Method and device for predicting state of miscible multiphase fluid, equipment and medium

Abstract

The invention provides a miscible multiphase fluid state prediction method and device, equipment and a medium, and the method comprises the steps: obtaining the state of a fluid system at the current moment, and obtaining at least one physical parameter used for controlling the dynamic evolution behavior of the fluid; according to the Newton's law of motion, calculating initial predicted values of the position and the speed of the fluid particles at the next moment; utilizing the trained multiphase fluid prediction network to predict the position correction of fluid particles and the volume fraction of each phase at the next moment; and obtaining the position and speed of the fluid particle at the next moment according to the position of the fluid particle at the current moment, the initial predicted value of the position and the position correction amount. Physical calculation and multiphase fluid prediction network correction are combined, and the network only needs to learn the deviation from physical basic motion, so that the learning difficulty is reduced, and the simulation process is more stable; and physical parameters are independently used as an input, so that the universality and the flexibility of the model are greatly enhanced.

Description

Technical Field

[0001] This invention relates to the fields of computer graphics and computational physics, and in particular to a method, apparatus, device, and medium for predicting the state of miscible multiphase fluids. Background Technology

[0002] Multiphase fluid phenomena are widespread in nature and industrial production, such as the mixing of two liquids and the diffusion of substances in chemical reactions. Accurate computer simulations of these phenomena are of great significance for scientific research, engineering design, and film and television special effects production. Traditional physics-based simulation methods, such as smoothed particle hydrodynamics (SPH) or the finite element method (FEM), while providing high-precision results, have high computational complexity, especially when dealing with large-scale particle systems or complex interactions, often making it difficult to meet real-time requirements.

[0003] In recent years, with the development of deep learning technology, data-driven physical simulation methods have become a research hotspot. These methods learn physical laws from large amounts of simulation data and build neural network models to replace traditional physical solvers, thereby significantly improving simulation speed. However, existing learning-based methods still face many challenges when dealing with miscible multiphase fluids.

[0004] First, there is the challenge of coupled prediction: the motion of fluid particles (i.e., position changes) and their composition changes (i.e., changes in the volume fraction of each phase, or simply changes in phase fraction) are closely coupled. If the position and phase fraction are directly output by a single network as a single vector, the learning objective of the network will be too complex and it will be difficult to converge to the ideal effect. If two completely independent networks are used to predict separately, the intrinsic relationship between the two will be ignored, and the repeated feature extraction process will waste computational resources.

[0005] Secondly, the generalization ability of physical parameters is poor: the mixing and diffusion process of fluids is significantly affected by physical parameters such as diffusion coefficient and cohesive / repulsive force. Traditional learning methods are usually trained on a fixed set of physical parameters. When the parameters change, the model performance will drop sharply and retraining is required. It lacks generalization ability and flexibility.

[0006] Therefore, there is an urgent need for a new method for predicting the state of multiphase fluids that can be calculated efficiently, accurately capture coupled physical processes, and have good generalization ability for different physical parameters. Summary of the Invention

[0007] To address the technical problems of slow processing speed and high computational load in existing physical methods for prediction, and unsatisfactory performance and poor generalization ability in neural network prediction, this invention provides a method, apparatus, device, and medium for predicting the state of miscible multiphase fluids. The technical solution is as follows:

[0008] On the one hand, a method for predicting the state of a miscible multiphase fluid is provided. This method is implemented by an electronic device and includes: acquiring the state of the fluid system at the current moment, the state including the position, velocity, and volume fraction of each phase of each fluid particle, and acquiring at least one physical parameter for controlling the dynamic evolution behavior of the fluid; calculating preliminary predicted values ​​of the position and velocity of the fluid particles at the next moment according to Newton's laws of motion; processing the preliminary predicted values ​​of the position and velocity of the fluid particles, the volume fraction of each phase at the current moment, and the physical parameter using a trained multiphase fluid prediction network to obtain a position correction amount of the fluid particles and the volume fraction of each phase at the next moment; and obtaining the position and velocity of the fluid particles at the next moment based on the position of the fluid particles at the current moment, the preliminary predicted value of the position, and the position correction amount.

[0009] On the other hand, a state prediction device for miscible multiphase fluids is provided. This device is applied to a state prediction method for miscible multiphase fluids. The device includes: a data acquisition unit for acquiring the position, velocity, and volume fraction of each phase of a fluid particle at the current moment, and acquiring at least one physical parameter for controlling the dynamic evolution behavior of the fluid; a first calculation unit for calculating preliminary predicted values ​​of the position and velocity of the fluid particle at the next moment according to Newton's laws of motion; a second calculation unit for processing the preliminary predicted values ​​of the position and velocity of the fluid particle, the volume fraction of each phase at the current moment, and the physical parameter using a trained multiphase fluid prediction network to obtain a position correction amount for the fluid particle and the volume fraction of each phase at the next moment; and a third calculation unit for obtaining the position and velocity of the fluid particle at the next moment based on the position of the fluid particle at the current moment, the preliminary predicted value of the position, and the position correction amount.

[0010] On the other hand, an electronic device is provided, comprising: a processor; and a memory storing computer-readable instructions, which, when executed by the processor, implement any of the methods described above for predicting the state of miscible multiphase fluids.

[0011] On the other hand, a computer-readable storage medium is provided, wherein program code is stored in the computer-readable storage medium, and the program code can be called by a processor to execute any of the above-described methods for predicting the state of miscible multiphase fluids.

[0012] The beneficial effects of the technical solution provided by the embodiments of the present invention include at least the following: by combining physical calculation and multiphase fluid prediction network correction, the network only needs to learn the deviation from the basic physical motion, which reduces the learning difficulty, makes the simulation process more stable, and can support a larger time step; at the same time, by using physical parameters as a separate input to the multiphase fluid prediction network, the trained single model can be directly applied to fluid systems with different physical properties without retraining, which greatly enhances the versatility and flexibility of the model. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a flowchart of a method for predicting the state of a miscible multiphase fluid provided in an embodiment of the present invention; Figure 2 This is an architecture diagram of a multiphase fluid prediction network provided in an embodiment of the present invention; Figure 3 This is a block diagram of a state prediction device for miscible multiphase fluids provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention; Figure 5 This is a rendering effect diagram of a canyon two-phase fluid scene prediction provided by an embodiment of the present invention. Detailed Implementation

[0015] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0016] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0017] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0018] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0019] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0020] This invention provides a method for predicting the state of miscible multiphase fluids. This method can be implemented by an electronic device, which can be a terminal or a server. Figure 1 The flowchart shown is a method for predicting the state of miscible multiphase fluids. The processing flow of this method may include steps S101 to S104.

[0021] Step S101: Obtain the state of the fluid system at the current moment, including the position, velocity and volume fraction of each phase of each fluid particle, and obtain at least one physical parameter for controlling the dynamic evolution behavior of the fluid.

[0022] Let the current time be t. Since each fluid particle has a position, velocity, and volume fraction of each phase, the state obtained in this step includes the set of positions P of all fluid particles. t Velocity set V t Volume fraction set of each phase VF t Assuming there are M fluid particles in a fluid system, then the set of positions P t The shape is [M,3], and the velocity set is V. t The shape is [M,3], and the volume fraction set of each phase is VF. t The shape is [M,N], where N is the number of phases in the fluid system, and the volume fraction of each particle is an N-dimensional vector whose components sum to 1.

[0023] In one specific embodiment of the present invention, the physical parameters include at least one of intrinsic physical properties for defining fluid constitutive relations and external driving parameters for defining external environmental conditions.

[0024] Physical parameters that define the intrinsic physical properties of fluid constitutive relations can be one or more of the following, or other coefficients: diffusion coefficient, cohesion coefficient, viscosity coefficient, surface tension coefficient, etc. Among these physical parameters, the diffusion coefficient controls the rate of diffusion and mixing between different phases; the cohesion coefficient describes the strength of the attractive or repulsive forces between different phases, affecting whether they tend to mix or separate; the viscosity coefficient affects the flow resistance of the fluid, thus affecting the rate of mixing and convection; and although the surface tension coefficient is more commonly used for immiscible fluids, it may also have an effect on miscible fluids in microscopic or transient processes.

[0025] External driving parameters that define external environmental conditions can be, for example, one or more of the following: boundary condition parameters, external force field parameters other than gravity, or other parameters. Boundary conditions can be, for example, a simple box, a complex canyon terrain, a pipe, etc., while other external force field parameters can be, for example, forces from other directions or lateral wind forces, etc.

[0026] Step S102: Calculate the preliminary predicted values ​​of the fluid particle's position and velocity at the next moment based on Newton's laws of motion. This step involves performing a physical integration of the particle's current position and velocity using Newton's laws of motion to calculate the preliminary predicted position and velocity considering only gravity and inertia. This step provides a stable benchmark that conforms to fundamental physical laws for subsequent neural network predictions, significantly reducing the learning difficulty of the neural network. It allows the network to focus solely on predicting corrections caused by complex local interactions, thereby greatly improving the numerical stability and computational efficiency of the entire simulation process.

[0027] In a specific embodiment of the present invention, taking gravity as an example, step S102 includes: calculating the preliminary predicted values ​​of the position and velocity of the fluid particle at the next moment using a physical integration method, based on the position and velocity of the fluid particle at the current moment and gravity. Let the next moment be t+Δt, where Δt is the time step. Then, the preliminary predicted values ​​of the position and velocity of the fluid particle at the next moment can be calculated using the following formula:

[0028] , , In the formula, i∈{1,2,…,M}, the velocity of each fluid particle is increased by g× This allows us to obtain a preliminary prediction of the velocity of each fluid particle at the next moment, V. t,i It is the set of velocities V t The velocity of the i-th fluid particle in the middle, It is a preliminary prediction of the velocity of the i-th fluid particle at the next moment. It is the set of preliminary predictions of the velocities of all fluid particles at the next moment; similarly, P t,i It is the set of locations Pt The position of the i-th fluid particle. It is a preliminary prediction of the position of the i-th fluid particle at the next moment. It is a set of preliminary predictions of the velocities of all fluid particles at the next moment; in the formula for calculating the position, the velocities of fluid particles within the time step are considered to be uniform, or the gravitational acceleration g can be taken into account for the calculation.

[0029] Step S103: Using the trained multiphase fluid prediction network, process the preliminary predicted values ​​of the position and velocity of the fluid particles, the volume fraction of each phase at the current moment, and the physical parameters to obtain the position correction of the fluid particles and the volume fraction of each phase at the next moment.

[0030] Please see Figure 2 In a specific embodiment of the present invention, the multiphase fluid prediction network includes an embedding layer, a feature splicing module, a shared feature extraction module, a position prediction head, and a phase fraction prediction head. The modules in the multiphase fluid prediction network are described below.

[0031] The embedding layer is used to convert physical parameters into physical feature embedding vectors. It can include one or more fully connected layers. Physical parameters are generally global scalars, with each parameter typically having a dimension of 1. Multiple physical parameters can be concatenated into a small input vector. For example, given two physical parameters, the diffusion coefficient (cd) and the cohesion coefficient (cf), concatenation yields a two-dimensional vector [cd, cf]. This two-dimensional vector is then processed by a fully connected layer to obtain a high-dimensional physical feature embedding vector. The dimension of this vector can be denoted as E, which is a hyperparameter that can be set as needed, such as 32, 64, or 128.

[0032] The feature concatenation module is used to concatenate the preliminary predicted values ​​of the position and velocity of fluid particles, the volume fraction of each phase at the current moment, and the physical feature embedding vector. For the i-th fluid particle, the dimension of its preliminary predicted position is 3; the dimension of its preliminary predicted velocity is also 3; the dimension of its volume fraction is N; and the dimension of its physical feature embedding vector is E. These vector features can be directly concatenated along the feature dimensions to form a long feature vector with a length of 3+3+N+E. After performing this operation on all M fluid particles, the concatenated features of all fluid particles can be obtained, with a shape of [M, (3+3+N+E)].

[0033] The shared feature extraction module aggregates the features of each fluid particle and other fluid particles within its neighborhood to obtain a shared feature vector. The shared feature extraction module takes the features of shape [M, (3+3+N+E)] output by the feature splicing module as input, and after processing, obtains the shared feature vector F.shared The shared feature extraction module comprises multiple layers of network consisting of consecutive convolutional layers and fully connected layers. Each consecutive convolutional layer aggregates features from other particles within its neighborhood radius R, thereby learning complex local fluid interaction patterns. The network layers in the shared feature extraction module also include activation functions and residual connections to enhance the module's learning ability. The high-dimensional shared features F, after multi-layer processing, are then obtained. shared It contains all the local environmental information needed to predict the state at the next moment.

[0034] The position prediction head predicts position corrections based on shared feature vectors. It typically includes convolutional layers and fully connected layers. The position prediction head outputs a three-dimensional position correction ΔP, with a shape of [M,3], corresponding one-to-one with the three dimensions of the fluid particle's position. For example, these three dimensions represent the position corrections in the x, y, and z directions, respectively.

[0035] The phase fraction prediction head predicts the volume fraction of each phase at the next time step based on shared eigenvectors. This is the set of phase volume fractions for all fluid particles at the next time step, with a shape of [M, N]. The phase fraction prediction head includes one or more convolutional layers and a normalization function. Its convolutional layer outputs an N-dimensional original prediction fraction vector, which is then processed by the normalization function to obtain the phase volume fractions at the next time step that satisfy the physical constraints, where N is the number of phases in the fluid system. The normalization function can be, for example, a softmax function, which ensures that the components of the output vector are non-negative and sum to 1, fully conforming to the physical definition of volume fractions and guaranteeing the conservation and stability of the simulation.

[0036] The aforementioned multiphase fluid prediction network avoids redundant calculations for position and phase fraction predictions by sharing a feature extraction network, significantly improving efficiency. At the same time, task-specific prediction heads can focus more on their respective learning objectives, improving prediction accuracy.

[0037] After the above-mentioned multiphase fluid prediction network is constructed, it needs to be trained. The model training process is the same as that of conventional models, and it is generally divided into the following steps: (1) Construct a training sample dataset. The complex physical prediction methods in the existing technology can be used to predict the position, velocity and volume fraction of each fluid particle in the fluid system at different times. The sampling step size is equal to the above-mentioned time step size. For any two adjacent time points, the input features of the multiphase fluid prediction network can be obtained according to the processing of steps S101 and S102. The position of the previous time point is subtracted from the position of the next time point to obtain the output of the position prediction head in the multiphase fluid prediction network. The volume fraction of each phase at the next time point is directly used as the output of the phase fraction prediction head in the multiphase fluid prediction network, thus obtaining a sample. When constructing the training sample dataset, data from multiple scenarios and multiple fluid particle miscibility should be selected. (2) Construct the above multiphase fluid prediction network. (3) Use the constructed training sample dataset to train the multiphase fluid prediction network until the training of the network model is completed.

[0038] Step S104: Based on the position of the fluid particle at the current moment, the preliminary predicted value of the position, and the position correction amount, obtain the position and velocity of the fluid particle at the next moment.

[0039] In a specific embodiment of the present invention, step S104 includes: (1) obtaining the position of the fluid particle at the next moment based on the preliminary predicted value of the fluid particle's position and the position correction amount. , It is the set of positions of all fluid particles at the next moment. For the i-th fluid particle, its position can be calculated using the following formula: ; (2) Based on the position of the fluid particle at the current moment and its position at the next moment, obtain the velocity of the fluid particle at the next moment. , It is the set of velocities of all fluid particles at the next moment. For the i-th fluid particle, its velocity can be calculated using the following formula: .

[0040] In this embodiment, the final position of the particle at the next moment is obtained by adding the position correction value output by the neural network to the preliminary predicted value of the position obtained by physical integration. Subsequently, the final velocity is calculated based on the position change between the current moment and the next moment. This ensures strict mathematical consistency between the particle's position and velocity, fundamentally avoiding the non-physical phenomena caused by the mismatch between the two in traditional methods, thereby greatly enhancing the numerical stability and physical realism of the simulation.

[0041] Based on the set of fluid particle positions obtained in step S104, the next time step Velocity set and the set of phase volume fractions obtained in step S103 for the next time step. This allows us to obtain the state of all fluid particles at the next moment. By continuing to use the state at the next moment as the new current state and repeating the above steps S101~S104, we can achieve continuous dynamic evolution simulation of the fluid system.

[0042] It should be noted that the steps of the various methods described above are only for clarity. In practice, they can be combined into one step or some steps can be split into multiple steps. As long as they contain the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, but without changing the core design of the algorithm and process, are also within the scope of protection of this patent.

[0043] To further verify the effectiveness of the method of the present invention under complex terrain and non-uniform initial conditions, this embodiment provides a simulation and rendering effect of a two-phase fluid mixing process in a canyon, the dynamic evolution process of which is shown in the appendix. Figure 5 As shown, six time snapshots illustrate the process by which the method of the present invention simulates the interaction of two miscible fluids in a canyon terrain.

[0044] (1) Initial state description (corresponding to the first sub-figure): The simulation scenario is a canyon model with complex elevation differences and a winding river channel. At the start of the simulation, two types of fluids are set up on the platform in the upper right of the canyon, serving as dynamic injection sources for subsequent processes at intervals. This setup is intended to test the ability of the method of the present invention to handle non-static, continuously flowing fluid sources.

[0045] (2) Dynamic process description (corresponding to the second to fifth sub-figures): As the simulation progresses, the fluid overflows from the platform under the influence of gravity and flows downward along the natural path of the canyon. The method of this invention can accurately predict the interaction between fluid particles and complex static boundaries (canyon walls), making the fluid exhibit flow around, convergence, and splashing behaviors that conform to physical laws.

[0046] As the fluid flows into the bottom of the canyon, the phase fraction prediction head of this invention becomes active. A clear miscibility of the two fluids can be observed. The liquid gradually diffuses in the bottom region, and its concentration (volume fraction) changes with increasing mixing, ultimately manifesting visually as a gradual and spreading color change. This demonstrates that the phase fraction prediction head of this invention can accurately learn and predict component changes driven by fluid dynamics.

[0047] (3) Final State and Effect Verification (corresponding to the sixth sub-figure): Finally, the system tends to a new equilibrium state. The fluid has gathered at the lowest point of the canyon, forming a mixed liquid pool, and its color distribution truly reflects the final volume fraction field. The whole process is smooth and natural, without any numerical instability.

[0048] The experiment demonstrates that the state prediction method for miscible multiphase fluids provided by this invention offers a complete and effective technical solution for real-time, high-fidelity simulation of miscible multiphase fluids through its innovative network architecture and prediction process.

[0049] Figure 3 This is a block diagram of a state prediction apparatus for a miscible multiphase fluid according to an exemplary embodiment, the apparatus being used to implement a state prediction method for a miscible multiphase fluid. (Refer to...) Figure 3 The device includes a data acquisition unit 301, a first calculation unit 302, a second calculation unit 303, and a third calculation unit 304. Specifically, the data acquisition unit 301 acquires the position, velocity, and phase volume fraction of the fluid particles at the current moment, and acquires at least one physical parameter used to control the dynamic evolution of the fluid. The first calculation unit 302 calculates preliminary predictions of the position and velocity of the fluid particles at the next moment based on Newton's laws of motion. The second calculation unit 303 uses a trained multiphase fluid prediction network to process the preliminary predictions of the fluid particles' position and velocity, the phase volume fraction at the current moment, and the physical parameter to obtain the position correction and phase volume fraction at the next moment. The third calculation unit 304 obtains the position and velocity of the fluid particles at the next moment based on the current position, the preliminary prediction, and the position correction.

[0050] It should be noted that the state prediction device for miscible multiphase fluids in this embodiment corresponds to the state prediction method for miscible multiphase fluids described above. The functional modules in the state prediction device for miscible multiphase fluids may correspond to specific steps in the state prediction method for miscible multiphase fluids. The state prediction device for miscible multiphase fluids in this embodiment can be implemented in conjunction with the state prediction method for miscible multiphase fluids. That is, where there is no conflict, the relevant technical details mentioned in the state prediction method for miscible multiphase fluids described above can also be applied to the state prediction device for miscible multiphase fluids in this embodiment.

[0051] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, such as... Figure 4 As shown, the electronic device may include the above-mentioned Figure 3 The illustration shows a state prediction device for a miscible multiphase fluid. Optionally, the electronic device 410 may include a first processor 2001.

[0052] Optionally, the electronic device 410 may also include a memory 2002 and a transceiver 2003.

[0053] The first processor 2001, memory 2002, and transceiver 2003 can be connected via a communication bus.

[0054] The following is combined Figure 4 A detailed description of each component of electronic device 410 is provided below: The first processor 2001 is the control center of the electronic device 410. It can be a single processor or a collective term for multiple processing elements. For example, the first processor 2001 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0055] Optionally, the first processor 2001 can perform various functions of the electronic device 410 by running or executing software programs stored in the memory 2002 and calling data stored in the memory 2002.

[0056] In a specific implementation, as one example, the first processor 2001 may include one or more CPUs, for example... Figure 4 CPU0 and CPU1 are shown in the diagram.

[0057] In a specific implementation, as one example, the electronic device 410 may also include multiple processors, for example... Figure 4 The first processor 2001 and the second processor 2004 are shown in the diagram. Each of these processors can be a single-core processor or a multi-core processor. Here, a processor can refer to one or more devices, circuits, and / or processing cores used to process data (such as computer program instructions).

[0058] The memory 2002 is used to store the software program that executes the present invention, and is controlled by the first processor 2001 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0059] Optionally, the memory 2002 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 2002 may be integrated with the first processor 2001 or may exist independently and be connected via the interface circuit of the electronic device 410. Figure 4 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0060] The transceiver 2003 is used to communicate with network devices or with terminal devices.

[0061] Alternatively, transceiver 2003 may include a receiver and a transmitter. Figure 4 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0062] Optionally, the transceiver 2003 can be integrated with the first processor 2001, or it can exist independently and be connected via the interface circuit of the electronic device 410. Figure 4 (Not shown in the image) is coupled to the first processor 2001, and this embodiment of the invention does not specifically limit this.

[0063] It should be noted that, Figure 4 The structure of the electronic device 410 shown does not constitute a limitation on the router. Actual knowledge structure identification devices may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0064] Furthermore, the technical effect of the electronic device 410 can be referred to the technical effect of the state prediction method for a miscible multiphase fluid described in the above method embodiments, and will not be repeated here.

[0065] It should be understood that the first processor 2001 in the embodiments of the present invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0066] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0067] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0068] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0069] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0070] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply 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 invention.

[0071] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0072] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0073] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0074] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0075] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0076] If the aforementioned functions are implemented as 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 this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0077] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for predicting the state of a miscible multiphase fluid, characterized in that, The method includes: The state of the fluid system at the current moment is obtained, including the position, velocity and volume fraction of each phase of each fluid particle, and at least one physical parameter is obtained for controlling the dynamic evolution behavior of the fluid. Calculate the preliminary predicted values ​​of the position and velocity of the fluid particle at the next moment based on Newton's laws of motion; Using a trained multiphase fluid prediction network, the preliminary predicted values ​​of the position and velocity of the fluid particles, the volume fraction of each phase at the current moment, and the physical parameters are processed to obtain the position correction of the fluid particles and the volume fraction of each phase at the next moment. Based on the current position of the fluid particle, the preliminary predicted position value, and the position correction amount, the position and velocity of the fluid particle at the next moment are obtained.

2. The method for predicting the state of miscible multiphase fluids according to claim 1, characterized in that, The physical parameters include at least one of intrinsic physical properties used to define fluid constitutive relations and extrinsic driving parameters used to define external environmental conditions.

3. The method for predicting the state of miscible multiphase fluids according to claim 1, characterized in that, Preliminary predictions of the position and velocity of a fluid particle at the next moment, calculated based on Newton's laws of motion, include: Based on the position and velocity of the fluid particle at the current moment, as well as gravity, a preliminary prediction of the fluid particle's position and velocity at the next moment is calculated using physical integration.

4. The method for predicting the state of miscible multiphase fluids according to claim 1, characterized in that, The multiphase fluid prediction network includes: An embedding layer is used to convert the physical parameters into physical feature embedding vectors; The feature splicing module is used to splice the preliminary predicted values ​​of the position and velocity of the fluid particles, the volume fraction of each phase at the current moment, and the physical feature embedding vector. The shared feature extraction module is used to aggregate the features of each fluid particle and other fluid particles in its neighborhood to obtain a shared feature vector. A position prediction head predicts the position correction amount based on the shared feature vector; and The phase fraction prediction head predicts the volume fraction of each phase at the next time step based on the shared feature vector.

5. The method for predicting the state of miscible multiphase fluids according to claim 4, characterized in that, The embedded layer includes one or more fully connected layers; The shared feature extraction module includes multiple network layers consisting of continuous convolutional layers and fully connected layers; The location prediction head includes convolutional layers and fully connected layers; The phase fraction prediction head includes one or more convolutional layers and a normalization function.

6. The method for predicting the state of miscible multiphase fluids according to claim 5, characterized in that, The network layer in the shared feature extraction module includes activation functions and residual connections; the position prediction head outputs a three-dimensional position correction; the convolutional layer in the phase fraction prediction head outputs an N-dimensional original prediction score vector, which is processed by the normalization function to obtain the volume fraction of each phase at the next time step that satisfies the physical constraints, where N is the number of phases in the fluid system.

7. The method for predicting the state of miscible multiphase fluids according to claim 1, characterized in that, Based on the current position of the fluid particle, the preliminary predicted position value, and the position correction amount, the position and velocity of the fluid particle at the next moment are obtained, including: Based on the preliminary predicted position of the fluid particle and the position correction amount, the position of the fluid particle at the next moment is obtained; The velocity of the fluid particle at the next moment can be obtained from its current position and its position at the next moment.

8. A state prediction device for a miscible multiphase fluid, the state prediction device for a miscible multiphase fluid being used to implement the state prediction method for a miscible multiphase fluid as described in any one of claims 1 to 7, characterized in that, The device includes: The data acquisition unit is used to acquire the position, velocity, and volume fraction of each phase of the fluid particles at the current moment, and to acquire at least one physical parameter for controlling the dynamic evolution behavior of the fluid. The first calculation unit is used to calculate the preliminary predicted values ​​of the position and velocity of the fluid particles at the next moment based on Newton's laws of motion. The second calculation unit is used to process the preliminary predicted values ​​of the position and velocity of the fluid particles, the volume fraction of each phase at the current moment, and the physical parameters using a trained multiphase fluid prediction network, to obtain the position correction of the fluid particles and the volume fraction of each phase at the next moment; and The third calculation unit obtains the position and velocity of the fluid particle at the next moment based on the current position of the fluid particle, the preliminary predicted value of the position, and the position correction amount.

9. An electronic device, characterized in that, The electronic device includes: processor; A memory storing computer-readable instructions that, when executed by the processor, implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program code that can be invoked by a processor to perform the method as described in any one of claims 1 to 7.

Images