Reservoir dispatching method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a reservoir dispatching method and device, electronic equipment and a storage medium. The reservoir dispatching method comprises the following steps: collecting reservoir area information of a target reservoir and section data of a connected river channel for the target reservoir; determining the number of divided sections of the connected river channel according to the section data; constructing a physical information neural network; determining a neural network output function for representing nitrate concentration through the physical information neural network based on the reservoir area information; constructing a reservoir one-dimensional nitrogen element transport model based on the reservoir area information, the division segment number and the neural network output function; constructing a residual function through the reservoir one-dimensional nitrogen element transport model; and determining a target weight, adjusting the residual function based on the target weight, and generating a scheduling scheme for the target reservoir, thereby improving the effectiveness and accuracy of scheduling for the target reservoir.

Description

Technical Field

[0001] The present invention relates to the technical field of reservoir dispatching, and in particular to a reservoir dispatching method, a reservoir dispatching device, an electronic device and a computer-readable storage medium. Background Art

[0002] Large reservoirs are the infrastructure and important means to achieve efficient use of water resources. Industrial and agricultural production and urban activities in the basin will discharge high-concentration nitrogen-containing wastewater, which increases the nitrogen content of the reservoir and poses a potential threat to water quality and environmental safety. The nitrogen element in the reservoir has the characteristics of non-uniform distribution, time variation, multi-source input and biological interaction, making it difficult to accurately predict its transport process. At the same time, due to the limited data monitoring points of the reservoir, it is difficult to obtain accurate simulation results, which brings difficulties to reservoir scheduling. Summary of the invention

[0003] The embodiments of the present invention provide a reservoir scheduling method, device, electronic device and computer-readable storage medium to overcome the above problems or at least partially solve the above problems.

[0004] The embodiment of the present invention discloses a reservoir scheduling method, which may include:

[0005] Collecting reservoir area information of the target reservoir and cross-sectional data of the river connecting the target reservoir;

[0006] Determining the number of sections for the connected river channel through the cross-section data;

[0007] Construct physical information neural network;

[0008] Determining a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information;

[0009] Constructing a one-dimensional nitrogen transport model for the reservoir based on the reservoir area information, the number of segment divisions and the neural network output function;

[0010] constructing a residual function through the one-dimensional nitrogen transport model of the reservoir;

[0011] A target weight is determined, and the residual function is adjusted based on the target weight, and a scheduling plan for the target reservoir is generated.

[0012] Optionally, the reservoir area information may at least include: flow parameters, diffusion coefficient, water flow direction coordinate parameters of the target reservoir, temperature, pH value, nitrogen content information and form information.

[0013] Optionally, the step of determining a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information may include:

[0014] Determining initial weights and detection time of an activation function for the physical information neural network;

[0015] The initial weight, the detection time, and the water flow direction coordinate parameters are input into Formula 1 to determine the neural network output function. Formula 1 is:

[0016] U(x, t)≈U s (x, t; θ)

[0017] Wherein, x is the water flow direction coordinate parameter, t is the detection time, θ is the initial weight; U is the nitrate concentration, U s is the predicted value of nitrate concentration by the physical information neural network.

[0018] Optionally, the step of constructing a one-dimensional nitrogen transport model for a reservoir based on the reservoir area information, the number of division segments and the neural network output function may include:

[0019] Determining the adsorption and desorption rates of the water body based on the temperature, the pH value, the content information and the form information of the nitrogen element;

[0020] The flow parameters, the nitrate concentration, the water body adsorption and desorption rates, the water flow direction coordinate parameters, and the diffusion coefficient are input into Formula 2 to construct a two-dimensional nitrate transport model for a reservoir. Formula 2 is:

[0021]

[0022] Where i is the number of equation groups, the number of equation groups is greater than or equal to the number of segment divisions and greater than 0, R i is the water adsorption and desorption rate of the i-th group of equations, D i is the diffusion coefficient of the i-th set of equations, and v is the flow parameter.

[0023] Optionally, the step of constructing a residual function through the one-dimensional nitrogen transport model of the reservoir may include:

[0024] The predicted value, the detection time, the water flow direction coordinate parameter, the initial weight, and the water adsorption and desorption rate of the i-th group of equations are input into Formula 3 to construct a residual function. Formula 3 is:

[0025]

[0026] Wherein, e is the residual function.

[0027] Optionally, the step of determining the target weight may include:

[0028] Determine the number of iterations and the learning rate, and input the number of iterations and the learning rate into Formula 4 to determine the target weight, which is:

[0029]

[0030] Wherein, n is the number of iterations, and α is the learning rate.

[0031] Optionally, the physical information neural network may include two hidden layers.

[0032] The embodiment of the present invention further discloses a reservoir dispatching device, which may include:

[0033] A cross-sectional data collection module, used to collect reservoir information of a target reservoir and cross-sectional data of a river channel connected to the target reservoir;

[0034] A module for determining the number of divided sections, used to determine the number of divided sections for the connected river channel through the cross-section data;

[0035] Physical information neural network building module, used to build physical information neural network;

[0036] A neural network output function determination module, used to determine a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information;

[0037] A reservoir one-dimensional nitrogen transport model construction module, used to construct a reservoir one-dimensional nitrogen transport model based on the reservoir area information, the number of division segments and the neural network output function;

[0038] A residual function construction module, used for constructing a residual function through the one-dimensional nitrogen transport model of the reservoir;

[0039] The scheduling scheme generating module is used to determine the target weight, adjust the residual function based on the target weight, and generate a scheduling scheme for the target reservoir.

[0040] The embodiment of the present invention further discloses an electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

[0041] The memory is used to store computer programs;

[0042] The processor is used to implement the method described in the embodiment of the present invention when executing the program stored in the memory.

[0043] The embodiment of the present invention further discloses a computer-readable storage medium having instructions stored thereon, which, when executed by one or more processors, enables the processors to execute the method described in the embodiment of the present invention.

[0044] The embodiments of the present invention include the following advantages:

[0045] The embodiment of the present invention collects reservoir information of a target reservoir and cross-sectional data of a connected river channel of the target reservoir; determines the number of divided sections for the connected river channel through the cross-sectional data; constructs a physical information neural network; determines a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir information; constructs a one-dimensional nitrogen transport model for the reservoir based on the reservoir information, the number of divided sections and the neural network output function; constructs a residual function through the one-dimensional nitrogen transport model for the reservoir; determines a target weight, adjusts the residual function based on the target weight, and generates a scheduling plan for the target reservoir, thereby improving the effectiveness and accuracy of scheduling for the target reservoir. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 is a flow chart of steps of a reservoir dispatching method provided in an embodiment of the present invention;

[0047] Figure 2 is a schematic diagram of a reservoir nitrogen transport process provided in an embodiment of the present invention;

[0048] Figure 3 is a schematic diagram of a topological structure of a distributed river section provided in an embodiment of the present invention;

[0049] Figure 4 is a structural block diagram of a reservoir dispatching device provided in an embodiment of the present invention;

[0050] Figure 5 is a hardware structure block diagram of an electronic device provided in an embodiment of the present invention;

[0051] Figure 6 is a schematic diagram of a computer-readable medium provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0053] In order to enable those skilled in the art to better understand the embodiments of the present invention, the technical terms involved in the embodiments of the present invention are explained below.

[0054] PINNs: Physics-Informed Neural Networks, also known as physical information neural networks, is a new machine learning method that combines physical laws with neural networks. It aims to address the limitations of traditional data-based machine learning methods in physical modeling.

[0055] In practical applications, simulating the nitrogen transport process is a very important link in reservoir scheduling. Through simulation, we can have a deeper understanding of the dynamic change law of nitrogen in the reservoir, so as to provide a decision-making basis for the scientific scheduling of the reservoir. By simulating the nitrogen transport process of the target reservoir, we can predict the nitrogen concentration at the reservoir outlet under different scheduling schemes, and provide a scientific basis for ensuring the safety of downstream water bodies. When the simulation results show that the nitrogen concentration may exceed the standard, measures can be taken in advance to avoid the occurrence of water pollution incidents. At present, the high-concentration nitrogen-containing wastewater discharged in the basin significantly increases the nitrogen content of the reservoir, posing a potential threat to water quality safety. The distribution of nitrogen in the reservoir is non-uniform, and is affected by time changes, multi-source input and complex biological effects. Its transportation process is difficult to accurately predict. In addition, due to the limited number of reservoir monitoring stations, Ogi also faces challenges in obtaining accurate simulation data. The embodiment of the present invention provides a reservoir scheduling method, which combines physical information neural network and reservoir one-dimensional nitrogen transport model to schedule the target reservoir, and realizes the simulation of nitrogen transport process to improve the accuracy and effectiveness of reservoir scheduling.

[0056] Reference Figure 1 , shows a flow chart of the steps of a reservoir scheduling method provided in an embodiment of the present invention, which may specifically include the following steps:

[0057] Step 101, collecting reservoir information of a target reservoir and cross-sectional data of a river channel connected to the target reservoir;

[0058] Step 102, determining the number of sections for the connected river channel through the cross-section data;

[0059] Step 103, constructing a physical information neural network;

[0060] Step 104, determining a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information;

[0061] Step 105, constructing a one-dimensional nitrogen transport model for the reservoir based on the reservoir area information, the number of divided segments and the neural network output function;

[0062] Step 106, constructing a residual function through the one-dimensional nitrogen transport model of the reservoir;

[0063] Step 107, determining a target weight, adjusting the residual function based on the target weight, and generating a scheduling plan for the target reservoir.

[0064] In practical applications, the physical information neural network in the embodiment of the present invention may be PINNs. PINNs can supplement information through physical laws when the amount of data is insufficient to improve the generalization ability of the model. It can also use existing observation data to train the neural network to ensure that the model can accurately fit the real data. By minimizing the loss function, the neural network can both satisfy the physical laws and fit the data. Guided by the physical laws, PINN can accelerate the convergence speed of the model.

[0065] In a specific implementation, the embodiment of the present invention can collect the reservoir information of the target reservoir and the cross-sectional data of the connected river channel of the target reservoir; determine the number of division sections for the connected river channel through the cross-sectional data; construct a physical information neural network; determine the neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir information; construct a one-dimensional nitrogen transport model for the reservoir based on the reservoir information, the number of division sections and the neural network output function; construct a residual function through the one-dimensional nitrogen transport model for the reservoir; determine the target weight, and adjust the residual function based on the target weight, and generate a scheduling plan for the target reservoir. For example, the target reservoir can be a reservoir that needs to be scheduled. As a research object, it can be obtained. Take the reservoir information for the target reservoir, for example, the historical data of the monitoring stations set up on the main and tributary rivers of the reservoir area, and the cross-sectional information of the connected river channel, such as the river connected to the target reservoir, including the cross-sectional information of the inflow river channel and the outflow river channel as the cross-sectional data. For example, the reservoir information may include the physical characteristics of the reservoir such as shape, capacity, water depth, sediment type, and environmental information such as land use and climate conditions around the reservoir. The cross-sectional data may include a river cross-sectional diagram, including geometric information such as cross-sectional shape, width, and water depth. Preferably, in order to facilitate the subsequent establishment of a mathematical model, the connected river channel can be divided into several sections based on the cross-sectional data according to certain rules, and the number of segments is used as the number of divided sections, providing a basis for the subsequent establishment of a one-dimensional model. The determination of the number of divided sections can comprehensively consider factors such as river channel length, hydrological characteristics, and pollution source distribution. This embodiment of the present invention does not impose any restrictions.

[0066] Then, a physical information neural network PINNs is constructed. Through the physical information neural network, the neural network output function is determined based on the reservoir information. The neural network output function can be used to predict the nitrate concentration in the target reservoir. PINNs can integrate the known physical laws of nitrate migration and transformation in water bodies (such as diffusion, convection, etc.) into the model to make the prediction results more realistic. The nitrate concentration in the reservoir is affected by many factors, and the relationships between these factors are complex. PINNs can learn complex relationships. By constructing the PINNs model, the reservoir information is used as input and the nitrate concentration is used as output. The PINNs model can embed known physical laws (such as convection-diffusion equations) into neural networks, thereby improving the prediction accuracy of the model. Next, a one-dimensional nitrogen transport model for the reservoir can be constructed based on the reservoir information, the number of segments, and the output function of the neural network. The one-dimensional nitrogen transport model for the reservoir can be a mathematical model used to describe the transport process of nitrogen along the reservoir and the connected river channel. It can be based on a one-dimensional partial differential equation. The one-dimensional nitrogen transport model for the reservoir can describe the migration and transformation process of nitrogen in the reservoir and the connected river channel. Through the one-dimensional nitrogen transport model for the reservoir, a residual function is constructed. The residual function can be used for the difference between the model prediction value and the actual observation value. Furthermore, the model parameters can be adjusted by minimizing the weighted residual function so that the model The prediction results of the model are more consistent with the actual observation values, which is a function for measuring the degree of fit of the model. At the same time, the target weights are determined. Preferably, a multi-objective optimization problem can be constructed by the weights of different objectives (such as water quality improvement and water volume scheduling). The residual function is adjusted based on the target weights to generate a scheduling plan for the target reservoir. After establishing a one-dimensional nitrogen transport model for the reservoir and calculating the error between the model prediction value and the actual observation value, the importance of the error can be weighted according to the preset target, and the weighted residual function can be adjusted to optimize the model, and finally a scheduling plan that can meet the target reservoir scheduling is generated. For example, according to the transport process of nitrogen, the reservoir scheduling plan is determined, for example, lowering the water level operation, and / or, layered scheduling, and / or, ecological scheduling, etc.

[0067] The embodiment of the present invention collects reservoir information of a target reservoir and cross-sectional data of a connected river channel of the target reservoir; determines the number of divided sections for the connected river channel through the cross-sectional data; constructs a physical information neural network; determines a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir information; constructs a one-dimensional nitrogen transport model for the reservoir based on the reservoir information, the number of divided sections and the neural network output function; constructs a residual function through the one-dimensional nitrogen transport model for the reservoir; determines a target weight, adjusts the residual function based on the target weight, and generates a scheduling plan for the target reservoir, thereby improving the effectiveness and accuracy of scheduling for the target reservoir.

[0068] Based on the above embodiment, a variant embodiment of the above embodiment is proposed. It should be noted that in order to make the description concise, only the differences from the above embodiment are described in the variant embodiment.

[0069] In an optional embodiment of the present invention, the reservoir area information includes at least: flow parameters, diffusion coefficient, water flow direction coordinate parameters, temperature, pH value, nitrogen content information and morphological information of the target reservoir.

[0070] In practical applications, when constructing a reservoir nitrogen transport model and conducting scheduling, it is very important to obtain comprehensive and accurate reservoir information, which can provide the necessary input data for the model and more accurately simulate the migration and transformation process of nitrogen in the reservoir, thus providing a basis for formulating scientific and reasonable scheduling plans.

[0071] In a specific implementation, the embodiment of the present invention can make the reservoir area information at least include: flow parameters, diffusion coefficient, water flow direction coordinate parameters of the target reservoir, temperature, pH value, nitrogen content information and morphological information. For example, flow parameters can refer to a set of parameters that describe the flow characteristics of the water body. These parameters directly affect the transport and transformation process of substances in the reservoir, especially nitrogen, an element that is easy to migrate and transform. The diffusion coefficient can be a physical quantity that characterizes the distance that a substance molecule migrates in a unit time due to the concentration gradient. In a water environment, the size of the diffusion coefficient reflects the speed of diffusion of a substance in a water body. For the transport of nitrogen in a reservoir, the diffusion coefficient directly affects the mixing degree of nitrogen in the water body, and then affects its bioavailability, conversion rate and distribution in the reservoir. The water flow direction coordinate parameters can be parameters indicating the flow direction and speed of water in the reservoir, such as flow rate, flow direction, etc. These parameters are crucial for simulating the impact of water flow on nitrogen transport. Temperature can be water temperature information, which affects the solubility of nitrogen, the activity of microorganisms, and the biogeochemical processes of nitrogen, such as nitrification and denitrification. pH can be the pH (pondus hydrogenii, hydrogen ion concentration index) value of the water body, which affects the form of nitrogen (such as ammonia nitrogen, nitrate nitrogen, etc.) and the biogeochemical processes of nitrogen. The content information of nitrogen can refer to the concentration of various forms of nitrogen in the water body, such as the content information of ammonia nitrogen, nitrate nitrogen, organic nitrogen, etc. The form information can refer to the form of nitrogen in the water body, such as ammonia nitrogen, nitrate nitrogen, organic nitrogen, etc. Different forms of nitrogen have different chemical properties and biological activities. Parameters such as temperature and pH reflect the physical and chemical characteristics of the reservoir, which will directly affect the form and conversion rate of nitrogen. The content information and form information of nitrogen reflect the nitrogen pollution status of the reservoir and are important indicators for evaluating water quality.

[0072] The embodiment of the present invention realizes comprehensive collection of reservoir information by making the reservoir area information include at least: flow parameters, diffusion coefficient, flow parameters of the target reservoir, water flow direction coordinate parameters, temperature, pH value, nitrogen content information and form information, thereby providing important data support for subsequent reservoir scheduling plans and further improving the accuracy and scientificity of reservoir scheduling.

[0073] In an optional embodiment of the present invention, the step of determining a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information comprises:

[0074] Determining initial weights and detection time of an activation function for the physical information neural network;

[0075] The initial weight, the detection time, and the water flow direction coordinate parameters are input into Formula 1 to determine the neural network output function. Formula 1 is:

[0076] U(x, t)≈U s (x, t, θ)

[0077] Wherein, x is the water flow direction coordinate parameter, t is the detection time, θ is the initial weight; U is the nitrate concentration, U s is the predicted value of nitrate concentration by the physical information neural network.

[0078] In practical applications, activation function is the key to nonlinear transformation in neural networks and can determine the expressive power of neural networks. In PINNs, the choice of activation function will directly affect the model's fitting effect on physical laws. The initial weight is the starting point of neural network training and can affect the model's convergence speed and final performance. The detection time can refer to the moment when the model is evaluated during the training process.

[0079] In a specific implementation, the embodiment of the present invention can determine the initial weight and detection time of the activation function for the physical information neural network; the initial weight, detection time, and water flow direction coordinate parameter are input into Formula 1 to determine the neural network output function. For example, x can be the water flow direction coordinate parameter, t is the detection time, and θ is the initial weight; U is the nitrate concentration, and Us is the predicted value of the physical information neural network for the nitrate concentration. The activation function can determine the nonlinear expression ability of the neural network. Different activation functions (such as ReLU, Rectified Linear Unit, corrected linear unit, such as Sigmoid function, S-type activation function) have different effects on the fitting ability and convergence speed of the model. Reasonable activation function, initial weight and detection time can improve the prediction accuracy of the model for nitrate concentration, accelerate the training process of the model, and reduce the training time. Through the detection time, overfitting can be discovered in time and measures can be taken. U(x, t) can represent the actual nitrate concentration and is the target output of the model. U s (x, t; θ) can be the nitrate concentration predicted by the neural network, θ as the initial weight, can be the parameter of the neural network, and can include weights and biases, x can be the coordinate parameter of the water flow direction, as the input of the neural network, t can be the detection time, as the input of the neural network. Through the above formula 1, the neural network can be linked to the actual problem, and a mathematical model for predicting nitrate concentration is established. The physical information neural network is applied to the prediction of nitrate concentration in the reservoir. By using the powerful fitting ability of the neural network and combining physical information such as water flow direction and time, a model that can accurately predict nitrate concentration is established. Furthermore, the predicted value U can be calculated. s The error between the true value U is used to evaluate the performance of the model.

[0080] In the embodiment of the present invention, by determining the initial weight and detection time of the activation function for the physical information neural network;

[0081] The initial weight, the detection time, and the water flow direction coordinate parameters are input into Formula 1 to determine the neural network output function. Formula 1 is:

[0082] U(x, t)≈U s (x, t; θ)

[0083] Wherein, x is the water flow direction coordinate parameter, t is the detection time, θ is the initial weight; U is the nitrate concentration, U s The predicted value of nitrate concentration by the physical information neural network is used to establish a model for accurately predicting nitrate concentration, provide support for subsequent data processing, and further improve the accuracy of determining the scheduling plan for the target reservoir.

[0084] In an optional embodiment of the present invention, the step of constructing a one-dimensional nitrogen transport model for a reservoir based on the reservoir area information, the number of division segments and the neural network output function comprises:

[0085] Determining the adsorption and desorption rates of the water body based on the temperature, the pH value, the content information and the form information of the nitrogen element;

[0086] The flow parameters, the nitrate concentration, the water body adsorption and desorption rates, the water flow direction coordinate parameters, and the diffusion coefficient are input into Formula 2 to construct a two-dimensional nitrate transport model for a reservoir. Formula 2 is:

[0087]

[0088] Where i is the number of equation groups, the number of equation groups is greater than or equal to the number of segment divisions and greater than 0, R i is the water adsorption and desorption rate of the i-th group of equations, D i is the diffusion coefficient of the i-th set of equations, and v is the flow parameter.

[0089] In practical applications, the adsorption and desorption process of nitrogen in water is closely related to the environmental conditions of the water. By analyzing factors such as temperature, pH, nitrogen content and form, the adsorption and desorption capacity of water for nitrogen can be more accurately quantified, thereby gaining a deeper understanding of the behavior of nitrogen in aquatic ecosystems.

[0090] In a specific implementation, the embodiment of the present invention can determine the water body adsorption and desorption rate based on temperature, pH value, nitrogen content information and form information; input the flow parameter, nitrate concentration, water body adsorption and desorption rate, water flow direction coordinate parameter, and diffusion coefficient into Formula 2 to construct a two-dimensional nitrate transport model for a reservoir. Exemplarily, i can be the number of equation groups, the number of equation groups is greater than or equal to the number of division segments and greater than 0, Ri can be the water body adsorption and desorption rate of the i-th group of equations, Di can be the diffusion coefficient of the i-th group of equations, and v can be the flow parameter, wherein Formula 2 can be a partial differential equation, which describes the rate of change of nitrate concentration U in time t and space x, It can be the rate of change of nitrate concentration U over time t, indicating the speed of change of nitrate concentration, can be a convection term, which indicates the advection transport of nitrate along the x direction due to water flow. v is a flow parameter, The concentration gradient in the x direction, Can be a diffusion term, representing the diffusion of nitrate in the x direction due to concentration difference. Di can be the diffusion coefficient, is the second-order derivative of concentration in the x direction, R i U can be a biochemical reaction term, which represents the biochemical reaction process of nitrate in water, such as nitrification, denitrification, etc. i It is the rate of adsorption and desorption of water bodies, which is related to factors such as temperature, pH, and nitrate form. Considering that the change of nitrate concentration is affected by multiple factors such as convection, diffusion, and biochemical reactions, the convection term describes the transport effect of water flow on nitrate, the diffusion term describes the diffusion process of nitrate due to uneven concentration, and the biochemical reaction term describes the transformation process of nitrate in water bodies. By constructing a two-dimensional nitrate transport model, the migration and transformation process of nitrate in the reservoir can be quantitatively described, providing a scientific basis for reservoir water quality management. The model construction process involves the discretization of equations, the setting of boundary conditions, the determination of parameters, and the solution of the model. The accuracy of the model depends on the structure of the model, the selection of parameters, and the quality of the input data.

[0091] In the embodiment of the present invention, the water body adsorption and desorption rates are determined based on the temperature, the pH value, the content information and the form information of the nitrogen element;

[0092] The flow parameters, the nitrate concentration, the water body adsorption and desorption rates, the water flow direction coordinate parameters, and the diffusion coefficient are input into Formula 2 to construct a two-dimensional nitrate transport model for a reservoir. Formula 2 is:

[0093]

[0094] Where i is the number of equation groups, the number of equation groups is greater than or equal to the number of segment divisions and greater than 0, R i is the water adsorption and desorption rate of the i-th group of equations, D i is the diffusion coefficient of the ith group of equations, and v is the flow parameter, so as to gain a deeper understanding of the migration and transformation law of nitrate in water bodies, provide a scientific basis for water quality management, and further improve the accuracy and rationality of the scheduling plan for the target reservoir.

[0095] In an optional embodiment of the present invention, the step of constructing a residual function through the one-dimensional nitrogen transport model of the reservoir includes:

[0096] The predicted value, the detection time, the water flow direction coordinate parameter, the initial weight, and the water adsorption and desorption rate of the i-th group of equations are input into Formula 3 to construct a residual function. Formula 3 is:

[0097]

[0098] Wherein, e is the residual function.

[0099] In a specific implementation, the embodiment of the present invention can input the predicted value, the detection time, the water flow direction coordinate parameter, the initial weight, and the water body adsorption and desorption rate of the i-th group of equations into Formula 3 to construct a residual function. Exemplarily, e can be the residual function, and e(θ) can represent the difference between the model predicted value and the true value. It can be the partial derivative of the physical quantity U (such as nitrate concentration) with respect to time t, which represents the rate of change of U with time, It can represent the partial derivative of the physical quantity U with respect to the spatial coordinate x, and represents the gradient of U in space. It can be the second-order partial derivative of the physical quantity U with respect to the spatial coordinate x, which represents the curvature of U in space. By using the one-dimensional nitrogen transport model, the predicted value at each time and each position is calculated according to the given input parameters, and the real observation value at the corresponding time and position is taken at the same time. For each time point and spatial position, the difference between the predicted value and the real value is calculated to obtain the local error. Then the calculated local error is substituted into e(θ) in Formula 3, and the predicted value is substituted into other items on the right side of Formula 3. The local errors of all time points and spatial positions are combined in a certain way to form an overall residual function. Preferably, the least squares method and the maximum likelihood estimation method can be used. By constructing the residual function, the difference between the model predicted value and the real value can be quantitatively measured, so as to evaluate the accuracy of the model. By minimizing the residual function, the model parameters can be optimized to improve the prediction accuracy of the model. By analyzing the residual function, the deficiencies of the model can be found, so as to improve the model. At the same time, it is used to analyze the uncertainty of the model.

[0100] In the embodiment of the present invention, the predicted value, the detection time, the water flow direction coordinate parameter, the initial weight, and the water adsorption and desorption rate of the i-th group of equations are input into Formula 3 to construct a residual function, and Formula 3 is:

[0101]

[0102] Wherein, e is the residual function, which further improves the accuracy of the model and provides support for subsequent data processing.

[0103] In an optional embodiment of the present invention, the step of determining the target weight includes:

[0104] Determine the number of iterations and the learning rate, and input the number of iterations and the learning rate into Formula 4 to determine the target weight, which is:

[0105]

[0106] Wherein, n is the number of iterations, and α is the learning rate.

[0107] In practical applications, by continuously updating the weight vector, the value of the residual function can be gradually reduced, thereby gradually reducing the error between the model's predicted value and the true value.

[0108] In a specific implementation, the embodiment of the present invention can determine the number of iterations and the learning rate, and input the number of iterations and the learning rate into Formula 4 to determine the target weight. For example, n can be the number of iterations, α can be the learning rate, and θ n can be the weight vector obtained by iterating the initial weights for the nth time, θ n+1 can be the weight vector at the n+1th iteration and used as the target weight, α can be the learning rate, and can be a hyperparameter that controls the step size of each update. It can be the gradient of the residual function e to the initial weight θ, indicating the rate of change of e in the direction of θ, that is, the direction in which the function drops fastest at the current point, so that each iteration updates the weight vector in the opposite direction of the gradient, so that the error between the predicted value of the model and the true value gradually decreases. Preferably, it can be determined whether the conditions for stopping iteration are met, such as setting the maximum number of iterations or the error is less than a certain threshold. If not, return to calculate the residual and continue iterating. When the conditions for stopping iteration are met, the current θ can be output. n+1 as the target weight of the model.

[0109] In the embodiment of the present invention, the number of iterations and the learning rate are determined, and the number of iterations and the learning rate are input into Formula 4 to determine the target weight. Formula 4 is:

[0110]

[0111] Wherein, n is the number of iterations, and α is the learning rate, thereby reducing the error between the predicted value and the true value of the model, and further improving the reliability and accuracy of determining the scheduling plan for the target reservoir.

[0112] In an optional embodiment of the present invention, the physical information neural network includes two hidden layers.

[0113] In a specific implementation, the physical information neural network in an embodiment of the present invention may include two hidden layers. For example, when the physical information neural network is PINNs, the first hidden layer may be mainly responsible for extracting low-level features from the input data, such as the relationship between simple physical quantities, and may learn the relationship between basic physical quantities such as speed, pressure, and density. The second hidden layer may further extract high-level features based on the features extracted by the first layer, capture more complex physical laws and nonlinear relationships, and may learn more complex fluid phenomena such as vorticity and boundary layers. The two hidden layers provide sufficient model capacity to learn complex physical laws. Too few hidden layers may cause the model to be underfitted and unable to capture the complexity of the data; too many hidden layers may cause overfitting and poor generalization ability. The first hidden layer may learn features related to physical laws, and the second hidden layer may combine these features to form a higher-level physical representation. In this way, PINNs can better integrate physical information and data information. By taking the partial differential equation for nitrogen transport as part of the loss function, the neural network can be forced to learn solutions that satisfy the laws of physics. This makes PINNs more accurate in predicting the temporal and spatial distribution of nitrogen concentration. Since the nitrogen transport process often involves complex nonlinear relationships, such as adsorption, desorption, and biodegradation, PINNs can effectively capture these nonlinear relationships through multi-layer neural networks. The neural network with two hidden layers can learn more complex nonlinear mapping relationships, thereby better fitting the complex process of nitrogen transport. Furthermore, by adjusting the number of neurons and activation functions in the hidden layer, the complexity and generalization ability of the model can be balanced.

[0114] In the embodiment of the present invention, by making the physical information neural network include two hidden layers, the physical information neural network can simulate complex water environment systems more accurately, thereby providing strong support for water resource management and water environment protection.

[0115] In order to enable those skilled in the art to better understand the embodiments of the present invention, the embodiments of the present invention are described below using a complete example.

[0116] refer to Figure 2 , showing a schematic diagram of a reservoir nitrogen transport process provided in an embodiment of the present invention; referring to Figure 3 , showing a schematic diagram of the topological structure of a distributed river section provided in an embodiment of the present invention.

[0117] Large reservoirs are an important guarantee of water resources, but they are facing an increasingly serious problem of nitrogen pollution. The discharge of a large amount of nitrogen-containing wastewater in the basin has led to an increase in the nitrogen content of the reservoir, which seriously threatens the safety of water quality. The distribution of nitrogen in the reservoir presents complex characteristics such as non-uniformity and temporal dynamics, and is affected by multiple sources of input and biological effects, making its transport process difficult to accurately simulate. In addition, the number of reservoir monitoring stations is limited, which limits the comprehensive grasp of water quality changes and increases the uncertainty of reservoir scheduling. Therefore, how to accurately predict the transport process of nitrogen in the reservoir and formulate a scientific reservoir scheduling plan based on this is an important challenge facing current water environment management. The specific process is as follows:

[0118] Collect the target reservoir area information, including the cross-section information of the main and tributary rivers along the way, and obtain the historical data of the monitoring stations set up on the main and tributary rivers in the reservoir area, including: flow, average flow rate, water temperature, nitrogen content, and preferably, the total length h of the reservoir area.

[0119] The river channel is divided into multiple sections according to the cross-sectional data, and a solver is established using a physical information neural network PINNs with stronger generalization ability. PINNs is built based on a deep feedforward neural network. The feedforward neural network consists of a series of nonlinear activation functions, which are ultimately combined into an output function. These output functions can approximate any function in terms of nonlinear function optimization. A neural network output function for the unknown nitrate concentration U is constructed here, as shown in the above formula 1, wherein x represents the spatial position along the flow direction, which can be a water flow direction coordinate parameter, t represents the detection time, θ can be a set of weights of the activation function in the neural network as the initial weight, and the structure of the neural network can include two hidden layers.

[0120] Can be Figure 3 A one-dimensional nitrogen transport model for a reservoir is established under the topological structure shown in the figure. The entire river channel is divided into i sections, and the number of control equations is i. The diffusion coefficient and water adsorption and desorption rates of the j-th section can be fixed values ​​Dj and Rj (0<j≤i) to represent the one-dimensional nitrogen convection diffusion of the j-th section. v is a flow parameter that can be taken based on the average flow velocity. The water adsorption and desorption rates can represent the activity of plankton and microorganisms in the water body, which is affected by the adsorption of oxygen content in the bottom mud of the water body. It can be measured by collecting water and bottom mud of the target reservoir in the laboratory. The diffusion coefficient can describe the diffusion capacity of nitrogen in the reservoir, as shown in the above formula 2.

[0121] By adding physical constraints to the neural network and constructing a residual term that does not depend on data, a continuous time format can be used to construct a residual function through a one-dimensional nitrogen transport model, which can be shown in the above formula 3, where e is the physical constraint residual function. This shows that the closer the value of the residual is to zero, the closer the approximate solution of PINNs is to the true solution. By adjusting the weight θ to reduce the value of the residual, the output function of the neural network is close to the above formula 1. A gradient algorithm suitable for convex optimization can be used, as shown in the above formula 4, α can be the learning rate; the Adam algorithm (Adaptive Moment Estimation) can be used to improve the learning efficiency through adaptive learning rate, and introduce momentum to enhance the stability of the algorithm, so that it is more suitable for complex non-convex optimization problems. Furthermore, when local convergence problems occur, global optimization algorithms such as GA (Genetic Algorithm) and PSO (Particle Swarm Optimization) can be used. After the residual converges, the output function of the neural network is the nitrogen transport process of the reservoir.

[0122] By introducing physical information neural networks, partial differential equations describing physical processes are integrated into the loss function of neural networks, and complex physical systems can be accurately learned and simulated with very little data. Not only can the nitrogen transport law be quickly extracted from limited monitoring data, but also the model prediction results can be highly consistent with the actual physical process, thus providing strong support for determining the final dispatching plan for the target reservoir.

[0123] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0124] Reference Figure 4 , shows a structural block diagram of a reservoir scheduling device provided in an embodiment of the present invention, which may specifically include the following modules:

[0125] The cross-sectional data collection module 401 is used to collect the reservoir area information of the target reservoir and the cross-sectional data of the river channel connected to the target reservoir;

[0126] A segment number determination module 402 is used to determine the segment number of the connected river channel according to the cross-section data;

[0127] A physical information neural network construction module 403 is used to construct a physical information neural network;

[0128] A neural network output function determination module 404 is used to determine a neural network output function for characterizing nitrate concentration based on the reservoir area information through the physical information neural network;

[0129] A reservoir one-dimensional nitrogen transport model construction module 405 is used to construct a reservoir one-dimensional nitrogen transport model based on the reservoir area information, the number of division segments and the neural network output function;

[0130] A residual function construction module 406 is used to construct a residual function through the one-dimensional nitrogen transport model of the reservoir;

[0131] The dispatching scheme generating module 407 is used to determine the target weight, adjust the residual function based on the target weight, and generate a dispatching scheme for the target reservoir.

[0132] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0133] In addition, an embodiment of the present invention further provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the various processes of the above-mentioned reservoir scheduling method embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0134] The embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, each process of the above-mentioned reservoir scheduling method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0135] Figure 5 A schematic diagram of the hardware structure of an electronic device for implementing various embodiments of the present invention.

[0136] The electronic device 500 includes but is not limited to: a radio frequency unit 501, a network module 502, an audio output unit 503, an input unit 504, a sensor 505, a display unit 506, a user input unit 507, an interface unit 508, a memory 509, a processor 510, and a power supply 511. Those skilled in the art will appreciate that Figure 5The electronic device structure shown in the figure does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. In the embodiments of the present invention, the electronic device includes but is not limited to a mobile phone, a tablet computer, a laptop computer, a PDA, a vehicle-mounted terminal, a wearable device, and a pedometer.

[0137] It should be understood that in the embodiment of the present invention, the radio frequency unit 501 can be used for receiving and sending signals during information transmission or communication. Specifically, after receiving downlink data from the base station, it is sent to the processor 510 for processing; in addition, uplink data is sent to the base station. Generally, the radio frequency unit 501 includes but is not limited to an antenna, at least one amplifier, a transceiver, a coupler, a low noise amplifier, a duplexer, etc. In addition, the radio frequency unit 501 can also communicate with the network and other devices through a wireless communication system.

[0138] The electronic device provides users with wireless broadband Internet access through the network module 502, such as helping users to send and receive emails, browse web pages, and access streaming media.

[0139] The audio output unit 503 can convert the audio data received by the RF unit 501 or the network module 502 or stored in the memory 509 into an audio signal and output it as sound. Moreover, the audio output unit 503 can also provide audio output related to a specific function performed by the electronic device 500 (for example, a call signal reception sound, a message reception sound, etc.). The audio output unit 503 includes a speaker, a buzzer, a receiver, etc.

[0140] The input unit 504 is used to receive audio or video signals. The input unit 504 may include a graphics processor (GPU) 5041 and a microphone 5042. The graphics processor 5041 processes the image data of a static picture or video obtained by an image capture device (such as a camera) in a video capture mode or an image capture mode. The processed image frame can be displayed on the display unit 506. The image frame processed by the graphics processor 5041 can be stored in the memory 509 (or other storage medium) or sent via the radio frequency unit 501 or the network module 502. The microphone 5042 can receive sound and can process such sound into audio data. The processed audio data can be converted into a format output that can be sent to a mobile communication base station via the radio frequency unit 501 in the case of a telephone call mode.

[0141] The electronic device 500 also includes at least one sensor 505, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor includes an ambient light sensor and a proximity sensor, wherein the ambient light sensor can adjust the brightness of the display panel 5061 according to the brightness of the ambient light, and the proximity sensor can turn off the display panel 5061 and / or the backlight when the electronic device 500 is moved to the ear. As a type of motion sensor, the accelerometer sensor can detect the magnitude of acceleration in each direction (generally three axes), and can detect the magnitude and direction of gravity when stationary, which can be used to identify the posture of the electronic device (such as horizontal and vertical screen switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc.; the sensor 505 can also include a fingerprint sensor, a pressure sensor, an iris sensor, a molecular sensor, a gyroscope, a barometer, a hygrometer, a thermometer, an infrared sensor, etc., which will not be repeated here.

[0142] The display unit 506 is used to display information input by the user or information provided to the user. The display unit 506 may include a display panel 5061, which may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like.

[0143] The user input unit 507 can be used to receive input digital or character information, and to generate key signal input related to user settings and function control of the electronic device. Specifically, the user input unit 507 includes a touch panel 5071 and other input devices 5072. The touch panel 5071, also known as a touch screen, can collect the user's touch operation on or near it (such as the user's operation on the touch panel 5071 or near the touch panel 5071 using any suitable object or accessory such as a finger, stylus, etc.). The touch panel 5071 may include two parts: a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch orientation, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into the contact point coordinates, and then sends it to the processor 510, receives the command sent by the processor 510 and executes it. In addition, the touch panel 5071 can be implemented using multiple types such as resistive, capacitive, infrared and surface acoustic waves. In addition to the touch panel 5071, the user input unit 507 may also include other input devices 5072. Specifically, other input devices 5072 may include but are not limited to a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, and a joystick, which are not described in detail here.

[0144] Furthermore, the touch panel 5071 may be overlaid on the display panel 5061. When the touch panel 5071 detects a touch operation on or near it, it transmits the information to the processor 510 to determine the type of the touch event. Then, the processor 510 provides a corresponding visual output on the display panel 5061 according to the type of the touch event. Figure 5 In the figure, the touch panel 5071 and the display panel 5061 are used as two independent components to realize the input and output functions of the electronic device. However, in some embodiments, the touch panel 5071 and the display panel 5061 can be integrated to realize the input and output functions of the electronic device, which is not limited here.

[0145] The interface unit 508 is an interface for connecting an external device to the electronic device 500. For example, the external device may include a wired or wireless headset port, an external power supply (or battery charger) port, a wired or wireless data port, a memory card port, a port for connecting a device with an identification module, an audio input / output (I / O) port, a video I / O port, a headphone port, etc. The interface unit 508 may be used to receive input (e.g., data information, power, etc.) from an external device and transmit the received input to one or more elements within the electronic device 500 or may be used to transmit data between the electronic device 500 and an external device.

[0146] The memory 509 can be used to store software programs and various data. The memory 509 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory 509 can include a high-speed random access memory, and can also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or a non-volatile solid-state storage device thereof.

[0147] The processor 510 is the control center of the electronic device. It uses various interfaces and lines to connect various parts of the entire electronic device. By running or executing software programs and / or modules stored in the memory 509 and calling data stored in the memory 509, it performs various functions of the electronic device and processes data, thereby monitoring the electronic device as a whole. The processor 510 may include one or more processing units; preferably, the processor 510 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 510.

[0148] The electronic device 500 may also include a power supply 511 (such as a battery) for supplying power to each component. Preferably, the power supply 511 may be logically connected to the processor 510 through a power management system, thereby implementing functions such as charging, discharging, and power consumption management through the power management system.

[0149] In addition, the electronic device 500 includes some functional modules not shown, which will not be described in detail here.

[0150] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0151] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0152] like Figure 6 As shown, in another embodiment provided by the present invention, a computer-readable storage medium 601 is also provided, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the reservoir scheduling method described in the above embodiment.

[0153] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

[0154] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the embodiments of the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0155] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0156] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0157] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0158] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0159] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0160] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A reservoir dispatching method, characterized in that: include: Collecting reservoir area information of the target reservoir and cross-sectional data of the river connecting the target reservoir; Determining the number of sections for the connected river channel through the cross-section data; Construct physical information neural network; Determining a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information; Constructing a one-dimensional nitrogen transport model for the reservoir based on the reservoir area information, the number of segment divisions and the neural network output function; constructing a residual function through the one-dimensional nitrogen transport model of the reservoir; A target weight is determined, and the residual function is adjusted based on the target weight, and a scheduling plan for the target reservoir is generated.

2. The method according to claim 1, characterized in that The reservoir area information includes at least: flow parameters, diffusion coefficient, water flow direction coordinate parameters of the target reservoir, temperature, pH value, nitrogen content information and form information.

3. The method according to claim 2, characterized in that The step of determining a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information comprises: Determining initial weights and detection time of an activation function for the physical information neural network; The initial weight, the detection time, and the water flow direction coordinate parameters are input into Formula 1 to determine the neural network output function. Formula 1 is: U(x,t)≈U s (x,t;θ) Among them, x is the water flow direction coordinate parameter, t is the detection time, θ is the initial weight; U is the nitrate concentration, and Us is the predicted value of the nitrate concentration by the physical information neural network.

4. The method according to claim 3, characterized in that: The step of constructing a one-dimensional nitrogen transport model for a reservoir based on the reservoir area information, the number of segment divisions and the neural network output function comprises: Determining the adsorption and desorption rates of the water body based on the temperature, the pH value, the content information and the form information of the nitrogen element; The flow parameters, the nitrate concentration, the water body adsorption and desorption rates, the water flow direction coordinate parameters, and the diffusion coefficient are input into Formula 2 to construct a two-dimensional nitrate transport model for a reservoir. Formula 2 is: Where i is the number of equation groups, the number of equation groups is greater than or equal to the number of segment divisions and greater than 0, R i is the water adsorption and desorption rate of the i-th group of equations, D i is the diffusion coefficient of the i-th set of equations, and v is the flow parameter.

5. The method according to claim 4, characterized in that The step of constructing a residual function through the one-dimensional nitrogen transport model of the reservoir comprises: The predicted value, the detection time, the water flow direction coordinate parameter, the initial weight, and the water adsorption and desorption rate of the i-th group of equations are input into Formula 3 to construct a residual function. Formula 3 is: Wherein, e is the residual function.

6. The method according to claim 5, characterized in that The step of determining the target weight comprises: Determine the number of iterations and the learning rate, and input the number of iterations and the learning rate into Formula 4 to determine the target weight, which is: Wherein, n is the number of iterations, and α is the learning rate.

7. The method according to any one of claims 1 to 6, characterized in that: The physical information neural network includes two hidden layers.

8. A reservoir dispatching device, characterized in that: include: A cross-sectional data collection module, used to collect reservoir information of a target reservoir and cross-sectional data of a river channel connected to the target reservoir; A module for determining the number of divided sections, used to determine the number of divided sections for the connected river channel through the cross-section data; Physical information neural network building module, used to build physical information neural network; A neural network output function determination module, used to determine a neural network output function for characterizing nitrate concentration through the physical information neural network based on the reservoir area information; A reservoir one-dimensional nitrogen transport model construction module, used to construct a reservoir one-dimensional nitrogen transport model based on the reservoir area information, the number of division segments and the neural network output function; A residual function construction module, used for constructing a residual function through the one-dimensional nitrogen transport model of the reservoir; The scheduling scheme generating module is used to determine the target weight, adjust the residual function based on the target weight, and generate a scheduling scheme for the target reservoir.

9. An electronic device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; The memory is used to store computer programs; The processor is used to implement the method according to any one of claims 1 to 7 when executing the program stored in the memory.

10. A computer-readable storage medium having instructions stored thereon, which, when executed by one or more processors, cause the processors to perform the method according to any one of claims 1 to 7.

Images