain an interpolated flood simulation result with the same resolution as the high-resolution terrain sample; The interpolation flood simulation result and the high-resolution terrain sample are respectively subjected to feature extraction by the shallow feature extraction module, so as to obtain flood simulation shallow features and terrain sample shallow features; The flood simulation shallow features and the terrain sample shallow features are respectively subjected to feature extraction by the deep feature extraction module, so as to obtain flood simulation deep features and terrain sample deep features; The flood simulation deep features and the terrain sample deep features are subjected to feature fusion by the feature fusion module, so as to obtain spatial detail residual data; The spatial detail residual data is injected into the interpolation flood simulation result, so as to obtain the high-resolution flood prediction result.

2. The high-precision flood simulation method according to claim 1, characterized in that, The hydrodynamic model is: where t denotes time, U is a vector of conserved variables, E and G are x and y fluxes in the x and y directions, S is a source term, h denotes water level, u and v denote x and y the average flow velocities in the x and y directions along the water level, q e is the mass source term per unit area, g denotes the gravitational acceleration, S 0x and S 0y are the bed slope terms in the x and y directions, x and y the bed slope terms in the x and y directions, S fx and S fy are the frictional slopes in the x and y directions, x and y the frictional slopes in the x and y directions, n denotes the Manning roughness coefficient.

3. The high-precision flood simulation method according to claim 1 or 2, characterized in that, The measured rainfall data is used as a boundary condition, and a hydrodynamic model is used to perform coarse-scale grid flood simulation on the target area according to the first hydrodynamic parameters, so as to obtain low-resolution flood simulation data of the target area, which specifically includes: The target area is divided into a plurality of coarse-scale grids; The measured rainfall data is used as a boundary condition, and a hydrodynamic model is used to calculate water level data and flow velocity data of each coarse-scale grid according to the first hydrodynamic parameters; The low-resolution flood simulation data is generated according to the water level data and the flow velocity data of each coarse-scale grid.

4. A high-precision simulation system for flood, characterized in that, A flood high-precision simulation method for realizing any one of claims 1 to 3, comprising: A data acquisition module for acquiring measured rainfall data and high-resolution terrain data of a target area, and determining first hydrodynamic parameters of the target area according to the high-resolution terrain data; A low-resolution flood simulation module for using a hydrodynamic model to perform coarse-scale grid flood simulation on the target area according to the first hydrodynamic parameters, with the measured rainfall data as a boundary condition, so as to obtain low-resolution flood simulation data of the target area; A high-resolution flood reconstruction module for inputting the low-resolution flood simulation data and the high-resolution terrain data into a pre-trained super-resolution generative adversarial network, so as to obtain high-resolution flood simulation data of the target area.

5. A high-precision simulation device for flood, characterized in that, Comprise: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a flood high-precision simulation method as claimed in any one of claims 1 to 3.

6. A computer readable storage medium having stored therein a program which is executable by a processor, characterized in that, The program executable by the processor when executed by the processor is used to perform a flood high-precision simulation method as claimed in any one of claims 1 to 3.

Images