Digital twinborn hybrid river channel hydrodynamic dimension reduction simulation and dimension rising visualization method

Abstract

The invention discloses a digital twinning-oriented hybrid riverway hydrodynamic dimension reduction simulation and dimension raising visualization method, and relates to the technical field of digital twinning, and the method comprises the following steps: S1, describing the complex form of a riverway according to the plait index and beach characteristic parameters of the riverway; s2, performing two-dimensional and three-dimensional simulation of working conditions in advance to obtain a refined flow field structure; s3, utilizing artificial intelligence training to establish a mapping relation between the complex form of the river channel and the two-dimensional and three-dimensional flow fields; the method is scientific and reasonable in structure and safe and convenient to use, the precision problem of a traditional one-dimensional model in a complex river channel is effectively solved by accurately describing and modeling the river channel form and the flow field structure, and the estimation method of the bed surface roughness and the vortex viscosity coefficient is improved by introducing local characteristics of the braided river channel, such as the beach density and the branch channel number. Therefore, the accuracy of the two-dimensional water flow simulation model is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital twinning, in particular to a river channel water power dimension reduction simulation and dimension increase visualization method for digital twinning. BACKGROUND

[0002] At present, for the water power simulation of long river channel, the method of numerically solving one-dimensional open channel Saint-Venant equation set is mainly adopted, which has been widely applied to the flow analysis of river channel. However, for complex river channels such as bifurcated and braided river channels, the water power characteristics have great complexity, and the one-dimensional model cannot effectively capture the complex flow field structure, island and beach distribution and local changes of water flow in the river channel, resulting in low simulation accuracy.

[0003] In addition, the parameters such as bed roughness and eddy viscosity coefficient commonly used in traditional models are difficult to accurately reflect the complex morphological characteristics of braided river channels, resulting in large errors in numerical simulation results. Although two-dimensional and three-dimensional models can better handle these complexities, they have huge calculation amount and low calculation efficiency, which is difficult to meet the real-time simulation demand in smart water conservancy.

[0004] Therefore, how to improve the calculation efficiency while ensuring the water power calculation accuracy and realize efficient and real-time visualization display on the digital twinning platform has become the focus of current research. SUMMARY

[0005] The present application provides a river channel water power dimension reduction simulation and dimension increase visualization method for digital twinning, which can effectively solve the problems raised in the background art.

[0006] To achieve the above purpose, the present application provides the following technical solution: a river channel water power dimension reduction simulation and dimension increase visualization method for digital twinning, comprising the following steps:

[0007] S1, describing the complex morphology of the river channel according to the braided index and island and beach characteristic parameters of the river channel;

[0008] S2, obtaining refined flow field structure by carrying out two-dimensional and three-dimensional simulation of working conditions in advance;

[0009] S3, establishing the mapping relationship between the complex morphology of the river channel and the two-dimensional and three-dimensional flow field by using artificial intelligence training;

[0010] S4, introducing plane morphology roughness, local roughness and horizontal eddy viscosity coefficient parameters in the water power model, improving the traditional two-dimensional flow simulation method and reflecting the flow characteristics of braided river channels;

[0011] S5, using the corrected water power model to simulate the one-dimensional simulation of long river section, obtaining the water power calculation result, and obtaining the two-dimensional flow field structure and island and beach submergence according to the water level data.

[0012] S6, a vertical logarithmic law distribution of water flow is introduced to further obtain a three-dimensional flow field structure to support the visualization display of the digital twin platform in smart water conservancy.

[0013] According to the above technical solution, the artificial intelligence training process of S3 uses a deep learning algorithm to establish a mapping relationship between the complex morphology of the river channel and the flow field;

[0014] By collecting a large amount of two-dimensional and three-dimensional flow field data and corresponding river channel morphology data of actual river channels, a training data set is constructed;

[0015] The convolutional neural network (CNN), recurrent neural network (RNN) or generative adversarial network (GAN) deep learning model is used to train the training data set, and the complex nonlinear relationship between the river channel morphology features and the flow field is automatically extracted.

[0016] According to the above technical solution, the three-dimensional flow field structure of S3 is calculated by the vertical logarithmic law distribution of water flow, and is used for three-dimensional visualization display of the digital twin platform;

[0017] The three-dimensional flow field structure is calculated by the vertical logarithmic law distribution of water flow, and is used for three-dimensional visualization display of the digital twin platform.

[0018] According to the above technical solution, the specific vertical logarithmic law distribution is usually represented as:

[0019]

[0020] u(z) is the water flow velocity at a depth z;

[0021] u ref is the water flow velocity at a reference height (usually the velocity of the water surface or a certain specific layer);

[0022] κ is the von Karman constant, equal to 0.41;

[0023] z is the vertical depth of the water flow;

[0024] z0 is the reference height of the friction layer, which is the roughness length on the bed surface.

[0025] According to the above technical solution, the visualization display of the digital twin platform includes dynamic display of the flow field structure, inundation of the beach, and real-time update of the hydrodynamic calculation results;

[0026] In addition, the platform can also update the hydrodynamic calculation results in real time, automatically adjust the displayed flow field data based on different working conditions and water flow states.

[0027] According to the technical scheme, the planar morphology roughness of S4 includes parameters determined by the bar distribution of the river channel, the river channel morphology and the water flow characteristics;

[0028] Specifically,

[0029] The area, shape and distribution density of the bars;

[0030] The longitudinal and transverse geometric characteristics of the river channel, including the width, depth and slope change of the river channel;

[0031] The local variation of the river channel inner surface roughness under the action of the water flow, including the influence of the water flow velocity and flow direction;

[0032] The spatial distribution of the bar and branch water power characteristics in the river channel and the influence on the flow field;

[0033] And the interaction characteristics of the water flow and the river bed, including the erosion and deposition of the water flow on the bar area.

[0034] According to the technical scheme, the two-dimensional water flow simulation model of S4 adds the local bar density and the branch number of the braided river channel characteristics, and improves the estimation method of the traditional bed roughness n and the horizontal eddy viscosity coefficient u;

[0035] The two-dimensional water flow simulation model adds the local bar density and the branch number of the braided river channel characteristics, and improves the estimation method of the traditional bed roughness n and the horizontal eddy viscosity coefficient u, specifically including:

[0036] In the traditional model, the bed roughness n and the horizontal eddy viscosity coefficient u are usually estimated depending on the macroscopic geometric morphology of the river bed, but this method cannot effectively consider the local feature change in the braided river channel;

[0037] Therefore, the unique characteristics of the local bar density and the branch number of the braided river channel are combined.

[0038] According to the technical scheme, the one-dimensional simulation model of S5 adopts an integral method based on the two-dimensional flow field structure for hydrodynamic calculation, and is used for further generating the hydrodynamic data of the long river section.

[0039] According to the technical scheme, the one-dimensional simulation model adopts an integral method based on the two-dimensional flow field structure for hydrodynamic calculation, and is used for further generating the hydrodynamic data of the long river section, specifically including:

[0040] Firstly, the water level and flow velocity parameters along the cross section of the river channel are obtained through the hydrodynamic calculation results of the two-dimensional flow field structure;

[0041] Then, the two-dimensional flow field data is integrated along the river width direction to obtain the hydrodynamic data along the river direction, including the water level distribution and flow one-dimensional information;

[0042] Through the integral method, complex flow characteristics in a two-dimensional flow field can be effectively converted into a one-dimensional hydrodynamic model suitable for a long river section, thereby reducing the calculation burden and improving the calculation efficiency.

[0043] Finally, based on the calculation results of the one-dimensional model, hydrodynamic data of the long river section are further generated.

[0044] According to the technical scheme, the modified one-dimensional model can adapt to complex river environments and various flow states without increasing too much calculation burden.

[0045] Compared with the prior art, the method has the beneficial effects that: the structure of the method is scientific and reasonable, and the method is safe and convenient to use; the method effectively solves the precision problem of a traditional one-dimensional model in a complex river by precisely describing and modeling the river morphology and flow field structure; the method improves the estimation method of the bed roughness and the eddy viscosity coefficient by introducing local characteristics of a braided river, such as the island density and the number of branches, thereby improving the precision of a two-dimensional flow simulation model; the method uses an integral method based on a two-dimensional flow field in hydrodynamic calculation, and obtains a three-dimensional flow field structure in combination with a vertical logarithmic law distribution model, thereby further improving the precision and calculation efficiency of simulation; the improved hydrodynamic model can quickly generate hydrodynamic data of a long river section, and meet the needs of a digital twin platform for real-time updating of a flow field structure, island submergence, and visualization of hydrodynamic calculation results; the method can provide high-precision hydrodynamic analysis results and real-time visualization display in a smart water conservancy system, and provide effective technical support for water resource management, river regulation, and environmental monitoring. BRIEF DESCRIPTION OF DRAWINGS

[0046] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, illustrate the present application, and are used to explain the present application together with the embodiments of the present application, and do not constitute a limitation on the present application.

[0047] In the drawings:

[0048] Figure 1 is a schematic diagram of the method steps of the present application. DETAILED DESCRIPTION

[0049] The preferred embodiments of the present application are described below in conjunction with the accompanying drawings, and it should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and do not limit the present application.

[0050] Embodiment: As shown in the accompanying drawings, the present application provides a technical scheme, a river channel hydrodynamic dimension reduction and dimension increase visualization method for a digital twin, including the following steps: Figure 1

[0051] S1, describing the complex morphology of the river according to the braided index and island characteristic parameters of the river; ​

[0052] S2, obtain a refined flow field structure by conducting two-dimensional and three-dimensional simulation of the working condition in advance;

[0053] S3, establish a mapping relationship between the complex morphology of the river channel and the two-dimensional and three-dimensional flow field by artificial intelligence training;

[0054] S4, introduce the plane morphology roughness, local roughness, and horizontal eddy viscosity coefficient parameters into the hydrodynamic model to improve the traditional two-dimensional flow simulation method and reflect the flow characteristics of the braided river channel;

[0055] S5, use the corrected hydrodynamic model to perform one-dimensional simulation of the long river section, obtain the hydrodynamic calculation results, and obtain the two-dimensional flow field structure and the inundation condition of the beach according to the water level data;

[0056] S6, introduce the vertical logarithmic law distribution of the flow to further obtain the three-dimensional flow field structure to support the visual display of the digital twin platform in smart water conservancy.

[0057] Further, the artificial intelligence training process of S3 uses a deep learning algorithm to establish a mapping relationship between the complex morphology of the river channel and the flow field;

[0058] By collecting a large amount of two-dimensional and three-dimensional flow field data and corresponding river channel morphology data of actual river channels, a training data set is constructed;

[0059] The deep learning model such as convolutional neural network (CNN), recurrent neural network (RNN), or generative adversarial network (GAN) is used to train the training data set to automatically extract the complex nonlinear relationship between the river channel morphology features and the flow field;

[0060] An appropriate loss function and optimization algorithm are used to optimize the parameters of the deep learning model to improve the prediction accuracy of the model under different river channel morphologies and flow conditions;

[0061] The deep learning model obtained by training can accurately predict the influence of complex morphology changes of the river channel on the flow field structure, thereby realizing efficient mapping between the river channel morphology and the flow field.

[0062] Further, the three-dimensional flow field structure of S3 is calculated by the vertical logarithmic law distribution of the flow and is used for three-dimensional visualization display of the digital twin platform;

[0063] The three-dimensional flow field structure of S3 is calculated by the vertical logarithmic law distribution of the flow and is used for three-dimensional visualization display of the digital twin platform, specifically including:

[0064] First, according to the vertical distribution characteristics of the flow, a vertical logarithmic law distribution model is used to calculate the flow velocity and flow field structure to accurately simulate the flow of the water body at different depths;

[0065] Then, using the water flow velocity gradient and turbulence information obtained from the logarithmic law distribution, a three-dimensional flow field structure model is constructed to accurately describe the vertical variation and three-dimensional flow characteristics of the water flow in the river channel. This three-dimensional flow field data can be further combined with two-dimensional flow field data to fully exhibit the spatial variation of the water flow;

[0066] Finally, the three-dimensional flow field structure is visualized through the digital twin platform, providing three-dimensional dynamic water flow information, flow field structure, island and beach inundation, and other hydrodynamic characteristics of the river channel, helping users more intuitively understand the river channel hydrodynamic characteristics and environmental changes.

[0067] Further, the specific vertical logarithmic law distribution is usually expressed as:

[0068]

[0069] u(z) is the water flow velocity at depth z;

[0070] u ref is the water flow velocity at the reference height (usually the water surface or the velocity of a certain specific layer);

[0071] κ is the von Karman constant, equal to 0.41;

[0072] z is the vertical depth of the water flow;

[0073] z0 is the reference height of the friction layer, which is the roughness length on the bed surface;

[0074] This formula is used to describe the vertical velocity distribution of the water flow and is suitable for simulating the water flow in the river channel, especially the flow rate variation at different depth positions.

[0075] Further, the visualization of the digital twin platform includes dynamic display of flow field structure, island and beach inundation, and real-time updating of hydrodynamic calculation results, specifically including:

[0076] By real-time acquisition and updating of hydrodynamic calculation data, the digital twin platform can dynamically display the changes of the flow field structure in the river channel, showing the water flow velocity, flow direction, vortex distribution, etc.

[0077] On this basis, the platform can accurately display the inundation of the island and beach area, helping users analyze the evolution of the island and beach under different water level conditions and the influence of the water flow on the island and beach;

[0078] In addition, the platform can also real-time update the hydrodynamic calculation results, automatically adjust the displayed flow field data based on different working conditions and water flow states, and ensure the timeliness and accuracy of the displayed information;

[0079] This dynamic display can provide more detailed water flow monitoring and decision support for the intelligent water conservancy system.

[0080] Further, the planform roughness of S4 includes parameters determined by the distribution of the river channel, the river channel morphology, and the water flow characteristics;

[0081] Specifically:

[0082] The area, shape, and distribution density of the river channel;

[0083] The longitudinal and transverse geometric characteristics of the river channel, including the width, depth, and slope changes of the river channel;

[0084] The local variation of the river channel surface roughness under the action of water flow, including the influence of water flow velocity and flow direction;

[0085] The spatial distribution of the river channel island, branch water power characteristics and their influence on the flow field;

[0086] And the interaction characteristics between water flow and riverbed, including the erosion and deposition of water flow on the island area;

[0087] These parameters jointly determine the influence of "planform roughness" on hydrodynamic simulation, thereby affecting the simulation accuracy.

[0088] Further, the two-dimensional flow simulation model of S4 improves the estimation method of the traditional bed roughness n and horizontal eddy viscosity coefficient u by adding local island density and branch number and other braided river characteristics;

[0089] The two-dimensional flow simulation model improves the estimation method of the traditional bed roughness n and horizontal eddy viscosity coefficient u by adding local island density and branch number and other braided river characteristics, specifically including:

[0090] In the traditional model, the bed roughness n and the horizontal eddy viscosity coefficient u are usually estimated depending on the macroscopic geometric shape of the riverbed, but this method cannot effectively consider the local feature changes in the braided river;

[0091] Therefore, combined with the unique characteristics of the braided river such as local island density, branch number, etc., more accurate simulation of water flow resistance, flow velocity distribution, and flow field diffusion, etc. is carried out;

[0092] By simulating different types of braided rivers, a local roughness model related to parameters such as island density and branch number is constructed, the calculation method of bed roughness n is optimized, and based on the complex morphology of the braided river, the estimation method of the horizontal eddy viscosity coefficient u is improved to better reflect the shear stress effect of water flow on the riverbed, thereby improving the simulation accuracy.

[0093] Further, the one-dimensional simulation model of S5 uses an integral method based on the two-dimensional flow field structure for hydrodynamic calculation, and is used to further generate hydrodynamic data for long river sections.

[0094] Further, the one-dimensional simulation model adopts an integral method based on the two-dimensional flow field structure for hydrodynamic calculation and is used to further generate hydrodynamic data of long river sections, specifically including:

[0095] Firstly, through the hydrodynamic calculation results of the two-dimensional flow field structure, the water level, flow velocity and other parameters along the cross section of the river are obtained;

[0096] Then, the two-dimensional flow field data is integrated along the river width direction to obtain the hydrodynamic data along the river direction, including water level distribution, flow and other one-dimensional information;

[0097] Through this integral method, the complex flow characteristics in the two-dimensional flow field can be effectively converted into a one-dimensional hydrodynamic model suitable for long river sections, thereby reducing the calculation burden and improving the calculation efficiency;

[0098] Finally, based on the calculation results of the one-dimensional model, the hydrodynamic data of the long river section is further generated, providing accurate support for subsequent flow prediction and digital twin platform display.

[0099] Further, the corrected one-dimensional model can provide high calculation accuracy without increasing excessive calculation burden, adapting to complex river environments and various flow states.

[0100] The hydrodynamic simulation method provided herein covers one-dimensional, two-dimensional and three-dimensional models, which can efficiently and accurately simulate and visualize complex river environments;

[0101] Firstly, in one-dimensional simulation, the water level and flow velocity along the cross section of the river are obtained through an integral method based on the two-dimensional flow field structure, and then the two-dimensional flow field data is integrated along the river width direction to generate hydrodynamic data suitable for long river sections, thereby effectively simplifying the calculation complexity and improving the efficiency and ensuring the calculation accuracy;

[0102] Secondly, the two-dimensional simulation obtains refined flow field structure through two-dimensional simulation of the working condition in advance, and introduces parameters such as planar form roughness, local roughness and horizontal eddy viscosity coefficient in the hydrodynamic model, improving the traditional flow simulation method. Specifically, combining the characteristics of braided river, such as local island shoal density and number of branches, the estimation method of bed roughness n and horizontal eddy viscosity coefficient u is optimized, thereby improving the accuracy of the two-dimensional flow simulation model;

[0103] Finally, the three-dimensional simulation establishes the mapping relationship between the complex morphology of the river channel and the two-dimensional and three-dimensional flow field through artificial intelligence algorithms such as convolutional neural network (CNN), recurrent neural network (RNN) or generative adversarial network (GAN) training, accurately simulates the three-dimensional flow field structure of the water flow, and performs calculation based on the vertical logarithmic law distribution of the water flow. The three-dimensional flow field structure can be used for real-time visual display of the digital twin platform, and detailed water flow dynamics, beach inundation conditions and hydrodynamic calculation results are provided.

[0104] Finally, it should be noted that the above is only a preferred example of the present application and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or equivalently replace some technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for dimensionality reduction simulation and dimensionality increase visualization of complex river hydrodynamics for digital twins, characterized by: The steps include: S1. Describe the complex morphology of the river channel based on the braiding index and shoal characteristic parameters; S2. Obtain a refined flow field structure by conducting two-dimensional and three-dimensional simulations of the working conditions in advance; S3. Use artificial intelligence training to establish a mapping relationship between the complex morphology of the river channel and the two-dimensional and three-dimensional flow fields; S4. Introducing the parameters of plane roughness, local roughness, and horizontal eddy viscosity coefficient into the hydrodynamic model to improve the traditional two-dimensional water flow simulation method and reflect the water flow characteristics of braided rivers; S5. Use the modified hydrodynamic model to perform one-dimensional simulation of the long river section to obtain hydrodynamic calculation results, and obtain the two-dimensional flow field structure and the inundation of the island and shoal based on the water level data; S6. Introduce the vertical logarithmic distribution of water flow to further obtain the three-dimensional flow field structure to support the visualization display of the digital twin platform in smart water conservancy.

2. A method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 1, characterized in that: The S3's artificial intelligence training process uses a deep learning algorithm to establish a mapping relationship between the complex morphology of the river channel and the flow field; By collecting a large amount of two-dimensional and three-dimensional flow field data of actual rivers and corresponding river morphology data, a training dataset is constructed; The training dataset is trained using convolutional neural networks, recursive neural networks or generative adversarial networks deep learning models to automatically extract the complex nonlinear relationship between river morphological characteristics and flow fields.

3. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 2 is characterized in that: The three-dimensional flow field structure of S3 is calculated by the vertical logarithmic distribution of water flow and used for the three-dimensional visualization of the digital twin platform; The three-dimensional flow field structure is obtained by calculating the vertical logarithmic distribution of water flow and is used for the three-dimensional visualization display of the digital twin platform.

4. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 3 is characterized in that: The specific vertical logarithmic distribution is usually expressed as: u(z) is the water velocity at depth z; u ref is the water velocity at a reference height (usually the velocity at the water surface or a specific layer); κ is the von Karman constant, which is equal to 0.41; z is the vertical depth of the water flow; z0 is the reference height of the friction layer and is the roughness length on the bed surface.

5. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 4 is characterized in that: The visualization of the digital twin platform includes dynamic display of flow field structure, inundation of islands and shoals, and real-time update of hydrodynamic calculation results; In addition, the platform can update the hydrodynamic calculation results in real time and automatically adjust the displayed flow field data based on different working conditions and water flow states.

6. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 1 is characterized in that: The plane roughness of S4 includes parameters determined by the distribution of river shoals, river morphology and water flow characteristics; Specifically: The area, shape and distribution density of beaches; The longitudinal and transverse geometric characteristics of the river channel, including its width, depth and slope changes; The local changes in the surface roughness of the river channel under the action of water flow, including the influence of water velocity and flow direction; The spatial distribution of hydrodynamic characteristics of shoals and distributaries within the river and their impact on the flow field; and the interaction characteristics of water flow and riverbed, including the erosion and sedimentation effects of water flow on beach areas.

7. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 6 is characterized in that: The two-dimensional flow simulation model of S4 improves the traditional estimation method of bed roughness n and horizontal eddy viscosity coefficient u by incorporating local beach density and braided channel characteristics; The two-dimensional flow simulation model improves the traditional estimation methods of bed roughness n and horizontal eddy viscosity coefficient u by incorporating local shoal density and braided channel characteristics. Specifically, In traditional models, the bed roughness n and horizontal eddy viscosity coefficient u are usually estimated based on the macroscopic geometry of the riverbed, but this method cannot effectively consider the local characteristic changes in braided channels. Therefore, combined with the unique characteristics of local shoal density and number of branches, braided rivers are considered.

8. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 1 is characterized in that: The one-dimensional simulation model of S5 uses an integration method based on a two-dimensional flow field structure to perform hydrodynamic calculations and is used to further generate hydrodynamic data for long river sections.

9. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 8 is characterized in that: The one-dimensional simulation model uses an integration method based on the two-dimensional flow field structure to calculate the hydrodynamics and is used to further generate hydrodynamic data for long river sections, including: First, the water level and flow velocity parameters at each section along the river are obtained through the hydrodynamic calculation results of the two-dimensional flow field structure; Then, these two-dimensional flow field data are integrated along the river width to obtain the hydrodynamic data along the river channel, including one-dimensional information of water level distribution and flow rate; Through this integration method, the complex flow characteristics in the two-dimensional flow field can be effectively converted into a one-dimensional hydrodynamic model suitable for long river sections, thereby reducing the computational burden and improving computational efficiency; Finally, based on the calculation results of the one-dimensional model, hydrodynamic data of the long river section are further generated.

10. The method for dimensionality reduction simulation and dimensionality increase visualization of hydrodynamics of a complex river channel for digital twins according to claim 9 is characterized in that: The modified one-dimensional model can adapt to complex river environments and various water flow states without increasing excessive computational burden.

Images