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. Reservoir area information of a target reservoir is collected; constructing a physical information neural network; determining a neural network output function for representing the phosphate concentration through the physical information neural network based on the reservoir area information; constructing a reservoir two-dimensional phosphate transport model based on the reservoir area information and the neural network output function; constructing a residual function through the reservoir two-dimensional phosphate 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 efficiency and accuracy of reservoir scheduling.

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] Phosphorus is an important biogenic element in water bodies and an essential element for the survival of aquatic organisms. It is mainly dissolved in water or precipitated in the form of phosphate in reservoirs. Its excessive enrichment will lead to eutrophication of water bodies, which will in turn cause a series of ecological and environmental problems such as blue algae blooms, threatening the water quality and ecosystem stability of reservoirs. Due to the influence of human activities, the enrichment of phosphorus in reservoirs has become a serious environmental problem. Phosphorus exists in water in dissolved and particulate states. Particulate phosphorus often attaches to suspended matter, so phosphorus is easy to accumulate and deposit in reservoir water bodies.

[0003] Phosphate in reservoirs is affected by factors such as water temperature, pH value, salinity, hydrodynamic conditions and phosphorus form. Its transport process is difficult to accurately predict. At the same time, due to the limited data monitoring points in the reservoir, it is difficult to obtain accurate simulation results, which brings difficulties to reservoir scheduling. Summary of the invention

[0004] 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.

[0005] The embodiment of the present invention discloses a reservoir scheduling method, comprising:

[0006] Collect reservoir area information of the target reservoir;

[0007] Construct physical information neural network;

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

[0009] Constructing a two-dimensional phosphate transport model for the reservoir based on the reservoir area information and the neural network output function;

[0010] constructing a residual function through the two-dimensional phosphate 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 includes at least: water flow direction coordinate parameters and vertical flow coordinate parameters of the target reservoir, temperature, pH value, and phosphorus content information and form information of the target reservoir, water flow direction parameters and vertical flow parameters.

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

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

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

[0016] U(x,y,t)≈U s (x, y, t; θ)

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

[0018] Optionally, the step of constructing a two-dimensional reservoir phosphate transport model based on the reservoir area information and the neural network output function includes:

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

[0020] The water flow direction parameter, the vertical flow parameter, the phosphate concentration, the water body adsorption and desorption rate, the water flow direction coordinate parameter, the vertical flow coordinate parameter, the water flow direction parameter and the vertical flow parameter are input into Formula 2 to construct a two-dimensional phosphate transport model of the reservoir. Formula 2 is:

[0021]

[0022] Among them, V x is the water flow direction parameter, V y is the vertical flow parameter, R is the water adsorption and desorption rate, D x is the water flow direction parameter, D y is the vertical flow parameter.

[0023] Optionally, the step of constructing a residual function through the reservoir two-dimensional phosphate transport model comprises:

[0024] The predicted value, the detection time, the water flow direction coordinate parameter, the vertical flow coordinate parameter, the initial weight, the water flow direction parameter, the vertical flow parameter, the water body adsorption and desorption rate 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 includes:

[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 includes multiple hidden layers.

[0032] The embodiment of the present invention also discloses a reservoir dispatching device, including:

[0033] A reservoir area information collection module is used to collect reservoir area information of the target reservoir;

[0034] Physical information neural network building module, used to build physical information neural network;

[0035] A neural network output function determination module, used to determine a neural network output function for characterizing phosphate concentration through the physical information neural network based on the reservoir area information;

[0036] A reservoir two-dimensional phosphate transport model construction module, used to construct a reservoir two-dimensional phosphate transport model based on the reservoir area information and the neural network output function;

[0037] The residual function framework is modular and used to construct the residual function through the two-dimensional phosphate transport model of the reservoir;

[0038] The target weight determination module is used to determine the target weight, adjust the residual function based on the target weight, and generate a scheduling plan for the target reservoir.

[0039] 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;

[0040] The memory is used to store computer programs;

[0041] The processor is used to implement the method described in the embodiment of the present invention when executing the program stored in the memory.

[0042] 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.

[0043] The embodiments of the present invention include the following advantages:

[0044] The embodiment of the present invention collects reservoir information of a target reservoir; constructs a physical information neural network; determines a neural network output function for characterizing phosphate concentration through the physical information neural network based on the reservoir information; constructs a two-dimensional phosphate transport model of the reservoir based on the reservoir information and the neural network output function; constructs a residual function through the two-dimensional phosphate transport model of 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 efficiency and accuracy of reservoir scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is a flow chart of steps of a reservoir dispatching method provided in an embodiment of the present invention;

[0046] Figure 2 is a structural schematic diagram of a dam provided in an embodiment of the present invention;

[0047] Figure 3 is a structural block diagram of a reservoir dispatching device provided in an embodiment of the present invention;

[0048] Figure 4 is a hardware structure block diagram of an electronic device provided in an embodiment of the present invention;

[0049] Figure 5 is a schematic diagram of a computer-readable medium provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] 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.

[0051] 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.

[0052] As an important nutrient element in aquatic ecosystems, excessive enrichment of phosphorus in reservoirs is the root cause of ecological problems such as eutrophication of water bodies and blue algae blooms. Phosphorus exists in water bodies mainly in two forms: dissolved and particulate. Particulate phosphorus is often adsorbed on suspended matter and easily accumulates in reservoir sediments. Due to the combined effects of hydrodynamic conditions such as water flow velocity and turbulent diffusion, as well as environmental factors such as water temperature, pH value, and salinity, the transport process of phosphorus in reservoirs is complex and changeable, and is difficult to accurately simulate due to the sparseness of monitoring data.

[0053] To solve the above problems, an embodiment of the present invention proposes a two-dimensional phosphate transport simulation method for reservoirs based on physical information neural networks (PINNs). PINNs combines the data-driven characteristics of deep learning and the physical constraints of traditional numerical simulations, and can accurately simulate the transport process of phosphorus in reservoirs with a small amount of training data. Compared with traditional artificial neural networks, PINNs can better adapt to the complexity and dynamics of reservoir environments and improve the reliability of simulation results. By training monitoring data, PINNs can effectively capture the spatiotemporal distribution of phosphorus in reservoirs, providing a scientific basis for reservoir management and ecological restoration.

[0054] Compared with artificial neural networks in related technologies, physical information neural networks (PINNs) incorporate the constraints of physical laws on the basis of deep feedforward neural networks. PINNs can learn patterns from data while satisfying the constraints of physical equations, realizing the combination of data-driven and physical-driven. This unique structure requires less labeled data during the training process of PINNs, and has strong generalization ability, which can adapt to the complexity and dynamic changes of reservoir environment. Through training on monitoring data, PINNs can accurately simulate the transport process of nitrogen in reservoirs, providing a reliable scientific basis for water environment management.

[0055] The embodiment of the present invention adopts the above method and combines data-driven with physical models through physical information neural networks (PINNs) to achieve accurate simulation of phosphate transport process. While ensuring that the model accurately reflects the actual physical process of water bodies, PINNs can use limited monitoring data to quickly obtain the phosphate transport characteristics under current water flow conditions.

[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 area information of the target reservoir;

[0058] Step 102, constructing a physical information neural network;

[0059] Step 103, determining a neural network output function for characterizing phosphate concentration through the physical information neural network based on the reservoir area information;

[0060] Step 104, constructing a two-dimensional phosphate transport model of the reservoir based on the reservoir area information and the neural network output function;

[0061] Step 105, constructing a residual function through the two-dimensional phosphate transport model of the reservoir;

[0062] Step 106, determining a target weight, adjusting the residual function based on the target weight, and generating a scheduling plan for the target reservoir.

[0063] Reservoir dispatching refers to the process of redistributing the inflow or determining the amount of water stored and released by the reservoir according to the reservoir development tasks and comprehensive utilization requirements. In simple terms, it is to reasonably arrange the storage and release of reservoirs according to different needs to achieve the best economic and social benefits.

[0064] In practical applications, hydrological elements such as rainfall and runoff have great uncertainty. Reservoir scheduling can help us deal with these uncertainties and ensure a stable supply of water resources.

[0065] Reservoirs usually have multiple functions, such as flood control, power generation, irrigation, water supply, etc. Reservoir scheduling needs to comprehensively consider these different goals to achieve the optimal configuration.

[0066] The target reservoir in the embodiment of the present invention may be a reservoir that needs to be dispatched.

[0067] refer to Figure 2 , Figure 2 It is a structural schematic diagram of a dam provided in an embodiment of the present invention.

[0068] The embodiment of the present invention can collect reservoir information of the target reservoir. Optionally, historical data of monitoring stations set up on the main and tributary rivers of the reservoir can be obtained as reservoir information, which may include but is not limited to: flow rate, water temperature, pH value, salinity, hydrodynamic conditions, phosphorus content and form, water flow direction coordinate parameter x and vertical flow coordinate parameter y of the target reservoir, water flow direction parameters and vertical flow parameters, etc.

[0069] The embodiment of the present invention can use physical information neural networks (PINNs) with stronger generalization ability to establish a solver. PINNs are built based on deep forward neural networks. The forward neural network consists of a series of nonlinear activation functions, which are ultimately combined into output functions. These output functions can achieve approximation of arbitrary functions in terms of nonlinear function optimization.

[0070] PINNs are a type of solver: they use deep learning techniques, specifically feed-forward neural networks, to solve complex physics problems.

[0071] PINNs have strong generalization ability: meaning that it can not only process known data, but also make accurate predictions on unknown data.

[0072] The working principle of PINNs: It constructs an output function that can approximate any function through a combination of a series of nonlinear activation functions. This output function can be used to represent complex physical phenomena.

[0073] Physical Information Neural Networks (PINNs):

[0074] Definition: PINNs are a new type of deep learning model that embeds the laws of physics into neural networks.

[0075] The characteristic is that PINNs can combine physical knowledge to not only learn patterns in data, but also use known physical equations as constraints to make the model more accurate and reliable.

[0076] Strong generalization ability: PINNs can provide accurate prediction results even when data is scarce.

[0077] Feedforward neural network: consists of multiple hidden layers, each containing multiple neurons.

[0078] Working principle: The input data is processed by neurons in each layer and passes through a nonlinear activation function to finally obtain the output result.

[0079] Function: Able to approximate arbitrarily complex functions.

[0080] The embodiment of the present invention introduces nonlinear factors, so that the neural network can learn more complex patterns.

[0081] Common types include activation functions such as ReLU, Sigmoid and Tanh.

[0082] After determining that the physical information neural network is the solver, an embodiment of the present invention can determine the neural network output function used to characterize the phosphate concentration through the physical information neural network based on the reservoir information. The output function can be the final output of the physical information neural network, which is used to solve the unknown phosphate concentration.

[0083] PINNs provides a new approach to solving complex physical problems by combining physical knowledge with deep learning. It uses the powerful fitting ability of feedforward neural networks and nonlinear activation functions to construct an output function that can approximate any function, thereby solving physical problems.

[0084] In a specific implementation, an embodiment of the present invention can construct a two-dimensional phosphate transport model of a reservoir based on reservoir information and a neural network output function; construct a residual function through the two-dimensional phosphate transport model of a reservoir; reduce the value of the residual by adjusting the initial weights so that the output function of the physical information neural network approximates the two-dimensional nitrogen convection-diffusion equation. After determining the target weights and achieving residual convergence, the function output by the physical information neural network can be uploaded to the reservoir dispatching department as a reference indicator for formulating dispatching plans.

[0085] The embodiment of the present invention collects reservoir information of a target reservoir; constructs a physical information neural network; determines a neural network output function for characterizing phosphate concentration through the physical information neural network based on the reservoir information; constructs a two-dimensional phosphate transport model of the reservoir based on the reservoir information and the neural network output function; constructs a residual function through the two-dimensional phosphate transport model of 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 efficiency and accuracy of reservoir scheduling.

[0086] 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.

[0087] In an optional embodiment of the present invention, the step of determining a neural network output function for characterizing phosphate concentration through the physical information neural network based on the reservoir area information comprises:

[0088] Determine weights and detection times for activation functions in neural networks;

[0089] The weight of the activation function in the neural network, the detection time, the water flow direction coordinate parameter and the vertical flow coordinate parameter are input into Formula 1 to determine the neural network output function. The Formula 1 is:

[0090] U(x,y,t)≈U s (x, y, t; θ)

[0091] Among them, x is the water flow direction coordinate parameter, y is the vertical flow coordinate parameter, t is the detection time, and θ is the weight of the activation function in the neural network.

[0092] U(x, y, t): actual phosphate concentration, which can be obtained from reservoir information. It is a function of spatial coordinates (water flow direction coordinate parameter x, vertical flow coordinate parameter y) and time (t).

[0093] U s(x, y, t; θ): Phosphate concentration output by the neural network, which is also a function of spatial coordinates (water flow direction coordinate parameter x, vertical flow coordinate parameter y), time (t) and physical information neural network residual function θ.

[0094] ≈: means approximately equal, that is, the output Us of the neural network is an approximation of the actual phosphate concentration U.

[0095] U s Represents the predicted value of phosphate concentration by the physical information neural network.

[0096] θ is a set of initial weights of the physical information neural network. The structure of the neural network contains multiple hidden layers. By adjusting the initial weights, U s Get as close to the real U as possible.

[0097] Physical information neural network determines U s The process:

[0098] Construct a neural network structure: Design a suitable neural network with input (x, y, t) and output Us.

[0099] Definition of loss function: The mean square error is usually used as the loss function, that is, the square difference between U and U is calculated and the sum is taken as the average.

[0100] Introducing physical constraints: Add relevant physical equations (such as diffusion equation, convection equation, etc.) as constraints to the loss function to ensure that the output of the neural network conforms to the laws of physics.

[0101] Training the neural network: Use optimization algorithms such as Adam (Adaptive Moment Estimation), SGD (Stochastic Gradient Descent), etc. to minimize the loss function and obtain the optimal neural network parameter θ.

[0102] Get U s : The trained neural network can predict the phosphate concentration Us based on the given (x, y, t).

[0103] Physical information: Incorporating physical laws into neural networks is the core idea of ​​PINN. By introducing physical constraints, the accuracy and generalization ability of the model can be improved.

[0104] Neural network parameter θ: θ is the "adjustable knob" of the neural network. By learning data, we can find the most suitable θ value so that U s It can fit U very well.

[0105] Loss function: Loss function is an important indicator to measure model performance. In addition to mean square error, other types of loss functions can also be considered, such as cross entropy loss.

[0106] In an optional embodiment of the present invention, the step of constructing a two-dimensional phosphate transport model of a reservoir based on the reservoir area information and the neural network output function comprises:

[0107] Determining the water body adsorption and desorption rates based on the temperature, pH value, content information and form information of the phosphorus element;

[0108] The water flow direction parameter, the vertical flow parameter, the phosphate concentration, the water body adsorption and desorption rate, the water flow direction coordinate parameter, the vertical flow coordinate parameter, the water flow direction parameter and the vertical flow parameter are input into Formula 2 to construct a two-dimensional phosphate transport model of the reservoir. Formula 2 is:

[0109]

[0110] Among them, V x is the water flow direction parameter, V y is the vertical flow parameter, R is the water adsorption and desorption rate, D x is the water flow direction parameter, D y is the vertical flow parameter.

[0111] U: Phosphate concentration, that is, the mass of phosphate per unit volume of water.

[0112] Vx, Vy: represent the velocity components of water flow in the x-direction and y-direction respectively, reflecting the advection effect of water flow on phosphate. Vx, Vy are hydrodynamic conditions that can be obtained from reservoir information.

[0113] Dx, Dy: represent the diffusion coefficients of phosphate in the x-direction and y-direction respectively, reflecting the diffusion movement of phosphate due to the concentration gradient.

[0114] R: The adsorption and desorption rate of phosphate by water reflects the exchange process of phosphate between water and sediment. It is affected by factors such as water temperature, pH value and phosphorus form, and can be measured by collecting target reservoir water in the laboratory.

[0115] x, y: represent the horizontal and vertical coordinates of the water flow respectively.

[0116] t: detection time.

[0117] U s : Prediction of phosphate concentration by physical information neural network.

[0118] The physical meaning of formula 2:

[0119] U / t represents the rate of change of phosphate concentration U with time t, that is, the speed of change of phosphate concentration.

[0120] The physical meanings of the parts on the right side of the equation are as follows:

[0121] - Represents the change in phosphate concentration due to the advection of water flow.

[0122] Represents the change in phosphate concentration due to molecular diffusion.

[0123] -RU: Indicates the process by which phosphate is removed from water bodies by adsorption.

[0124] The embodiment of the present invention describes the change of phosphate concentration in water body by the above formula 2, which is jointly affected by factors such as water flow, diffusion and adsorption. By solving formula 2, the temporal and spatial distribution of phosphate in water body can be simulated, so as to evaluate the pollution degree of water body and predict the migration and transformation trend of pollutants.

[0125] Relationship with neural networks:

[0126] U in Formula 2 represents the predicted value of phosphate concentration by the physical information neural network, which shows that the model combines the advantages of traditional physical models and neural networks. Neural networks can improve the prediction accuracy of the model by learning a large amount of data and establishing complex nonlinear relationships. The physical model ensures the physical interpretability of the model, allowing the model to better reflect the real physical process.

[0127] In an optional embodiment of the present invention, the step of constructing a residual function through the reservoir two-dimensional phosphate transport model comprises:

[0128] The predicted value, the detection time, the water flow direction coordinate parameter, the vertical flow coordinate parameter, the initial weight, the water flow direction parameter, the vertical flow parameter, the water flow direction parameter, the vertical flow parameter, the water body adsorption and desorption rate are input into Formula 3 to construct a residual function. Formula 3 is:

[0129]

[0130] Among them, e is the residual function; e(θ): physical constraint residual function; represents the error between the model prediction value and the true value; Us: phosphate concentration predicted by the neural network; t: detection time; x, y: water flow direction and vertical coordinates; θ: weight parameter of the neural network; Vx, Vy: velocity components of water flow in the x and y directions; Dx, Dy: diffusion coefficients in the x and y directions; R: water adsorption and desorption rates.

[0131] In a specific implementation, the embodiment of the present invention constructs a data-independent residual term by adding physical constraints based on the two-dimensional nitrogen convection-diffusion equation (Formula 1) to the neural network, where a continuous time format is used.

[0132] Formula 3 is a partial differential equation, which shows the process of phosphate concentration changing over time in a two-dimensional water body. The left side of Formula 3 shows the rate of change of phosphate concentration, and the right side shows various factors that affect the change of phosphate concentration, including:

[0133] The rate of change of phosphate concentration U with time t indicates the growth or decay rate of phosphate concentration.

[0134] It represents the concentration change of phosphate due to the advection of water flow. Vx and Vy represent the velocity components of water flow in the x direction and y direction respectively. and Represents the gradient of concentration in the x and y directions.

[0135] It represents the concentration change of phosphate due to molecular diffusion. x and D y represent the diffusion coefficients of phosphate in the x and y directions, respectively, and Represents the second derivative of concentration in the x and y directions.

[0136] RU: represents the process of phosphate removal from water due to adsorption. R is the rate of adsorption and desorption of water, and Us is the phosphate concentration predicted by the neural network.

[0137] Changes in phosphate concentration are affected by:

[0138] 1. Water flow transport: Water flow will carry phosphate particles and cause changes in concentration distribution.

[0139] 2. Molecular diffusion. Due to uneven concentration, phosphate molecules will spontaneously diffuse from high concentration areas to low concentration areas.

[0140] 3. Adsorption / desorption rate. Solid particles in the water will adsorb phosphates, thereby reducing the phosphate concentration in the water.

[0141] Relationship with neural networks:

[0142] U in Formula 3 s is the predicted value of phosphate concentration by the physical information neural network. Substituting this predicted value into formula 3, the error e between the predicted value and the actual observed value can be calculated. This error can be used as the loss function of neural network training. By minimizing this loss function, the parameters of the neural network can be continuously optimized to make the predicted value closer to the actual value.

[0143] The embodiment of the present invention describes the mathematical model of the phosphate transport process in water bodies through Formula 3. By combining the physical model and the neural network, the temporal and spatial distribution of phosphate in water bodies can be simulated more accurately, providing a scientific basis for water environment management.

[0144] In an optional embodiment of the present invention, the step of determining the target weight includes:

[0145] 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:

[0146]

[0147] Wherein, n is the number of iterations, and α is the learning rate.

[0148] e is the residual function in formula 3, and α is the learning rate; the Adam algorithm uses an adaptive learning rate to improve learning efficiency, and introduces momentum to enhance the stability of the algorithm, making it more suitable for complex non-convex optimization problems. Optionally, if local convergence problems occur, global optimization algorithms such as genetic algorithms (GA) and particle swarm optimization methods (PSO) can be used. After the residual converges, the output function of the neural network is the phosphate transport process of the reservoir.

[0149] In a specific implementation, the embodiment of the present invention can determine the learning rate and the number of iterations; the learning rate, the residual function and the number of iterations are input into Formula 4 to determine the target weight. For example, θ can be the initial weight, n can be the number of iterations, α can be the learning rate, e can be the residual function, θ n+1 can be the target weight, and the weight formula can be updated by the gradient descent method to optimize the neural network model, θ n+1 is the target weight, which can be the weight after the n+1th iteration, θ nis the weight after the nth iteration. The α learning rate can control the step size of each update. In the process of updating parameters, if the step size is too large, the model may jump in the parameter space, resulting in failure to converge to the optimal solution or even divergence. If the step size is too small, the model will be more inclined to use the existing information and gradually approach the local optimal solution, but it may fall into the local minimum. Selecting the optimal learning rate can make the model converge to the optimal solution quickly. Through multiple experiments, the learning rate can be gradually adjusted to find the optimal value. As the training progresses, the learning rate can be gradually reduced to improve the stability of convergence. It represents the gradient of the residual function e(θ) with respect to the initial weight θ, and identifies the steepest descent direction of the loss function at the current weight point. By finding an optimal set of weights θ as the target weights, the error between the model output and the true value is minimized. Through continuous iteration, the model weights will gradually adjust in the direction of minimizing the loss function, thereby improving the performance of the model. When the residual converges, the output function of the neural network is the phosphate transport process of the reservoir.

[0150] In an optional embodiment of the present invention, the physical information neural network includes multiple hidden layers.

[0151] Exemplarily, when the physical information neural network is PINNs, the first hidden layer can be mainly responsible for extracting low-level features from the input data, such as the relationship between simple physical quantities, and can learn the relationship between basic physical quantities such as speed, pressure, and density. The second hidden layer can further extract high-level features based on the features extracted by the first layer to capture more complex physical laws and nonlinear relationships. It can 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 can learn features related to physical laws, and the second hidden layer can combine these features to form a more advanced physical representation. In this way, PINNs can better integrate physical information and data information.

[0152] In practical applications, physical information neural networks contain two hidden layers, which can work together to improve the performance of the model by extracting features, introducing nonlinearity, increasing model capacity, and fusing physical information. Reasonable design of the structure and parameters of the hidden layer is crucial to building an efficient and accurate PINNs model.

[0153] 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.

[0154] Reference Figure 3 , 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:

[0155] The reservoir area information collection module 301 is used to collect the reservoir area information of the target reservoir;

[0156] A physical information neural network construction module 302 is used to construct a physical information neural network;

[0157] A neural network output function determination module 303, used to determine a neural network output function for characterizing phosphate concentration through the physical information neural network based on the reservoir area information;

[0158] A reservoir two-dimensional phosphate transport model construction module 304 is used to construct a reservoir two-dimensional phosphate transport model based on the reservoir area information and the neural network output function;

[0159] The residual function framework is configured as module 305 for constructing a residual function through the two-dimensional phosphate transport model of the reservoir;

[0160] The target weight determination module 306 is used to determine the target weight, adjust the residual function based on the target weight, and generate a scheduling plan for the target reservoir.

[0161] 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.

[0162] 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.

[0163] 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.

[0164] Figure 4 A schematic diagram of the hardware structure of an electronic device for implementing various embodiments of the present invention.

[0165] The electronic device 400 includes but is not limited to: a radio frequency unit 401, a network module 402, an audio output unit 403, an input unit 404, a sensor 405, a display unit 406, a user input unit 407, an interface unit 408, a memory 409, a processor 410, and a power supply 411. Those skilled in the art will appreciate that Figure 4 The 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.

[0166] It should be understood that in the embodiment of the present invention, the radio frequency unit 401 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 410 for processing; in addition, uplink data is sent to the base station. Generally, the radio frequency unit 401 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 401 can also communicate with the network and other devices through a wireless communication system.

[0167] The electronic device provides users with wireless broadband Internet access through the network module 402, such as helping users to send and receive emails, browse web pages, and access streaming media.

[0168] The audio output unit 403 can convert the audio data received by the RF unit 401 or the network module 402 or stored in the memory 409 into an audio signal and output it as sound. Moreover, the audio output unit 403 can also provide audio output related to a specific function performed by the electronic device 400 (for example, a call signal reception sound, a message reception sound, etc.). The audio output unit 403 includes a speaker, a buzzer, a receiver, etc.

[0169] The input unit 404 is used to receive audio or video signals. The input unit 404 may include a graphics processor (GPU) 4041 and a microphone 4042, and the graphics processor 4041 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 406. The image frame processed by the graphics processor 4041 can be stored in the memory 409 (or other storage medium) or sent via the radio frequency unit 401 or the network module 402. The microphone 4042 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 401 in the case of a telephone call mode.

[0170] The electronic device 400 also includes at least one sensor 405, 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 4061 according to the brightness of the ambient light, and the proximity sensor can turn off the display panel 4061 and / or the backlight when the electronic device 400 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 405 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.

[0171] The display unit 406 is used to display information input by the user or information provided to the user. The display unit 406 may include a display panel 4061, which may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like.

[0172] The user input unit 407 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 407 includes a touch panel 4071 and other input devices 4072. The touch panel 4071, 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 4071 or near the touch panel 4071 using any suitable object or accessory such as a finger, stylus, etc.). The touch panel 4071 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 410, receives the command sent by the processor 410 and executes it. In addition, the touch panel 4071 can be implemented using various types such as resistive, capacitive, infrared and surface acoustic waves. In addition to the touch panel 4071, the user input unit 407 may also include other input devices 4072. Specifically, other input devices 4072 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.

[0173] Furthermore, the touch panel 4071 may be overlaid on the display panel 4061. When the touch panel 4071 detects a touch operation on or near it, it is transmitted to the processor 410 to determine the type of the touch event. Then, the processor 410 provides a corresponding visual output on the display panel 4061 according to the type of the touch event. Figure 4 In the figure, the touch panel 4071 and the display panel 4061 are two independent components to realize the input and output functions of the electronic device. However, in some embodiments, the touch panel 4071 and the display panel 4061 can be integrated to realize the input and output functions of the electronic device, which is not limited here.

[0174] The interface unit 408 is an interface for connecting an external device to the electronic device 400. 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 408 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 400 or may be used to transmit data between the electronic device 400 and an external device.

[0175] The memory 409 can be used to store software programs and various data. The memory 409 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a 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 409 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 other volatile solid-state storage devices.

[0176] The processor 410 is the control center of the electronic device. It uses various interfaces and lines to connect various parts of the entire electronic device. It executes various functions of the electronic device and processes data by running or executing software programs and / or modules stored in the memory 409, and calling data stored in the memory 409, so as to monitor the electronic device as a whole. The processor 410 may include one or more processing units; preferably, the processor 410 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 410.

[0177] The electronic device 400 may also include a power supply 411 (such as a battery) for supplying power to each component. Preferably, the power supply 411 may be logically connected to the processor 410 through a power management system, thereby implementing functions such as charging, discharging, and power consumption management through the power management system.

[0178] In addition, the electronic device 400 includes some functional modules not shown, which will not be described in detail here.

[0179] 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 includes 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.

[0180] 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.

[0181] like Figure 5 As shown, in another embodiment provided by the present invention, a computer-readable storage medium 501 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.

[0182] 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.

[0183] 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.

[0184] 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.

[0185] 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.

[0186] 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.

[0187] 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.

[0188] 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.

[0189] 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: Collect reservoir area information of the target reservoir; Construct physical information neural network; Determining a neural network output function for characterizing phosphate concentration through the physical information neural network based on the reservoir area information; Constructing a two-dimensional phosphate transport model for the reservoir based on the reservoir area information and the neural network output function; constructing a residual function through the two-dimensional phosphate 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: water flow direction coordinate parameters and vertical flow coordinate parameters of the target reservoir, temperature, pH value, and phosphorus content information and form information of the target reservoir, water flow direction parameters and vertical flow parameters.

3. The method according to claim 2, characterized in that The step of determining a neural network output function for characterizing phosphate 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, the water flow direction coordinate parameter and the vertical flow coordinate parameter are input into Formula 1 to determine the neural network output function. The Formula 1 is: U(x,y,t)≈U s (x,y,t,θ) Among them, x is the water flow direction coordinate parameter, y is the vertical flow coordinate parameter, t is the detection time, θ is the initial weight; U is the phosphate concentration, and Us is the predicted value of the phosphate concentration by the physical information neural network.

4. The method according to claim 3, characterized in that The step of constructing a two-dimensional phosphate transport model for a reservoir based on the reservoir area information and the neural network output function comprises: Determining the water body adsorption and desorption rate based on the temperature, the pH value, the content information and the form information of the phosphorus element; The water flow direction parameter, the vertical flow parameter, the phosphate concentration, the water body adsorption and desorption rate, the water flow direction coordinate parameter, the vertical flow coordinate parameter, the water flow direction parameter and the vertical flow parameter are input into Formula 2 to construct a two-dimensional phosphate transport model of the reservoir. Formula 2 is: Among them, V x is the water flow direction parameter, V y is the vertical flow parameter, R is the water adsorption and desorption rate, D x is the water flow direction parameter, D y is the vertical flow parameter.

5. The method according to claim 4, characterized in that The step of constructing a residual function through the reservoir two-dimensional phosphate transport model comprises: The predicted value, the detection time, the water flow direction coordinate parameter, the vertical flow coordinate parameter, the initial weight, the water flow direction parameter, the vertical flow parameter, the water body adsorption and desorption rate 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 multiple hidden layers.

8. A reservoir dispatching device, characterized in that: include: A reservoir area information collection module is used to collect reservoir area information of the target reservoir; 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 phosphate concentration through the physical information neural network based on the reservoir area information; A reservoir two-dimensional phosphate transport model construction module, used to construct a reservoir two-dimensional phosphate transport model based on the reservoir area information and the neural network output function; The residual function framework is modular and used to construct the residual function through the two-dimensional phosphate transport model of the reservoir; The target weight determination module is used to determine the target weight, adjust the residual function based on the target weight, and generate a scheduling plan 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