A method and device for calibrating roughness of a hydrodynamic model and an electronic device

Abstract

This application provides a method, apparatus, and electronic device for calibrating the roughness of a hydrodynamic model. The method includes: detecting hydrodynamic element information corresponding to each initial roughness array in a preset roughness sample set; constructing a hydrodynamic proxy model based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; iteratively optimizing the preset roughness sample set, and obtaining the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set based on the hydrodynamic proxy model; selecting target roughness arrays according to the prediction accuracy represented by the hydrodynamic element prediction results, and calibrating the roughness of the preset hydrodynamic model based on the target roughness arrays. By using a hydrodynamic proxy model to replace the hydrodynamic model for simulation calculation, the hydrodynamic proxy model can determine the hydrodynamic element prediction results of the entire river channel based on a single simulation calculation, reducing the number of simulation calculations and saving time costs.

Description

Technical Field

[0001] This application relates to the field of hydrodynamic analysis technology, and in particular to a method, apparatus and electronic equipment for calibrating the roughness of a hydrodynamic model. Background Technology

[0002] Hydrodynamic models are used to reenact and predict the flow processes of rivers, lakes, reservoirs, and flood storage areas, characterizing the spatiotemporal variations of important hydraulic elements such as water level and flow rate. They are widely applied in many fields, including flood control and disaster reduction, hydrological forecasting, and hydraulic engineering design. Calibration of the roughness of a hydrodynamic model is a crucial step in its final construction; therefore, how to calibrate the roughness of a hydrodynamic model has become a key research focus.

[0003] In existing technologies, hydrodynamic models are typically used to simulate hydrodynamic elements such as water level and flow rate according to different combinations of roughness, and the corresponding simulation results are obtained. Based on the accuracy of the model as represented by the simulation results of the hydrodynamic model, the optimal roughness combination is selected from many combinations of roughness.

[0004] However, for large natural river channels, it is necessary to divide the river channel into sections and zones according to geological characteristics and give corresponding roughness coefficients. If the optimal combination of roughness coefficients is determined by repeated simulation calculations of the hydrodynamic model of the natural river channel, a massive amount of simulation calculations will be required, which will consume a lot of time. Summary of the Invention

[0005] This application provides a method, apparatus, and electronic device for calibrating the roughness of a hydrodynamic model, in order to address the shortcomings of existing technologies that require massive simulation calculations, which consume a lot of time and resources.

[0006] The first aspect of this application provides a method for calibrating the roughness of a hydrodynamic model, comprising:

[0007] Obtain a preset roughness sample set, which includes multiple sets of initial roughness arrays for the target river channel;

[0008] Detect the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set;

[0009] Based on the preset roughness sample set and the hydrodynamic element information corresponding to each set of initial roughness arrays in the preset roughness sample set, a hydrodynamic proxy model is constructed.

[0010] The preset roughness sample set is iteratively optimized, and based on the hydrodynamic proxy model, the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set are obtained.

[0011] Based on the prediction accuracy characterized by the prediction results of the hydrodynamic elements, a target roughness array is selected from the optimized roughness sample set to calibrate the roughness of the preset hydrodynamic model according to the target roughness array.

[0012] Optionally, detecting the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set includes:

[0013] Each initial roughness array in the preset roughness sample set is input into the preset hydrodynamic model to obtain the hydrodynamic element information corresponding to each initial roughness array.

[0014] The hydrodynamic information includes at least water level and flow rate.

[0015] Optionally, the step of constructing a hydrodynamic proxy model based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set includes:

[0016] Based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, a hydrodynamic proxy model training sample is generated.

[0017] Using the training samples of the hydrodynamic proxy model, an initial hydrodynamic proxy model is trained to obtain a fully trained hydrodynamic proxy model.

[0018] Optionally, the initial hydrodynamic proxy model includes several initial hydrodynamic proxy sub-models. The step of training the initial hydrodynamic proxy model using the training samples of the hydrodynamic proxy model to obtain the fully trained hydrodynamic proxy model includes:

[0019] Using the training samples of the hydrodynamic proxy model, train the several initial hydrodynamic proxy sub-models to obtain several hydrodynamic proxy sub-models that have been trained.

[0020] Based on the model accuracy of each hydrodynamic proxy sub-model, the coupling weight of each hydrodynamic proxy sub-model is determined;

[0021] Based on the coupling weights of each hydrodynamic proxy sub-model, the plurality of hydrodynamic proxy sub-models are coupled to obtain the hydrodynamic proxy model.

[0022] Optionally, the initial hydrodynamic proxy sub-models include an initial radial basis function model, an initial multinomial regression model, and an initial kriging model;

[0023] The sum of the coupling weights of the hydrodynamic surrogate sub-model is equal to 1.

[0024] Optionally, the iterative optimization of the preset roughness sample set includes:

[0025] Based on a preset genetic algorithm, the preset roughness sample set is optimized and mutated to generate a new roughness sample set.

[0026] Based on the hydrodynamic proxy model, the corresponding hydrodynamic element prediction results are determined according to the new roughness sample set.

[0027] Based on the prediction accuracy characterized by the hydrodynamic element prediction results, the optimization mutation direction is determined, and the process returns to the step of optimizing and mutating the preset roughness sample set based on the preset genetic algorithm to generate a new roughness sample set.

[0028] Optionally, the step of selecting a target roughness array from the optimized roughness sample set based on the prediction accuracy characterized by the hydrodynamic element prediction results includes:

[0029] For any set of roughness arrays, the prediction accuracy of the hydrodynamic element prediction result is determined based on the difference between the predicted hydrodynamic element result and the measured value of the hydrodynamic element of the target river channel.

[0030] The roughness array with the highest prediction accuracy is determined as the target roughness array in the optimized roughness sample set.

[0031] A second aspect of this application provides a roughness calibration device for a hydrodynamic model, comprising:

[0032] The acquisition module is used to acquire a preset roughness sample set, which includes multiple sets of initial roughness arrays for the target river channel.

[0033] The detection module is used to detect the hydrodynamic element information corresponding to each set of initial roughness arrays in the preset roughness sample set;

[0034] The proxy model construction module is used to construct a hydrodynamic proxy model based on the preset roughness sample set and the hydrodynamic element information corresponding to each set of initial roughness arrays in the preset roughness sample set.

[0035] The optimization module is used to iteratively optimize the preset roughness sample set and, based on the hydrodynamic proxy model, obtain the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set.

[0036] The calibration module is used to select a target roughness array from the optimized roughness sample set based on the prediction accuracy characterized by the prediction results of the hydrodynamic elements, so as to calibrate the roughness of the preset hydrodynamic model according to the target roughness array.

[0037] Optionally, the detection module is specifically used for:

[0038] Each initial roughness array in the preset roughness sample set is input into the preset hydrodynamic model to obtain the hydrodynamic element information corresponding to each initial roughness array.

[0039] The hydrodynamic information includes at least water level and flow rate.

[0040] Optionally, the proxy model construction module is specifically used for:

[0041] Based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, a hydrodynamic proxy model training sample is generated.

[0042] Using the training samples of the hydrodynamic proxy model, an initial hydrodynamic proxy model is trained to obtain a fully trained hydrodynamic proxy model.

[0043] Optionally, the initial hydrodynamic proxy model includes several initial hydrodynamic proxy sub-models, and the proxy model construction module is specifically used for:

[0044] Using the training samples of the hydrodynamic proxy model, train the several initial hydrodynamic proxy sub-models to obtain several hydrodynamic proxy sub-models that have been trained.

[0045] Based on the model accuracy of each hydrodynamic proxy sub-model, the coupling weight of each hydrodynamic proxy sub-model is determined;

[0046] Based on the coupling weights of each hydrodynamic proxy sub-model, the plurality of hydrodynamic proxy sub-models are coupled to obtain the hydrodynamic proxy model.

[0047] Optionally, the initial hydrodynamic proxy sub-models include an initial radial basis function model, an initial multinomial regression model, and an initial kriging model;

[0048] The sum of the coupling weights of the hydrodynamic surrogate sub-model is equal to 1.

[0049] Optionally, the optimization module is specifically used for:

[0050] Based on a preset genetic algorithm, the preset roughness sample set is optimized and mutated to generate a new roughness sample set.

[0051] Based on the hydrodynamic proxy model, the corresponding hydrodynamic element prediction results are determined according to the new roughness sample set.

[0052] Based on the prediction accuracy characterized by the hydrodynamic element prediction results, the direction of optimization mutation is determined, and the preset roughness sample set is re-optimized and mutated based on the preset genetic algorithm to generate a new roughness sample set.

[0053] Optionally, the calibration module is specifically used for:

[0054] For any set of roughness arrays, the prediction accuracy of the hydrodynamic element prediction result is determined based on the difference between the predicted hydrodynamic element result and the measured value of the hydrodynamic element of the target river channel.

[0055] The roughness array with the highest prediction accuracy is determined as the target roughness array in the optimized roughness sample set.

[0056] A third aspect of this application provides an electronic device, comprising: at least one processor and a memory;

[0057] The memory stores computer-executed instructions;

[0058] The at least one processor executes computer execution instructions stored in the memory, causing the at least one processor to perform the method described in the first aspect above and various possible designs of the first aspect.

[0059] The fourth aspect of this application provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the method described in the first aspect above and various possible designs of the first aspect.

[0060] The technical solution of this application has the following advantages:

[0061] This application provides a method, apparatus, and electronic device for calibrating the roughness of a hydrodynamic model. The method includes: acquiring a preset roughness sample set, which includes multiple initial roughness arrays for a target river channel; detecting hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; constructing a hydrodynamic proxy model based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; iteratively optimizing the preset roughness sample set, and obtaining the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set based on the hydrodynamic proxy model; and selecting target roughness arrays in the optimized roughness sample set according to the prediction accuracy represented by the hydrodynamic element prediction results, so as to calibrate the roughness of the preset hydrodynamic model based on the target roughness arrays. The method provided above, by constructing a hydrodynamic proxy model, uses the hydrodynamic proxy model to replace the hydrodynamic model for simulation calculation. The hydrodynamic proxy model can determine the hydrodynamic element prediction results of the entire river channel based on a single simulation calculation, thereby reducing the number of simulation calculations and saving the time cost of calibrating the roughness of the hydrodynamic model. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0063] Figure 1 This is a schematic diagram of the structure of the hydrodynamic model roughness calibration system based on the embodiments of this application;

[0064] Figure 2 A schematic flowchart illustrating the hydrodynamic model roughness calibration method provided in this application embodiment;

[0065] Figure 3 A comparison diagram of the iterative convergence of the hydrodynamic model roughness calibration method provided in the embodiments of this application and the prior art;

[0066] Figure 4 A schematic diagram illustrating the simulation accuracy achieved by the hydrodynamic model roughness calibration method provided in the embodiments of this application;

[0067] Figure 5 A schematic diagram of the structure of the hydrodynamic model roughness calibration device provided in the embodiments of this application;

[0068] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0069] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the present disclosure in any way, but rather to illustrate the concepts of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0071] Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. In the following descriptions of embodiments, "a plurality of" means two or more, unless otherwise explicitly defined.

[0072] In existing technologies, hydrodynamic models are typically used to simulate hydrodynamic elements such as water level and flow rate using different combinations of roughness. The simulation results are then used to determine the optimal roughness combination from among numerous options, based on the model's accuracy as indicated by these results. However, for large natural rivers, it is necessary to segment and zone the river according to geological characteristics and assign corresponding roughness coefficients. Determining the optimal roughness combination through repeated simulations using hydrodynamic models of natural rivers would require a massive amount of computation and consume a significant amount of time.

[0073] To address the aforementioned issues, the hydrodynamic model roughness calibration method, apparatus, and electronic equipment provided in this application involve: acquiring a preset roughness sample set, which includes multiple initial roughness arrays for the target river channel; detecting the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; constructing a hydrodynamic proxy model based on the preset roughness sample set and the corresponding hydrodynamic element information; iteratively optimizing the preset roughness sample set and obtaining the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set based on the hydrodynamic proxy model; and selecting target roughness arrays from the optimized roughness sample set based on the prediction accuracy represented by the hydrodynamic element prediction results, thereby calibrating the roughness of the preset hydrodynamic model according to the target roughness arrays. The method provided above, by constructing a hydrodynamic proxy model to replace the hydrodynamic model for simulation calculations, allows the hydrodynamic proxy model to determine the hydrodynamic element prediction results for the entire river channel based on a single simulation calculation, thereby reducing the number of simulation calculations and saving the time cost of calibrating the roughness of the hydrodynamic model.

[0074] The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of the present invention will now be described with reference to the accompanying drawings.

[0075] First, the structure of the hydrodynamic model roughness calibration system on which this application is based will be described:

[0076] The hydrodynamic model roughness calibration method, apparatus, and electronic equipment provided in this application are applicable to calibrating the roughness parameters of hydrodynamic models of natural river channels. Figure 1 The diagram shown is a structural schematic of the hydrodynamic model roughness calibration system based on the embodiments of this application. It mainly includes a preset hydrodynamic model and a hydrodynamic model roughness calibration device for calibrating the roughness of the preset hydrodynamic model. Specifically, the hydrodynamic model roughness calibration device first acquires a preset roughness sample set, and then determines the target roughness array of the preset hydrodynamic model through iterative optimization of the preset roughness sample set.

[0077] This application provides a method for calibrating the roughness parameters of a hydrodynamic model, used to calibrate the roughness parameters of a hydrodynamic model of a natural river channel. The execution subject of this application is an electronic device, such as a server, desktop computer, laptop computer, tablet computer, or other electronic devices that can be used to call up and adjust parameters of a preset hydrodynamic model.

[0078] like Figure 2 The diagram shown is a flowchart illustrating the hydrodynamic model roughness calibration method provided in this application embodiment. The method includes:

[0079] Step 201: Obtain a preset roughness sample set.

[0080] The preset roughness sample set includes multiple sets of initial roughness arrays for the target river channel, which is based on the natural river channel applicable to the preset hydrodynamic model.

[0081] Step 202: Detect the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set.

[0082] Specifically, for any initial roughness array in the preset roughness sample set, the hydrodynamic element information corresponding to the roughness attribute of the target river channel when set to any initial roughness array can be determined by combining other attribute information of the target river channel.

[0083] Specifically, in one embodiment, each initial roughness array in the preset roughness sample set can be input into the preset hydrodynamic model to obtain the hydrodynamic element information corresponding to each initial roughness array.

[0084] Among them, hydrodynamic element information includes at least water level and flow rate.

[0085] It should be noted that the preset hydrodynamic model can be constructed based on the topography and hydrological data of the target river channel. The expression of the preset hydrodynamic model is as follows:

[0086]

[0087]

[0088] Where x represents the river location, t represents time, Q represents the river cross-sectional flow, Z represents the river cross-sectional water level, and B... T α represents the equivalent river width, q represents the lateral inflow per unit river length, α represents the dynamic correction coefficient, A represents the cross-sectional area of ​​the river channel, g represents the gravitational acceleration, c represents the roughness value, and R represents the hydraulic radius. The preset roughness array includes the roughness values ​​of each cross-section of the river channel.

[0089] Specifically, for any set of preset roughness arrays, each roughness value in the preset roughness array can be input into the preset hydrodynamic model in sequence to obtain the corresponding hydrodynamic element information, which includes at least the river cross-sectional flow and the river cross-sectional water level.

[0090] Step 203: Based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, construct a hydrodynamic proxy model.

[0091] Specifically, a hydrodynamic proxy model is constructed based on a preset roughness sample set and the hydrodynamic element information corresponding to each set of initial roughness arrays in the preset roughness sample set. The resulting hydrodynamic proxy model can predict the corresponding hydrodynamic element information based on a set of roughness arrays, so that the hydrodynamic proxy model and the preset hydrodynamic model are equivalent.

[0092] Step 204: Iteratively optimize the preset roughness sample set, and based on the hydrodynamic proxy model, obtain the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set.

[0093] It should be noted that the simulation calculation logic of the hydrodynamic proxy model provided in this application embodiment is different from that of the preset hydrodynamic model. As shown in the above expression, the preset hydrodynamic model can only determine the hydrodynamic element information of one river section in one simulation calculation, while the hydrodynamic proxy model can determine the hydraulic element information of all river sections of the target river in one simulation calculation. That is, the simulation calculation efficiency of the hydrodynamic proxy model is much higher than that of the preset hydrodynamic model.

[0094] Specifically, by iteratively optimizing a preset roughness sample set, while continuously expanding the sample capacity of the roughness sample set, a better roughness array can be obtained. During the iterative optimization process, based on the hydrodynamic proxy model, the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set can be obtained.

[0095] Step 205: Based on the prediction accuracy represented by the prediction results of hydrodynamic elements, select a target roughness array from the optimized roughness sample set, so as to calibrate the roughness of the preset hydrodynamic model according to the target roughness array.

[0096] The prediction accuracy can be specifically evaluated based on the difference between the predicted hydrodynamic element results and the measured hydrodynamic element values.

[0097] Specifically, in one embodiment, for any set of roughness arrays corresponding to the hydrodynamic element prediction results, the prediction accuracy corresponding to the set of roughness arrays can be determined based on the difference between the predicted hydrodynamic element results and the measured values ​​of the hydrodynamic elements of the target river channel; the roughness array with the highest prediction accuracy is determined as the target roughness array in the optimized roughness sample set.

[0098] Specifically, the prediction accuracy corresponding to each roughness coefficient array can be determined according to the following formula:

[0099]

[0100] in, This represents the measured values ​​of the hydrodynamic elements of the target river channel at time t. This represents the predicted hydrodynamic elements of the target river channel at time t. This represents the average value of all measured values ​​of hydrodynamic elements.

[0101] Specifically, the value of E is [0, +∞). The closer E is to 0, the better the simulation effect of the hydrodynamic surrogate model is, and the higher the credibility of the corresponding roughness condition, that is, the higher the prediction accuracy corresponding to the roughness array. When E is much greater than 0, it means that the roughness condition is unreliable, that is, the prediction accuracy corresponding to the roughness array is low.

[0102] Based on the above embodiments, as an implementable approach, in one embodiment, a hydrodynamic proxy model is constructed based on a preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, including:

[0103] Step 2031: Based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, generate hydrodynamic proxy model training samples.

[0104] Step 2032: Use the training samples of the hydrodynamic proxy model to train the initial hydrodynamic proxy model and obtain the trained hydrodynamic proxy model.

[0105] Specifically, in one embodiment, in order to improve the proxy accuracy of the hydrodynamic proxy model, the initial hydrodynamic proxy model may include several initial hydrodynamic proxy sub-models. Using the training samples of the hydrodynamic proxy model, several initial hydrodynamic proxy sub-models are trained to obtain several hydrodynamic proxy sub-models that have completed training. Based on the model accuracy of each hydrodynamic proxy sub-model, the coupling weight of each hydrodynamic proxy sub-model is determined. Based on the coupling weight of each hydrodynamic proxy sub-model, the several hydrodynamic proxy sub-models are coupled to obtain the hydrodynamic proxy model.

[0106] Among them, several initial hydrodynamic surrogate sub-models include the initial radial basis function model, the initial multinomial regression model, and the initial kriging model; the sum of the coupling weights of the hydrodynamic surrogate sub-models is equal to 1.

[0107] It should be noted that the Radial Basis Function (RBF) model, the Multinomial Regression (PR) model, and the Kriging model are all surrogate models with strong generalization ability and fast convergence speed for complex nonlinear problems. The RBF model is a single-hidden-layer feedforward network consisting of an input layer, hidden layers, and an output layer. The hidden layers contain radial basis functions, including Gaussian functions, reflected sigmoid functions, and inverse quadratic functions. The PR model is a multinomial regression model with a series of unknowns x as its basis. The Kriging model is a regression algorithm that uses covariance functions to spatially model and predict (interpolate) stochastic processes / random fields. Therefore, coupling these three models in a weighted manner can fully leverage their respective advantages, which is beneficial for enhancing the predictive performance of complex hydrodynamic models.

[0108] For example, the formula for constructing the initial radial basis function model is as follows:

[0109]

[0110] φ(d)=d 3

[0111] Where x represents an unknown sample, that is, the roughness array of the corresponding hydrodynamic elements to be simulated; x i λ represents the i-th hydrodynamic surrogate model training sample; φ(·) represents the radial basis function; λ i ...

[0112] Accordingly, the formula for constructing the initial multinomial regression model is as follows:

[0113]

[0114] Where m represents the series terms of x; b m This represents the weighting coefficient.

[0115] Accordingly, the formula for constructing the initial Kriging model is as follows:

[0116]

[0117]

[0118] Where R represents the correlation matrix and r represents the correlation function.

[0119] Specifically, after model training to obtain several hydrodynamic proxy sub-models, the model accuracy of each hydrodynamic proxy sub-model can be determined based on the prediction accuracy determination formula provided in the above embodiments. Then, coupling weights are allocated according to the model accuracy of each hydrodynamic proxy sub-model. The hydrodynamic proxy sub-model with higher model accuracy receives a larger coupling weight to ensure the model accuracy of the coupled hydrodynamic proxy model.

[0120] The expression for the coupled hydrodynamic surrogate model is as follows:

[0121]

[0122] Where, α RBF α PR and α KR This represents the coupling weights of the hydrodynamic surrogate sub-models corresponding to the radial basis function model, the multinomial regression model, and the Kriging model, respectively.

[0123] Based on the above embodiments, as an implementable approach, in one embodiment, iterative optimization of a preset roughness sample set includes:

[0124] Step 2041: Based on a preset genetic algorithm, optimize and mutate the preset roughness sample set to generate a new roughness sample set;

[0125] Step 2042: Based on the hydrodynamic proxy model, determine the corresponding hydrodynamic element prediction results according to the new roughness sample set;

[0126] Step 2043 involves determining the direction of optimization mutation based on the prediction accuracy represented by the hydrodynamic element prediction results, and then returning to the step of optimizing and mutating the preset roughness sample set based on the preset genetic algorithm to generate a new roughness sample set. The parameters of the genetic algorithm include population size (set to 100), number of generations (set to 500), mutation rate (set to 0.4), and recombination / crossover rate (set to 0.01), and initializing the population individuals.

[0127] Specifically, a genetic algorithm can be used to perform selection, crossover, and mutation steps on a preset roughness array in a preset roughness sample set. The selection step uses roulette wheel selection, the crossover step uses a single-point crossover strategy, and the mutation step uses basic position mutation. Each update of the roughness sample set allows the hydrodynamic proxy model to predict the corresponding hydrodynamic elements based on the newly added roughness array, thereby determining the corresponding prediction accuracy. The prediction accuracy can also be considered the fitness of the roughness array. When the number of iterations, which is also the number of times the hydrodynamic proxy model is called, reaches a preset threshold, the iteration ends.

[0128] Specifically, in one embodiment, after each prediction of the corresponding hydrodynamic element based on the new roughness sample set using the hydrodynamic proxy model, the new roughness sample set is input into a preset hydrodynamic model to obtain the hydrodynamic element information corresponding to the new roughness sample set. Based on the new roughness sample set and the corresponding hydrodynamic element information, new training samples for the hydrodynamic proxy model are generated, and then the hydrodynamic proxy model is optimized based on these training samples. That is, the model accuracy of the hydrodynamic proxy model can be improved while iteratively optimizing the roughness array.

[0129] Among them, such as Figure 3 The figure shown is a comparison of the iterative convergence of the hydrodynamic model roughness calibration method provided in this application embodiment and the prior art. Figure 3 The upper curve represents the roughness calibration efficiency of the prior art, and the lower curve represents the calibration efficiency of the hydrodynamic model roughness calibration method provided in this application embodiment. The number of times the numerical model is called represents the number of times the preset hydrodynamic model is called. Figure 3 The comparison results show that the hydrodynamic model roughness calibration method provided in this application embodiment can achieve high prediction accuracy after 800-1000 calls, while the prediction accuracy achieved by existing technologies after 800-1000 calls is significantly insufficient. The hydrodynamic model roughness calibration method provided in this application embodiment shows a trend of efficient convergence, and under the same computational load, the solution accuracy is improved by 10 times compared with existing technologies. The purpose of calling the preset hydrodynamic model in this application embodiment is to generate the above-mentioned training samples, which are used to optimize the hydrodynamic proxy model while iteratively optimizing the roughness array.

[0130] Among them, such as Figure 4 The diagram illustrates the simulation accuracy achieved by the hydrodynamic model roughness calibration method provided in this embodiment. After inputting the converged roughness values ​​(target roughness array) into the preset hydrodynamic model for numerical simulation, the simulated and measured values ​​of water level changes at four typical hydrological stations (station A, station B, station C, and station D) along the target river channel are compared. The results show higher simulation accuracy. Figure 4 The fitted value is the simulated value, and the observed value is the measured value.

[0131] The hydrodynamic model roughness calibration method provided in this application involves: acquiring a preset roughness sample set, which includes multiple initial roughness arrays for the target river channel; detecting the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; constructing a hydrodynamic proxy model based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; iteratively optimizing the preset roughness sample set and obtaining the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set based on the hydrodynamic proxy model; and selecting target roughness arrays from the optimized roughness sample set according to the prediction accuracy represented by the hydrodynamic element prediction results, so as to calibrate the roughness of the preset hydrodynamic model based on the target roughness arrays. The method provided above, by constructing a hydrodynamic proxy model, uses the hydrodynamic proxy model to replace the hydrodynamic model for simulation calculations. The hydrodynamic proxy model can determine the hydrodynamic element prediction results for the entire river channel based on a single simulation calculation, thereby reducing the number of simulation calculations and saving the time cost of calibrating the roughness of the hydrodynamic model. Furthermore, by coupling multiple hydrodynamic surrogate sub-models, a hydrodynamic surrogate model is obtained, ensuring the model accuracy of the hydrodynamic surrogate model.

[0132] This application provides a hydrodynamic model roughness calibration device for performing the hydrodynamic model roughness calibration method provided in the above embodiments.

[0133] like Figure 5 The diagram shown is a structural schematic of the hydrodynamic model roughness calibration device provided in an embodiment of this application. The hydrodynamic model roughness calibration device 50 includes: an acquisition module 501, a detection module 502, a proxy model construction module 503, an optimization module 504, and a calibration module 505.

[0134] The system comprises the following modules: an acquisition module for acquiring a preset roughness sample set, which includes multiple initial roughness arrays for the target river channel; a detection module for detecting the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; a surrogate model construction module for constructing a hydrodynamic surrogate model based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; an optimization module for iteratively optimizing the preset roughness sample set and obtaining the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set based on the hydrodynamic surrogate model; and a calibration module for selecting target roughness arrays in the optimized roughness sample set according to the prediction accuracy represented by the hydrodynamic element prediction results, so as to calibrate the roughness of the preset hydrodynamic model based on the target roughness arrays.

[0135] Specifically, in one embodiment, the detection module is specifically used for:

[0136] Input each initial roughness array from the preset roughness sample set into the preset hydrodynamic model to obtain the hydrodynamic element information corresponding to each initial roughness array;

[0137] Among them, hydrodynamic information includes at least water level and flow rate.

[0138] Specifically, in one embodiment, the proxy model construction module is specifically used for:

[0139] Based on the preset roughness sample set and the hydrodynamic element information corresponding to each set of initial roughness arrays in the preset roughness sample set, a hydrodynamic proxy model training sample is generated.

[0140] Using training samples from the hydrodynamic proxy model, an initial hydrodynamic proxy model is trained, resulting in a fully trained hydrodynamic proxy model.

[0141] Specifically, in one embodiment, the initial hydrodynamic proxy model includes several initial hydrodynamic proxy sub-models, and the proxy model construction module is specifically used for:

[0142] Using training samples from the hydrodynamic proxy model, several initial hydrodynamic proxy sub-models are trained, resulting in several completed hydrodynamic proxy sub-models.

[0143] Based on the model accuracy of each hydrodynamic proxy sub-model, determine the coupling weight of each hydrodynamic proxy sub-model;

[0144] Based on the coupling weights of each hydrodynamic proxy sub-model, several hydrodynamic proxy sub-models are coupled to obtain a hydrodynamic proxy model.

[0145] Specifically, in one embodiment, the initial hydrodynamic surrogate sub-models include an initial radial basis function model, an initial multinomial regression model, and an initial kriging model;

[0146] The sum of the coupling weights of the hydrodynamic surrogate sub-model is equal to 1.

[0147] Specifically, in one embodiment, the optimization module is specifically used for:

[0148] Based on a pre-defined genetic algorithm, a pre-defined roughness sample set is optimized and mutated to generate a new roughness sample set.

[0149] Based on the hydrodynamic proxy model, the corresponding hydrodynamic element prediction results are determined according to the new roughness sample set.

[0150] Based on the prediction accuracy represented by the hydrodynamic element prediction results, the direction of optimization mutation is determined, and the preset roughness sample set is optimized and mutated again based on the preset genetic algorithm to generate a new roughness sample set.

[0151] Specifically, in one embodiment, the calibration module is specifically used for:

[0152] For any set of roughness arrays, the prediction accuracy of the hydrodynamic elements is determined based on the difference between the predicted hydrodynamic elements and the measured values ​​of the hydrodynamic elements of the target river channel.

[0153] The roughness array with the highest prediction accuracy is determined as the target roughness array in the optimized roughness sample set.

[0154] Regarding the hydrodynamic model roughness calibration device in this embodiment, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated here.

[0155] The hydrodynamic model roughness calibration device provided in this application embodiment is used to execute the hydrodynamic model roughness calibration method provided in the above embodiment. Its implementation method and principle are the same, and will not be described again.

[0156] This application provides an electronic device for executing the hydrodynamic model roughness calibration method provided in the above embodiments.

[0157] like Figure 6 The diagram shown is a structural schematic of an electronic device provided in an embodiment of this application. The electronic device 60 includes at least one processor 61 and a memory 62.

[0158] The memory stores computer-executable instructions; at least one processor executes the computer-executable instructions stored in the memory, causing the at least one processor to execute the hydrodynamic model roughness calibration method provided in the above embodiment.

[0159] This application provides an electronic device for executing the hydrodynamic model roughness calibration method provided in the above embodiments. Its implementation method and principle are the same, and will not be described again.

[0160] This application provides a computer-readable storage medium storing computer-executable instructions. When a processor executes the computer-executable instructions, it implements the hydrodynamic model roughness calibration method provided in any of the above embodiments.

[0161] The storage medium containing computer-executable instructions in this embodiment can be used to store the computer-executable instructions for the hydrodynamic model roughness calibration method provided in the foregoing embodiments. Its implementation method and principle are the same, and will not be described again.

[0162] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0163] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0164] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in a combination of hardware and software functional units.

[0165] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0166] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional modules is merely an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0167] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for calibrating the roughness of a hydrodynamic model, characterized in that, include: Obtain a preset roughness sample set, which includes multiple sets of initial roughness arrays for the target river channel, and the target river channel is a natural river channel applicable to the preset hydrodynamic model. Detect the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; Based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, a hydrodynamic proxy model is constructed; wherein, the hydrodynamic proxy model is a model obtained by weighted coupling of radial basis function model, multinomial regression model and Kriging model; The preset roughness sample set is iteratively optimized, and based on the hydrodynamic proxy model, the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set are obtained. Based on the prediction accuracy characterized by the prediction results of the hydrodynamic elements, a target roughness array is selected from the optimized roughness sample set to calibrate the roughness of the preset hydrodynamic model according to the target roughness array. The step of selecting a target roughness array from the optimized roughness sample set based on the prediction accuracy characterized by the hydrodynamic element prediction results includes: For any set of roughness arrays, the prediction accuracy of the hydrodynamic element prediction result is determined based on the difference between the predicted hydrodynamic element result and the measured value of the hydrodynamic element of the target river channel. The roughness array with the highest prediction accuracy is determined as the target roughness array in the optimized roughness sample set. The prediction accuracy for each roughness array is determined based on the following formula: in, This represents the measured values ​​of the hydrodynamic elements of the target river channel at time t. This represents the predicted hydrodynamic elements of the target river channel at time t. This represents the average value of all measured values ​​of hydrodynamic elements; The process of constructing a hydrodynamic proxy model based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set includes: Based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, a hydrodynamic proxy model training sample is generated. Using the training samples of the hydrodynamic proxy model, an initial hydrodynamic proxy model is trained to obtain a fully trained hydrodynamic proxy model. The initial hydrodynamic proxy model includes several initial hydrodynamic proxy sub-models. The process of training the initial hydrodynamic proxy model using the training samples of the hydrodynamic proxy model to obtain the trained hydrodynamic proxy model includes: Using the training samples of the hydrodynamic proxy model, train the several initial hydrodynamic proxy sub-models to obtain several hydrodynamic proxy sub-models that have been trained. Based on the model accuracy of each hydrodynamic proxy sub-model, the coupling weight of each hydrodynamic proxy sub-model is determined; Based on the coupling weights of each hydrodynamic proxy sub-model, the plurality of hydrodynamic proxy sub-models are coupled to obtain the hydrodynamic proxy model.

2. The method according to claim 1, characterized in that, The detection of hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set includes: Each initial roughness array in the preset roughness sample set is input into the preset hydrodynamic model to obtain the hydrodynamic element information corresponding to each initial roughness array. The hydrodynamic information includes at least water level and flow rate.

3. The method according to claim 1, characterized in that, The initial hydrodynamic proxy sub-models include the initial radial basis function model, the initial multinomial regression model, and the initial kriging model; The sum of the coupling weights of the hydrodynamic surrogate sub-model is equal to 1.

4. The method according to claim 1, characterized in that, The iterative optimization of the preset roughness sample set includes: Based on a preset genetic algorithm, the preset roughness sample set is optimized and mutated to generate a new roughness sample set. Based on the hydrodynamic proxy model, the corresponding hydrodynamic element prediction results are determined according to the new roughness sample set. Based on the prediction accuracy characterized by the hydrodynamic element prediction results, the optimization mutation direction is determined, and the process returns to the step of optimizing and mutating the preset roughness sample set based on the preset genetic algorithm to generate a new roughness sample set.

5. A roughness calibration device for a hydrodynamic model, characterized in that, include: The acquisition module is used to acquire a preset roughness sample set, which includes multiple sets of initial roughness arrays for the target river channel, and the target river channel is a natural river channel applicable to the preset hydrodynamic model. The detection module is used to detect the hydrodynamic element information corresponding to each set of initial roughness arrays in the preset roughness sample set; The surrogate model construction module is used to construct a hydrodynamic surrogate model based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set; wherein, the hydrodynamic surrogate model is a model obtained by weighted coupling of radial basis function model, multinomial regression model and Kriging model; The optimization module is used to iteratively optimize the preset roughness sample set and, based on the hydrodynamic proxy model, obtain the hydrodynamic element prediction results corresponding to each roughness array in the optimized roughness sample set. The calibration module is used to select a target roughness array from the optimized roughness sample set based on the prediction accuracy characterized by the prediction results of the hydrodynamic elements, so as to calibrate the roughness of the preset hydrodynamic model according to the target roughness array. Specifically, the calibration module is used for: For any set of roughness arrays, the prediction accuracy of the hydrodynamic element prediction result is determined based on the difference between the predicted hydrodynamic element result and the measured value of the hydrodynamic element of the target river channel. The roughness array with the highest prediction accuracy is determined as the target roughness array in the optimized roughness sample set. The prediction accuracy for each roughness array is determined based on the following formula: in, This represents the measured values ​​of the hydrodynamic elements of the target river channel at time t. This represents the predicted hydrodynamic elements of the target river channel at time t. This represents the average value of all measured values ​​of hydrodynamic elements; The proxy model construction module is specifically used for: Based on the preset roughness sample set and the hydrodynamic element information corresponding to each initial roughness array in the preset roughness sample set, a hydrodynamic proxy model training sample is generated. Using the training samples of the hydrodynamic proxy model, an initial hydrodynamic proxy model is trained to obtain a fully trained hydrodynamic proxy model. The initial hydrodynamic proxy model includes several initial hydrodynamic proxy sub-models. The proxy model construction module is specifically used for: Using the training samples of the hydrodynamic proxy model, train the several initial hydrodynamic proxy sub-models to obtain several hydrodynamic proxy sub-models that have been trained. Based on the model accuracy of each hydrodynamic proxy sub-model, the coupling weight of each hydrodynamic proxy sub-model is determined; Based on the coupling weights of each hydrodynamic proxy sub-model, the plurality of hydrodynamic proxy sub-models are coupled to obtain the hydrodynamic proxy model.

6. An electronic device, characterized in that, include: At least one processor and memory; The memory stores computer-executed instructions; The at least one processor executes computer execution instructions stored in the memory, causing the at least one processor to perform the method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, implement the method as described in any one of claims 1 to 4.

Images