Rough factor parameter rapid estimation method and device and storage medium

By replacing the solution of the traditional hydrodynamic river model by replacing the response surface replacement model, the problem of large and time-consuming calculation of roughness parameter estimation is solved, and the rapid estimation of roughness parameter is realized, which is suitable for short-term flood rapid forecasting.

CN118886353AActive Publication Date: 2024-11-01CHINA THREE GORGES CORPORATION

Patent Information

Application Number
CN202410923144.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-10
Publication Date
2024-11-01
Estimated Expiration
2044-07-10

AI Technical Summary

Technical Problem

The existing roughness parameter estimation method is computationally expensive and time-consuming in hydrodynamic river models, which limits its application with high real-time requirements such as rapid forecasting of short-term floods.

Method used

Using the response surface replacement model, the training set and verification set are generated and optimized, and the solution of the traditional one-dimensional hydrodynamic river model is replaced by replacing the solution of the traditional one-dimensional hydrodynamic river model. Only a few solutions are required to obtain the optimal roughness parameter.

Benefits of technology

It significantly reduces the estimation time of roughness parameters, improves the estimation rate, and can be applied more quickly to production work with high real-time requirements such as rapid forecasting of short-term floods.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118886353A_ABST
    Figure CN118886353A_ABST
Patent Text Reader

Abstract

The invention provides a roughness factor parameter rapid estimation method and device and a storage medium, and relates to the technical field of hydrological simulation. And acquiring water channel data, constructing a corresponding one-dimensional hydrodynamic simulation model according to the water channel data, generating a roughness parameter set, substituting the roughness parameter set into the one-dimensional hydrodynamic simulation model, and solving the one-dimensional hydrodynamic simulation model to obtain water flow simulation data of the observation section. Calculating a simulation error target value corresponding to each element in the roughness parameter set in combination with the water flow simulation data to form a simulation error target value set; generating, optimizing and verifying a response surface substitution model; and applying the response surface substitution model meeting the precision requirement to the roughness parameter set to obtain an optimal roughness parameter. According to the method, the optimal roughness parameter is accurately obtained, the calculated amount is small, the consumed time is short, the estimation rate of the roughness parameter is greatly improved, and the optimal roughness parameter can be estimated more quickly, so that the method can be better applied to production work with higher real-time requirement, such as short-term flood quick forecasting and the like.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of hydrological simulation, and in particular to a method, a device and a storage medium for quickly estimating a roughness parameter. Background Art

[0002] The roughness parameter used to describe the riverbed roughness in current one-dimensional hydrodynamic river models cannot usually be obtained directly through field measurements. The estimation of the roughness parameter is usually a manual identification process that relies on the modeler's experience and expertise.

[0003] At present, the commonly used roughness parameter estimation method is the sensitivity analysis method, which samples the parameter values ​​within a given range, performs a large number of model runs, and fits the actual observed values ​​to obtain the parameter value with the best fit. However, the existing roughness parameter estimation method based on hydrodynamic model simulation has a large amount of calculation and a long time consumption, which limits its application in production work with high real-time requirements such as short-term flood rapid forecasting. Therefore, it is urgent to propose a roughness parameter rapid estimation method with small calculation amount and short time consumption, so as to better apply it to production work with high real-time requirements such as short-term flood rapid forecasting. Summary of the invention

[0004] In view of the above problems, the present invention proposes a method, device and storage medium for quickly estimating roughness parameters.

[0005] An embodiment of the present invention provides a method for quickly estimating a roughness parameter, and the method for quickly estimating a roughness parameter includes:

[0006] Acquire waterway data, and construct a corresponding one-dimensional hydrodynamic simulation model according to the waterway data, wherein the waterway includes: a river or a channel;

[0007] Generate a roughness parameter set and substitute it into the one-dimensional hydrodynamic simulation model, solve the one-dimensional hydrodynamic simulation model with the flow rate of the inflow and outflow sections of the waterway as boundary conditions, and obtain water flow simulation data of the observation section;

[0008] Acquire the water flow observation data of the observation section, and calculate the simulation error target value corresponding to each element in the roughness parameter set in combination with the water flow simulation data to form a simulation error target value set;

[0009] Based on the roughness parameter set, a training set of a response surface substitution model is generated, and the training set is used as input, and the simulation error target value corresponding to the element in the training set is output, so as to optimize the response surface substitution model;

[0010] Based on the roughness parameter set, generating a validation set of the response surface substitution model, and verifying the simulation accuracy of the response substitution model to meet the accuracy requirement;

[0011] A response surface substitution model that meets the accuracy requirement is applied to the roughness parameter set to obtain an optimal roughness parameter.

[0012] Optionally, acquiring waterway data and constructing a corresponding one-dimensional hydrodynamic simulation model according to the waterway data includes:

[0013] Acquiring dimension data of the river channel or the channel, wherein the dimension data includes: cross-sectional dimensions and slope;

[0014] Based on the cross-sectional dimensions and the slope, the one-dimensional hydrodynamic simulation model is constructed.

[0015] Optionally, generate a roughness parameter set, including:

[0016] The roughness parameter set is generated within a preset value range by a random sampling method.

[0017] Optionally, the roughness parameter set is generated within a preset value range by a random sampling method, including:

[0018] Set the value range of roughness parameters according to experience;

[0019] The roughness parameter set is generated within the value range by a Latin hypercube sampling method.

[0020] Optionally, calculating a simulation error target value corresponding to each element in the roughness parameter set to form a simulation error target value set includes:

[0021] A root mean square error calculation is performed on the water flow simulation data and the water flow observation data to obtain a simulation error target value corresponding to each element in the roughness parameter set to form the simulation error target value set.

[0022] Optionally, generating a training set of a response surface substitution model based on the roughness parameter set includes:

[0023] The training set is obtained by sampling in the roughness parameter set using a Latin hypercube sampling method, and the number of elements included in the training set does not exceed the number of elements included in the validation set.

[0024] Optionally, taking the training set as input and the simulation error target value corresponding to the element in the training set as output, optimizing the response surface substitution model includes:

[0025] The Gaussian process algorithm was used to generate the initial response surface surrogate model;

[0026] Taking the training set as the input of the Gaussian process algorithm, taking the simulation error target value corresponding to the elements in the training set as the output of the Gaussian process algorithm, and optimizing the hyperparameters to be optimized in the Gaussian process algorithm by using a particle swarm optimization algorithm;

[0027] The response surface substitution model is obtained based on the optimized hyperparameters.

[0028] Optionally, generating a validation set of the response surface substitution model based on the roughness parameter set includes:

[0029] The validation set is obtained by sampling in a target roughness parameter set using a Latin hypercube sampling method, wherein the target roughness parameter set is a set formed by remaining elements in the roughness parameter set after removing elements in the training set.

[0030] Optionally, verifying the simulation accuracy of the response surrogate model to meet the accuracy requirement includes:

[0031] If the simulation accuracy of the response substitution model does not meet the accuracy requirement, the number of elements in the training set is increased based on the roughness parameter to form a new training set, and the steps of optimizing the response surface and verifying the simulation accuracy of the response substitution model are iteratively performed until the simulation accuracy of the response substitution model meets the accuracy requirement.

[0032] Optionally, applying a response surface substitution model that meets the accuracy requirement to the roughness parameter set to obtain an optimal roughness parameter includes:

[0033] The roughness parameter set is used as an input of the response surface substitution model that meets the accuracy requirement, and the response surface substitution model that meets the accuracy requirement outputs a simulation error target value corresponding to each element in the roughness parameter set;

[0034] Determine the minimum simulation error target value among the simulation error target values ​​corresponding to all elements in the roughness parameter set as the target simulation error target value;

[0035] The element corresponding to the target simulation error target value is determined as the optimal roughness parameter.

[0036] Optionally, constructing the one-dimensional hydrodynamic simulation model includes:

[0037] The constructed one-dimensional hydrodynamic simulation model is expressed using the Saint-Venant equations as follows:

[0038]

[0039]

[0040] In the above two formulas, Q is the flow rate; x is the direction of water flow; B is the width of the waterway; Z is the water level; t is the time; A is the cross-sectional area of ​​the water flow; g is the gravitational acceleration; n is the roughness parameter; R is the hydraulic radius; V is the average flow velocity of the section; ql and ul are the lateral inflow of a single field length and the component of the lateral inflow in the x direction, respectively.

[0041] Optionally, the expression of the root mean square error operation is:

[0042]

[0043] In the above formula, RMSE is the root mean square error; N is the number of observation points; Qs is the water flow simulation data; Qo is the water flow observation data.

[0044] Optionally, the Gaussian process algorithm is represented by a mean function μ(m) and a covariance function k(m, m'), and the mean function μ(m) is selected to be 0;

[0045] The diagonal square exponential covariance function is selected as the covariance function k(m, m'), and its expression is: In this formula, σf and λ are the hyperparameters to be optimized in the Gaussian process algorithm.

[0046] An embodiment of the present invention provides a roughness parameter fast estimation device, the roughness parameter fast estimation device comprising:

[0047] A model building module is used to request waterway data and build a corresponding one-dimensional hydrodynamic simulation model according to the waterway data. The waterway includes a river or a channel.

[0048] A water flow simulation data solving module is used to generate a roughness parameter set and substitute it into the one-dimensional hydrodynamic simulation model, and solve the one-dimensional hydrodynamic simulation model with the flow of the inflow and outflow sections of the waterway as boundary conditions to obtain the water flow simulation data of the observation section;

[0049] A simulation error target value calculation module is used to obtain the water flow observation data of the observation section, and calculate the simulation error target value corresponding to each element in the roughness parameter set in combination with the water flow simulation data to form a simulation error target value set;

[0050] A response surface substitution model generation and optimization module is used to generate a training set of the response surface substitution model based on the roughness parameter set, and use the training set as input, the simulation error target value corresponding to the element in the training set as output, and optimize the response surface substitution model;

[0051] A verification module, used to generate a verification set of the response surface substitution model based on the roughness parameter set, and verify the simulation accuracy of the response substitution model to meet the accuracy requirement;

[0052] The optimal roughness parameter module is used to apply the response surface substitution model that meets the accuracy requirements to the roughness parameter set to obtain the optimal roughness parameter.

[0053] An embodiment of the present invention further provides a computer-readable storage medium, wherein a computer program stored in the storage medium enables a processor to execute the method for rapid estimation of roughness parameters as described in any one of the above items.

[0054] The roughness parameter rapid estimation method provided by the present invention first obtains waterway data, and constructs a corresponding one-dimensional hydrodynamic simulation model based on the waterway data, then generates a roughness parameter set and substitutes it into the one-dimensional hydrodynamic simulation model, and uses the flow rates of the inflow and outflow sections of the waterway as boundary conditions to solve the one-dimensional hydrodynamic simulation model and obtain water flow simulation data of the observed section.

[0055] Then, the water flow observation data of the observation section is obtained, and combined with the water flow simulation data, the simulation error target value corresponding to each element in the roughness parameter set is calculated to form a simulation error target value set; based on the roughness parameter set, a training set of the response surface substitution model is generated, and the training set is used as input, and the simulation error target value corresponding to the element in the training set is output to optimize the response surface substitution model;

[0056] Based on the roughness parameter set, a validation set of the response surface substitution model is generated, and the simulation accuracy of the response surface substitution model is tested to meet the accuracy requirements; finally, the response surface substitution model that meets the accuracy requirements is applied to the roughness parameter set to obtain the optimal roughness parameters.

[0057] The present invention aims to solve the problems of large amount of calculation and long time consumption of the traditional roughness parameter estimation method based on the hydrodynamic model simulation. The method of rapid estimation of roughness parameters is proposed. The method trains and replaces the solution of the traditional one-dimensional hydrodynamic river model with the response surface substitution model. It only needs to solve a few one-dimensional hydrodynamic river models instead of solving the one-dimensional hydrodynamic river model for each element in the roughness parameter set. While accurately obtaining the optimal roughness parameter, the calculation amount is small and the time consumption is short, which greatly improves the estimation rate of the roughness parameter. The optimal roughness parameter can be estimated more quickly, so that it can be better applied to production work with high real-time requirements such as short-term flood rapid forecasting. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0059] Figure 1 is a flow chart of a method for quickly estimating roughness parameters according to an embodiment of the present invention;

[0060] Figure 2 It is a block diagram of a roughness parameter fast estimation device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0061] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, are only part of the embodiments of the present invention, not all of the embodiments, and are not used to limit the present invention.

[0062] Reference Figure 1 , shows a flow chart of a method for quickly estimating roughness parameters according to an embodiment of the present invention, the method for quickly estimating roughness parameters comprising:

[0063] Step 101: Acquire waterway data, and construct a corresponding one-dimensional hydrodynamic simulation model based on the waterway data. The waterway includes: a river or a channel.

[0064] The roughness parameter rapid estimation method proposed in the present invention first obtains waterway data and constructs a corresponding one-dimensional hydrodynamic simulation model based on the waterway data, wherein the waterway includes: rivers or channels and other waterways involving hydrological simulation and roughness parameters.

[0065] A better method for constructing a one-dimensional hydrodynamic simulation model includes: first obtaining the dimension data of a river or channel, the dimension data including: cross-sectional dimensions and slope; and then constructing a one-dimensional hydrodynamic simulation model based on the cross-sectional dimensions and slope.

[0066] In order to better understand the constructed one-dimensional hydrodynamic simulation model, the Saint-Venant equations are preferably used to express the constructed one-dimensional hydrodynamic simulation model, and its expression is as follows:

[0067]

[0068] In the above two formulas, Q is the flow rate; x is the direction of water flow; B is the width of the waterway; Z is the water level; t is the time; A is the cross-sectional area of ​​the water flow; g is the gravitational acceleration; n is the roughness parameter; R is the hydraulic radius; V is the average flow velocity of the section; ql and ul are the lateral inflow of a single field length and the component of the lateral inflow in the x direction, respectively.

[0069] Step 102: Generate a roughness parameter set and substitute it into a one-dimensional hydrodynamic simulation model, use the flow rates at the inflow and outflow sections of the waterway as boundary conditions, solve the one-dimensional hydrodynamic simulation model, and obtain water flow simulation data of the observation section.

[0070] After the one-dimensional hydrodynamic simulation model is constructed, a roughness parameter set is generated, and then the roughness parameters are substituted into the one-dimensional hydrodynamic simulation model. The flow rates at the inflow and outflow sections of the waterway are used as boundary conditions to solve the one-dimensional hydrodynamic simulation model and obtain the water flow simulation data of the observed section.

[0071] For the generation of the roughness parameter set, the preferred method is to generate the roughness parameter set within a preset value range by random sampling. Specifically:

[0072] The value range of the roughness parameter is set according to the empirical value, for example, the value range is (0.05, 0.5), and then the Latin hypercube sampling method is used to generate a roughness parameter set within the value range. For example, 100 values ​​are randomly sampled as 100 elements in the value range (0.05, 0.5), and these 100 elements form the roughness parameter set.

[0073] Step 103: Obtain the water flow observation data of the observation section, and calculate the simulation error target value corresponding to each element in the roughness parameter set in combination with the water flow simulation data to form a simulation error target value set.

[0074] After obtaining the water flow simulation data of the observation section, the water flow observation data of the observation section is obtained, and then combined with the calculated water flow simulation data, the simulation error target value corresponding to each element in the roughness parameter set is calculated to form a simulation error target value set.

[0075] A better method for obtaining a simulation error target value includes: performing a root mean square error calculation on water flow simulation data and water flow observation data to obtain a simulation error target value corresponding to each element in a roughness parameter set to form a simulation error target value set.

[0076] For the root mean square error operation, the expression is:

[0077]

[0078] In the above formula, RMSE is the root mean square error; N is the number of observation points; Qs is the water flow simulation data; Qo is the water flow observation data.

[0079] Step 104: Based on the roughness parameter set, a training set of the response surface substitution model is generated, and the training set is used as input, and the simulation error target value corresponding to the elements in the training set is output to optimize the response surface substitution model.

[0080] After obtaining the set of simulation error target values, a training set of the response surface substitution model is generated based on the roughness parameter set. Preferably, the Latin hypercube sampling method can be used to sample the roughness parameter set to obtain a training set, and the number of elements contained in the training set does not exceed the number of elements contained in the validation set. For example, 30 elements can be sampled from the 100 elements of the roughness parameter set by the Latin hypercube sampling method to obtain a training set, and the remaining 70 elements are waiting to form a validation set later. In general, the number of elements contained in the training set does not exceed the number of elements contained in the validation set, that is, taking the roughness parameter as an example with 100 elements, the training set has a maximum of 50 elements, and the remaining 50 elements are the elements in the validation set.

[0081] After obtaining the training set, the training set is used as input, and the simulation error target value corresponding to the elements in the training set is output. The better methods for optimizing the response surface substitution model include:

[0082] First, the Gaussian process algorithm is used to generate the initial response surface substitution model; then the training set is used as the input of the Gaussian process algorithm, and the simulation error target value corresponding to the elements in the training set is used as the output of the Gaussian process algorithm. The particle swarm optimization algorithm (PSO algorithm) is used to optimize the hyperparameters to be optimized in the Gaussian process algorithm; finally, the response surface substitution model is obtained based on the optimized hyperparameters.

[0083] Preferably, the Gaussian process algorithm can be represented by a mean function μ(m) and a covariance function k(m, m'). In the embodiment of the present invention, the mean function μ(m) is preferably selected to be 0.

[0084] A better choice is to choose the diagonal square exponential covariance function as the covariance function k(m, m'), which is expressed as: In this formula, σf and λ are the hyperparameters to be optimized in the Gaussian process algorithm.

[0085] Step 105: Based on the roughness parameter set, a validation set of the response surface substitution model is generated, and the simulation accuracy of the response substitution model is tested to meet the accuracy requirement.

[0086] After the response surface substitution model is generated, it needs to be verified. Therefore, it is necessary to generate a verification set of the response surface substitution model based on the roughness parameter set, and use the verification set to test the simulation accuracy of the response substitution model to meet the accuracy requirements.

[0087] Similarly, the target roughness parameter set can be sampled by the Latin hypercube sampling method to obtain a validation set. The so-called target roughness parameter set is a set formed by the remaining elements after removing the elements in the training set from the roughness parameter set. Using the above example: 30 elements are sampled from 100 elements of the roughness parameter set by the Latin hypercube sampling method to obtain a training set, and 70 elements remain. Among the remaining 70 elements, 50 elements are sampled by the Latin hypercube sampling method as a validation set. Naturally, it can be understood that the number of elements in the roughness parameter set, the number of elements in the training set, and the number of elements in the validation set are all determined by technical personnel in the field according to empirical values. The more elements there are, the higher the accuracy of the response surface substitution model, but the time to generate and verify the response surface substitution model is relatively long. The fewer elements there are, the lower the accuracy of the response surface substitution model, but the time to generate and verify the response surface substitution model is relatively short. After a large number of tests and studies, the inventors found that the optimal number of elements in the roughness parameter set is 100, the number of elements in the training set is 30, and the number of elements in the verification set is 50, which can not only ensure the accuracy of the response surface substitution model, but also optimize the time for generating and verifying the response surface substitution model.

[0088] For checking whether the simulation accuracy of the response substitution model meets the accuracy requirement, if the simulation accuracy of the response substitution model does not meet the accuracy requirement, the number of elements in the training set is increased based on the roughness parameter to form a new training set, and iteratively executes: the steps of optimizing the response surface and checking the simulation accuracy of the response substitution model, until the simulation accuracy of the response substitution model meets the accuracy requirement, that is, iteratively executes steps 104 to 105.

[0089] For example, if a response substitution model is generated with 30 elements in the training set, but the simulation accuracy of the response substitution model cannot meet the accuracy requirements, then the number of elements in the training set is increased to 40 based on the roughness parameter to form a new training set. Then, the 40 elements in the new training set are used as input, and the simulation error target value corresponding to the 40 elements in the new training set is output. The response surface substitution model is optimized again, and a new validation set of 50 elements of the response surface substitution model is generated. The simulation accuracy of the response substitution model is tested using the 50 elements in the new validation set to determine whether it meets the accuracy requirements. This process is repeated iteratively until the simulation accuracy of the response substitution model meets the accuracy requirements.

[0090] Step 106: Apply the response surface substitution model that meets the accuracy requirement to the roughness parameter set to obtain the optimal roughness parameter.

[0091] After obtaining the response surface substitution model that meets the accuracy requirements, the response surface substitution model that meets the accuracy requirements is finally applied to the roughness parameter set to obtain the optimal roughness parameter. A better method for obtaining the optimal roughness parameter includes: determining the minimum simulation error target value among the simulation error target values ​​corresponding to all elements in the roughness parameter set as the target simulation error target value; and determining the element corresponding to the target simulation error target value as the optimal roughness parameter.

[0092] The above-mentioned method of rapid estimation of roughness parameters can be better understood by taking the simulation of the advancement and retreat process of water flow in a water conveyance channel as an example. Assuming that the channel length is 350m, an observation section is set every 50m to measure the flow rate of the section. After the gate at the head of the channel is opened, water is released at a flow rate of 3.5L / s, the release time is 110min, and the tail of the channel is free to flow out.

[0093] According to the data of the above channel, assuming that the channel cross-section shape is trapezoidal and the slope is 0.0025, a one-dimensional hydrodynamic simulation model of the channel is constructed, and the Saint-Venant equations are expressed as follows:

[0094]

[0095] In the formula: Q is the flow rate; x is the direction of water flow; B is the channel width; Z is the water level; t is the time; A is the cross-sectional area of ​​the water flow; g is the gravitational acceleration; n is the roughness parameter; R is the hydraulic radius; V is the average flow velocity of the section; ql and ul are the lateral inflow of a single field length and the component of the lateral inflow in the x direction, respectively.

[0096] According to experience, the value range of the roughness parameter is set to (0.05, 0.5), and the roughness parameter set is generated by the Latin hypercube sampling method (the number of set elements is set to 100). It is substituted into the aforementioned one-dimensional hydrodynamic simulation model, and the flow rates at the inflow and outflow sections of the channel are used as boundary conditions. The upper boundary condition is the flow boundary of 3.5 L / s, and the lower boundary condition is the free outflow boundary. The finite volume method is used to solve the Saint-Venant equations to obtain the flow simulation data Qs of the observation section.

[0097] Obtain the water flow observation data Qo at the observation section, calculate the simulation error target value set corresponding to each element in the roughness set, and the simulation error uses the root mean square error RMSE, which is expressed as follows:

[0098]

[0099] Where N is the number of observation points; Qs and Qo are the simulated water flow observation data and water flow observation data, respectively.

[0100] The Latin hypercube sampling method is used to sample the roughness parameter set to obtain the training set of the response surface surrogate model (the number of elements in the training set is set to 30). The response surface surrogate model uses the Gaussian process algorithm, which is represented by the mean function μ(m) and the covariance function k(m, m'), and the mean function is selected as 0.

[0101] The diagonal square exponential covariance function is selected as the covariance function, and the expression is: Where σf and λ are the hyperparameters to be optimized of the Gaussian process algorithm. The training set is used as input, and the simulation error target value corresponding to the 30 elements in the training set is output. The Particle Swarm Optimization (PSO) algorithm is used to optimize the hyperparameters of the Gaussian process algorithm.

[0102] The Latin hypercube sampling method is used to sample the roughness parameter set of 70 elements remaining after removing 30 elements from the training set to obtain the validation set of the response surface substitution model (the number of elements in the validation set is set to 50). The optimized response surface substitution model is input to test the simulation accuracy of the response substitution model. If the accuracy is not met, the training sample size is increased (the number of samples increased by 5 each time), and the above method is repeated until the accuracy requirements are met.

[0103] The response surface substitution model that meets the accuracy requirements is applied to the roughness parameter set (100 elements), and the minimum simulation error target value among the simulation error target values ​​corresponding to the 100 elements in the roughness parameter set is determined as the target simulation error target value. Finally, the element corresponding to the target simulation error target value is determined as the optimal roughness parameter.

[0104] In order to verify the effectiveness of the roughness parameter fast estimation method proposed in the present invention, the time for solving the Saint-Venant equations to obtain the optimal roughness parameters in a single time on a laptop computer with an Intel Core i7 CPU is about 150 seconds. Using the method proposed in the present invention, the time for obtaining the optimal roughness parameters using the response surface substitution model is about 2 seconds, which shows that the estimation rate of the roughness parameters is greatly improved.

[0105] In the embodiment of the present invention, based on the above roughness parameter rapid estimation method, a roughness parameter rapid estimation device is also proposed. Figure 2 The block diagram of the roughness parameter fast estimation device shown in FIG. 1 comprises:

[0106] The model building module 210 requests to obtain waterway data and builds a corresponding one-dimensional hydrodynamic simulation model according to the waterway data, wherein the waterway includes: a river or a channel;

[0107] The water flow simulation data solving module 220 is used to generate a roughness parameter set and substitute it into the one-dimensional hydrodynamic simulation model, and solve the one-dimensional hydrodynamic simulation model with the flow of the inflow and outflow sections of the waterway as boundary conditions to obtain the water flow simulation data of the observation section;

[0108] The simulation error target value calculation module 230 is used to obtain the water flow observation data of the observation section, and calculate the simulation error target value corresponding to each element in the roughness parameter set in combination with the water flow simulation data to form a simulation error target value set;

[0109] A response surface substitution model generation and optimization module 240 is used to generate a training set of a response surface substitution model based on the roughness parameter set, and use the training set as input, the simulation error target value corresponding to the element in the training set as output, and optimize the response surface substitution model;

[0110] A verification module 250 is used to generate a verification set of the response surface substitution model based on the roughness parameter set, and verify the simulation accuracy of the response substitution model to meet the accuracy requirement;

[0111] The optimal roughness parameter module 260 is used to apply the response surface substitution model that meets the accuracy requirement to the roughness parameter set to obtain the optimal roughness parameter.

[0112] Optionally, the model building module 210 is specifically used for:

[0113] Acquiring dimension data of the river channel or the channel, wherein the dimension data includes: cross-sectional dimensions and slope;

[0114] Based on the cross-sectional dimensions and the slope, construct the one-dimensional hydrodynamic simulation model;

[0115] Wherein, constructing the one-dimensional hydrodynamic simulation model includes:

[0116] The constructed one-dimensional hydrodynamic simulation model is expressed using the Saint-Venant equations as follows:

[0117]

[0118] In the above two formulas, Q is the flow rate; x is the direction of water flow; B is the width of the waterway; Z is the water level; t is the time; A is the cross-sectional area of ​​the water flow; g is the gravitational acceleration; n is the roughness parameter; R is the hydraulic radius; V is the average flow velocity of the section; ql and ul are the lateral inflow of a single field length and the component of the lateral inflow in the x direction, respectively.

[0119] Optionally, the water flow simulation data solving module 220 includes:

[0120] The generation set submodule is used to generate the roughness parameter set within a preset value range by a random sampling method.

[0121] Optionally, the generating set submodule is specifically used for:

[0122] Set the value range of roughness parameters according to experience;

[0123] The roughness parameter set is generated within the value range by a Latin hypercube sampling method.

[0124] Optionally, the simulation error target value calculation module 230 is specifically used for:

[0125] Performing a root mean square error calculation on the water flow simulation data and the water flow observation data to obtain a simulation error target value corresponding to each element in the roughness parameter set to form the simulation error target value set;

[0126] The expression of the root mean square error operation is:

[0127]

[0128] In the above formula, RMSE is the root mean square error; N is the number of observation points; Qs is the water flow simulation data; Qo is the water flow observation data.

[0129] Optionally, the generating and optimizing response surface surrogate model module 240 includes:

[0130] The training set submodule is used to sample the roughness parameter set by a Latin hypercube sampling method to obtain the training set, wherein the number of elements contained in the training set does not exceed the number of elements contained in the validation set.

[0131] Optionally, the generating and optimizing response surface surrogate model module 240 is specifically used for:

[0132] The Gaussian process algorithm was used to generate the initial response surface surrogate model;

[0133] Taking the training set as the input of the Gaussian process algorithm, taking the simulation error target value corresponding to the elements in the training set as the output of the Gaussian process algorithm, and optimizing the hyperparameters to be optimized in the Gaussian process algorithm by using a particle swarm optimization algorithm;

[0134] Obtaining the response surface substitution model based on the optimized hyperparameters;

[0135] Wherein, the Gaussian process algorithm is represented by a mean function μ(m) and a covariance function k(m, m'), and the mean function μ(m) is selected to be 0;

[0136] The diagonal square exponential covariance function is selected as the covariance function k(m, m'), and its expression is: In this formula, σf and λ are the hyperparameters to be optimized in the Gaussian process algorithm.

[0137] Optionally, the inspection module 250 includes:

[0138] The validation set submodule is used to sample the target roughness parameter set by Latin hypercube sampling method to obtain the validation set. The target roughness parameter set is a set formed by the remaining elements in the roughness parameter set after removing the elements in the training set.

[0139] Optionally, the inspection module 250 is specifically used for:

[0140] If the simulation accuracy of the response substitution model does not meet the accuracy requirement, the number of elements in the training set is increased based on the roughness parameter to form a new training set, and the steps of optimizing the response surface and verifying the simulation accuracy of the response substitution model are iteratively performed until the simulation accuracy of the response substitution model meets the accuracy requirement.

[0141] Optionally, the optimal roughness parameter module 260 is specifically used for:

[0142] The roughness parameter set is used as an input of the response surface substitution model that meets the accuracy requirement, and the response surface substitution model that meets the accuracy requirement outputs a simulation error target value corresponding to each element in the roughness parameter set;

[0143] Determine the minimum simulation error target value among the simulation error target values ​​corresponding to all elements in the roughness parameter set as the target simulation error target value;

[0144] The element corresponding to the target simulation error target value is determined as the optimal roughness parameter.

[0145] In an embodiment of the present invention, based on the above-mentioned roughness parameter rapid estimation method, a computer-readable storage medium is further provided, wherein the computer program stored in the medium enables a processor to execute the roughness parameter rapid estimation method as described in any one of steps 101 to 105.

[0146] In summary, the roughness parameter rapid estimation method of the present invention first obtains waterway data, and constructs a corresponding one-dimensional hydrodynamic simulation model based on the waterway data, then generates a roughness parameter set and substitutes it into the one-dimensional hydrodynamic simulation model, and uses the flow rates of the inflow and outflow sections of the waterway as boundary conditions to solve the one-dimensional hydrodynamic simulation model and obtain the water flow simulation data of the observed section.

[0147] Then, the water flow observation data of the observation section is obtained, and combined with the water flow simulation data, the simulation error target value corresponding to each element in the roughness parameter set is calculated to form a simulation error target value set; based on the roughness parameter set, a training set of the response surface substitution model is generated, and the training set is used as input, and the simulation error target value corresponding to the element in the training set is output to optimize the response surface substitution model;

[0148] Based on the roughness parameter set, a validation set of the response surface substitution model is generated, and the simulation accuracy of the response surface substitution model is tested to meet the accuracy requirements; finally, the response surface substitution model that meets the accuracy requirements is applied to the roughness parameter set to obtain the optimal roughness parameters.

[0149] The present invention aims to solve the problems of large amount of calculation and long time consumption of the traditional roughness parameter estimation method based on the hydrodynamic model simulation. The method of rapid estimation of roughness parameters is proposed. The method trains and replaces the solution of the traditional one-dimensional hydrodynamic river model with the response surface substitution model. It only needs to solve a few one-dimensional hydrodynamic river models instead of solving the one-dimensional hydrodynamic river model for each element in the roughness parameter set. While accurately obtaining the optimal roughness parameter, the calculation amount is small and the time consumption is short, which greatly improves the estimation rate of the roughness parameter. The optimal roughness parameter can be estimated more quickly, so that it can be better applied to production work with high real-time requirements such as short-term flood rapid forecasting.

[0150] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0151] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0152] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A method for rapid estimation of roughness parameters, characterized in that: The roughness parameter rapid estimation method comprises: Acquire waterway data, and construct a corresponding one-dimensional hydrodynamic simulation model according to the waterway data, wherein the waterway includes: a river or a channel; Generate a roughness parameter set and substitute it into the one-dimensional hydrodynamic simulation model, solve the one-dimensional hydrodynamic simulation model with the flow rate of the inflow and outflow sections of the waterway as boundary conditions, and obtain water flow simulation data of the observation section; Acquire the water flow observation data of the observation section, and calculate the simulation error target value corresponding to each element in the roughness parameter set in combination with the water flow simulation data to form a simulation error target value set; Based on the roughness parameter set, a training set of a response surface substitution model is generated, and the training set is used as input, and the simulation error target value corresponding to the element in the training set is output, so as to optimize the response surface substitution model; Based on the roughness parameter set, a validation set of the response surface substitution model is generated, and the simulation accuracy of the response substitution model is tested to meet the accuracy requirement; A response surface substitution model that meets the accuracy requirement is applied to the roughness parameter set to obtain an optimal roughness parameter.

2. The roughness parameter rapid estimation method according to claim 1, characterized in that: Acquiring waterway data and constructing a corresponding one-dimensional hydrodynamic simulation model according to the waterway data, including: Acquiring dimension data of the river channel or the channel, wherein the dimension data includes: cross-sectional dimensions and slope; Based on the cross-sectional dimensions and the slope, the one-dimensional hydrodynamic simulation model is constructed.

3. The roughness parameter rapid estimation method according to claim 1, characterized in that: Generates a set of roughness parameters, including: The roughness parameter set is generated within a preset value range by a random sampling method.

4. The method for rapid estimation of roughness parameters according to claim 3, characterized in that: The roughness parameter set is generated within a preset value range by a random sampling method, including: Set the value range of roughness parameters according to experience; The roughness parameter set is generated within the value range by a Latin hypercube sampling method.

5. The method for rapid estimation of roughness parameters according to claim 1, characterized in that: Calculating the simulation error target value corresponding to each element in the roughness parameter set to form a simulation error target value set, including: A root mean square error calculation is performed on the water flow simulation data and the water flow observation data to obtain a simulation error target value corresponding to each element in the roughness parameter set to form the simulation error target value set.

6. The method for rapid estimation of roughness parameters according to claim 1, characterized in that: Based on the roughness parameter set, a training set of a response surface substitution model is generated, including: The training set is obtained by sampling in the roughness parameter set using a Latin hypercube sampling method, and the number of elements included in the training set does not exceed the number of elements included in the validation set.

7. The method for rapid estimation of roughness parameters according to claim 1, characterized in that: Taking the training set as input and the simulation error target value corresponding to the elements in the training set as output, optimizing the response surface substitution model includes: The Gaussian process algorithm was used to generate the initial response surface surrogate model; Taking the training set as the input of the Gaussian process algorithm, taking the simulation error target value corresponding to the elements in the training set as the output of the Gaussian process algorithm, and optimizing the hyperparameters to be optimized in the Gaussian process algorithm by using a particle swarm optimization algorithm; The response surface substitution model is obtained based on the optimized hyperparameters.

8. The method for rapid estimation of roughness parameters according to claim 1, characterized in that: Based on the roughness parameter set, generating a validation set of the response surface substitution model, comprising: The validation set is obtained by sampling in a target roughness parameter set using a Latin hypercube sampling method, wherein the target roughness parameter set is a set formed by remaining elements in the roughness parameter set after removing elements in the training set.

9. The method for rapid estimation of roughness parameters according to claim 1, characterized in that: Verify the simulation accuracy of the response surrogate model to meet accuracy requirements, including: If the simulation accuracy of the response substitution model does not meet the accuracy requirement, the number of elements in the training set is increased based on the roughness parameter to form a new training set, and the steps of optimizing the response surface and verifying the simulation accuracy of the response substitution model are iteratively performed until the simulation accuracy of the response substitution model meets the accuracy requirement.

10. The roughness parameter rapid estimation method according to claim 1, characterized in that: Applying a response surface substitution model that meets the accuracy requirements to the roughness parameter set to obtain the optimal roughness parameter includes: The roughness parameter set is used as an input of the response surface substitution model that meets the accuracy requirement, and the response surface substitution model that meets the accuracy requirement outputs a simulation error target value corresponding to each element in the roughness parameter set; Determine the minimum simulation error target value among the simulation error target values ​​corresponding to all elements in the roughness parameter set as the target simulation error target value; The element corresponding to the target simulation error target value is determined as the optimal roughness parameter.

11. The roughness parameter rapid estimation method according to claim 2, characterized in that: Constructing the one-dimensional hydrodynamic simulation model includes: The constructed one-dimensional hydrodynamic simulation model is expressed using the Saint-Venant equations as follows: In the above two formulas, Q is the flow rate; x is the direction of water flow; B is the width of the waterway; Z is the water level; t is the time; A is the cross-sectional area of ​​the water flow; g is the gravitational acceleration; n is the roughness parameter; R is the hydraulic radius; V is the average flow velocity of the section; ql and ul are the lateral inflow of a single field length and the component of the lateral inflow in the x direction, respectively.

12. The roughness parameter rapid estimation method according to claim 5, characterized in that: The expression of the root mean square error operation is: In the above formula, RMSE is the root mean square error; N is the number of observation points; Qs is the water flow simulation data; Qo is the water flow observation data.

13. The roughness parameter rapid estimation method according to claim 7, characterized in that: The Gaussian process algorithm is represented by a mean function μ(m) and a covariance function k(m, m'), and the mean function μ(m) is selected to be 0; The diagonal square exponential covariance function is selected as the covariance function k(m, m'), and its expression is: In this formula, σf and λ are the hyperparameters to be optimized in the Gaussian process algorithm.

14. A roughness parameter rapid estimation device, characterized in that: The roughness parameter rapid estimation device comprises: A model building module is used to request waterway data and build a corresponding one-dimensional hydrodynamic simulation model according to the waterway data. The waterway includes a river or a channel. A water flow simulation data solving module is used to generate a roughness parameter set and substitute it into the one-dimensional hydrodynamic simulation model, and solve the one-dimensional hydrodynamic simulation model with the flow of the inflow and outflow sections of the waterway as boundary conditions to obtain the water flow simulation data of the observation section; A simulation error target value calculation module is used to obtain the water flow observation data of the observation section, and calculate the simulation error target value corresponding to each element in the roughness parameter set in combination with the water flow simulation data to form a simulation error target value set; A response surface substitution model generation and optimization module is used to generate a training set of the response surface substitution model based on the roughness parameter set, and use the training set as input, the simulation error target value corresponding to the element in the training set as output, and optimize the response surface substitution model; A verification module, used to generate a verification set of the response surface substitution model based on the roughness parameter set, and verify the simulation accuracy of the response substitution model to meet the accuracy requirement; The optimal roughness parameter module is used to apply the response surface substitution model that meets the accuracy requirements to the roughness parameter set to obtain the optimal roughness parameter.

15. A computer-readable storage medium, characterized in that: The computer program stored therein enables the processor to execute the roughness parameter rapid estimation method as claimed in any one of claims 1 to 13.

Citation Information

Patent Citations

  • Giant river network hydrodynamic mathematical model calibration method and system and storage medium

    CN114818320A

  • River missing section interpolation and unknown roughness calibration method

    CN117669415A

  • Ecological flow calibration method and system based on three-dimensional hydrodynamic model

    CN118294002A

  • Method for acquiring anisotropic basin surface roughness and use thereof

    US20180038688A1

  • Hydrodynamic model parameter optimization method and apparatus, and water level and flow change process simulation method and apparatus

    WO2024098950A1

Cited By

  • Method for predicting adhesion roughness of limnoperna fortunei in water transfer project

    CN120562329A

  • A method for predicting the roughness of a zebra mussel attachment for water diversion projects

    CN120562329B

  • River channel evolution model construction method and device, electronic equipment and readable storage medium

    CN122021073A