Method and device for determining two-fluid parameters of nuclear reactor, medium and equipment

Abstract

The invention discloses a method and a device for determining two fluid parameters of a nuclear reactor, a medium and equipment. The method comprises the following steps: determining a preset constraint condition and a two-fluid control equation of a nuclear reactor according to a working condition variable of the nuclear reactor; constructing a target loss function according to a preset constraint condition and the two-fluid control equation; iteratively training by minimizing the target loss function to obtain a parameter prediction model; and inputting the target working condition parameters into the parameter prediction model to obtain characteristic parameters of the two-phase flow behavior of the nuclear reactor under the target working condition parameters. According to the method provided by the invention, the control equation is embedded into the network model, so that the complexity of a problem solution space is remarkably reduced, the requirement on training sample data is reduced, the quick solution of the two-fluid equation of the nuclear reactor in a complex and changeable environment is realized, and meanwhile, the prediction reliability and generalization ability of two-phase characteristic parameters are considered.

Description

Technical Field

[0001] The present application relates to the field of nuclear reactor engineering and computer science technology, and in particular to a method, device, medium and equipment for determining parameters of two fluids in a nuclear reactor. Background Art

[0002] Gas-liquid two-phase flow phenomenon is widely present in various industrial equipment of nuclear power plants. Research on gas-liquid two-phase flow phenomenon is helpful to improve the safety and reliability of evaluation systems and equipment. At present, in the process of nuclear reactor system safety analysis, the simulation and analysis of gas-liquid two-phase flow behavior characteristics are mainly achieved through nuclear reactor system analysis programs (such as RELAP5, etc.). For complex accident conditions such as large LOCA of multi-loop pressurized water reactor systems, when using nuclear reactor system analysis programs for modeling and analysis, due to the division of a large number of control nodes, a large sparse matrix is ​​usually formed for overall calculation. However, this method will reduce the calculation speed of the program, thereby affecting the progress of design analysis. Summary of the invention

[0003] In view of this, the present application provides a method, device, storage medium and computer equipment for determining the two-fluid parameters of a nuclear reactor, which can reduce the consumption of computing resources and realize the intelligent solution of the two-fluid equations of a nuclear reactor system with fast calculation.

[0004] According to one aspect of the present application, a method for determining parameters of two fluids in a nuclear reactor is provided, comprising:

[0005] Determining preset constraint conditions and two-fluid control equations of the nuclear reactor according to the operating variables of the nuclear reactor;

[0006] Constructing a target loss function according to the preset constraint conditions and the two fluid control equations;

[0007] Iterative training is performed by minimizing the objective loss function to obtain a parameter prediction model;

[0008] The target operating condition parameters are input into the parameter prediction model to obtain characteristic parameters of the two-phase flow behavior of the nuclear reactor under the target operating condition parameters.

[0009] Optionally, constructing a target loss function according to preset constraints of the nuclear reactor and two-fluid control equations includes:

[0010] Establishing the preset constraint conditions and the loss functions of the two fluid control equations respectively;

[0011] Performing weighted summation on the loss function of the preset constraint condition and the loss function of the two fluid control equations to determine the target loss function;

[0012] The weight value corresponding to the loss function of the momentum conservation equation in the two fluid control equations is smaller than the weight value corresponding to the loss function of the mass conservation equation or the energy conservation equation in the two fluid control equations; the target loss function is encoded in the parameter prediction model using the physical information neural network framework.

[0013] Optionally, the establishing of the preset constraint conditions and the loss functions of the two fluid control equations respectively includes:

[0014] Determining the preset constraint conditions and the residual terms of the two fluid control equations by an automatic differentiation algorithm;

[0015] Taking the mean square error as the target, a loss function of the preset constraint condition is constructed according to the residual term of the preset constraint condition, and a loss function of the two fluid control equation is constructed according to the residual term of the two fluid control equation.

[0016] Optionally, the operating condition variables include the operating time of the nuclear reactor and / or the spatial coordinates of a one-dimensional flow conduit in the nuclear reactor.

[0017] Optionally, the flow characteristic parameters include at least one of the following: flow velocity of the two fluids, pressure exerted on the two fluids, gravitational acceleration of the two fluids, angle between the flow channel and the horizontal position, flow resistance of the two fluids, internal energy, heat source energy, and density of the mixture of the two fluids.

[0018] Optionally, the preset constraints include initial conditions and boundary conditions, the initial conditions include constant internal energy and a cavitation fraction of 0, and the boundary conditions include a constant inlet pressure of a one-dimensional flow pipe in the nuclear reactor and a constant inlet velocity of a one-dimensional flow pipe in the nuclear reactor.

[0019] Optionally, the two fluid control equations include: a mass conservation equation, a momentum conservation equation and / or an energy conservation equation;

[0020] The mass conservation equation is expressed as:

[0021] The momentum conservation equation is expressed as:

[0022] The energy conservation equation is expressed as:

[0023] Wherein, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ mis the density of the mixture of the two fluids.

[0024] Optionally, the parameter prediction model adopts a physical information neural network framework, and the target loss function is encoded in the physical information neural network framework.

[0025] Optionally, the residual term of the mass conservation equation is expressed as:

[0026]

[0027] The loss function of the mass conservation equation is expressed as:

[0028]

[0029] The residual term of the momentum conservation equation is expressed as:

[0030]

[0031] The loss function of the momentum conservation equation is expressed as:

[0032]

[0033] The residual term of the energy conservation equation is expressed as:

[0034]

[0035] The loss function of the energy conservation equation is expressed as:

[0036]

[0037] Wherein, Φ is the neural network parameter, L1, L2, L3 are the loss functions of the control equations of the two fluids, f1, f2, f3 are the residual terms of the control equations of the two fluids, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids, N b is the number of samples at the boundary points; N0 is the number of samples at the initial points; N f is the number of midpoint samples.

[0038] Optionally, the residual term of the initial condition is expressed as:

[0039] f4(t,z;Φ)=U(z)-U0;

[0040] f5(t,z;Φ)=α g (z);

[0041] The loss function of the initial condition is expressed as:

[0042]

[0043] The residual term of the boundary condition is expressed as:

[0044] f6(t,z;Φ)=P(t)-P0;

[0045] f7(t,z;Φ)=v(t)-v0;

[0046] The loss function of the boundary condition is expressed as:

[0047]

[0048] Wherein, L4 and L5 are the loss functions of the preset constraint conditions, f4 and f5 are the residual terms of the initial conditions, f6 and f7 are the residual terms of the boundary conditions, t is the operation time of the nuclear reactor, z is the spatial coordinate of the one-dimensional flow pipe in the nuclear reactor, U0 is the constant value of internal energy, v0 is the constant value of flow velocity, p0 is the constant value of pressure, N b is the number of samples at the boundary points; N0 is the number of samples at the initial points; N f is the number of midpoint samples.

[0049] According to another aspect of the present application, a device for determining parameters of two fluids in a nuclear reactor is provided, comprising:

[0050] A construction module is used to determine the preset constraint conditions and two-fluid control equations of the nuclear reactor according to the operating variables of the nuclear reactor; and

[0051] Constructing a target loss function according to the preset constraint conditions and the two fluid control equations;

[0052] A training module, used for iterative training by minimizing the target loss function to obtain a parameter prediction model;

[0053] The prediction module is used to input the target operating parameters into the parameter prediction model to obtain the characteristic parameters of the two-phase flow behavior of the nuclear reactor under the target operating parameters.

[0054] Optionally, the construction module is specifically used to establish the preset constraint conditions and the loss function of the two fluid control equations respectively; perform weighted summation on the loss function of the preset constraint conditions and the loss function of the two fluid control equations to determine the target loss function; wherein the weight value corresponding to the loss function of the momentum conservation equation in the two fluid control equations is less than the weight value corresponding to the loss function of the mass conservation equation or the energy conservation equation in the two fluid control equations; the target loss function is encoded in the parameter prediction model using the physical information neural network framework

[0055] Optionally, the construction module is specifically used to determine the preset constraints and the residual terms of the two-fluid control equations through an automatic differentiation algorithm; taking the mean square error as the target, constructing the loss function of the preset constraints based on the residual terms of the preset constraints, and constructing the loss function of the two-fluid control equations based on the residual terms of the two-fluid control equations.

[0056] Optionally, the parameter prediction model adopts a physical information neural network framework, and the target loss function is encoded in the physical information neural network framework.

[0057] Optionally, the operating condition variables include the operating time of the nuclear reactor and / or the spatial coordinates of a one-dimensional flow conduit in the nuclear reactor.

[0058] Optionally, the flow characteristic parameters include at least one of the following: flow velocity of the two fluids, pressure exerted on the two fluids, gravitational acceleration of the two fluids, angle between the flow channel and the horizontal position, flow resistance of the two fluids, internal energy, heat source energy, and density of the mixture of the two fluids.

[0059] Optionally, the preset constraints include initial conditions and boundary conditions, the initial conditions include constant internal energy and a cavitation fraction of 0, and the boundary conditions include a constant inlet pressure of a one-dimensional flow pipe in the nuclear reactor and a constant inlet velocity of a one-dimensional flow pipe in the nuclear reactor.

[0060] Optionally, the two fluid control equations include: a mass conservation equation, a momentum conservation equation and / or an energy conservation equation;

[0061] The mass conservation equation is expressed as:

[0062] The momentum conservation equation is expressed as:

[0063] The energy conservation equation is expressed as:

[0064] Wherein, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids.

[0065] Optionally, the residual term of the mass conservation equation is expressed as:

[0066]

[0067] The loss function of the mass conservation equation is expressed as:

[0068]

[0069] The residual term of the momentum conservation equation is expressed as:

[0070]

[0071] The loss function of the momentum conservation equation is expressed as:

[0072]

[0073] The residual term of the energy conservation equation is expressed as:

[0074]

[0075] The loss function of the energy conservation equation is expressed as:

[0076]

[0077] Wherein, Φ is the neural network parameter, L1, L2, L3 are the loss functions of the control equations of the two fluids, f1, f2, f3 are the residual terms of the control equations of the two fluids, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids, N b is the number of samples at the boundary points; N0 is the number of samples at the initial points; N f is the number of midpoint samples.

[0078] Optionally, the residual term of the initial condition is expressed as:

[0079] f4(t,z;Φ)=U(z)-U0;

[0080] f5(t,z;Φ)=α g (z);

[0081] The loss function of the initial condition is expressed as:

[0082]

[0083] The residual term of the boundary condition is expressed as:

[0084] f6(t,z;Φ)=P(t)-P0;

[0085] f7(t,z;Φ)=v(t)-v0;

[0086] The loss function of the boundary condition is expressed as:

[0087]

[0088] Wherein, L4 and L5 are the loss functions of the preset constraint conditions, f4 and f5 are the residual terms of the initial conditions, f6 and f7 are the residual terms of the boundary conditions, t is the operation time of the nuclear reactor, z is the spatial coordinate of the one-dimensional flow pipe in the nuclear reactor, U0 is the constant value of internal energy, v0 is the constant value of flow velocity, p0 is the constant value of pressure, N b is the number of samples at the boundary points; N0 is the number of samples at the initial points; N f is the number of midpoint samples.

[0089] According to another aspect of the present application, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method for determining the two fluid parameters of the nuclear reactor are implemented.

[0090] According to another aspect of the present application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, and when the processor executes the program, the steps of the method for determining the two fluid parameters of the nuclear reactor are implemented.

[0091] By means of the above technical scheme, the preset constraints and two-fluid control equations related to the gas-liquid two-phase flow and heat transfer process of the nuclear reactor are obtained, and the target loss function is established accordingly. Furthermore, the parameter prediction model is trained by machine learning with the target loss function related to the gas-liquid two-phase flow and heat transfer process, and the parameter prediction model is used to solve the characteristic parameters of the two-phase flow behavior related to the two-phase flow and heat transfer process of the nuclear reactor system. On the one hand, embedding the control equation into the network model significantly reduces the complexity of the problem solution space, which helps to reduce the demand for training sample data. On the other hand, combining different control algorithms based on the neural network model can effectively solve the problem of weak generalization ability in a single model or homogeneous integrated model, which not only realizes the rapid solution of the two-fluid equations of the nuclear reactor in a complex and changeable environment, but also takes into account the prediction reliability and generalization ability of the two-phase characteristic parameters.

[0092] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0094] Figure 1 A schematic flow chart of a method for determining parameters of two fluids in a nuclear reactor provided in an embodiment of the present application is shown;

[0095] Figure 2 A schematic diagram of a one-dimensional flow pipeline provided by an embodiment of the present application is shown;

[0096] Figure 3 A structural block diagram of a device for determining parameters of two fluids in a nuclear reactor provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0097] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0098] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application, and cannot be interpreted as limiting the present application.

[0099] Those skilled in the art will appreciate that, unless expressly stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present application refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, or there may be intermediate elements. In addition, the "connection" or "connection" used herein may include wireless connection or wireless fusion. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0100] Now, exemplary embodiments according to the present application will be described in more detail with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in a variety of different forms and should not be interpreted as being limited to the embodiments set forth herein. It should be understood that these embodiments are provided to make the disclosure of the present application thorough and complete, and to fully convey the concepts of these exemplary embodiments to those of ordinary skill in the art.

[0101] In this embodiment, a method for determining parameters of two fluids in a nuclear reactor is provided, such as Figure 1 As shown, the method includes:

[0102] Step 101, determining preset constraints of the nuclear reactor and two-fluid control equations according to the operating variables of the nuclear reactor;

[0103] The operating variables are specific simulation conditions of the nuclear reactor system under the conditions of simulating high-pressure and high-velocity bubbling flow, including the spatial coordinates of the nuclear reactor's one-dimensional flow pipe, the simulation running time of the nuclear reactor, the constant temperature and steady flow of incompressible fluid, etc. Figure 2 As shown, with a one-dimensional circular flow pipe with a length of R, a constant temperature steady flow of an incompressible fluid, and a calculation time of 100s as simulation conditions, the three-equation homogeneous flow theoretical model of the nuclear reactor system is solved when two fluids flow through a one-dimensional flow pipe under high pressure and high flow rate bubbly flow conditions.

[0104] In this embodiment, the operating conditions of the reactor and the key factors affecting the fluid behavior and heat exchange process can be matched through the operating variables of the nuclear reactor, and the relevant preset constraints and two-fluid control equations can be screened out based on this. This can improve the simulation accuracy of the fluid flow and heat transfer process, thereby better predicting the behavior of the reactor and ensuring stability and safety under various operating conditions.

[0105] It is understandable that the correspondence between different operating condition variables, preset constraints and two-fluid control equations can be established in advance, and the system can automatically match the relevant preset constraints and two-fluid control equations through the correspondence.

[0106] In actual application scenarios, the preset constraints are the solution constraints of the two-fluid control equations, which include initial conditions and boundary conditions, and can be reasonably set according to the needs of the nuclear reactor structure.

[0107] For example, for the constant temperature and steady flow of an incompressible fluid in a one-dimensional pipe, the initial conditions include: t = 0, U(z) = U0, that is, at the initial moment, no matter where the calculation is performed in the one-dimensional flow pipe, the internal energy U is a constant value of U0; t = 0, α g (z) = 0, that is, at the initial moment, no matter where the calculation is performed in the one-dimensional flow channel, the cavitation fraction αg is 0. The boundary conditions include: z = 0, P(t) = P0, that is, at the inlet of the one-dimensional flow channel in the nuclear reactor, the pressure P is a constant pressure value P0 at any time point; z = 0, v(t) = v0, that is, at the inlet of the one-dimensional flow channel in the nuclear reactor, the velocity v is a constant velocity force value v0 at any time point; where t is the operating time of the nuclear reactor, and z is the spatial coordinate of the one-dimensional flow channel in the nuclear reactor.

[0108] In practical application scenarios, the two-fluid control equations are mathematical models that describe the interaction, flow, and interface behavior of two different fluids (e.g., gas and liquid). The two-fluid control equations include mass conservation equations, momentum conservation equations, and / or energy conservation equations. Mass conservation equation: describes how the mass of each fluid changes over time. Momentum conservation equation: describes how the momentum of the fluid changes as it flows and interacts. Energy conservation equation: describes the energy transfer and conversion process of the fluid.

[0109] The mass conservation equation is expressed as:

[0110] The momentum conservation equation is expressed as:

[0111] The energy conservation equation is expressed as:

[0112] In the formula, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids.

[0113] Step 102, constructing a target loss function according to preset constraints and two fluid control equations;

[0114] In this embodiment, by constructing the target loss function, multiple constraints and two-fluid control equations can be integrated together, and multiple factors such as fluid dynamics and thermodynamics can be comprehensively considered to form an overall optimization framework. This optimizes the deviation and error of model training, so as to improve the accuracy of model prediction by adjusting model parameters or optimizing design, thereby realizing the rapid solution of the two-fluid equations of nuclear reactors in a complex and changeable environment, while taking into account the prediction reliability and generalization ability of the two-phase characteristic parameters.

[0115] In one embodiment, step 102, i.e., constructing a target loss function based on preset constraints and the two-fluid control equations, specifically includes the following steps:

[0116] Step 102-1, respectively establishing preset constraint conditions and loss functions of the two fluid control equations;

[0117] Step 102-2, performing weighted summation on the loss function of the preset constraint condition and the loss function of the two fluid control equations to determine the target loss function.

[0118] In this embodiment, different weight values ​​are set for the loss functions corresponding to different items to weight them into a unified target loss function, so that the weights of each loss can be flexibly adjusted according to the specific application scenarios and needs, which helps to achieve multi-objective optimization and reduces the complexity of the solution process, so as to quickly find a comprehensive optimal solution, thereby improving the balance and training efficiency of subsequent parameter prediction model training.

[0119] Exemplarily, the one or two fluid control equations include: mass conservation equation, momentum conservation equation and energy conservation equation, and the preset constraint conditions include initial conditions and boundary conditions as an example. The target loss function is expressed as:

[0120] L(Φ)=ψ1L1(Φ)+ψ2L2(Φ)+ψ3L3(Φ)+ψ4L4(Φ)+ψ5L5(Φ);

[0121] Wherein, L is the target loss function, ψ1, ψ2, ψ3, ψ4, ψ5 are the weight values ​​of the mass conservation equation, momentum conservation equation, energy conservation equation, initial conditions and boundary conditions respectively, Φ is the neural network parameter, L1, L2, L3 are the loss functions of the two fluid control equations (mass, momentum, energy conservation equations), L4, L5 are the loss functions of the preset constraints (initial and boundary conditions).

[0122] Furthermore, the adaptive weight method can be used to balance the various loss functions, and the weight values ​​corresponding to the loss functions of the preset constraints and the loss functions of the two fluid control equations can be adaptively adjusted. Specifically, the initialization weight values ​​of the multiple loss functions that need to be optimized are determined, and in each optimization iteration, the weights are adjusted according to the gradient information of the loss function, or according to the rule that the weights are inversely proportional to the value of the loss function. It should be noted that in order to ensure that the weight values ​​of each item remain in a reasonable proportion during the optimization process, all weights can be normalized so that the sum of the weight values ​​of all the functions shown is 1.

[0123] It is worth mentioning that the weight value corresponding to the loss function of the momentum conservation equation in the two-fluid control equation is smaller than the weight value corresponding to the loss function of the mass conservation equation or the energy conservation equation in the two-fluid control equation; the target loss function is encoded in the parameter prediction model using the physical information neural network framework, thereby reducing the sensitivity to the momentum indicator with greater volatility when training the parameter prediction model, and reducing its impact on the overall target loss function, thereby enhancing the robustness of the parameter prediction model to uncertainty or disturbance.

[0124] In one embodiment, step 102-1, respectively establishes preset constraints and loss functions of the two-fluid control equations, including: determining the preset constraints and the residual terms of the two-fluid control equations through an automatic differentiation algorithm; taking the mean square error as the target, constructing the loss function of the preset constraints based on the residual terms of the preset constraints, and constructing the loss function of the two-fluid control equations based on the residual terms of the two-fluid control equations.

[0125] In this embodiment, automatic differentiation is used for calculation, which avoids the calculation error caused by the discrete equation in the solution process of numerical differentiation, improves the calculation accuracy, makes the value after gradient calculation more accurate, and then quickly updates the neural network model parameters. At the same time, by constructing the residual term, the deviation of the preset constraints and the two-fluid control equation under specific input can be accurately captured, and the deviation of the model can be quantified and adjusted, thereby improving the optimization accuracy in the model training process and improving the robustness of the model in the face of input uncertainty.

[0126] In practical application scenarios, the residual term of the mass conservation equation is expressed as:

[0127]

[0128] The residual term of the momentum conservation equation is expressed as:

[0129]

[0130] The residual term of the energy conservation equation is expressed as:

[0131]

[0132] Among them, the residual term of the two-fluid control equation is the quadratic gradient differential of the two-fluid control equation. The residual term of the initial condition is expressed as:

[0133] f4(t,z;Φ)=U(z)-U0;

[0134] f5(t,z;Φ)=α g (z);

[0135] The residual term of the boundary condition is expressed as:

[0136] f6(t,z;Φ)=P(t)-P0;

[0137] f7(t,z;Φ)=v(t)-v0;

[0138] The loss function of the mass conservation equation is expressed as:

[0139]

[0140] The loss function of the momentum conservation equation is expressed as:

[0141]

[0142] The loss function of the energy conservation equation is expressed as:

[0143]

[0144] The loss function of the initial condition is expressed as:

[0145]

[0146] The loss function of the boundary condition is expressed as:

[0147]

[0148] Wherein, Φ is the neural network parameter, L1, L2, L3 are the loss functions of the two fluid control equations, L4, L5 are the loss functions of the preset constraint conditions, f1, f2, f3 are the residual terms of the two fluid control equations, f4, f5 are the residual terms of the initial conditions, f6, f7 are the residual terms of the boundary conditions, t is the operation time of the nuclear reactor, z is the spatial coordinate of the one-dimensional flow pipe in the nuclear reactor, t is the operation time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ mis the density of the mixture of the two fluids, U0 is the constant value of internal energy, v0 is the constant value of flow velocity, p0 is the constant value of pressure, N b is the number of samples at the boundary points, N0 is the number of samples at the initial points, N f is the number of midpoint samples.

[0149] The mixture density can be expressed as: m =α g ρ g +(1-α g )ρ l , gas and liquid phase density ρ g , ρl can be expressed as a function of pressure and temperature (or internal energy), which can be obtained by looking up the physical property table. The boundary point samples, initial point samples, and intermediate point samples are used as sample data to train the parameter prediction model. Different operating condition parameters and corresponding historical characteristic parameters are obtained based on the operating condition variables according to the experiment or simulation program as sample data.

[0150] Step 103, performing iterative training by minimizing the target loss function to obtain a parameter prediction model;

[0151] Among them, the parameter prediction model can be trained based on the Physical Informed Neural Networks (PINNs) framework. The Physical Informed Neural Networks framework helps to encode specific physical equations into the network model without the need for grid generation, avoiding dimensionality explosion in complex scenarios, and significantly reducing the complexity of the problem solution space, thereby greatly reducing the demand for training data.

[0152] In this embodiment, sample data including boundary point samples, initial point samples, and intermediate point samples are input into a preset neural network model based on a physical information neural network architecture for preliminary training to establish a parameter prediction model. The optimization algorithm is used to obtain the optimal loss function weight and neuron number bias hyperparameter until the target loss function approaches the minimum value to complete the optimization of the parameter prediction model. This improves the accuracy and computational efficiency of the parameter prediction model while ensuring low computational and memory overhead.

[0153] For example, in the time-space sample domain, that is, the space domain [0, R] and the time domain [0, 100], select the corresponding training set data, N b 、N0、N f They are 100, 120, and 8000 respectively, and the learning rate is set to 0.001 to train the neural network.

[0154] Step 104, inputting the target operating parameters into the parameter prediction model to obtain characteristic parameters of the two-phase flow behavior of the nuclear reactor under the target operating parameters.

[0155] Among them, the characteristic parameters of the two-phase flow behavior include multiple combinations of parameters such as pressure, phase internal energy, temperature, steam cavitation fraction, non-condensable gas content, boron concentration, etc., which are not listed one by one in this embodiment.

[0156] The method for determining the two-fluid parameters of a nuclear reactor provided in an embodiment of the present application obtains preset constraints and two-fluid control equations related to the gas-liquid two-phase flow and heat transfer process of the nuclear reactor, and establishes a target loss function accordingly. Furthermore, a parameter prediction model is trained through machine learning with the target loss function related to the gas-liquid two-phase flow and heat transfer process, and the parameter prediction model is used to solve the characteristic parameters of the two-phase flow behavior related to the two-phase flow and heat transfer process of the nuclear reactor system. On the one hand, embedding the control equation into the network model significantly reduces the complexity of the problem solution space, which helps to reduce the demand for training sample data. On the other hand, combining different control algorithms based on the neural network model can effectively solve the problem of weak generalization ability in a single model or homogeneous integrated model, which not only realizes the rapid solution of the two-fluid equations of the nuclear reactor in a complex and changeable environment, but also takes into account the prediction reliability and generalization ability of the two-phase characteristic parameters.

[0157] The method for determining the parameters of the two fluids of a nuclear reactor provided in the embodiment of the present application can be applied to a terminal, can be applied to a server, and can also be software running in a terminal or a server. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the method for determining the parameters of the two fluids of a nuclear reactor, etc., but is not limited to the above forms.

[0158] It should be noted that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0159] Furthermore, if Figure 3 As shown, as a specific implementation of the above-mentioned method for determining the parameters of two fluids in a nuclear reactor, an embodiment of the present application provides a device 300 for determining the parameters of two fluids in a nuclear reactor. The device 300 for determining the parameters of two fluids in a nuclear reactor includes: a construction module 301, a training module 302 and a prediction module 303.

[0160] The construction module 301 is used to determine the preset constraints and two-fluid control equations of the nuclear reactor according to the operating variables of the nuclear reactor; and to construct the target loss function according to the preset constraints and the two-fluid control equations;

[0161] A training module 302 is used to perform iterative training by minimizing a target loss function to obtain a parameter prediction model;

[0162] The prediction module 303 is used to input the target operating parameters into the parameter prediction model to obtain the characteristic parameters of the two-phase flow behavior of the nuclear reactor under the target operating parameters.

[0163] In this embodiment, the preset constraints and two-fluid control equations related to the gas-liquid two-phase flow and heat transfer process of the nuclear reactor are obtained, and the target loss function is established accordingly. Further, the target loss function related to the gas-liquid two-phase flow and heat transfer process is used to train the parameter prediction model through machine learning, and the parameter prediction model is used to solve the characteristic parameters of the two-phase flow behavior related to the two-phase flow and heat transfer process of the nuclear reactor system. On the one hand, embedding the control equation into the network model significantly reduces the complexity of the problem solution space, which helps to reduce the demand for training sample data. On the other hand, combining different control algorithms based on the neural network model can effectively solve the problem of weak generalization ability in a single model or homogeneous integrated model, which not only realizes the rapid solution of the two-fluid equations of the nuclear reactor in a complex and changeable environment, but also takes into account the prediction reliability and generalization ability of the two-phase characteristic parameters.

[0164] Further, the construction module 301 is specifically used to establish the preset constraint conditions and the loss function of the two fluid control equations respectively; the loss function of the preset constraint conditions and the loss function of the two fluid control equations are weighted summed to determine the target loss function; wherein the weight value corresponding to the loss function of the momentum conservation equation in the two fluid control equations is less than the weight value corresponding to the loss function of the mass conservation equation or the energy conservation equation in the two fluid control equations; the target loss function is encoded in the parameter prediction model using the physical information neural network framework

[0165] Furthermore, a construction module 301 is specifically used to determine the preset constraints and the residual terms of the two-fluid control equations through an automatic differentiation algorithm; taking the mean square error as the target, a loss function of the preset constraints is constructed based on the residual terms of the preset constraints, and a loss function of the two-fluid control equations is constructed based on the residual terms of the two-fluid control equations.

[0166] Furthermore, the parameter prediction model adopts a physical information neural network framework, and the target loss function is encoded in the physical information neural network framework.

[0167] Further, the operating condition variables include the operating time of the nuclear reactor and / or the spatial coordinates of a one-dimensional flow channel in the nuclear reactor.

[0168] Furthermore, the flow characteristic parameters include at least one of the following: flow velocity of the two fluids, pressure exerted on the two fluids, gravitational acceleration of the two fluids, angle between the flow channel and the horizontal position, flow resistance of the two fluids, internal energy, heat source energy, and density of the mixture of the two fluids.

[0169] Furthermore, the preset constraints include initial conditions and boundary conditions. The initial conditions include constant internal energy and a cavitation fraction of 0. The boundary conditions include constant inlet pressure and constant inlet velocity of a one-dimensional flow pipe in a nuclear reactor.

[0170] Further, the two fluid control equations include: a mass conservation equation, a momentum conservation equation and / or an energy conservation equation;

[0171] The mass conservation equation is expressed as:

[0172] The momentum conservation equation is expressed as:

[0173] The energy conservation equation is expressed as:

[0174] In the formula, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids.

[0175] Furthermore, the residual term of the mass conservation equation is expressed as:

[0176]

[0177] The loss function of the mass conservation equation is expressed as:

[0178]

[0179]

[0180]

[0181] The loss function of the momentum conservation equation is expressed as:

[0182]

[0183] The residual term of the energy conservation equation is expressed as:

[0184]

[0185] The loss function of the energy conservation equation is expressed as:

[0186]

[0187] Where Φ is the neural network parameter, L1, L2, L3 are the loss functions of the two fluid control equations, f1, f2, f3 are the residual terms of the two fluid control equations, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids, N b is the number of samples at the boundary points; N0 is the number of samples at the initial points; N f is the number of midpoint samples.

[0188] Furthermore, the residual term of the initial condition is expressed as:

[0189] f4(t,z;Φ)=U(z)-U0;

[0190] f5(t,z;Φ)=α g (z);

[0191] The loss function of the initial condition is expressed as:

[0192]

[0193] The residual term of the boundary condition is expressed as:

[0194] f6(t,z;Φ)=P(t)-P0;

[0195] f7(t,z;Φ)=v(t)-v0;

[0196] The loss function of the boundary condition is expressed as:

[0197]

[0198] Where L4 and L5 are loss functions of preset constraints, f4 and f5 are residual terms of initial conditions, f6 and f7 are residual terms of boundary conditions, t is the operation time of the nuclear reactor, z is the spatial coordinate of the one-dimensional flow pipe in the nuclear reactor, U0 is the constant value of internal energy, v0 is the constant value of flow velocity, p0 is the constant value of pressure, and N b is the number of samples at the boundary points; N0 is the number of samples at the initial points; N f is the number of midpoint samples.

[0199] The specific definition of the device for determining the parameters of the two fluids of the nuclear reactor can be found in the definition of the method for determining the parameters of the two fluids of the nuclear reactor mentioned above, which will not be repeated here. Each module in the above-mentioned device for determining the parameters of the two fluids of the nuclear reactor can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0200] Based on the above Figure 1 The method shown in the embodiment of the present application accordingly provides a readable storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned Figure 1 The method for determining the two fluid parameters of a nuclear reactor is shown.

[0201] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0202] Based on the above Figure 1 The method shown, and Figure 3 In order to achieve the above-mentioned purpose, the embodiment of the present application further provides a computer device, which can be a personal computer, a server, a network device, etc. The computer device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figure 1 The method for determining the two fluid parameters of a nuclear reactor is shown.

[0203] Optionally, the computer device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a Bluetooth interface, a WI-FI interface), etc.

[0204] Those skilled in the art will appreciate that the computer device structure provided in this embodiment does not limit the computer device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0205] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages and saves the hardware and software resources of the computer device, and supports the operation of information processing programs and other software and / or programs. The network communication module is used to realize communication between the components inside the storage medium, and communication with other hardware and software in the physical device.

[0206] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform, or by hardware.

[0207] Those skilled in the art will appreciate that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily necessary for implementing the present application. Those skilled in the art will appreciate that the modules in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the description of the implementation scenario, or can be changed accordingly and located in one or more devices different from the present implementation scenario. The modules of the above-mentioned implementation scenario can be combined into one module, or can be further split into multiple submodules.

[0208] The above serial numbers of this application are only for description and do not represent the advantages and disadvantages of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of this application, but this application is not limited to them, and any changes that can be thought of by technicians in this field should fall within the scope of protection of this application.

Claims

1. A method for determining parameters of two fluids in a nuclear reactor, characterized in that: The method comprises: Determining preset constraint conditions and two-fluid control equations of the nuclear reactor according to the operating variables of the nuclear reactor; Constructing a target loss function according to the preset constraint conditions and the two fluid control equations; Iterative training is performed by minimizing the objective loss function to obtain a parameter prediction model; The target operating condition parameters are input into the parameter prediction model to obtain characteristic parameters of the two-phase flow behavior of the nuclear reactor under the target operating condition parameters.

2. The method for determining the parameters of two fluids in a nuclear reactor according to claim 1, characterized in that: The target loss function is constructed according to the preset constraints of the nuclear reactor and the two-fluid control equation, including: Establishing the preset constraint conditions and the loss functions of the two fluid control equations respectively; Performing weighted summation on the loss function of the preset constraint condition and the loss function of the two fluid control equations to determine the target loss function; Among them, the weight value corresponding to the loss function of the momentum conservation equation in the two fluid control equations is smaller than the weight value corresponding to the loss function of the mass conservation equation or the energy conservation equation in the two fluid control equations.

3. The method for determining the parameters of two fluids in a nuclear reactor according to claim 2, characterized in that: The loss functions of the preset constraint conditions and the two fluid control equations are established respectively, including: Determining the preset constraint conditions and the residual terms of the two fluid control equations by an automatic differentiation algorithm; Taking the mean square error as the target, a loss function of the preset constraint condition is constructed according to the residual term of the preset constraint condition, and a loss function of the two fluid control equation is constructed according to the residual term of the two fluid control equation.

4. The method for determining the parameters of two fluids in a nuclear reactor according to any one of claims 1 to 3, characterized in that: The operating variables include the operating time of the nuclear reactor and / or the spatial coordinates of a one-dimensional flow conduit in the nuclear reactor; The flow characteristic parameters include at least one of the following: flow velocity of the two fluids, pressure of the two fluids, gravitational acceleration of the two fluids, angle between the flow channel and the horizontal position, flow resistance of the two fluids, internal energy, heat source energy, and density of the mixture of the two fluids; The preset constraint conditions include initial conditions and boundary conditions, wherein the initial conditions include constant internal energy and cavitation fraction being 0, and the boundary conditions include constant inlet pressure of the one-dimensional flow pipe in the nuclear reactor and constant inlet velocity of the one-dimensional flow pipe in the nuclear reactor; The two fluid control equations include: a mass conservation equation, a momentum conservation equation and / or an energy conservation equation; The mass conservation equation is expressed as: The momentum conservation equation is expressed as: The energy conservation equation is expressed as: Wherein, t is the running time, z is the spatial coordinate, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, and F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids.

5. The method for determining the parameters of two fluids in a nuclear reactor according to any one of claims 1 to 3, characterized in that: The parameter prediction model adopts a physical information neural network framework, and the target loss function is encoded in the physical information neural network framework.

6. The method for determining the parameters of two fluids in a nuclear reactor according to claim 3, characterized in that: The residual term of the mass conservation equation is expressed as: The loss function of the mass conservation equation is expressed as: The residual term of the momentum conservation equation is expressed as: The loss function of the momentum conservation equation is expressed as: The residual term of the energy conservation equation is expressed as: The loss function of the energy conservation equation is expressed as: Wherein, Φ is the neural network parameter, L1, L2, L3 are the loss functions of the two fluid control equations, f1, f2, f3 are the residual terms of the two fluid control equations, t is the operation time of the nuclear reactor, z is the spatial coordinate of the one-dimensional flow pipe in the nuclear reactor, v is the flow velocity of the two fluids, p is the pressure of the two fluids, g is the gravitational acceleration of the two fluids, θ is the angle between the flow channel and the horizontal position, F w is the flow resistance of the two fluids, U is the internal energy, q is the heat source energy, ρ m is the density of the mixture of the two fluids, N b is the number of samples at the boundary points, N0 is the number of samples at the initial points, N f is the number of midpoint samples.

7. The method for determining the parameters of two fluids in a nuclear reactor according to claim 3, characterized in that: The residual term of the initial condition is expressed as: f4(t,z;Φ)=U(z)-U0; f5(t,z;Φ)=α g (from); The loss function of the initial condition is expressed as: The residual term of the boundary condition is expressed as: f6(t,z;Φ)=P(t)-P0; f7(t,z;Φ)=v(t)-v0; The loss function of the boundary condition is expressed as: Wherein, L4 and L5 are the loss functions of the preset constraints, f4 and f5 are the residual terms of the initial conditions in the preset constraints, f6 and f7 are the residual terms of the boundary conditions in the preset constraints, t is the operation time of the nuclear reactor, z is the spatial coordinate of the one-dimensional flow pipe in the nuclear reactor, U0 is the constant value of internal energy, v0 is the constant value of flow velocity, p0 is the constant value of pressure, N b is the number of samples at the boundary points, N0 is the number of samples at the initial points, N f is the number of midpoint samples.

8. A device for determining parameters of two fluids in a nuclear reactor, characterized in that: The device comprises: A construction module is used to determine the preset constraint conditions and two-fluid control equations of the nuclear reactor according to the operating variables of the nuclear reactor; and Constructing a target loss function according to the preset constraint conditions and the two fluid control equations; A training module, used for iterative training by minimizing the target loss function to obtain a parameter prediction model; The prediction module is used to input the target operating parameters into the parameter prediction model to obtain the characteristic parameters of the two-phase flow behavior of the nuclear reactor under the target operating parameters.

9. A readable storage medium having a program or instruction stored thereon, characterized in that: When the program or instruction is executed by a processor, the steps of the method for determining two fluid parameters of a nuclear reactor as described in any one of claims 1 to 7 are implemented.

10. A computer device comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that: When the processor executes the program, the method for determining the two fluid parameters of a nuclear reactor as described in any one of claims 1 to 7 is implemented.

Images