Sediment concentration prediction method and system based on freeze-thaw coupling physical constraint model

Abstract

The invention provides a sediment concentration prediction method and system based on a freeze-thaw coupling physical constraint model, and relates to the technical field of sediment concentration prediction, and the method comprises the steps: obtaining hydrological data, meteorological data and various original data of soil and vegetation, and carrying out the preprocessing; constructing a benchmark sediment concentration physical model based on the multiple kinds of preprocessed original data, and constructing an input feature vector; inputting the input feature vector into a sediment concentration prediction ANN model, and outputting to obtain the sediment concentration; wherein in the training process of the sediment concentration prediction ANN model, reference sediment prediction and sediment mass conservation constraint of the physical model are introduced into an ANN model loss function, the loss function is enhanced by adopting a freeze-thaw weighting strategy, and the sediment concentration prediction ANN model in which the physical model and the ANN model are deeply fused is obtained. According to the invention, the stability and accuracy of prediction during extreme rainfall, runoff surge or frequent freezing and thawing are improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of sediment concentration prediction, in particular to a sediment concentration prediction method and system based on a freeze-thaw coupled physical constraint model. BACKGROUND

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art.

[0003] The spatiotemporal variation of sediment not only affects the stability of river morphology, but also directly relates to the safety of water conservancy operation, the health of water ecological system, and the risk of downstream flood disasters. Therefore, accurately predicting the variation law of sediment concentration is of great significance for water and sediment management and ecological restoration in a basin. In recent years, with global warming and frequent extreme weather events, freeze-thaw cycles in mid-high latitudes have become increasingly intense. The freeze-thaw process can significantly change the soil structure and permeability, leading to complex nonlinear changes in surface runoff and sediment production mechanisms. In the spring and early winter, repeated freeze-thaw of the ground surface causes the soil structure to be loose, and the erosion resistance decreases, so that rainfall or snowmelt can cause significant sediment surge. Traditional sediment prediction models are difficult to depict the influence of freeze-thaw erosion on sediment concentration in this process. In addition, the development of remote sensing monitoring, automatic hydrological monitoring, and soil sensing technology enables the continuous acquisition of multi-source data (such as air temperature, rainfall, flow, soil moisture, snowmelt, soil temperature, etc.), providing an accurate and comprehensive data basis for establishing a high-precision and dynamically responsive sediment concentration prediction model.

[0004] Although there are many existing sediment concentration prediction methods, the existing methods still have many shortcomings: (1) Traditional empirical models and multiple linear regression models, such as , rely on fixed parameters for fitting and are difficult to reflect the characteristics of hydrological response under freeze-thaw conditions, and have poor applicability in different seasons and years; (2) Secondly, physical models can better describe the basic transport mechanisms of water flow dynamics and sediment, but lack model parameters related to the freeze-thaw process when depicting freeze-thaw frequency and soil erosion characteristics during the freeze-thaw period, and the initial conditions are difficult to obtain, resulting in significant deviations between the model and actual observations; (3) In addition, pure data-driven artificial neural networks can use historical data to fit sediment concentration, but lack physical constraints, are prone to overfitting, have insufficient extrapolation ability for physical conservation and extreme conditions, and the prediction results may appear unreasonable negative values or do not conform to the conservation of mass. SUMMARY

[0005] The present disclosure proposes a silt concentration prediction method and system based on a freeze-thaw coupled physical constraint model to solve the above problems. The freeze-thaw physical process is explicitly modeled and integrated into the ANN model (physical model and artificial neural network). The benchmark silt prediction of the physical model and the silt mass conservation constraint are introduced into the ANN loss function. The output is constrained by the physical constraint loss function, so that the model data fitting satisfies the physical law. The freeze-thaw weighting strategy is used to enhance the loss function, and the stability and accuracy of the model in extreme rainfall, runoff surge or frequent freeze-thaw are improved.

[0006] According to some embodiments, the present disclosure adopts the technical scheme as follows: The silt concentration prediction method based on a freeze-thaw coupled physical constraint model comprises: Obtaining hydrological data, meteorological data, and various original data of soil and vegetation, and preprocessing; Constructing an input feature vector based on the preprocessed various original data; Inputting the input feature vector into the silt concentration prediction ANN model to output the silt concentration; In the training process of the silt concentration prediction ANN model, a benchmark silt concentration physical model is constructed. The benchmark silt prediction of the physical model and the silt mass conservation constraint are introduced into the ANN model loss function. The freeze-thaw weighting strategy is used to enhance the loss function, and the silt concentration prediction ANN model of the deep fusion of the physical model and the ANN model is obtained.

[0007] As an embodiment, the obtaining of the hydrological data, meteorological data, and various original data of soil and vegetation comprises: Aligning the data to a uniform time step to obtain the hydrological data, meteorological data, and various original data of soil and vegetation; The hydrological data includes runoff, runoff surge, observed silt concentration, and snowmelt. The meteorological data includes air temperature, rainfall, snowfall, and extreme rainfall. The soil and vegetation data includes soil temperature, soil volume water content, and freeze-thaw frequency.

[0008] As an embodiment, the preprocessing process comprises time series alignment, monthly extraction, outlier removal, and normalization processing.

[0009] As an embodiment, the construction of the benchmark silt concentration physical model comprises: Calculating the runoff surge based on the runoff; Constructing an indicator function, and calculating the freeze-thaw frequency based on the indicator function and the soil temperature; Constructing a rainfall indicator based on the rainfall, and calculating the soil temperature gradient based on the soil temperature; A benchmark sediment concentration physical model is constructed based on runoff, runoff surge, observed sediment concentration and snowmelt.

[0010] As an embodiment, the input feature vector is constructed based on runoff, snowmelt, air temperature, rainfall, snowfall, soil temperature, soil volume water content and calculated runoff surge, freeze-thaw frequency, calculated rainfall indicator and soil temperature gradient.

[0011] As an embodiment, the total loss function of the sediment concentration prediction ANN model includes data fitting loss, physical boundary constraint, sediment mass conservation constraint and freeze-thaw weighted constraint. Through the physical constraint loss and the sediment mass conservation constraint, the negative value or unreasonable extreme value of the prediction result of the pure data-driven model is avoided. Through the freeze-thaw weighting strategy, the response ability of the model to the freeze-thaw alternating period and the extreme runoff event is enhanced.

[0012] According to some embodiments, the present disclosure adopts the following technical solutions: The sediment concentration prediction system based on the freeze-thaw coupled physical constraint model comprises: A data acquisition module is configured to acquire hydrological data, meteorological data and various original data of soil and vegetation, and to preprocess the data. A feature construction module is configured to construct an input feature vector based on the preprocessed various original data. A sediment concentration prediction module is configured to input the input feature vector into a sediment concentration prediction ANN model and output a sediment concentration. In the training process of the sediment concentration prediction ANN model, a benchmark sediment concentration physical model is constructed, the benchmark sediment prediction of the physical model and the sediment mass conservation constraint are introduced into the loss function of the ANN model, the loss function is enhanced by using the freeze-thaw weighting strategy, and a sediment concentration prediction ANN model with deep fusion of the physical model and the ANN model is obtained.

[0013] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product comprises a computer program, which, when executed by a processor, implements the sediment concentration prediction method based on the freeze-thaw coupled physical constraint model.

[0014] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer readable storage medium is configured to store computer instructions, which, when executed by a processor, implement the sediment concentration prediction method based on the freeze-thaw coupled physical constraint model.

[0015] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device comprises a processor, a memory and a computer program; wherein the processor is connected with the memory, and the computer program is stored in the memory; when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the method for predicting sediment concentration based on a freeze-thaw coupled physical constraint model.

[0016] Compared with the prior art, the beneficial effects of the present disclosure are: The method for predicting sediment concentration based on a freeze-thaw coupled physical constraint model of the present disclosure is used for predicting the sediment concentration of a river in an alpine freeze-thaw area, so that the model prediction accuracy is high, the physical consistency is good, the uncertainty is assessable, and the extrapolation ability is good especially for extreme or transitional freeze-thaw periods under the influence of freeze-thaw processes.

[0017] The method for predicting sediment concentration based on a freeze-thaw coupled physical constraint model of the present disclosure constructs a benchmark sediment concentration physical model, explicitly models the freeze-thaw physical process and integrates it into an ANN model, improves the freeze-thaw alternating period sediment concentration prediction accuracy, introduces the benchmark sediment prediction of the physical model and the sediment mass conservation constraint into the ANN model loss function, avoids the negative value or exceeds the physical upper limit in the prediction, keeps the ANN model fitting ability for complex nonlinear relationship, and through the physical constraint loss function, the model data fitting meets the physical law; through the freeze-thaw weighting strategy, the loss function is enhanced, and the stability and accuracy of the model in the extreme rainfall, runoff surge or frequent freeze-thaw are improved.

[0018] The method for predicting sediment concentration based on a freeze-thaw coupled physical constraint model of the present disclosure deeply integrates the physical model and the artificial neural network (ANN), and introduces the freeze-thaw frequency, soil temperature gradient and other key features, which can significantly improve the prediction accuracy while ensuring the physical consistency of sediment transport. Through the physical constraint loss and the sediment mass conservation constraint, the negative value or unreasonable extreme value of the prediction result of the pure data-driven model is avoided; through the freeze-thaw weighting strategy, the response ability of the model to the freeze-thaw alternating period and the extreme runoff event is enhanced, and the robustness and generalization ability of the prediction are improved. The present disclosure can accurately fit the historical observation, meet the mass conservation and physical boundary requirements of sediment transport, and balance the explainability and engineering applicability, and can be widely applied to actual scenes such as basin sediment early warning, soil and water conservation and reservoir scheduling. BRIEF DESCRIPTION OF DRAWINGS

[0019] The drawings accompanying the specification of the present disclosure serve to provide a further understanding of the present disclosure, and the illustrative embodiments of the present disclosure and the description thereof serve to explain the present disclosure, and do not constitute an improper limitation on the present disclosure.

[0020] Figure 1A flow chart of a sediment concentration prediction method based on a freeze-thaw coupled physical constraint model is provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0021] The present disclosure is further described below with reference to the accompanying drawings and embodiments.

[0022] It should be noted that the following detailed description is illustrative only, and is intended to provide further description of the present disclosure. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present disclosure belongs.

[0023] It should be noted that the terms used herein are only intended to describe specific embodiments and are not intended to limit exemplary embodiments according to the present disclosure. As used herein, the singular form is intended to include the plural form unless the context clearly indicates otherwise, and it should also be understood that the terms "comprise" and / or "include" as used in the specification indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0024] Embodiment 1 In an embodiment of the present disclosure, a sediment concentration prediction method based on a freeze-thaw coupled physical constraint model is provided, and the method steps include: Step 1: Obtain hydrological data, meteorological data, and various original data of soil and vegetation, and preprocess; Step 2: Construct an input feature vector based on the preprocessed various original data; Step 3: Input the input feature vector into the sediment concentration prediction ANN model, and output the sediment concentration; In the training process of the sediment concentration prediction ANN model, a benchmark sediment concentration physical model is constructed, the benchmark sediment prediction of the physical model and the sediment mass conservation constraint are introduced into the loss function of the ANN model, the loss function is enhanced by using a freeze-thaw weighting strategy, and a sediment concentration prediction ANN model with deep fusion of the physical model and the ANN model is obtained.

[0025] As an embodiment, the sediment concentration prediction method based on the freeze-thaw coupled physical constraint model of the present disclosure deeply fuses the physical model and the artificial neural network (ANN), and introduces key features such as freeze-thaw frequency and soil temperature gradient, which can significantly improve the prediction accuracy while ensuring the physical consistency of sediment transport. Through the physical constraint loss and the sediment mass conservation constraint, the situation of negative values or unreasonable extreme values in the prediction results of the pure data-driven model is avoided; through the freeze-thaw weighting strategy, the response ability of the model to the freeze-thaw alternating period and extreme runoff events is enhanced, and the robustness and generalization ability of the prediction are improved. The specific implementation training process of the present disclosure is as follows: Step 1: Obtain hydrological data, meteorological data, and various original data of soil and vegetation, and preprocess; Specifically, first, the data is time-aligned to a uniform time step T, and various original data is collected, including hydrological data, meteorological data, and soil and vegetation data, as follows: Hydrological data includes: runoff (m³ / s), runoff surge (m³ / s), observed sediment concentration (kg / m³), snowmelt (mm / month); Meteorological data includes: air temperature (℃), rainfall (mm / month), snowfall (mm / month), extreme rainfall , (mm / month); Soil and vegetation data includes: soil temperature (℃), soil volume water content , freeze-thaw frequency (times).

[0026] Second, the various original data obtained is preprocessed, including time series alignment, monthly extraction, removal of outliers, and normalization, as follows: (1) Time series processing: Align the data from different sources, including runoff, sediment concentration SSC, runoff surge, snowmelt, air temperature, rainfall, snowfall, extreme rainfall, soil temperature, soil volume water content, and freeze-thaw frequency, to daily scale. Linear interpolation is used to fill in missing time points to ensure that all variables are within the same study time range.

[0027] (2) Monthly alignment: Extract daily scale data to monthly scale, with the following specific requirements: Calculate monthly average for runoff, runoff surge, sediment concentration SSC, air temperature, soil temperature, and soil volume water content. Calculate monthly cumulative for snowmelt, rainfall, snowfall, and extreme rainfall. Calculate freeze-thaw frequency by calculating temperature that meets freeze-thaw conditions from hourly soil temperature and then calculating cumulative on daily scale and monthly scale. Align monthly data to the same monthly scale.

[0028] (3) Remove outliers: Search for data with large values and verify whether they are caused by extreme events or supported by reliable data.

[0029] (4) Normalization: Before inputting data into the model, perform Z-score standardization:

[0030] wherein, X i is the minimum value of the original data, X min is the maximum value of the original data, so that the data is mapped to [0, 1). X max is the maximum value of the original data, so that the data is mapped to [0, 1).

[0031] Further, the runoff surge is calculated based on the runoff, including:

[0032] wherein, is the runoff, is the t flow at time -1.

[0033] Further, an indicator function is constructed, and the freeze-thaw frequency is calculated based on the indicator function and the soil temperature;

[0034] wherein, is the indicator function, which is 1 when the soil temperature crosses 0℃; is the soil temperature at time t, in ℃; t is the soil temperature at time -1, in ℃. N is the statistical window length.

[0035] Further, a rainfall indicator is constructed based on the rainfall, including:

[0036] wherein, is the rainfall, is a certain quantile of the historical rainfall, and in this study, the 95th quantile, i.e., P 95 .

[0037] Further, the soil temperature gradient is calculated based on the soil temperature, including:

[0038] wherein, is the soil temperature at time t, t is the soil temperature at time -1. t is the soil temperature at time -1.

[0039] Step 2: Based on the pre-processed multiple original data, a benchmark sediment concentration physical model and an input feature vector are constructed; (1) First, based on the runoff, runoff surge, observed sediment concentration, and snowmelt, a benchmark sediment concentration physical model is constructed: ​

[0040] In the formula, .

[0041] (2) Freeze-thaw correction based on the reference sediment concentration physical model, as follows:

[0042] In the formula, , φ is the freeze-thaw influence correction coefficient; k is the freeze-thaw sensitive coefficient.

[0043] The purpose of the freeze-thaw correction of the present disclosure is because SSC The prediction is related to runoff Q and rainfall P and other indicators, but in the seasonal frozen soil area or alpine basin, these mechanisms will change significantly with the freeze-thaw state, when freeze-thaw does not occur, the soil is in a frozen state, the soil erosion is small, SSC The output will decrease, when freeze-thaw occurs, the soil is in a loose state, and it is easy to be washed away, SSC The output will increase; when freeze-thaw alternately occurs, the soil structure is destroyed, and the soil erosion resistance is weakened, SSC The output increases, therefore, in the seasonal frozen soil basin or alpine basin, the same runoff and rainfall and other indicators, if the freeze-thaw frequency increases, SSC It will also increase, through correction, to compensate for the problems of traditional models and increase the physical interpretability.

[0044] (3) Construct an input feature vector; Based on the runoff, snowmelt, air temperature, rainfall, snowfall, soil temperature, soil volume water, and calculated runoff surge, freeze-thaw frequency, calculated rainfall indicator, and soil temperature gradient, an input feature vector is constructed:

[0045] Step 3: Construct a sediment concentration prediction ANN model; (1) ANN predicts sediment concentration output:

[0046] In the formula, W All combinations of weights and biases in the model.

[0047]

[0048] Wherein: is the weight matrix of the first layer; is the bias vector of the first layer; is the total number of neural network layers.

[0049] (2) Construct the loss function: 1) Data fitting loss:

[0050] In the formula, is the ANN predicted sediment concentration output, is the observed sediment concentration.

[0051] 2) Physical boundary constraints:

[0052] In the formula, The historical maximum observed value or engineering upper limit can be taken.

[0053] 3) Sediment mass conservation constraint:

[0054] 4) Freeze-thaw weighted constraint:

[0055]

[0056] In the formula, is the freeze-thaw period sample set, is the freeze-thaw sensitive weighting coefficient.

[0057] Further, the weight coefficient includes and wherein, is calculated by the validation set:

[0058] is calculated by the mass conservation deviation <5%:

[0059] In summary, the total loss function of the present disclosure is:

[0060] Step 4: Model training and prediction; (1) Use the Adam optimizer, learning rate 1e-3, sliding window training, and train the optimal weight:

[0061] (2) The prediction formula is:

[0062] (3) Residual analysis:

[0063] The present disclosure introduces the reference sediment prediction of the physical model and the sediment mass conservation constraint into the ANN model loss function, constructs an artificial neural network with 11 input nodes as the input layer, two hidden layers and one output layer to capture the nonlinear sediment transport characteristics under freeze-thaw conditions. Considering the mean square error, physical constraint and mass conservation constraint, a total loss function is constructed, and the neural network parameters and physical weight parameters are updated through the back propagation algorithm α 、 β 、 λ 1、 λ 2 are optimized synchronously to form a parameter adaptive update cycle until the model converges. A freeze-thaw weighting strategy is used to enhance the loss function, and through the iterative training process, the sediment concentration prediction is performed until the set iteration number or optimization condition is reached, and the prediction result is obtained. The result can be dynamically adjusted according to the freeze-thaw period conditions to realize the physical consistency and data-driven fusion prediction of the complex surface sediment transport process. The present disclosure uses a freeze-thaw weighting strategy to enhance the loss function, and through the iterative training process, until the set iteration number or optimization condition is reached, a trained sediment concentration prediction ANN model is obtained for sediment concentration prediction.

[0064] Embodiment 2 In an embodiment of the present disclosure, a sediment concentration prediction system based on a freeze-thaw coupled physical constraint model is provided, comprising: A data acquisition module for acquiring hydrological data, meteorological data and various original data of soil and vegetation, and preprocessing; A feature construction module for constructing a reference sediment concentration physical model based on the preprocessed various original data, and constructing an input feature vector; A sediment concentration prediction module for inputting the input feature vector into the sediment concentration prediction ANN model to output the sediment concentration; Wherein, in the training process of the sediment concentration prediction ANN model, the reference sediment prediction of the physical model and the sediment mass conservation constraint are introduced into the ANN model loss function, and a freeze-thaw weighting strategy is used to enhance the loss function to obtain a sediment concentration prediction ANN model deeply fused with the physical model and the ANN model.

[0065] Embodiment 3 In an embodiment of the present disclosure, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the sediment concentration prediction method based on the freeze-thaw coupled physical constraint model.

[0066] Embodiment 4 In an embodiment of the present disclosure, a non-transitory computer readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the method for predicting sediment concentration based on a freeze-thaw coupled physical constraint model.

[0067] Embodiment 5 In an embodiment of the present disclosure, an electronic device is provided, comprising a processor, a memory, and a computer program; wherein the processor is connected with the memory, and the computer program is stored in the memory; when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device implements the method for predicting sediment concentration based on a freeze-thaw coupled physical constraint model.

[0068] The present disclosure is described with reference to the flowcharts and / or block diagrams of the method, device (system), and computer program product according to embodiments of the present disclosure. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The means for implementing the functions specified in one or more flows and / or blocks.

[0069] These computer program instructions can also be loaded onto a computer or other programmable data processing devices, so that a series of operation steps are performed on the computer or other programmable data processing devices to generate a computer-implemented process, so that the instructions executed on the computer or other programmable data processing devices provide a means for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The steps for implementing the functions specified in one or more flows and / or blocks.

[0070] Although the specific embodiments of the present disclosure are described above with reference to the accompanying drawings, the present disclosure is not limited thereto, and various modifications or variations can be made by those skilled in the art without creative labor, which are still within the protection scope of the present disclosure.

Claims

1. A method for predicting sediment concentration based on a freeze-thaw coupled physical constraint model, characterized in that, include: Acquire and preprocess various raw data, including hydrological data, meteorological data, and soil and vegetation data; Construct input feature vectors based on various preprocessed raw data; The input feature vector is fed into the sediment concentration prediction ANN model, and the output is the sediment concentration. In the training process of the sediment concentration prediction ANN model, a benchmark sediment concentration physical model is constructed. The benchmark sediment prediction of the physical model and the sediment mass conservation constraint are introduced into the loss function of the ANN model. The loss function is enhanced by a freeze-thaw weighting strategy, resulting in a sediment concentration prediction ANN model that deeply integrates the physical model and the ANN model.

2. The sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in claim 1, characterized in that, The acquisition of various raw data, including hydrological data, meteorological data, and soil and vegetation data, includes: Data is time-aligned to a unified time step to acquire various raw data, including hydrological data, meteorological data, and soil and vegetation data. The hydrological data includes runoff, runoff surge, observed sediment concentration, and snowmelt; the meteorological data includes temperature, rainfall, snowfall, and extreme rainfall; and the soil and vegetation data includes soil temperature, soil volumetric water, and freeze-thaw frequency.

3. The sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in claim 1, characterized in that, The preprocessing process includes time-series alignment, monthly extraction, outlier removal, and normalization.

4. The sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in claim 1, characterized in that, The construction of the benchmark sediment concentration physical model includes: Calculate runoff surge based on runoff volume; Construct an indicator function, and calculate the freeze-thaw frequency based on the indicator function and soil temperature; A rainfall indicator is constructed based on rainfall data, and a soil temperature gradient is calculated based on soil temperature. A physical model of baseline sediment concentration was constructed based on runoff, runoff surge, observed sediment concentration, and snowmelt.

5. The sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in claim 4, characterized in that, An input feature vector is constructed based on runoff, snowmelt, air temperature, rainfall, snowfall, soil temperature, soil volumetric water, as well as the calculated runoff surge, freeze-thaw frequency, calculated rainfall indicator, and soil temperature gradient.

6. The sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in claim 1, characterized in that, The total loss function of the sediment concentration prediction ANN model includes data fitting loss, physical boundary constraints, sediment mass conservation constraints, and freeze-thaw weighted constraints. Through physical constraint loss and sediment mass conservation constraints, the model avoids negative values ​​or unreasonable extreme values ​​in the prediction results of the pure data-driven model. Through the freeze-thaw weighted strategy, the model's response capability to freezing-thawing transition periods and extreme runoff events is enhanced.

7. A sediment concentration prediction system based on a freeze-thaw coupled physical constraint model, specifically implementing the sediment concentration prediction method based on a freeze-thaw coupled physical constraint model as described in any one of claims 1-6, characterized in that, include: The data acquisition module is used to acquire and preprocess various raw data, including hydrological data, meteorological data, and soil and vegetation data. The feature construction module is used to construct input feature vectors based on various preprocessed raw data. The sediment concentration prediction module is used to input the input feature vector into the sediment concentration prediction ANN model and output the sediment concentration. In the training process of the sediment concentration prediction ANN model, a benchmark sediment concentration physical model is constructed. The benchmark sediment prediction of the physical model and the sediment mass conservation constraint are introduced into the loss function of the ANN model. The loss function is enhanced by a freeze-thaw weighting strategy, resulting in a sediment concentration prediction ANN model that deeply integrates the physical model and the ANN model.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in any one of claims 1-6.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in any one of claims 1-6.

10. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the sediment concentration prediction method based on the freeze-thaw coupling physical constraint model as described in any one of claims 1-6.

Images