A water level measurement system based on water environment simulation

Abstract

The invention belongs to the technical field of water level measurement, and specifically refers to a water level measurement system based on water environment simulation, comprising a data integration module, a data transmission module, a water body simulation module and a water level display module; the system sets a network topology structure, uses a variety of water body sensors to collect water body data, pre-processes the water body data, ensures data quality, and provides a data basis for water level measurement; establishes a Saint-Venant equation model to simulate water body flow, combines a spectral density function to simulate water body surface waves, establishes a water level measurement model, extracts water level characteristic information, obtains water level measurement results, compares and verifies with actual water levels according to evaluation indicators, and improves the accuracy of model prediction; the system can reflect the actual water body conditions, facilitates users to obtain water level measurement results and related water body data, and provides an efficient and accurate solution for water level measurement.

Description

Technical Field

[0001] The invention belongs to the technical field of water level measurement, and specifically refers to a water level measurement system based on water environment simulation. Background Art

[0002] Water level is an important indicator reflecting the amount of water resources. By continuously measuring the water level, we can understand the dynamic changes of water resources, including the increase and decrease trends and seasonal changes of water resources. This is crucial for the scientific management and planning of water resources and is conducive to the rational planning of water resources development and utilization. However, the current water level measurement has the following shortcomings:

[0003] (1) Only focusing on a single type of data, traditional measurement methods may only measure water level height, while ignoring other related data such as water flow temperature, flow rate, and pressure. Temperature affects water density and thus affects water level. Flow rate and pressure are related to the flow state and energy of the water body and have an important impact on water level changes. There is a lack of effective processing methods for the collected water data. Abnormal values ​​generated by sensor failures and inaccurate data caused by other external interference factors are difficult to effectively identify, which is not conducive to subsequent water level analysis;

[0004] (2) There is a lack of physical models that describe the flow characteristics of water bodies under different conditions. Traditional water level measurement methods do not take into account key factors such as the continuity of water flow and momentum changes, resulting in the inability to fully and accurately simulate water flow conditions, thereby affecting the accuracy of water level measurement; there are deficiencies in simulating water surface waves. It is impossible to accurately simulate water surface waves and consider the influence of waves in different directions. The influence of water surface waves is ignored, making the model inconsistent with the actual water conditions, affecting the accuracy of water level measurement; the water level measurement process uses a fixed grid division method, and it is impossible to use grids of different precisions in different areas according to the water flow velocity and water resistance. If the grid is not fine enough in areas with large water flow changes, it will not be able to capture the detailed changes in the water flow. If the grid is too fine in areas with stable water flow, the workload of water level measurement will increase and resources will be wasted, affecting the calculation efficiency and simulation accuracy;

[0005] (3) Traditional measurement methods are unable to automatically learn characteristic patterns in water body data, lack in-depth data analysis and feature extraction, are unable to improve water level measurement capabilities by adjusting parameters, and are difficult to quantitatively evaluate the ability to extract water level characteristic information and the accuracy of prediction. Summary of the invention

[0006] In view of the above situation, the present invention sets up a network topology structure, utilizes a variety of water body sensors to collect water body data, pre-processes the water body data, ensures data quality, and provides a data basis for water level measurement; establishes a Saint-Venant equation model to simulate water flow, combines the spectral density function to simulate water surface waves, establishes a water level measurement model, extracts water level characteristic information, obtains water level measurement results, and compares and verifies with the actual water level based on evaluation indicators to improve the accuracy of model prediction; the system can reflect the actual water level, facilitates users to obtain water level measurement results and related water body data, and provides an efficient and accurate solution for water level measurement.

[0007] A water level measurement system based on water environment simulation, comprising a data integration module, a data transmission module, a water body simulation module and a water level display module;

[0008] The data integration module is connected to the data transmission module, the data transmission module is connected to the water body simulation module and the water level display module, and the water body simulation module is connected to the water level display module.

[0009] The data integration module collects and fuses water body data to obtain a simulation signal of the water body data;

[0010] The data transmission module converts the analog signal of the water body data into a digital signal, and transmits the digital signal to the water body simulation module;

[0011] The water body simulation module establishes a water level measurement model, extracts water level characteristic information, and generates water level measurement results based on the water level characteristic information;

[0012] The water level display module displays water body data, water level characteristic information and water level measurement results.

[0013] The data integration module collects and fuses water body data to obtain a simulated signal of the water body data, specifically including the following steps:

[0014] Step G1: Set the network topology, select the measurement location as the network node, set the network node address, and connect each network node for direct communication;

[0015] Step G2: According to the network topology, a coordinator, a router and a water sensor are set up, the water sensor is used to collect water data, and the water data is sent to the network topology, the coordinator is used to manage the network nodes, the water data collected by the water sensor is received and processed, and the router is used for signal relay;

[0016] Step G3: Set a threshold range for water body data, remove abnormal values ​​of water body data that exceed the threshold range, use the average value of adjacent data to fill in the missing values ​​of water body data, and obtain a simulated signal of water body data.

[0017] Furthermore, in step G2, water body data is collected using a water body sensor, which specifically includes the following steps:

[0018] Step G21: using a temperature sensor to measure the water temperature and obtain water temperature data;

[0019] Step G22: using a flow velocity sensor to detect the induced electromotive force generated by the water flow, determining the water flow velocity according to the induced electromotive force, and obtaining water flow velocity data;

[0020] Step G23: Using a pressure sensor to measure water pressure at different depths to obtain water pressure data;

[0021] Step G24: Use a total station to measure the geometric parameters of the river channel, including the riverbed slope and cross-sectional shape, calculate the cross-sectional area and roughness coefficient of the water flow, reflect the resistance of the river channel wall to the water flow, and obtain the water resistance data.

[0022] The data transmission module converts the analog signal of the water body data into a digital signal and transmits the digital signal to the water body simulation module, specifically comprising the following steps:

[0023] Step T1: calibrate the analog signal of the water body data, and use an analog-to-digital converter to convert the analog signal of the water body data into a digital signal;

[0024] Step T2: Encapsulate the digital signal of the water body data into a data packet, use the water body sensor to send the data packet to the network topology, and forward the data packet to the coordinator through the router;

[0025] Step T3: The coordinator extracts the data packet and sends it to the water body simulation module through the serial port.

[0026] The water body simulation module establishes a water level measurement model, extracts water level characteristic information, and generates water level measurement results according to the water level characteristic information, specifically including the following steps:

[0027] Step S1: Establish the Saint-Venant equation model to simulate the water flow conditions of rivers and lakes. The formula used is as follows:

[0028] Continuity equation: ;

[0029] Momentum equation: ;

[0030] Among them, A represents the cross-sectional area of ​​the water flow, t represents time, Q represents flow rate, x represents the distance along the river channel, h represents water level, g represents gravity acceleration, and C represents the Xie Cai coefficient. represents the partial derivative operation;

[0031] Step S2: The Saint-Venant equation model is discretized using the finite difference method, and the river channel is divided into calculation units in space. The length of each calculation unit is , the time is divided into time steps, each time step is t;

[0032] For the continuity equation, the central difference format method is used for discretization, and the formula used is as follows:

[0033] ;

[0034] Among them, j represents the number of the computational unit in space, and n represents the number of the time step;

[0035] For the momentum equation, the upwind format method is used for discretization;

[0036] Step S3: using the spectral density function to simulate the initial water surface waves, and using the directional diffusion function to calculate the water surface waves in different directions;

[0037] Step S4: according to the initial water surface wave, boundary conditions and initial conditions are set, the boundary conditions include inlet boundary conditions, outlet boundary conditions and wall boundary conditions, and the initial conditions include initial water level setting and initial flow velocity setting;

[0038] Inlet boundary conditions: determine the calculation unit of the water inflow position, water velocity data and water pressure data;

[0039] Outlet boundary conditions: calculation unit that determines the water outflow location, water velocity data, and water resistance data;

[0040] Wall boundary conditions: For the contact position between the water body and the solid wall, the no-slip boundary condition is adopted, assuming that the velocity of the water body on the wall is zero;

[0041] Step S5: Adopting adaptive meshing technology, in each calculation unit, based on water velocity data and water resistance data, the river channel is meshed and refined;

[0042] Use fine meshes at river bends and near sluice gates, and coarse meshes in areas with smooth water flow;

[0043] Step S6: Establish a convolutional neural network as a water level measurement model, input water body data into the water level measurement model, and obtain water level measurement results.

[0044] Furthermore, step S3 specifically includes the following steps:

[0045] Step S31: Use the spectral density function to simulate the initial water surface wave, create a two-dimensional initial wave state, and obtain the nominal wave amplitude. The formula used is as follows:

[0046] ;

[0047] in, represents the vertical displacement on the horizontal axis of the water surface, Y represents the horizontal axis of the water surface, t represents time, i represents the parameter point of the water surface wave, represents the nominal wave amplitude, represents the cosine function, T represents the matrix transpose operation, represents the wave vector, X represents the horizontal axis position of the water surface wave, represents the angular frequency, represents the displacement of water surface waves, represents the angular frequency of the i-th water surface wave parameter point, represents the displacement of the i-th water surface wave parameter point;

[0048] Step S32: propagate the initial wave state forward, use the directional wave spectrum, simulate the wavelength and relative speed of the water surface wave, and use the following formula:

[0049] ;

[0050] in, represents the change in angular frequency of water surface waves, represents the displacement change of the water surface wave, represents a directional spectrum;

[0051] Step S33: In the time step, the nominal wave amplitude is linked to the directional wave spectrum to describe the relationship between the vertical displacement of the water surface wave and the directional wave spectrum in each calculation unit. The formula used is as follows:

[0052] ;

[0053] Step S34: define expansion factors in different directions, use the directional diffusion function combined with the expansion factors, calculate the influence of the expansion factors in different directions on the water surface wave, and obtain the calculation result of the water surface wave. The formula used is as follows:

[0054] ;

[0055] in, represents the directional diffusion function, s represents the parameter related to the angular frequency, is the gamma function, represents the directional characteristics of the water surface wave, p represents the expansion factor of the water surface wave, represents the directional characteristics of the i-th water surface wave parameter point, Represents the directional characteristics of the expansion factor of the water surface wave.

[0056] Furthermore, step S6 specifically includes the following steps:

[0057] Step S61: Establishing a convolutional neural network: Establishing a convolutional neural network as a water level measurement model, the convolutional neural network includes an input layer, a convolution layer, a feature extraction layer, a pooling layer and an output layer;

[0058] Step S62: Divide the training set and the validation set: Divide the training set and the validation set, input the water pressure data, the water flow velocity data, and the water temperature data into the convolutional neural network as different channels, set the parameters of the convolutional neural network, the parameters include weights and biases, use the training set for forward propagation, and obtain the water level measurement results on each computing unit;

[0059] Step S63: Determine the loss function: use the loss function to calculate the loss value, calculate the gradient of each layer of the convolutional neural network through the back propagation algorithm, and update the parameters of the convolutional neural network;

[0060] Step S64: Perform iterative training: set the number of training rounds and control parameters, perform iterative training, and repeat steps S362 to S363;

[0061] Step S65: Water level measurement: Use the validation set to evaluate the water level measurement model, calculate the evaluation index, adjust the control parameters according to the evaluation index, and obtain the final water level measurement result.

[0062] Further, in step S62, performing forward propagation specifically includes the following steps:

[0063] Step S621: using convolution kernels of different sizes to perform multi-scale feature extraction, extracting water level feature information of water body data, and obtaining a feature map;

[0064] Step S622: Use the SIREN periodic activation function to extract the periodic changes of the water body data, perform weighted summation on each calculation unit and add the bias, and finally obtain the activated feature map through the activation function;

[0065] Step S623: Introduce the attention mechanism in the pooling layer, downsample the activated feature map, calculate the attention weight of each computing unit, perform weighted pooling on the water level feature information of different network nodes, and give high attention weight to fine grids.

[0066] Further, step S63 specifically includes the following steps:

[0067] Step S631: using the mean square error loss function to provide constraints on the water level measurement results of each computing unit;

[0068] Step S632: Use the Adam optimization algorithm to adaptively adjust the learning rate update parameter. The formula used is as follows:

[0069] ;

[0070] in, Indicates the number of iterations of training. Indicates The parameter value at the time of training iteration, Indicates The parameter value at the time of training iteration, represents the learning rate, represents the first-order moment estimate of the gradient, represents the second-order moment estimate of the gradient, Represents a small constant.

[0071] The beneficial effects achieved by the present invention are as follows:

[0072] (1) The system collects a variety of water body data, providing an information basis for subsequent accurate water level measurement. Temperature data affects water body density, which in turn affects water level. Flow rate and pressure data are directly related to the flow state and energy of the water body, and have an important impact on water level changes. By integrating these different types of water body data, we can have a more comprehensive understanding of the actual situation of the water body, thereby improving the accuracy and reliability of water level measurement; outliers are removed and missing values ​​are filled in the collected data, so that the water body data can be better used for model training; by setting the network topology structure, the data can be transmitted stably and efficiently to avoid data loss or delay;

[0073] (2) The system establishes the Saint-Venant equation model to simulate the water flow conditions of rivers and lakes, which can accurately describe the flow characteristics of water under different conditions, including changes in continuity and momentum, and convert continuous physical processes into discrete mathematical models for easy solution and simulation; uses spectral density functions to simulate initial water surface waves, and calculates water surface waves in different directions through directional diffusion functions, while setting boundary conditions and initial conditions, which can make the model more consistent with actual water conditions; uses adaptive meshing technology to improve the calculation efficiency of the model while ensuring simulation accuracy, making water simulation more efficient and accurate;

[0074] (3) The system establishes a convolutional neural network as a water level measurement model, which can automatically learn the characteristic patterns in water body data and perform in-depth analysis and feature extraction on the input water body data. Secondly, by dividing the training set and the validation set, the network is trained and optimized using the loss function and the back propagation algorithm, and the parameters can be continuously adjusted to improve the prediction ability of the model. The convolution kernels of different sizes are used for multi-scale feature extraction, the neurons of the periodic activation function are used to extract periodic changes, and the innovative method of introducing the attention mechanism in the pooling layer is used to further improve the model's ability to extract water level feature information and the accuracy of the predicted water level measurement results;

[0075] (4) The system can intuitively display water body data, water level characteristic information and water level measurement results, so that users can obtain relevant information without the need for complex data processing and analysis. Water conservancy project managers can understand the water level situation in real time through the water level display module, make reasonable decisions in a timely manner, and adjust the opening of the river sluice. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 This is an architecture diagram of a water level measurement system based on water environment simulation proposed by the present invention;

[0077] Figure 2 This is a flow chart of step S6 proposed by the present invention. DETAILED DESCRIPTION

[0078] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0079] Example 1: Reference Figure 1 , a water level measurement system based on water environment simulation, including a data integration module, a data transmission module, a water body simulation module and a water level display module;

[0080] The data integration module is connected to the data transmission module, the data transmission module is connected to the water body simulation module and the water level display module, and the water body simulation module is connected to the water level display module.

[0081] Embodiment 2: This embodiment is based on the above embodiment, with reference to Figure 2 The data integration module collects and fuses water body data to obtain a simulation signal of the water body data, which specifically includes the following steps:

[0082] Step G1: Set the network topology, select the measurement location as the network node, set the network node address, and connect each network node for direct communication;

[0083] Step G2: According to the network topology, a coordinator, a router and a water sensor are set up, the water sensor is used to collect water data, and the water data is sent to the network topology, the coordinator is used to manage the network nodes, the water data collected by the water sensor is received and processed, and the router is used for signal relay;

[0084] Step G3: Set a threshold range for water body data, remove abnormal values ​​of water body data that exceed the threshold range, use the average value of adjacent data to fill in the missing values ​​of water body data, and obtain a simulated signal of water body data.

[0085] Embodiment 3: This embodiment is based on the above embodiment. In step G2, water body data is collected using a water body sensor, which specifically includes the following steps:

[0086] Step G21: using a temperature sensor to measure the water temperature and obtain water temperature data;

[0087] Step G22: using a flow velocity sensor to detect the induced electromotive force generated by the water flow, determining the water flow velocity according to the induced electromotive force, and obtaining water flow velocity data;

[0088] Step G23: Using a pressure sensor to measure water pressure at different depths to obtain water pressure data;

[0089] Step G24: Use a total station to measure the geometric parameters of the river channel, including the riverbed slope and cross-sectional shape, calculate the cross-sectional area and roughness coefficient of the water flow, reflect the resistance of the river channel wall to the water flow, and obtain the water resistance data.

[0090] Embodiment 4: This embodiment is based on the above embodiment. The data transmission module converts the analog signal of the water body data into a digital signal and transmits the digital signal to the water body simulation module. Specifically, the following steps are included:

[0091] Step T1: calibrate the analog signal of the water body data, and use an analog-to-digital converter to convert the analog signal of the water body data into a digital signal;

[0092] Step T2: Encapsulate the digital signal of the water body data into a data packet, use the water body sensor to send the data packet to the network topology, and forward the data packet to the coordinator through the router;

[0093] Step T3: The coordinator extracts the data packet and sends it to the water body simulation module through the serial port.

[0094] Embodiment 5: This embodiment is based on the above embodiment. The water body simulation module establishes a water level measurement model, extracts water level characteristic information, and generates a water level measurement result according to the water level characteristic information. Specifically, the following steps are included:

[0095] Step S1: Establish the Saint-Venant equation model to simulate the water flow conditions of rivers and lakes. The formula used is as follows:

[0096] Continuity equation: ;

[0097] Momentum equation: ;

[0098] Among them, A represents the cross-sectional area of ​​the water flow, t represents time, Q represents flow rate, x represents the distance along the river channel, h represents water level, g represents gravity acceleration, and C represents the Xie Cai coefficient. represents the partial derivative operation;

[0099] Step S2: The Saint-Venant equation model is discretized using the finite difference method, and the river channel is divided into calculation units in space. The length of each calculation unit is , the time is divided into time steps, each time step is t;

[0100] For the continuity equation, the central difference format method is used for discretization, and the formula used is as follows:

[0101] ;

[0102] Among them, j represents the number of the computational unit in space, and n represents the number of the time step;

[0103] For the momentum equation, the upwind format method is used for discretization;

[0104] Step S3: using the spectral density function to simulate the initial water surface waves, and using the directional diffusion function to calculate the water surface waves in different directions;

[0105] Step S4: according to the initial water surface wave, boundary conditions and initial conditions are set, the boundary conditions include inlet boundary conditions, outlet boundary conditions and wall boundary conditions, and the initial conditions include initial water level setting and initial flow velocity setting;

[0106] Step S5: Adopting adaptive meshing technology, in each calculation unit, based on water velocity data and water resistance data, the river channel is meshed and refined;

[0107] Step S6: Establish a convolutional neural network as a water level measurement model, input water body data into the water level measurement model, and obtain water level measurement results.

[0108] Embodiment 6: This embodiment is based on the above embodiment, and step S3 specifically includes the following steps:

[0109] Step S31: Use the spectral density function to simulate the initial water surface wave, create a two-dimensional initial wave state, and obtain the nominal wave amplitude. The formula used is as follows:

[0110] ;

[0111] in, represents the vertical displacement on the horizontal axis of the water surface, Y represents the horizontal axis of the water surface, t represents time, i represents the parameter point of the water surface wave, represents the nominal wave amplitude, represents the cosine function, T represents the matrix transpose operation, represents the wave vector, X represents the horizontal axis position of the water surface wave, represents the angular frequency, represents the displacement of water surface waves, represents the angular frequency of the i-th water surface wave parameter point, represents the displacement of the i-th water surface wave parameter point;

[0112] Step S32: propagate the initial wave state forward, use the directional wave spectrum, simulate the wavelength and relative speed of the water surface wave, and use the following formula:

[0113] ;

[0114] in, represents the change in angular frequency of water surface waves, represents the displacement change of the water surface wave, represents a directional spectrum;

[0115] Step S33: In the time step, the nominal wave amplitude is linked to the directional wave spectrum to describe the relationship between the vertical displacement of the water surface wave and the directional wave spectrum in each calculation unit. The formula used is as follows:

[0116] ;

[0117] Step S34: define expansion factors in different directions, use the directional diffusion function combined with the expansion factors, calculate the influence of the expansion factors in different directions on the water surface wave, and obtain the calculation result of the water surface wave. The formula used is as follows:

[0118] ;

[0119] in, represents the directional diffusion function, s represents the parameter related to the angular frequency, is the gamma function, represents the directional characteristics of the water surface wave, p represents the expansion factor of the water surface wave, represents the directional characteristics of the i-th water surface wave parameter point, Represents the directional characteristics of the expansion factor of the water surface wave.

[0120] Embodiment 7: This embodiment is based on the above embodiment, and step S6 specifically includes the following steps:

[0121] Step S61: Establishing a convolutional neural network: Establishing a convolutional neural network as a water level measurement model, the convolutional neural network includes an input layer, a convolution layer, a feature extraction layer, a pooling layer and an output layer;

[0122] Step S62: Divide the training set and the validation set: Divide the training set and the validation set, input the water pressure data, the water flow velocity data, and the water temperature data into the convolutional neural network as different channels, set the parameters of the convolutional neural network, the parameters include weights and biases, use the training set for forward propagation, and obtain the water level measurement results on each computing unit;

[0123] Step S63: Determine the loss function: use the loss function to calculate the loss value, calculate the gradient of each layer of the convolutional neural network through the back propagation algorithm, and update the parameters of the convolutional neural network;

[0124] Step S64: Perform iterative training: set the number of training rounds and control parameters, perform iterative training, and repeat steps S362 to S363;

[0125] Step S65: Water level measurement: Use the validation set to evaluate the water level measurement model, calculate the evaluation index, adjust the control parameters according to the evaluation index, and obtain the final water level measurement result.

[0126] Embodiment 8: This embodiment is based on the above embodiment. In step S62, forward propagation specifically includes the following steps:

[0127] Step S621: using convolution kernels of different sizes to perform multi-scale feature extraction, extracting water level feature information of water body data, and obtaining a feature map;

[0128] In this embodiment, 3×3 and 5×5 convolution kernels are used simultaneously. The 3×3 convolution kernel captures the changes in water body data in a short period of time, and the 5×5 convolution kernel captures the trend of water body data changes on a long time scale.

[0129] Step S622: Use the SIREN periodic activation function to extract the periodic changes of the water body data, perform weighted summation on each calculation unit and add the bias, and finally obtain the activated feature map through the activation function;

[0130] Step S623: Introduce the attention mechanism in the pooling layer, downsample the activated feature map, calculate the attention weight of each computing unit, perform weighted pooling on the water level feature information of different network nodes, and give high attention weight to fine grids.

[0131] Embodiment 9: This embodiment is based on the above embodiment, and step S63 specifically includes the following steps:

[0132] Step S631: using the mean square error loss function to provide constraints on the water level measurement results of each computing unit;

[0133] Step S632: Use the Adam optimization algorithm to adaptively adjust the learning rate update parameter. The formula used is as follows:

[0134] ;

[0135] in, Indicates the number of iterations of training. Indicates The parameter value at the time of training iteration, Indicates The parameter value at the time of training iteration, represents the learning rate, represents the first-order moment estimate of the gradient, represents the second-order moment estimate of the gradient, Represents a small constant.

[0136] The present invention and its implementation methods are described above. Such description is not restrictive. If a person skilled in the art is inspired by it and creatively designs structures and implementation methods similar to the technical solution without departing from the purpose of the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A water level measurement system based on water environment simulation, characterized in that: It includes data integration module, data transmission module, water body simulation module and water level display module; The data integration module collects and fuses water body data to obtain a simulation signal of the water body data; The data transmission module converts the analog signal of the water body data into a digital signal, and transmits the digital signal to the water body simulation module; The water body simulation module establishes a water level measurement model, extracts water level characteristic information, and generates water level measurement results based on the water level characteristic information; The water level display module displays water body data, water level characteristic information and water level measurement results; The data integration module collects and fuses water body data to obtain a simulation signal of the water body data, specifically including the following steps: Step G1: Set the network topology, select the measurement location as the network node, set the network node address, and connect each network node for direct communication; Step G2: According to the network topology, a coordinator, a router and a water sensor are set up, and the water sensor is used to collect water data; Step G3: setting a threshold range for water body data, removing abnormal values ​​of water body data that exceed the threshold range, using the average value of adjacent data to fill in the missing values ​​of water body data, and obtaining a simulated signal of water body data; The water body simulation module establishes a water level measurement model, extracts water level characteristic information, and generates water level measurement results according to the water level characteristic information, specifically including the following steps: Step S1: Establish a Saint-Venant equation model to simulate the water flow conditions of rivers and lakes; Step S2: using the finite difference method to discretize the Saint-Venant equation model, dividing the river channel into calculation units in space and dividing the time steps in time; Step S3: using the spectral density function to simulate the initial water surface waves, and using the directional diffusion function to calculate the water surface waves in different directions; Step S4: according to the initial water surface wave, boundary conditions and initial conditions are set, the boundary conditions include inlet boundary conditions, outlet boundary conditions and wall boundary conditions, and the initial conditions include initial water level setting and initial flow velocity setting; Step S5: Adopting adaptive meshing technology, in each calculation unit, based on water velocity data and water resistance data, the river channel is meshed and refined; Step S6: Establish a convolutional neural network as a water level measurement model, input water body data into the water level measurement model, and obtain a water level measurement result; Step S3 specifically includes the following steps: Step S31: using the spectral density function to simulate the initial water surface wave, create a two-dimensional initial wave state, and obtain the nominal wave amplitude; Step S32: propagate the initial wave state forward, using a directional wave spectrum to simulate the wavelength and relative speed of the water surface wave; Step S33: in the time step, the nominal wave amplitude is linked to the directional wave spectrum to describe the relationship between the vertical displacement of the water surface wave and the directional wave spectrum in each calculation unit; Step S34: define expansion factors in different directions, use the directional diffusion function in combination with the expansion factors, calculate the influence of the expansion factors in different directions on the water surface wave, and obtain the calculation result of the water surface wave.

2. A water level measurement system based on water environment simulation according to claim 1, characterized in that: In step G2, water body data is collected using a water body sensor, which specifically includes the following steps: Step G21: using a temperature sensor to measure the water temperature and obtain water temperature data; Step G22: using a flow velocity sensor to detect the induced electromotive force generated by the water flow, determining the water flow velocity according to the induced electromotive force, and obtaining water flow velocity data; Step G23: Using a pressure sensor to measure water pressure at different depths to obtain water pressure data; Step G24: Use a total station to measure the geometric parameters of the river channel, calculate the water-passing cross-sectional area and roughness coefficient, and obtain water resistance data.

3. A water level measurement system based on water environment simulation according to claim 2, characterized in that: The data transmission module converts the analog signal of the water body data into a digital signal and transmits the digital signal to the water body simulation module, specifically comprising the following steps: Step T1: calibrate the analog signal of the water body data, and use an analog-to-digital converter to convert the analog signal of the water body data into a digital signal; Step T2: Encapsulate the digital signal of the water body data into a data packet, use the water body sensor to send the data packet to the network topology, and forward the data packet to the coordinator through the router; Step T3: The coordinator extracts the data packet and sends it to the water body simulation module through the serial port.

4. The water level measurement system based on water environment simulation according to claim 3 is characterized in that: Step S6 specifically includes the following steps: Step S61: Establishing a convolutional neural network: Establishing a convolutional neural network as a water level measurement model, the convolutional neural network includes an input layer, a convolution layer, a feature extraction layer, a pooling layer and an output layer; Step S62: Divide the training set and the validation set: Divide the training set and the validation set, input the water pressure data, water flow velocity data, and water temperature data into the water level measurement model as different channels, set the parameters of the convolutional neural network, including weights and biases, use the training set for forward propagation, and obtain the water level measurement results on each computing unit; Step S63: Determine the loss function: use the loss function to calculate the loss value, calculate the gradient of each layer of the convolutional neural network through the back propagation algorithm, and update the parameters of the convolutional neural network; Step S64: Perform iterative training: set the number of training rounds and control parameters, perform iterative training, and repeat steps S362 to S363; Step S65: Water level measurement: Use the validation set to evaluate the water level measurement model, calculate the evaluation index, adjust the control parameters according to the evaluation index, and obtain the final water level measurement result.

5. The water level measurement system based on water environment simulation according to claim 4 is characterized in that: In step S62, forward propagation specifically includes the following steps: Step S621: using convolution kernels of different sizes to perform multi-scale feature extraction, extracting water level feature information of water body data, and obtaining a feature map; Step S622: Use the SIREN periodic activation function to extract the periodic changes of the water body data, perform weighted summation on each calculation unit and add the bias, and finally obtain the activated feature map through the activation function; Step S623: Introduce the attention mechanism in the pooling layer, downsample the activated feature map, calculate the attention weight of each computing unit, and perform weighted pooling on the water level feature information of different network nodes.

Images