Intelligent reduced-order modeling method for liquid rocket engine combustion chamber head cavity

Abstract

The application provides an intelligent reduction method for a liquid rocket engine combustion chamber head cavity simulation model, comprising the following steps: establishing a numerical simulation model of a liquid rocket engine combustion chamber head cavity based on the structure and working principle of the liquid rocket engine combustion chamber head cavity; performing numerical simulation calculation based on the numerical simulation model and a preset boundary condition, and constructing a data set based on the numerical simulation calculation result; training a preset neural network model based on the data set to obtain a target reduction model of the liquid rocket engine combustion chamber head cavity. The application can realize intelligent reduction of the combustion chamber head cavity simulation model, improve the simulation efficiency of the liquid rocket engine combustion chamber head cavity, and ensure the simulation accuracy of the liquid rocket engine combustion chamber head cavity.

Description

Technical Field

[0001] This invention relates to the field of liquid rocket engine simulation technology, and in particular to an intelligent reduction method for simulation models of the combustion chamber head cavity of liquid rocket engines. Background Technology

[0002] The filling flow process of the combustion chamber head cavity has a crucial impact on the start-up characteristics of liquid rocket engines. The operating conditions of liquid rocket engines can be simulated by replicating this filling flow process. Model order reduction, as a key means to achieve high-efficiency and high-precision simulation calculations, is a core technology in the digitalization process of rocket engines. Traditional model order reduction methods mainly include the Krylov subspace method, equilibrium truncation method, orthogonal decomposition method, response surface model, and polynomial fitting method. However, traditional model order reduction methods have significant shortcomings in both computational accuracy and efficiency, and are not suitable for the strong transient characteristics of the non-azeotropic two-phase filling flow heat transfer process in the combustion chamber head cavity of liquid rocket engines, resulting in poor model order reduction performance. Summary of the Invention

[0003] In view of this, the purpose of this invention is to provide an intelligent reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine, so as to alleviate the above-mentioned problems existing in related technologies.

[0004] In a first aspect, embodiments of the present invention provide an intelligent order reduction method for a simulation model of a liquid rocket engine combustion chamber head cavity, comprising: establishing a numerical simulation model of the liquid rocket engine combustion chamber head cavity based on its structure and working principle; the numerical simulation model is used to describe the two-phase flow and evaporation-boiling phenomena during the two-phase filling process of the cryogenic propellant in the liquid rocket engine combustion chamber head cavity based on a multiphase flow model and an evaporation-condensation model; the numerical simulation model includes a continuity equation, a momentum equation, a phase equation, and an evaporation-condensation model; performing numerical simulation calculations based on the numerical simulation model and preset boundary conditions, and constructing a dataset based on the numerical simulation calculation results; training a preset neural network model based on the dataset to obtain a target order reduction model of the liquid rocket engine combustion chamber head cavity; wherein, the input variables of the preset neural network model include the inlet mass flow rate, inlet temperature, outlet pressure, and initial wall temperature of the liquid rocket engine combustion chamber head cavity, and the output variables of the preset neural network model include the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy, and outlet temperature of the liquid rocket engine combustion chamber head cavity; the evaporation-condensation model adopts the LEE model, and the evaporation-condensation model is as follows: ,in, For gradient, For density, t For time, It is the phase volume fraction. It is a vapor phase. It is a liquid phase. and These represent the mass transfer rates of evaporation and condensation, respectively; when the liquid phase temperature... Above saturation temperature At that time, the expression for the mass transfer rate of evaporation is: ,at this time The evaporation coefficient is... It is the liquid volume fraction. The density is the liquid phase density; when the vapor phase temperature is... Below saturation temperature When the mass transfer rate of condensation is expressed as: ,at this time The condensation coefficient is . This represents the volume fraction of the vapor phase. ρ is the density of the vapor phase.

[0005] Secondly, embodiments of the present invention also provide an intelligent order reduction device for a simulation model of a liquid rocket engine combustion chamber head cavity, comprising: a construction module, used to establish a numerical simulation model of the liquid rocket engine combustion chamber head cavity based on its structure and working principle; the numerical simulation model is used to describe the two-phase flow and evaporation-boiling phenomena during the two-phase filling process of the cryogenic propellant in the liquid rocket engine combustion chamber head cavity based on a multiphase flow model and an evaporation-condensation model, the numerical simulation model including a continuity equation, a momentum equation, a phase equation, and an evaporation-condensation model; and a simulation module, used to perform simulation based on the numerical simulation model and preset boundary conditions. Numerical simulation calculations are performed, and a dataset is constructed based on the numerical simulation results. A training module is used to train a preset neural network model based on the dataset to obtain a target reduced-order model of the liquid rocket engine combustion chamber head cavity. The input variables of the preset neural network model include the inlet mass flow rate, inlet temperature, outlet pressure, and initial wall temperature of the liquid rocket engine combustion chamber head cavity. The output variables of the preset neural network model include the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy, and outlet temperature of the liquid rocket engine combustion chamber head cavity. The evaporation-condensation model adopts the LEE model, which is as follows: ,in, For gradient, For density, t For time, It is the phase volume fraction. It is a vapor phase. It is a liquid phase. and These represent the mass transfer rates of evaporation and condensation, respectively; when the liquid phase temperature... Above saturation temperature At that time, the expression for the mass transfer rate of evaporation is: ,at this time The evaporation coefficient is... It is the liquid volume fraction. The density is the liquid phase density; when the vapor phase temperature is... Below saturation temperature When the mass transfer rate of condensation is expressed as: ,at this time The condensation coefficient is . This represents the volume fraction of the vapor phase. ρ is the density of the vapor phase.

[0006] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the intelligent reduction method for the simulation model of the combustion chamber head cavity of the liquid rocket engine described in the first aspect.

[0007] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are invoked and executed by a processor, the computer-executable instructions cause the processor to implement the intelligent reduction method for the simulation model of the combustion chamber head cavity of the liquid rocket engine described in the first aspect.

[0008] This invention provides an intelligent order reduction method for a liquid rocket engine combustor head cavity simulation model. First, a numerical simulation model of the liquid rocket engine combustor head cavity is established based on its structure and working principle. Then, numerical simulation calculations are performed based on the numerical simulation model and preset boundary conditions, and a dataset is constructed based on the simulation results. Finally, a preset neural network model is trained using this dataset to obtain a target order-reduced model of the liquid rocket engine combustor head cavity. Using this technique, the complex non-azeotropic two-phase filling flow and heat transfer mechanism inside the combustor head cavity can be characterized based on a deep recurrent neural network architecture, thereby achieving intelligent order reduction of the combustor head cavity simulation model. Compared to traditional model order reduction methods, the target order-reduced model exhibits superior computational efficiency and accuracy. Furthermore, the input and output variables of the target order-reduced model correspond one-to-one with the interface variables of the combustor head cavity in the engine system-level simulation model. Therefore, the intelligent order reduction method for the combustor head cavity simulation model can be seamlessly integrated into the liquid rocket engine system-level simulation, significantly improving the computational accuracy of the liquid rocket engine combustor head cavity simulation.

[0009] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0010] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0011] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0012] Figure 1 This is a flowchart illustrating an intelligent order reduction method for a simulation model of the combustion chamber head cavity of a liquid rocket engine, as described in an embodiment of the present invention.

[0013] Figure 2 This is a schematic diagram of the structure of the combustion chamber head cavity of the liquid rocket engine in an embodiment of the present invention;

[0014] Figure 3 This is a flowchart illustrating the intelligent order reduction process of the simulation model of the combustion chamber head cavity of a liquid rocket engine in an embodiment of the present invention.

[0015] Figure 4 This is a schematic diagram of the structure of an intelligent order reduction device for a simulation model of the combustion chamber head cavity of a liquid rocket engine in an embodiment of the present invention;

[0016] Figure 5 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Currently, traditional model order reduction methods mainly include the Krylov subspace method, equilibrium truncation method, orthogonal decomposition method, response surface model, and polynomial fitting method. These traditional methods suffer from significant shortcomings in both computational accuracy and efficiency, and are unsuitable for the strong transient characteristics of the non-azeotropic two-phase filling flow heat transfer process in the combustion chamber head cavity of liquid rocket engines, resulting in poor model order reduction performance. Therefore, this invention provides an intelligent model order reduction method for the simulation model of the combustion chamber head cavity of liquid rocket engines, which can alleviate the aforementioned problems in related technologies.

[0019] To facilitate understanding of this embodiment, a detailed description of the intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine, as disclosed in this embodiment, will be provided first. (See [link to relevant documentation]). Figure 1 As shown, the method may include the following steps:

[0020] Step S102: Based on the structure and working principle of the combustion chamber head cavity of the liquid rocket engine, establish a numerical simulation model of the combustion chamber head cavity of the liquid rocket engine.

[0021] Figure 2 The main structure of the combustion chamber head cavity of a liquid rocket engine is shown.

[0022] Step S104: Perform numerical simulation calculations based on the numerical simulation model and preset boundary conditions, and construct a dataset based on the numerical simulation calculation results.

[0023] Step S106: Train the preset neural network model based on the dataset to obtain the target reduced-order model of the combustion chamber head cavity of the liquid rocket engine.

[0024] The input variables of the preset neural network model may include the inlet mass flow rate, inlet temperature, outlet pressure, and initial wall temperature of the liquid rocket engine combustion chamber head cavity, and the output variables of the preset neural network model may include the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy, and outlet temperature of the liquid rocket engine combustion chamber head cavity.

[0025] The aforementioned intelligent reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine can characterize the complex non-azeotropic two-phase filling flow and heat transfer mechanism inside the combustion chamber head cavity based on a deep recurrent neural network architecture. This achieves intelligent reduction of the simulation model of the combustion chamber head cavity. Compared with traditional model reduction methods, the target reduction model shows excellent performance in both computational efficiency and accuracy. Furthermore, the input and output variables of the target reduction model correspond one-to-one with the interface variables of the combustion chamber head cavity in the engine system-level simulation model. Therefore, the intelligent reduction method for the simulation model of the combustion chamber head cavity can be seamlessly integrated into the system-level simulation of liquid rocket engines, significantly improving the computational accuracy of the simulation of the combustion chamber head cavity of liquid rocket engines.

[0026] As one possible implementation, step S102 (i.e., establishing a numerical simulation model of the liquid rocket engine combustion chamber head cavity based on the structure and working principle of the liquid rocket engine combustion chamber head cavity) may include: establishing a geometric model of the liquid rocket engine combustion chamber head cavity based on the structure and working principle of the liquid rocket engine combustion chamber head cavity; extracting the target fluid domain based on the geometric model, and establishing a numerical simulation model based on the target fluid domain and the fluid dynamic parameters of the liquid rocket engine combustion chamber head cavity.

[0027] The target fluid domain includes the solid and fluid domains of the rocket engine combustion chamber inlet section, the rocket engine combustion chamber head cavity, the rocket engine combustion chamber bottom cavity, and the rocket engine combustion chamber injector.

[0028] In practical applications, the above geometric model can be a CAD geometric model. For example, a CAD geometric model of the combustion chamber head cavity can be established first based on the combustion chamber head cavity filling process; then, based on the established head cavity geometric model, the fluid domain is extracted, including the solid domain and fluid domain of the combustion chamber inlet section, combustion chamber head cavity, combustion chamber bottom cavity, and combustion chamber injector; finally, a numerical simulation model of the combustion chamber head cavity is established based on the CFD method.

[0029] For example, the above numerical simulation model can describe the two-phase flow and evaporation boiling phenomenon during the two-phase filling process of the cryogenic propellant in the combustion chamber head cavity based on the multiphase flow model and the evaporation-condensation model. Taking the VOF (Volume of Fluid) model as an example, the above numerical simulation model can include the continuity equation, momentum equation and phase equation.

[0030] The VOF model is based on an interface tracing method using an Eulerian grid. In this method, incompatible fluid components share a common momentum equation, and phase volume fraction is introduced to achieve this. This variable is used to track the interfaces between phases within the computational domain. It represents the ratio of the volume of a phase to the volume of the grid in which it resides.

[0031]

[0032] Phase A represents phase A; Phase B represents phase B.

[0033] The VOF model can construct the interface by calculating the phase volume fraction of each grid cell in the entire computational domain.

[0034] In the governing equations of the VOF model, the continuity equation can be:

[0035]

[0036] in, For speed, For gradient.

[0037] The momentum equation can be:

[0038]

[0039] in, For density, t For time, It is the acceleration due to gravity. Viscosity, This refers to the force source item of the interface. This is equivalent to the pressure increment caused by interfacial tension at the interface.

[0040] .

[0041] in, For surface tension, This represents the phase volume fraction. Indicates curvature, combined with interfacial tension. This can be represented as the pressure difference across the interface. The total interface force source term can be expressed as the differential form of the interface pressure.

[0042] The phase equation can be: .

[0043] For example, the numerical simulation model described above may also include an evaporation-condensation model. The evaporation-condensation model can employ the LEE model, which can fit the saturation temperature under different pressure ranges through piecewise linear interpolation. It can perform calculations on evaporation and condensation problems.

[0044] In the LEE model, the mass transfer of the evaporation-condensation process is determined by the steam transport equation, which can be expressed as:

[0045]

[0046] Among them, subscript For vapor phase, subscript It is a liquid phase. and These represent the mass transfer rates of evaporation and condensation, respectively.

[0047] When the liquid phase temperature Above saturation temperature At that time, the expression for the mass transfer rate of evaporation is:

[0048]

[0049] at this time The evaporation coefficient is... It is the liquid volume fraction. ρ is the density of the liquid phase.

[0050] When the gas phase temperature Below saturation temperature At that time, the expression for the mass transfer rate of condensation is:

[0051]

[0052] at this time The condensation coefficient is . It is the liquid volume fraction. This refers to the density of the liquid phase. Numerical or physical experiments are needed to determine this. Perform calibration.

[0053] As one possible implementation, the aforementioned boundary conditions may include inlet boundary conditions, outlet boundary conditions, and wall boundary conditions. Based on this, the numerical simulation calculation in step S104, which involves performing numerical simulation calculations based on the numerical simulation model and preset boundary conditions, may include: setting the inlet boundary condition to a mass flow inlet, setting the outlet boundary condition to a pressure outlet, and setting the wall boundary condition to a fluid-structure interaction heat transfer surface; and performing numerical simulation calculations of the numerical simulation model by simulating the inlet mass flow rate, wall temperature, and outlet pressure under different operating conditions, and using the inlet boundary conditions, outlet boundary conditions, and wall boundary conditions as constraints.

[0054] As one possible implementation, the numerical simulation results may include first characteristic time series of inlet mass flow rate, inlet temperature, outlet pressure, and initial wall temperature, and second characteristic time series of inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy, and outlet temperature. Based on this, the step S104 of constructing a dataset based on the numerical simulation results may include: generating multiple time series samples based on the first and second characteristic time series, and combining the obtained time series samples into a dataset.

[0055] Each time series sample has its own corresponding input information and label information. The input information includes the inlet mass flow rate, inlet temperature, outlet pressure and initial wall temperature corresponding to the corresponding time series sample. The label information includes the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy and outlet temperature corresponding to the corresponding time series sample.

[0056] The numerical simulation model established in step S102 above requires the combination of inlet boundary conditions, outlet boundary conditions, and wall boundary conditions to perform numerical simulation calculations. For example, the inlet boundary condition can be set as a mass flow rate inlet, the outlet boundary condition as a pressure outlet, and the boundary conditions of the fluid-structure interaction wall (i.e., wall boundary conditions) can be defined as the fluid-structure interaction heat transfer surface. The core calculation process of the numerical simulation model is based on the finite volume method for numerical solution. In this calculation process, physical quantities are integrated and discretized within a discrete control volume, thereby transforming the partial differential equation system into an algebraic-differential equation system, which is then solved iteratively. To achieve this calculation process, a pressure-based solver is used, and the SIMPLE algorithm is selected for calculation. To comprehensively simulate the two-phase filling process of the combustion chamber head cavity, various different operating conditions (including different inlet mass flow rates, wall temperatures, and outlet pressures) are set for numerical simulation calculations of the liquid rocket engine combustion chamber head cavity. Specifically, the operating conditions cover different engine start-up sequences, different start-up methods, and combustion chamber head cavity temperatures, etc. Through numerical simulation calculations under different working conditions, a series of characteristic time series (in the form of transient time series data) were extracted. After standardization and normalization, a dataset for subsequent analysis was finally constructed.

[0057] As one possible implementation, the aforementioned pre-defined neural network model may include a Gated Recurrent Unit (GRU), a Recurrent Neural Network (RNN), or a Long Short-Term Memory (LSTM) network. Term Memory (LSTM), Temporal Convolutional Network (TCN), Transformer neural network model based on self-attention mechanism, etc., are not limited to this.

[0058] Taking GRU as an example, step S106 above (i.e., training a preset neural network model based on the dataset to obtain the target reduced-order model of the liquid rocket engine combustion chamber head cavity) may include: constructing a training set and a test set for GRU based on the dataset; using the mean squared error (MSE) as the loss function of GRU, using Adam as the optimizer, training GRU using the training set, and adjusting the parameters of GRU using the test set during the training process, and obtaining the target reduced-order model after training.

[0059] GRU is a gated recurrent neural network model, where gating is a mechanism that captures short-term and long-term dependencies in a sequence. In GRU, the input at each time step controls the flow of information through update and reset gates, thus constructing a dynamically dependent network architecture. Compared to traditional RNNs, GRU's structure reduces the number of parameters by simplifying the gating mechanism, while improving computational efficiency. Using the GRU architecture, the complex non-azeotropic two-phase filling flow and heat transfer characteristics inside the combustion chamber head cavity can be extracted to achieve model order reduction.

[0060] GRU controls the flow of information through reset and update gates, effectively capturing short-term and long-term dependencies in a sequence. The core formula of GRU includes the calculation of reset gate, update gate, candidate hidden state, and current hidden state. The final output of GRU is obtained by mapping the hidden states.

[0061] The formula for calculating the reset gate is as follows:

[0062]

[0063] To reset the value of the door; For activation function, To reset the weight matrix of the gate, This is the hidden state from the previous moment. For the current input, To reset the door's bias. Used to control the hidden state of the previous moment. Candidate hidden state at the current moment The impact.

[0064] The formula for calculating the update gate is as follows:

[0065]

[0066] To update the value of the gate, To update the gate weight matrix, To update the bias term of the gate. Used to control the hidden state of the previous moment. and the candidate hidden state at the current moment The weight.

[0067] The formula for calculating the candidate hidden state is as follows:

[0068]

[0069] Let be the candidate hidden state at the current time, and let represent the potential hidden state at the current time. The hyperbolic tangent activation function is used. Let be the weight matrix of the candidate hidden states. For Hadamard product, The bias term for the candidate hidden state.

[0070] The formula for calculating the current hidden state is as follows:

[0071]

[0072] The current hidden state is the same as the previous hidden state. and candidate hidden state The weighted sum; To update the complement value of the gate, it means retaining the hidden state from the previous moment. The proportion; To update the value of the gate, we need to introduce the candidate hidden state at the current time step. The proportion.

[0073] The input and output of the GRU are both sequential data. This embodiment of the invention proposes using the inlet mass flow rate, inlet temperature, outlet pressure, and initial wall temperature of the combustion chamber head cavity as the GRU input, and the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy, and outlet temperature as the GRU output. The time step is the number of numerical simulation time steps contained in a training sample. The GRU input and output are data from the same time step; that is, the GRU does not predict future data, and the time step can be set as needed. Training data is transmitted in the network in batches, with each batch containing multiple training samples.

[0074] For example, MSE can be used as the loss function for GRU, and Adam can be selected as the optimizer. Training and testing sets for GRU can be constructed based on the pre-built dataset. This process is essentially an intelligent model reduction process. The specific training process of GRU is described below:

[0075] 1) Perform forward propagation: The input sequence is passed through the update gate and reset gate of GRU to calculate the hidden state step by step, thereby obtaining the output value.

[0076] The output value of GRU is calculated using the following formula:

[0077]

[0078] This is the output value at the current moment. This is the weight matrix of the output layer. This is the bias term for the output layer.

[0079] 2) Calculate the loss function (Compute Loss): Compare the output value of GRU (corresponding to the corresponding input sample) with the actual label value (included in the label information of the corresponding input sample) to calculate the value of the loss function.

[0080] 3) Perform backpropagation: Calculate the derivative of the GRU parameters with respect to the loss function to determine the degree of influence of each parameter on the loss function, and use the chain rule to propagate the gradient to each time step.

[0081] 4) Update Parameters: With the goal of minimizing the loss function, the values ​​of each parameter are adjusted according to the preset step size, following the gradient descent method.

[0082] The process is repeated from step 1) to step 4) until the goal of minimizing the loss function is achieved, at which point training ends. However, during the training of the GRU, due to its recursive computation and nonlinear activation functions, problems such as vanishing or exploding gradients can easily occur, affecting the training effect. To solve these problems, gradient clipping, weight initialization, and batch normalization can be used. During the training of the GRU, the test set is used to tune the GRU parameters. The GRU with the best performance obtained after training is the reduced-order model of the combustion chamber head cavity.

[0083] For example, Figure 3 This paper presents an intelligent reduction process for the simulation model of the combustion chamber head cavity of a liquid rocket engine. First, a numerical simulation model of the combustion chamber head cavity is established. Then, boundary conditions are set to perform numerical simulation calculations under different operating conditions and a dataset is established. Finally, a reduced-order model of the combustion chamber head cavity is determined by using a deep learning framework, a recursive network architecture, determining the input and output format, and model training and evaluation.

[0084] In summary, this invention proposes an intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine. This method is based on a deep recurrent neural network architecture to characterize the complex non-azeotropic two-phase filling flow heat transfer mechanism inside the combustion chamber head cavity. Compared with traditional order reduction methods, the intelligent order reduction model obtained by this method shows excellent performance in both computational efficiency and accuracy. Since this method sets clear input and output variables for the intelligent order reduction model used to simulate the two-phase filling flow heat transfer process of the combustion chamber head cavity, and these variables correspond one-to-one with the interface variables of the combustion chamber head cavity in the engine system-level simulation model, this method is suitable for engine system-level simulation and can significantly improve the computational accuracy of the combustion chamber head cavity components in the engine system-level simulation model.

[0085] Based on the above-mentioned intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine, this embodiment of the invention also provides an intelligent order reduction device for the simulation model of the combustion chamber head cavity of a liquid rocket engine, see [link to relevant documentation]. Figure 4 As shown, the device may include the following modules:

[0086] Module 402 is used to build a numerical simulation model of the combustion chamber head cavity of a liquid rocket engine based on the structure and working principle of the combustion chamber head cavity.

[0087] The simulation module 404 is used to perform numerical simulation calculations based on the numerical simulation model and preset boundary conditions, and to construct a dataset based on the numerical simulation calculation results.

[0088] Training module 406 is used to train a preset neural network model based on the dataset to obtain a target reduced-order model of the liquid rocket engine combustion chamber head cavity; wherein, the input variables of the preset neural network model include the inlet mass flow rate, inlet temperature, outlet pressure and initial wall temperature of the liquid rocket engine combustion chamber head cavity, and the output variables of the preset neural network model include the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy and outlet temperature of the liquid rocket engine combustion chamber head cavity.

[0089] The aforementioned construction module 402 can also be used to: establish a geometric model of the liquid rocket engine combustion chamber head cavity based on the structure and working principle of the liquid rocket engine combustion chamber head cavity; extract the target fluid domain based on the geometric model; and establish the numerical simulation model based on the target fluid domain and the fluid dynamic parameters of the liquid rocket engine combustion chamber head cavity.

[0090] The aforementioned boundary conditions may include inlet boundary conditions, outlet boundary conditions, and wall boundary conditions. Based on this, the simulation module 404 can also be used to: set the inlet boundary condition as a mass flow inlet, set the outlet boundary condition as a pressure outlet, and set the wall boundary condition as a fluid-structure interaction heat transfer surface; and perform numerical simulation calculations of the numerical simulation model by simulating the inlet mass flow rate, wall temperature, and outlet pressure under different operating conditions, and using the inlet boundary conditions, the outlet boundary conditions, and the wall boundary conditions as constraints.

[0091] The numerical simulation results described above can include first characteristic time series of inlet mass flow rate, inlet temperature, outlet pressure, and initial wall temperature, and second characteristic time series of inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy, and outlet temperature, respectively. Based on this, the simulation module 404 can also be used to: generate multiple time series samples based on the first and second characteristic time series, and combine the obtained time series samples into the dataset. Each time series sample has corresponding input information and label information. The input information includes the inlet mass flow rate value, inlet temperature value, outlet pressure value, and initial wall temperature value corresponding to the corresponding time series sample. The label information includes the inlet pressure value, outlet mass flow rate value, outlet liquid oxygen volume fraction value, outlet specific enthalpy value, and outlet temperature value corresponding to the corresponding time series sample.

[0092] The aforementioned preset neural network model may include a gated recurrent unit; based on this, the aforementioned training module 406 may also be used to: construct a training set and a test set for the gated recurrent unit based on the dataset; use the average error as the loss function of the gated recurrent unit, use Adam as the optimizer, train the gated recurrent unit using the training set, and adjust the parameters of the gated recurrent unit using the test set during the training process, and obtain the target reduced-order model after training.

[0093] The intelligent reduction device for the simulation model of the liquid rocket engine combustion chamber head cavity provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned embodiment of the intelligent reduction method for the simulation model of the liquid rocket engine combustion chamber head cavity. For the sake of brevity, any parts not mentioned in the embodiment of the intelligent reduction device for the simulation model of the liquid rocket engine combustion chamber head cavity can be referred to the corresponding content in the aforementioned embodiment of the intelligent reduction method for the simulation model of the liquid rocket engine combustion chamber head cavity.

[0094] This invention also provides an electronic device, such as... Figure 5The diagram shows the structure of the electronic device, which includes a processor 51 and a memory 50. The memory 50 stores computer-executable instructions that can be executed by the processor 51. The processor 51 executes the computer-executable instructions to implement the intelligent reduction method for the simulation model of the combustion chamber head cavity of the liquid rocket engine described above.

[0095] exist Figure 5 In the illustrated embodiment, the electronic device further includes a bus 52 and a communication interface 53, wherein the processor 51, the communication interface 53, and the memory 50 are connected via the bus 52.

[0096] The memory 50 may include high-speed random access memory (RAM) and may also include 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 53 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 52 may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 52 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0097] Processor 51 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 51 or by instructions in software form. Processor 51 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this invention can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can 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 the memory. The processor 51 reads the information in the memory and, in conjunction with its hardware, completes the steps of the intelligent reduction method for the simulation model of the liquid rocket engine combustion chamber head cavity in the aforementioned embodiment.

[0098] This invention also provides a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are called and executed by a processor, they cause the processor to implement the aforementioned intelligent reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine. For specific implementation details, please refer to the foregoing method embodiments, which will not be repeated here.

[0099] The computer program product of the intelligent reduction method, device and electronic device for the simulation model of the combustion chamber head cavity of the liquid rocket engine provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0100] Unless otherwise specifically stated, the relative steps, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention.

[0101] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion 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.

[0102] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0103] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered 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. An intelligent order reduction method for simulation models of the combustion chamber head cavity of a liquid rocket engine, characterized in that, include: Based on the structure and working principle of the combustion chamber head cavity of a liquid rocket engine, a numerical simulation model of the combustion chamber head cavity of a liquid rocket engine is established. The numerical simulation model is used to describe the two-phase flow and evaporation boiling phenomena during the two-phase filling process of cryogenic propellant in the combustion chamber head cavity of a liquid rocket engine based on a multiphase flow model and an evaporation-condensation model. The numerical simulation model includes a continuity equation, a momentum equation, a phase equation, and an evaporation-condensation model. Numerical simulation calculations are performed based on the numerical simulation model and preset boundary conditions, and a dataset is constructed based on the numerical simulation calculation results. The preset neural network model is trained based on the dataset to obtain the target reduced-order model of the liquid rocket engine combustion chamber head cavity; wherein, the input variables of the preset neural network model include the inlet mass flow rate, inlet temperature, outlet pressure and initial wall temperature of the liquid rocket engine combustion chamber head cavity, and the output variables of the preset neural network model include the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy and outlet temperature of the liquid rocket engine combustion chamber head cavity; The evaporation-condensation model adopts the LEE model, which is as follows: in, For gradient, For density, t For time, It is the phase volume fraction. It is a vapor phase. It is a liquid phase. and These are the mass transfer rates of evaporation and condensation, respectively. When the liquid phase temperature Above saturation temperature At that time, the expression for the mass transfer rate of evaporation is: at this time The evaporation coefficient is... It is the liquid volume fraction. The density of the liquid phase; When the vapor phase temperature Below saturation temperature At that time, the expression for the mass transfer rate of condensation is: at this time The condensation coefficient is . This represents the volume fraction of the vapor phase. ρ is the density of the vapor phase.

2. The intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine according to claim 1, characterized in that, Based on the structure and working principle of the combustion chamber head cavity of a liquid rocket engine, a numerical simulation model of the combustion chamber head cavity of a liquid rocket engine is established, including: Based on the structure and working principle of the combustion chamber head cavity of a liquid rocket engine, a geometric model of the combustion chamber head cavity of a liquid rocket engine is established. The target fluid domain is extracted based on the geometric model, and the numerical simulation model is established based on the target fluid domain and the fluid dynamic parameters of the liquid rocket engine combustion chamber head cavity.

3. The intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine according to claim 2, characterized in that, The continuity equation is: in, For speed, For gradient; The momentum equation is: in, For density, t For time, It is the acceleration due to gravity. Viscosity, For interface force source item, For surface tension, It is the phase volume fraction; The phase equation is as follows: .

4. The intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine according to claim 2, characterized in that, The boundary conditions include inlet boundary conditions, outlet boundary conditions, and wall boundary conditions; Numerical simulation calculations are performed based on the numerical simulation model and preset boundary conditions, including: Set the inlet boundary condition to mass flow inlet, the outlet boundary condition to pressure outlet, and the wall boundary condition to fluid-structure interaction heat transfer surface. The numerical simulation model is performed by simulating the inlet mass flow rate, wall temperature, and outlet pressure under different operating conditions, and by using the inlet boundary conditions, the outlet boundary conditions, and the wall boundary conditions as constraints.

5. The intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine according to claim 2, characterized in that, The numerical simulation results include first characteristic time series of inlet mass flow rate, inlet temperature, outlet pressure and initial wall temperature, and second characteristic time series of inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy and outlet temperature. A dataset was constructed based on the results of numerical simulation calculations, including: Multiple time series samples are generated based on the first feature time series and the second feature time series, and the obtained time series samples are combined into the dataset; wherein, each time series sample has corresponding input information and label information, the input information includes the inlet mass flow rate value, inlet temperature value, outlet pressure value and initial wall temperature value corresponding to the corresponding time series sample, and the label information includes the inlet pressure value, outlet mass flow rate value, outlet liquid oxygen volume fraction value, outlet specific enthalpy value and outlet temperature value corresponding to the corresponding time series sample.

6. The intelligent order reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine according to claim 5, characterized in that, The preset neural network model includes a gated recurrent unit; the preset neural network model is trained based on the dataset to obtain a target reduced-order model of the liquid rocket engine combustion chamber head cavity, including: The training and test sets of the gated recurrent unit are constructed based on the dataset. The average error is used as the loss function of the gated recurrent unit, Adam is used as the optimizer, the gated recurrent unit is trained using the training set, and the parameters of the gated recurrent unit are adjusted using the test set during the training process. After training, the target reduced-order model is obtained.

7. An intelligent order reduction device for a simulation model of the combustion chamber head cavity of a liquid rocket engine, characterized in that, include: The module is used to build a numerical simulation model of the combustion chamber head cavity of a liquid rocket engine based on the structure and working principle of the combustion chamber head cavity. The numerical simulation model is used to describe the two-phase flow and evaporation boiling phenomena during the two-phase filling process of cryogenic propellant in the combustion chamber head cavity of a liquid rocket engine based on a multiphase flow model and an evaporation-condensation model. The numerical simulation model includes a continuity equation, a momentum equation, a phase equation, and an evaporation-condensation model. The simulation module is used to perform numerical simulation calculations based on the numerical simulation model and preset boundary conditions, and to construct a dataset based on the numerical simulation calculation results. The training module is used to train a preset neural network model based on the dataset to obtain a target reduced-order model of the liquid rocket engine combustion chamber head cavity; wherein, the input variables of the preset neural network model include the inlet mass flow rate, inlet temperature, outlet pressure and initial wall temperature of the liquid rocket engine combustion chamber head cavity, and the output variables of the preset neural network model include the inlet pressure, outlet mass flow rate, outlet liquid oxygen volume fraction, outlet specific enthalpy and outlet temperature of the liquid rocket engine combustion chamber head cavity; The evaporation-condensation model adopts the LEE model, which is as follows: in, For gradient, For density, t For time, It is the phase volume fraction. It is a vapor phase. It is a liquid phase. and These are the mass transfer rates of evaporation and condensation, respectively. When the liquid phase temperature Above saturation temperature At that time, the expression for the mass transfer rate of evaporation is: at this time The evaporation coefficient is... It is the liquid volume fraction. The density of the liquid phase; When the vapor phase temperature Below saturation temperature At that time, the expression for the mass transfer rate of condensation is: at this time The condensation coefficient is . This represents the volume fraction of the vapor phase. ρ is the density of the vapor phase.

8. An electronic device, characterized in that, The system includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the intelligent reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the intelligent reduction method for the simulation model of the combustion chamber head cavity of a liquid rocket engine as described in any one of claims 1 to 6.

Images