Construction method of high-pressure fluid model

Abstract

The invention relates to the technical field of fluid mechanics, in particular to a method for constructing a high-pressure fluid model, which comprises the following steps: acquiring multi-scale fluid data including molecular dynamics data, experimental scale data and in-situ detection data; migrating the medium-low pressure domain data to an ultrahigh pressure domain through a pressure grading migration learning network, and outputting a physical property parameter prediction value; receiving in-situ data in real time based on embedded hardware, and calculating a correction factor; dynamically reconstructing a fluid control equation source item based on the correction factor; and when the average relative error exceeds a threshold value, triggering a reinforcement learning optimizer to reconstruct a model equation. The invention aims to provide the construction method of the high-pressure fluid model, the computer equipment and the storage medium, solves many problems in the prior art, shows remarkable advantages in the aspects of accuracy, real-time performance and adaptability of the fluid model, and provides powerful technical support for research and application in related fields.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fluid mechanics, in particular to a construction method of a high-pressure fluid model. BACKGROUND

[0002] Currently, the construction of high-pressure fluid models has important application value in the fields of chemical industry, petroleum, natural gas, etc. Traditional fluid model construction methods usually rely on experimental data and theoretical derivation. However, these methods have certain limitations when facing complex fluid behavior and high-pressure environments. In the prior art, the construction of fluid models mainly relies on single-scale data, which is difficult to comprehensively reflect the physical properties of fluids under different pressure conditions. In addition, the existing models have low prediction accuracy under high-pressure conditions, which cannot meet the requirements of high precision and real-time performance for industrial applications.

[0003] In terms of multi-scale fluid data acquisition, the existing technology often lacks effective integration mechanisms, resulting in information island phenomenon between different scale data, affecting the overall performance of the model. At the same time, traditional model updating methods are usually static and cannot respond to changes in fluid state in real time, resulting in insufficient adaptability of the model. Especially in the super-high pressure domain, the physical property parameters of the fluid change dramatically, and traditional methods are difficult to accurately predict.

[0004] Therefore, there is an urgent need for a new high-pressure fluid model construction method that can effectively integrate multi-scale fluid data and realize data migration and optimization through advanced learning algorithms to improve the prediction accuracy and real-time performance of the model. SUMMARY

[0005] In order to overcome the shortcomings and deficiencies in the prior art, the purpose of the present application is to provide a construction method of a high-pressure fluid model, a computer device and a storage medium, which solves many problems in the prior art and also exhibits significant advantages in the accuracy, real-time performance and adaptability of the fluid model, providing strong technical support for research and application in related fields.

[0006] The present application is realized by the following technical solutions: In a first aspect, the present application discloses a construction method of a high-pressure fluid model, comprising the following steps: S100. Obtain multi-scale fluid data, including molecular dynamics data, experimental scale data and in-situ detection data; S200. Migrate low-pressure domain data to super-high pressure domain through a pressure hierarchical transfer learning network, and output physical property parameter prediction values; S300. Real-time receive in-situ data based on embedded hardware, and calculate correction factors; S400. Dynamically reconstruct source terms of fluid control equations based on correction factors; S500. Trigger the reinforcement learning optimizer to reconstruct the model equation when the average relative error exceeds the threshold value.

[0007] In combination with the first aspect, further, in step S100, the pressure-temperature mapping function is set as: Data alignment is performed:

[0008] wherein, is the original pressure value, is the original temperature value, is the learnable weight matrix, is the bias vector; The pressure-temperature mapping function is: Training through a generative adversarial network:

[0009] wherein, is the generator, is the discriminator, is the molecular dynamics data; and generate distribution alignment data .

[0010] In combination with the first aspect, further, in step S200, the calculation formula of the physical property parameter prediction value is:

[0011] wherein, is the calculated intermolecular distance in the high-pressure model, is the measured intermolecular distance in the low-pressure model, is the pre-trained neural network model based on medium-low data ; is the high-pressure adapter module, inputting molecular scale input data ; is a feature concatenation operation; The calculation formula of the high-pressure adapter module is:

[0012] wherein, is the molecular scale feature vector, is the gating vector, is the layer normalization operator, is the Gaussian error linear unit activation.

[0013] In combination with the first aspect, further, between steps S200 and S300, there is also a step S201: uncertainty quantification step:

[0014] wherein, is the model variance, is the data uncertainty, is the density sensitivity to pressure, is the pressure control accuracy.

[0015] In combination with the first aspect, further, in step S300, the in-situ data is received in real time by the embedded hardware , and the correction factor is calculated:

[0016] wherein, is the predicted property vector, is the in-situ probe inversion value, is the sensor noise standard deviation, is the sensitivity coefficient;

[0017]

[0018] wherein, is the current measurement value, is the observation matrix, is the kalman gain, is the sensor noise covariance, is the estimation error covariance.

[0019] In combination with the first aspect, further, in step S400, the correction formula of the fluid control equation source term is:

[0020] wherein, is the density field, is the velocity vector field, is the time correction weight; The time correction weight is optimized by variational inference:

[0021]

[0022] wherein, is the time series error vector, is the multi-layer perception module, is the standard deviation.

[0023] In combination with the first aspect, further, in step S500, when the average relative error when the average relative error of the model exceeds a set threshold, triggering a reinforcement learning optimizer:

[0024] wherein, is a reward function, is an action value function, is a first error distribution state, is a first modified network layer action, is a regularization coefficient; the reward function is defined as:

[0025] wherein, is an error reduction amount, is a computational resource increment, is a weight coefficient, is a decay factor, is a resource threshold.

[0026] In a second aspect, the present application also discloses a computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the high-pressure fluid model construction method as described above.

[0027] In a third aspect, the present application also discloses a computer-readable storage medium having a computer program stored thereon, wherein the program is executed by a processor to implement the high-pressure fluid model construction method as described above.

[0028] The present application has the following beneficial effects: The high-pressure fluid model construction method provided by the present application significantly improves the construction efficiency and accuracy of the fluid model through a multi-step process design. First, the multi-scale fluid data acquisition method is adopted, which covers molecular scale, experimental scale and in-situ detection data, ensuring the comprehensiveness and accuracy of the model under different pressure conditions. Second, the pressure grading transfer learning network effectively transfers the data in the low-pressure domain to the super-high pressure domain, overcoming the prediction deficiency of traditional methods in high-pressure environments.

[0029] In addition, based on the embedded hardware, the in-situ data are received in real time and the correction factor is calculated, so that the model can dynamically adapt to the changes of the fluid state, improving the real-time performance and response ability of the model. When the average relative error of the model exceeds a set threshold, a reinforcement learning optimizer is triggered to reconstruct the model equation, further improving the accuracy and reliability of the model.

[0030] In summary, the high-pressure fluid model construction method of the present invention not only solves many problems in the prior art, but also shows significant advantages in the accuracy, real-time performance and adaptability of the fluid model, providing strong technical support for research and application in related fields. Attached Figure Description

[0031] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0032] Figure 1 This is a flowchart of the method for constructing the high-pressure fluid model of the present invention.

[0033] Figure 2 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0034] Figure 3 This is a schematic diagram of a storage medium provided in an embodiment of the present invention. Detailed Implementation

[0035] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0036] Example 1 Due to insufficient experimental data (>50 GPa), the extrapolation of traditional state equations is distorted, and offline correction models cannot adapt to real-time changes in high-pressure environments (such as pipeline pressure fluctuations). Therefore, as... Figure 1 As shown, this embodiment discloses a method for constructing a high-pressure fluid model. Through cross-domain knowledge transfer via transfer learning and embedded hardware closed-loop correction, a high-precision, low-latency intelligent modeling system is constructed. It includes the following steps: S100. Acquire multi-scale fluid data, including molecular dynamics data, experimental-scale data, and in-situ detection data; S200. By using a pressure grading transfer learning network, data from the medium and low pressure domains are transferred to the ultra-high pressure domain, and predicted values ​​of physical property parameters are output. S300. Receives in-situ data in real time based on embedded hardware and calculates correction factors; S400. Dynamically reconstructing the source terms of the fluid control equations based on correction factors; S500. Trigger reinforcement learning optimizer to reconstruct model equation when average relative error exceeds threshold value.

[0037] In the embodiment, S100-S200 solve initial modeling accuracy, S300-S400 implement runtime dynamic calibration, and S500 guarantee long-term reliability, forming a complete chain of "perception-decision-execution-optimization".

[0038] Embodiment 2 Step S100: Obtain multi-scale fluid data, including molecular dynamics data, experimental scale data, and in-situ detection data, and set a pressure-temperature mapping function Align the data:

[0039] wherein, is the original pressure value, unit: GPa, the static pressure suffered by the fluid, and the typical value of the ultra-high pressure scene is 10-100 GPa; is the original temperature value, unit: K, the thermodynamic temperature of the fluid, and the typical range of the supercritical state is 300-1000 K; is the natural logarithm of pressure, to compress the numerical span of high pressure area; is the 0.7 power of temperature, and the experience transformation eliminates the skewness of temperature distribution; is dimension 2 is the learnable weight matrix of 2, to establish the mapping relationship between the molecular scale (Å level) and the engineering scale (m level); is dimension 2 is the bias vector of 1, to compensate for the system error of experimental equipment; Pressure-temperature mapping function Through the training of the generative adversarial network:

[0040] and generate distribution alignment data .

[0041] wherein, is the generator, which maps the molecular dynamics data to the experimental scale data . ; is the discriminator of the 3-layer convolutional network, which outputs the probability of authenticity. The embodiment provides distribution alignment input for transfer learning, and provides an estimation basis for data uncertainty for uncertainty quantification.

[0042] Embodiment 3 Step S200: migrating the medium-low pressure domain data to the ultra-high pressure domain by pressure staged transfer learning network, outputting the property parameter prediction value, and the calculation formula of the property parameter prediction value is:

[0043] wherein, is the calculated intermolecular distance in the high-pressure model, is the measured intermolecular distance in the low-pressure model, is the medium-low data-based The pre-trained 4-layer ResNet neural network model is trained on <10 GPa experimental data. is a high-pressure adapter module, and a gating mechanism prevents negative transfer; the input molecular-scale input data , the DFT calculated electron density matrix; is a feature splicing operation; high-pressure adapter module The calculation formula is:

[0044] wherein, is a molecular-scale feature vector with dimension d = 512; is a gating vector with dimension d 1 to control the knowledge transfer strength; is a layer normalization operator to stabilize the high-pressure feature distribution; is a Gaussian error linear unit activation, which is more suitable for non-linear high-pressure features.

[0045] The gating vector is determined by the gradient of the feature with respect to the pressure, realizing physically inspired adaptive transfer, and high-pressure sensitive features (such as electron density) automatically obtain high gating values .

[0046] The following is one of the experimental data of an embodiment (80 GPa metal hydrogen transport modeling):

[0047] The experimental data shows that the 80 GPa conductivity prediction is: Without using the adapter: error 38.7% (phase transition causes the traditional model to fail); Using the adapter of the present embodiment: error 3.1% (accurately capturing the electron delocalization effect); The 10 GPa viscosity prediction error remains 0.8%.

[0048] Embodiment 4 Between steps S200 and S300, there is also a step S201: uncertainty quantification step:

[0049] where, is the model variance, is the data uncertainty, is the density sensitivity to pressure, is the pressure control accuracy.

[0050] In one of the 55 GPa supercritical water delivery system embodiments, the following scenario parameters are set: Target property: density ; Working point: , ; Equipment: Diamond anvil cell synchrotron measurement.

[0051] Uncertainty component calculation: Model variance by Dropout Monte Carlo sampling (100 forward propagations):

[0052] X-ray diffraction calibration certificate gives data uncertainty

[0053] Pressure sensitivity ; Pressure control accuracy ;

[0054] Combined uncertainty

[0055] Predicted value , final output .

[0056] Experimental results prove:

[0057] The confidence interval width is reduced by 42% (0.12 → 0.07) In 20 high-pressure fluctuations, the emergency shutdown due to density misjudgment is reduced from 5 to 1.

[0058] This embodiment unifies the model randomness ( ), data error ( ), parameter sensitivity ( ) in a single formula, and calculates the standard deviation of the predicted value

[0059] Embed error propagation theory into real-time fluid model control loop; confidence interval coverage rate increases by 22%, directly reducing the risk of high-pressure engineering accidents.

[0060] Embodiment 5 Step S300: Real-time receive in-situ data based on embedded hardware, calculate correction factor, real-time receive in-situ data through embedded hardware , calculate correction factor The correction factor is used to scale the space-time term of the mass conservation equation:

[0061] Wherein, is the predicted physical property vector, consisting of ; is the in-situ detection inversion value, derived from diamond anvil cell Raman spectrum inversion value; is the standard deviation of sensor noise, typical value: 0.03-0.05; is the sensitivity coefficient, ;

[0062]

[0063] Wherein, is the current measurement value; is the observation matrix, usually the unit matrix; is the kalman gain; is the sensor noise covariance, typical value: ; is the estimated error covariance.

[0064] Embodiment 6 Step S400: Based on the correction factor, dynamically reconstruct the source term of the fluid control equation, and the correction formula of the source term of the fluid control equation is:

[0065] Wherein, is the density field, with the unit of ; is the velocity vector field, with the velocity of ; is the time correction weight, updated by variational inference; The time correction weight is optimized by variational inference:

[0066]

[0067] ​in, This is the timing error vector; It is a multilayer perceptual module with 2 fully connected layers and 64 hidden neurons; Standard deviation, constraints .

[0068] Example 7 Further, in step S500: when the average relative error exceeds a threshold, the reinforcement learning optimizer is triggered to reconstruct the model equation. At that time, the reinforcement learning optimizer is triggered:

[0069] in, For the reward function; The action value function takes the input state as its input. ,action ; For the first Error distribution status, For the first The action of modifying the number of network layers in the item. This is the regularization coefficient, with a default value of 0. To prevent overfitting; reward function Defined as:

[0070] in, The amount of error reduction is calculated using the following formula: ; To calculate the incremental increase in computing resources, the equivalent CPU core time; These are the weighting coefficients. As the attenuation factor, Set as a resource threshold .

[0071] Example 8 Please refer to Figure 2 This document illustrates a schematic diagram of the structure of a computer device provided by some embodiments of this application. The electronic device 10 includes: a processor 100, a memory 101, a bus 102, and a communication interface 103. The processor 100, the communication interface 103, and the memory 101 are connected via the bus 102. The memory 101 stores a computer program that can run on the processor 100. The processor 100 is used to operate according to the instructions to execute the steps of the method described in any one of embodiments 1-7.

[0072] The memory 101 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 103 (which can be wired or wireless), such as the Internet, wide area network, local area network, or metropolitan area network.

[0073] Bus 102 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. Memory 101 is used to store programs. After receiving an execution instruction, processor 100 executes the program. The high-pressure fluid model construction method disclosed in any of the foregoing embodiments of this application can be applied to processor 100, or implemented by processor 100.

[0074] Processor 100 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 100 or by instructions in software form. Processor 100 may be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it may also be a Digital Signal Processor (DSP), an Application-Specific Integrated Circuit (ASIC), an Off-the-shelf Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 101. The processor 100 reads the information in memory 101 and, in conjunction with its hardware, completes the steps of the above method.

[0075] Example 9 This embodiment also provides a computer-readable storage medium corresponding to the high-pressure fluid model construction method provided in the foregoing embodiments. Please refer to [link / reference]. Figure 3The computer-readable storage medium shown is an optical disc, on which a computer program (i.e., program product 20) is stored. When the computer program is run by a processor, it executes the power scheduling method provided in any of the foregoing embodiments.

[0076] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.

[0077] The computer-readable storage medium provided in the above embodiments of this application and the method for constructing a high-pressure fluid model provided in the embodiments of this application are based on the same application concept and have the same beneficial effects as the methods adopted, run or implemented by the applications stored therein.

[0078] It should be noted that in the above text, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0079] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0080] The embodiments of this application have been described above with reference to the accompanying drawings. These are merely specific implementations of this application, but this application is not limited to the specific implementations described above. The specific implementations described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for constructing a high-pressure fluid model, characterized in that, Includes the following steps: S100. Acquire multi-scale fluid data, including molecular dynamics data, experimental-scale data, and in-situ detection data; S200. By using a pressure grading transfer learning network, data from the medium and low pressure domains are transferred to the ultra-high pressure domain, and predicted values ​​of physical property parameters are output. S300. Receives in-situ data in real time based on embedded hardware and calculates correction factors; S400. Dynamically reconstructing the source terms of the fluid control equations based on correction factors; S500. When the average relative error exceeds the threshold, the reinforcement learning optimizer is triggered to reconstruct the model equation.

2. The method for constructing a high-pressure fluid model according to claim 1, characterized in that, In step S100, the pressure-temperature mapping function is set. Perform data alignment: in, The original pressure value. This is the original temperature value. For learnable weight matrix, It is the bias vector; Pressure-temperature mapping function Training via adversarial generative networks: in, For generator, For discriminator, For molecular dynamics data, For experimental data; And generate distribution-aligned data .

3. The method for constructing a high-pressure fluid model according to claim 1, characterized in that, In step S200, the formula for calculating the predicted values ​​of physical property parameters is as follows: in, For the intermolecular distances calculated in the high-pressure model, This represents the measured intermolecular distance in the low-pressure model. Based on low to medium data Pre-trained neural network models; For the high-voltage adapter module, input molecular-scale data. ; For feature splicing operations; High voltage adapter module The calculation formula is: in, These are molecular-scale eigenvectors. For the gated vector, For layer normalization operators, Activate the Gaussian error linear unit.

4. The method for constructing a high-pressure fluid model according to claim 3, characterized in that, Between steps S200 and S300, step S201 is also included: uncertainty quantification step: in, For model variance, For data uncertainty, For the sensitivity of density to pressure, For pressure control accuracy.

5. The method for constructing a high-pressure fluid model according to claim 1, characterized in that, In step S300, in-situ data is received in real time via embedded hardware. Calculate the correction factor : in, To predict the material property vector, These are in-situ detection inversion values. The standard deviation of sensor noise. This is the sensitivity coefficient; in, This is the current measurement value. For the observation matrix, For Kalman gain, For sensor noise covariance, To estimate the error covariance.

6. The method for constructing a high-pressure fluid model according to claim 1, characterized in that, In step S400, the source terms of the fluid control equations The corrected formula is: in, For density field, For velocity vector field, Adjust the weights for time; Time-corrected weights Optimization through variational inference: in, This is the timing error vector. It is a multi-layer sensing module. The standard deviation is denoted as .

7. The method for constructing a high-pressure fluid model according to claim 1, characterized in that, In step S500, when the average relative error At that time, the reinforcement learning optimizer is triggered: in, For the reward function, Let the action value function be... For the first Error distribution status, For the first The action of modifying the number of network layers in the item. The regularization coefficient is used. reward function Defined as: in, For the amount of error reduction, To calculate resource increments, These are the weighting coefficients. As the attenuation factor, This represents the resource threshold.

8. A computer device, characterized in that, Including processor and storage media; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by a processor to implement the method for constructing a high-pressure fluid model as described in any one of claims 1-7.

Images