An engine plume flow field generation method and device based on characteristic equation guidance

Abstract

The application provides an engine plume flow field generation method and device based on characteristic equation guidance, and relates to the technical field of plume flow field prediction. The method comprises the following steps: inputting fuel components, engine parameters and current working conditions of an obtained target engine into a pre-trained prediction model to output a predicted plume flow field distribution; acquiring physical information of the target engine in the running process; wherein the physical information comprises a plurality of characteristic equations; evaluating the predicted plume flow field distribution according to the characteristic equations to obtain an evaluation result; and when the evaluation result meets a preset condition, determining that the predicted plume flow field distribution is a target plume flow field. The scheme realizes rapid and accurate growth of the engine plume flow field, and significantly improves the accuracy and interpretability of the flow field generation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of plume flow field prediction, in particular to an engine plume flow field generation method and device based on characteristic equation guidance. BACKGROUND

[0002] Engine plume flow field calculation is an important research topic in the field of aerospace, which is complex in mechanism and usually involves multiple disciplines such as fluid mechanics, thermodynamics and chemical kinetics. It is usually calculated by fluid dynamics learning software CFD, but the calculation speed is slow, and the calculation efficiency is usually hours, which cannot meet the rapid and high-precision application requirements of obtaining target plume in military scenarios.

[0003] In order to realize the efficient and high-precision generation of engine plume flow field, the existing methods include: (1) Increase the computing power, use distributed or high-performance cluster computers for joint calculation, and provide the possibility of calculation speed from the hardware; But this method needs to build large-scale cluster equipment which is difficult to maintain, and on the other hand, it also introduces complex operations in cluster computing, which improves the professional ability requirements of practitioners, and the maintenance cost is high. More importantly, for some operators that do not support distributed computing, they still do not have fast computing power. (2) Use engineering experience method, summarize some typical engine plume radiation fast calculation method under certain scene; But this method has strong conditional assumption and use condition, poor practicability, and often at the expense of calculation accuracy. Therefore, it is urgent to provide a method for quickly and accurately generating engine plume flow field. SUMMARY

[0004] The present application provides an engine plume flow field generation method and device based on characteristic equation guidance, which can quickly and accurately generate engine plume flow field, and has strong interpretability, does not require a large number of training samples, and has strong generalization ability, meeting real-time calculation requirements.

[0005] In the first aspect, the present application provides an engine plume flow field generation method based on characteristic equation guidance, comprising:

[0006] Input the fuel components, engine parameters and current working conditions of the target engine obtained into the pre-trained prediction model, and output the predicted plume flow field distribution;

[0007] Obtain the physical information of the target engine during operation; wherein the physical information includes a plurality of characteristic equations;

[0008] According to the characteristic equation, the predicted plume flow field distribution is evaluated to obtain an evaluation result;

[0009] determine the predicted plume flow field distribution as the target plume flow field when the evaluation result meets a preset condition.

[0010] Optionally, the pre-trained prediction model is trained by the following method:

[0011] The plume flow field data of the target engine under any working condition is unfolded in spatial dimension to obtain structured flow field data; wherein the structured flow field data includes position coordinates of each spatial node, and velocity, temperature and pressure at each spatial node;

[0012] The structured flow field data and corresponding fuel components, engine parameters and current working condition of the target engine are taken as a sample set; wherein each sample set includes fuel components, engine parameters and current working condition as input, and structured flow field data as output;

[0013] At least two sample sets are used to train the prediction model.

[0014] Optionally, the engine parameters include combustion chamber pressure, combustion chamber temperature and nozzle configuration; and the current working condition includes ambient pressure, ambient temperature, flight Mach number and ambient air humidity.

[0015] Optionally, the characteristic equation includes:

[0016]

[0017] wherein ρ represents density; u represents velocity field; ∇ represents divergence operator symbol; p represents pressure field; ∇ 2 represents Laplace operator symbol; μ represents dynamic viscosity; γ represents volume viscosity; f represents external force; φ represents internal energy; k represents thermal conductivity; T represents temperature; represents viscous dissipation; N represents the number of exhaust components; W k represents the molar mass of the kth exhaust component, and R is the gas constant.

[0018] Optionally, the evaluation of the predicted plume flow field distribution according to the characteristic equation obtains an evaluation result, including:

[0019] Obtaining velocity, temperature and pressure at each spatial node of the predicted plume flow field distribution;

[0020] Inputting the velocity, temperature and pressure at each spatial node into the characteristic equation, and determining that the evaluation result meets the preset condition when the characteristic equation is established.

[0021] Optionally, the training of the prediction model using at least two sample sets includes:

[0022] For each of the sample sets, the following are performed:

[0023] The fuel components, engine parameters and current operating conditions in the sample set are input into the prediction model to output predicted structured flow field data; and real structured flow field data corresponding to the sample set is obtained;

[0024] Frequency domain feature extraction is performed on the predicted structured flow field data and the real structured flow field data distribution to obtain predicted frequency domain features and real frequency domain features;

[0025] The predicted frequency domain features and the real frequency domain features are compared to obtain a first loss value;

[0026] The predicted structured flow field data and the real structured flow field data are compared to obtain a second loss value;

[0027] The predicted structured flow field data is extracted from all sample sets and input into the characteristic equation for evaluation to obtain an evaluation value;

[0028] The first loss value, the second loss value and the evaluation value are used to calculate a total loss value;

[0029] If the total loss value is less than a preset loss threshold, the pre-trained prediction model is obtained; if the total loss value is not less than the preset loss threshold, the total loss value is transmitted through back propagation to adjust the parameters of the prediction model for further training.

[0030] In a second aspect, the present application further provides an engine plume flow field generation device based on characteristic equation guidance, comprising:

[0031] A prediction generation module is configured to input the fuel components, engine parameters and current operating conditions of a target engine obtained into a prediction model to output predicted plume flow field distribution;

[0032] A characteristic acquisition module is configured to acquire physical information of the target engine during operation; wherein the physical information comprises a plurality of characteristic equations;

[0033] An evaluation module is configured to evaluate the predicted plume flow field distribution according to the characteristic equations to obtain an evaluation result; and when the evaluation result meets a preset condition, the predicted plume flow field distribution is determined as a target plume flow field.

[0034] In a third aspect, the present application further provides a computing device comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the engine plume flow field generation method based on characteristic equation guidance as described in any of the above aspects.

[0035] In a fourth aspect, the present application also provides a computer readable storage medium having stored thereon a computer program which, when executed in a computer, causes the computer to perform the method of any one of the preceding aspects.

[0036] In a fifth aspect, the present application also provides a computer program product comprising computer instructions which, when executed by a processor, implement the steps of the method of any one of the first aspects.

[0037] The present application provides a method and device for generating engine plume flow field based on characteristic equation guidance. The method inputs fuel components, engine parameters and current working conditions of a target engine into a pre-trained prediction model, outputs a predicted plume flow field distribution, and then evaluates the predicted plume flow field distribution through each characteristic equation contained in physical information to determine a target plume flow field. In this way, the present application not only can quickly generate a target plume flow field, but also ensures the accuracy and interpretability of the generated target plume flow field. BRIEF DESCRIPTION OF DRAWINGS

[0038] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0039] Figure 1 is a flow chart of a method for generating engine plume flow field based on characteristic equation guidance provided by an embodiment of the present application;

[0040] Figure 2 is a distribution map of engine plume pressure field provided by an embodiment of the present application;

[0041] Figure 3 is a distribution map of engine plume temperature field provided by an embodiment of the present application;

[0042] Figure 4 is a distribution map of engine plume velocity field provided by an embodiment of the present application;

[0043] Figure 5 is a hardware architecture diagram of a computing device provided by an embodiment of the present application;

[0044] Figure 6 is a structure diagram of a device for generating engine plume flow field based on characteristic equation guidance provided by an embodiment of the present application. DETAILED DESCRIPTION

[0045] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0046] The existing engine plume flow field intelligent generation method based on deep learning can consider flow field elements as a picture directly through the powerful fitting capability of a neural network by using a computer vision idea, and generate by training an adversarial neural network, but this method needs a large number of paired training samples on the one hand, and the generated results have poor interpretability and poor generalization capability due to the black box characteristics of the neural network, which cannot meet the application occasions with strict requirements on physical characteristics. In view of the above problems, the present application provides an engine plume flow field intelligent generation method based on characteristic equation guidance, which constrains and adjusts the generation parameters of the data-driven branch by introducing an equation describing the radiation law, reduces the demand of the intelligent network for training samples, and improves the accuracy and interpretability of the flow field generation.

[0047] The concept of the present application will be described below, please refer to Figure 1 The embodiment of the present application provides an engine plume flow field generation method based on characteristic equation guidance, which comprises:

[0048] Step 100, input the fuel components, engine parameters and current working conditions of the target engine into the pre-trained prediction model, and output the predicted plume flow field distribution;

[0049] Step 102, acquiring physical information of the target engine during operation; wherein the physical information comprises a plurality of characteristic equations;

[0050] Step 104, evaluating the predicted plume flow field distribution according to the characteristic equations to obtain an evaluation result;

[0051] Step 106, when the evaluation result meets the preset condition, determining that the predicted plume flow field distribution is the target plume flow field.

[0052] In the embodiment of the present application, the fuel components, engine parameters and current working conditions of the target engine are input into the pre-trained prediction model, and the predicted plume flow field distribution is output, and then the predicted plume flow field distribution is evaluated by the characteristic equations contained in the physical information to determine the target plume flow field. In this way, the present application not only can quickly generate the target plume flow field, but also ensures the accuracy and interpretability of the generated target plume flow field.

[0053] The concept of the present application will be described below, please refer toFigure 1 the execution mode of each step.

[0054] Firstly, in step 100, the engine parameters include the combustion chamber pressure, the combustion chamber temperature and the nozzle configuration; the current working conditions include the ambient pressure, the ambient temperature, the flight Mach number and the ambient air humidity.

[0055] In a preferred embodiment, the pre-trained prediction model is trained by the following method:

[0056] The plume flow field data of the target engine under any working condition is expanded in spatial dimension to obtain structured flow field data; wherein the structured flow field data includes the position coordinates of each spatial node, and the velocity, temperature and pressure at each spatial node;

[0057] The structured flow field data and the fuel components, engine parameters and current working conditions of the corresponding target engine are taken as a sample set; wherein each sample set includes the fuel components, engine parameters and current working conditions as input, and the structured flow field data as output;

[0058] At least two sample sets are used to train the prediction model.

[0059] In the embodiment of the present application, the plume flow field data generated by CFD cannot be directly used for the training of neural network, and further information extraction is required to form a representation form that can be used to generate a neural network framework. Specifically, considering the grid problem required for CFD calculation, the data is discretely represented, and the plume flow field data is structured as point-by-point spatial and characteristic data combination (x, y, z, u, T, p); wherein x, y, z represent the spatial coordinate position, and u, T, p represent the velocity, temperature and pressure at x, y, z spatial coordinates respectively. It should be noted that a typical engine type forms a corresponding structured representation set, forming a sample set that can be used for training and testing.

[0060] In the embodiment of the present application, the sample set includes the velocity, temperature and pressure of the current engine plume, so that the prediction model trained can simultaneously predict three kinds of flow field information, rather than being limited to only a single velocity flow field or a single temperature flow field.

[0061] In a more preferred embodiment, at least two sample sets are used to train the prediction model, including:

[0062] For each sample set, the following is performed:

[0063] The fuel components, engine parameters and current working conditions in the sample set are input into the prediction model to output the predicted structured flow field data; and the real structured flow field data corresponding to the sample set is obtained;

[0064] extracting frequency domain features from the predicted structured flow field data and the real structured flow field data distribution to obtain predicted frequency domain features and real frequency domain features;

[0065] comparing the predicted frequency domain features and the real frequency domain features to obtain a first loss value;

[0066] comparing the predicted structured flow field data and the real structured flow field data to obtain a second loss value;

[0067] extracting the predicted structured flow field data from all sample sets to input into a characteristic equation for evaluation to obtain an evaluation value;

[0068] calculating a total loss value according to the first loss value, the second loss value and the evaluation value;

[0069] If the total loss value is less than a preset loss threshold, a pre-trained prediction model is obtained; if the total loss value is not less than the preset loss threshold, the total loss value is transmitted through back propagation to adjust the parameters of the prediction model for further training.

[0070] In the embodiment of the present application, the training process needs to collect information from both the data-driven branch and the knowledge-driven branch, wherein the data-driven branch uses sample sets to predict structured flow field data through a neural network, and the knowledge-driven branch extracts the predicted structured flow field data from the sample sets predicted by the neural network and calculates the residual error between the predicted structured flow field data and the characteristic equation, i.e. the evaluation value. Then, the second loss value between the real and predicted structured flow field data and the first loss value between the real and predicted frequency domain features are calculated, the total loss of the neural network is calculated by dynamic weighted summation, and the total loss is transmitted to the neurons in the network architecture through back propagation and the parameters are adjusted. Then, the iteration is repeated until convergence, and the final pre-trained prediction model is obtained. In this way, in the embodiment of the present application, the evaluation value is calculated by the knowledge-driven branch, which does not increase the calculation amount of the data-driven branch, i.e. the data-driven branch and the knowledge-driven branch are parallel, which can further reduce the calculation amount of the data-driven branch. At the same time, since the characteristic equation describing the radiation law is introduced to constrain and adjust the predicted structured flow field data generated by the data-driven branch, the fast and accurate generation of the engine plume flow field is finally realized by the neural network fusion method. The introduction of the characteristic equation greatly reduces the demand for training samples of the neural network, significantly improves the accuracy and interpretability of the plume flow field generation, and reduces the calculation efficiency from hours to seconds, which has important significance for supporting the fast, accurate and intelligent generation of radiation. It should be noted that the evaluation value is calculated by the calculation module of the knowledge-driven branch, the first loss value and the second loss value are calculated by the calculation module of the data-driven branch, and the two calculation modules run in parallel.

[0071] Specifically, the data-driven branch can adopt the form of an encoding and decoding framework or an adversarial neural network, while the knowledge-driven branch is in the form of a fully connected neural network, and the evaluation value of the prediction structured flow field data output by the data-driven branch is calculated, and the evaluation value is fused with the loss function in the training process to obtain the final total loss value.

[0072] Specifically, the total loss value is determined by the following formula:

[0073]

[0074] Wherein, L toatl is the total loss value; λ is the first attenuation coefficient; d is the current iteration number in the training process, that is, the current training round; loss1 is the first loss value; loss2 is the second loss value; S c is the evaluation value. For example, λ = 0.02.

[0075] In the embodiment of the application, the weight of the first loss value increases with the training round, which can automatically adapt to the training accuracy; the dynamic weight value of the second loss value can automatically respond to the loss ratio and adjust the influence of the loss value brought by the frequency domain feature; and the evaluation value can maintain the constraint of physical information. In this way, the application introduces a dynamic weight mechanism, which not only avoids the extreme of the weight value while maintaining the definition of the original loss term, but also significantly enhances the expression ability and optimization effect of the loss function.

[0076] In step 102, since the flight speed of the aircraft to which the engine belongs is high enough, a compressible fluid needs to be modeled, and for the compressible fluid, the characteristic equation includes:

[0077]

[0078] Wherein, ρ represents density; u represents velocity field; ▽ represents divergence operator symbol; p represents pressure field; ▽ 2 represents Laplace operator symbol; μ represents dynamic viscosity; γ represents volume viscosity; f represents external force; φ represents internal energy; k represents thermal conductivity; T represents temperature; represents viscous dissipation; N represents the number of exhaust components; W k represents the molar mass of the kth exhaust component, and R is the gas constant.

[0079] Specifically, the present application describes the physical information by using partial differential equations, and performs non-dimensional processing, which is convenient for converting radiation into knowledge structure that can be recognized by neural networks. For fast analysis targets, a compressible fluid is used for modeling, which is much more complex than an incompressible fluid in the description process, and the density is no longer a constant. Through the mass conservation equation, the following can be obtained:

[0080]

[0081] Wherein, ρ represents density;u represents velocity field;▽ represents divergence operator symbol;The first term is used for describing the density change, and the second term is used for describing the coupling relationship between density and velocity field;

[0082] But only by the above equation cannot solve two parameters, need further introduction momentum conservation:

[0083]

[0084] Wherein, p represents pressure field;▽ 2 Laplace operator symbol;μ represents dynamic viscosity;γ represents volume viscosity;f represents external force (including gravity, etc.);In the equation Convection term, Pressure term, for describing the pressure distribution of the plume fluid; Diffusion term, for describing the complex diffusion physical process of flow field, f for describing the action of external force on the whole flow field;

[0085] Considering the effect of temperature in the whole flow field, further introduce energy conservation equation:

[0086]

[0087] Wherein, φ represents internal energy;k represents thermal conductivity;T represents temperature; Viscous dissipation;Wherein, internal energy includes combustion heat and heat energy inside the engine;

[0088] At the same time, considering that the engine plume is a complex distribution composed of multiple components, therefore, further consider the equation of state of multiple component gas:

[0089]

[0090] Wherein, p represents total pressure of mixed gas;N represents the number of tail gas components;W k The molar mass of the kth tail gas component, R is the gas constant, R=8.314;Tail gas components are determined by fuel components, engine parameters and current working condition.

[0091] In the embodiment of the application, by guiding the radiation knowledge of compressible flow, the deduction and dimensionless conversion of mass conservation, energy conservation, momentum conservation, multi-component gas state equation and other description operators are carried out, the core structure of engine plume flow field intelligent generation based on characteristic equation guidance is formed, the constraint and quality evaluation of predicted structured flow field data are realized, and the theoretical basis for realizing accurate generation is laid.

[0092] In a preferred embodiment, the exhaust plume of a slow engine is modeled using an incompressible fluid with a constant density. This can be obtained using the mass conservation equation:

[0093] By conservation of momentum, we can obtain:

[0094] In step 104, the predicted exhaust flow field distribution is evaluated based on the characteristic equation, and the evaluation results are obtained, including:

[0095] Obtain the velocity, temperature, and pressure at each spatial node of the predicted exhaust flow field distribution;

[0096] The velocity, temperature, and pressure at each space node are input into the characteristic equation. When the characteristic equation is valid, the evaluation result is determined to meet the preset conditions.

[0097] In this embodiment of the invention, in order to further confirm that the predicted exhaust flow field distribution conforms to the physical constraints, further verification is performed based on the characteristic equation to obtain the target exhaust flow field that conforms to the physical constraints.

[0098] In one specific embodiment Figures 2 to 4 The diagram shows the exhaust pressure field, exhaust temperature field, and exhaust velocity field generated by the engine exhaust flow field of the present invention for aircraft A under operating condition a. The generation time of each flow field distribution using the method of the present invention is within 3 seconds. Compared with the traditional CFD method, its accuracy error is within 5%, which can meet the needs of most applications.

[0099] like Figure 5 , Figure 6 As shown, this embodiment of the invention provides a device for generating engine exhaust flow field based on characteristic equations. The device embodiment can be implemented through software, hardware, or a combination of both. From a hardware perspective, as... Figure 5 The diagram shown is a hardware architecture diagram of a computing device for an engine exhaust flow field generation device guided by characteristic equations, provided in an embodiment of the present invention. (Except for...) Figure 5 In addition to the processor, memory, network interface, and non-volatile memory shown, the computing device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing packets. Taking software implementation as an example, such as... Figure 6 As shown, as a logical device, it is formed by the CPU of its computing device reading the corresponding computer program from the non-volatile memory into memory and running it. This embodiment provides a device for generating engine exhaust flow field based on characteristic equations, comprising:

[0100] The prediction generation module 600 is configured to input the obtained fuel composition, engine parameters and current working condition of the target engine into a pre-trained prediction model, and output a predicted plume flow field distribution.

[0101] The characteristic acquisition module 602 is configured to acquire physical information of the target engine during operation, wherein the physical information comprises a plurality of characteristic equations.

[0102] The evaluation module 604 is configured to evaluate the predicted plume flow field distribution according to the characteristic equations to obtain an evaluation result, and determine that the predicted plume flow field distribution is the target plume flow field when the evaluation result meets a preset condition.

[0103] In some specific embodiments, the prediction generation module 600 can be configured to perform the step 100, the characteristic acquisition module 602 can be configured to perform the step 102, and the evaluation module 604 can be configured to perform the steps 104 and 106.

[0104] In some specific embodiments, the engine parameters comprise combustion chamber pressure, combustion chamber temperature and nozzle configuration, and the current working condition comprises ambient pressure, ambient temperature, flight Mach number and ambient air humidity.

[0105] In some specific embodiments, the training module is further configured to perform the following operations:

[0106] The plume flow field data of the target engine under any working condition is expanded in spatial dimension to obtain structured flow field data, wherein the structured flow field data comprises position coordinates of each spatial node, and velocity, temperature and pressure at each spatial node;

[0107] The structured flow field data and corresponding fuel composition, engine parameters and current working condition of the target engine are taken as a sample set, wherein each sample set comprises, as input, the fuel composition, engine parameters and current working condition, and as output, the structured flow field data;

[0108] At least two sample sets are used to train the prediction model.

[0109] In some specific embodiments, the training module is further configured to perform the following operations:

[0110] For each sample set, the following operations are performed:

[0111] The fuel composition, engine parameters and current working condition in the sample set are input into the prediction model to output predicted structured flow field data, and real structured flow field data corresponding to the sample set is acquired;

[0112] The frequency domain feature extraction is performed on the predicted structured flow field data and the real structured flow field data distribution to obtain predicted frequency domain features and real frequency domain features.

[0113] The first loss value is obtained by comparing the predicted frequency domain features and the real frequency domain features.

[0114] The second loss value is obtained by comparing the predicted structured flow field data and the real structured flow field data.

[0115] The evaluation value is obtained by extracting the predicted structured flow data from all sample sets and inputting the predicted structured flow data into the characteristic equation for evaluation.

[0116] The total loss value is calculated according to the first loss value, the second loss value and the evaluation value.

[0117] If the total loss value is less than the preset loss threshold, the pre-trained prediction model is obtained; if the total loss value is not less than the preset loss threshold, the total loss value is transmitted through back propagation to adjust the parameters of the prediction model for further training.

[0118] In some specific embodiments, the characteristic equation used by the characteristic acquisition module 602 includes:

[0119]

[0120]

[0121] wherein, ρ represents density; u represents velocity field; ∇ represents divergence operator symbol; p represents pressure field; ∇ 2 represents Laplace operator symbol; μ represents dynamic viscosity; γ represents volume viscosity; f represents external force; φ represents internal energy; k represents thermal conductivity; T represents temperature; represents viscous dissipation; N represents the number of exhaust components; W k represents the molar mass of the kth exhaust component, and R is the gas constant.

[0122] In some specific embodiments, the evaluation module 604 is further configured to perform the following operations:

[0123] The velocity, temperature and pressure of each spatial node of the predicted exhaust plume flow field distribution are obtained.

[0124] The velocity, temperature and pressure of each spatial node are input into the characteristic equation, and when the characteristic equation is established, the evaluation result is determined to meet the preset condition.

[0125] It can be understood that the structural schematic of the embodiments of the present application does not constitute a specific limitation of the engine plume flow field generation device based on characteristic equation guidance. In other embodiments of the present application, the engine plume flow field generation device based on characteristic equation guidance can include more or fewer components than the schematic, or combine certain components, or split certain components, or different component arrangement. The components shown can be implemented in hardware, software, or a combination of software and hardware.

[0126] The information interaction, execution process and the like between the modules in the device are based on the same concept as the method embodiments of the present application, and the specific content can be referred to the description in the method embodiments of the present application, which will not be described here.

[0127] The embodiments of the present application also provide a computing device, including a memory and a processor, the memory stores a computer program, and the processor executes the computer program to realize the engine plume flow field generation method based on characteristic equation guidance in any of the embodiments of the present application.

[0128] The embodiments of the present application also provide a computer readable storage medium, which stores a computer program, and the computer program, when executed by a processor, causes the processor to execute the engine plume flow field generation method based on characteristic equation guidance in any of the embodiments of the present application.

[0129] The embodiments of the present application also provide a computer program product, which includes a computer program, and a processor of a computer device reads the computer program from a computer readable storage medium, and the processor executes the computer program to cause the computer device to execute the engine plume flow field generation method based on characteristic equation guidance in any of the embodiments of the present application.

[0130] Specifically, a system or device provided with a storage medium can be provided, and the storage medium stores a software program code for realizing the functions of any of the above embodiments, and the computer (or CPU or MPU) of the system or device reads and executes the program code stored in the storage medium.

[0131] In this case, the program code read from the storage medium itself can realize the functions of any of the above embodiments, and therefore the program code and the storage medium storing the program code constitute a part of the present application.

[0132] The storage medium for providing the program code includes a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as a CD-ROM, a CD-R, a CD-RW, a DVD-ROM, a DVD-RAM, a DVD- RW, a DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer via a communication network.

[0133] Further, it should be understood that not only the program code read by the computer is executed, but also the operating system or the like operating on the computer is caused to perform part or all of the actual operation based on the instructions of the program code, thereby realizing the function of any one of the above-described embodiments.

[0134] Further, it should be understood that the program code read by the storage medium is written into the memory provided in the expansion board inserted into the computer or the memory provided in the expansion module connected to the computer, and then part or all of the actual operation is performed based on the instructions of the program code by the CPU or the like mounted on the expansion board or the expansion module, thereby realizing the function of any one of the above-described embodiments.

[0135] It should be noted that, in this document, the terms "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Also, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus including a list of elements does not only include those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or apparatus. Without more limitations, the element defined by the statement "comprises a" does not exclude the presence of additional identical elements in the process, method, article or apparatus including the element.

[0136] It should be understood by those of ordinary skill in the art that all or part of the steps of the above-described method embodiments can be completed by program instruction-related hardware, and the aforementioned program can be stored in a computer-readable storage medium, and the program performs the steps of the above-described method embodiments when executed; and the aforementioned storage medium includes various storage media that can store program codes, such as ROM, RAM, magnetic disk or optical disk.

[0137] It should be pointed out finally that the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit the same; and although the present application has been described in detail with reference to the foregoing embodiments, it should be appreciated by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features thereof can be replaced equivalently; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An engine plume flow field generation method based on characteristic equation guidance, characterized by, The method comprises the following steps: inputting the obtained fuel components, engine parameters and current working conditions of the target engine into a pre-trained prediction model to output a predicted plume flow field distribution; obtaining physical information of the target engine during operation; wherein the physical information comprises a plurality of characteristic equations; evaluating the predicted plume flow field distribution according to the characteristic equations to obtain an evaluation result; when the evaluation result meets a preset condition, determining that the predicted plume flow field distribution is a target plume flow field; the pre-trained prediction model is trained by the following method: spatial dimension expansion is performed on the plume flow field data of the target engine under any working condition to obtain structured flow field data; wherein the structured flow field data comprises position coordinates of each spatial node, and velocity, temperature and pressure at each spatial node; the structured flow field data and corresponding fuel components, engine parameters and current working conditions of the target engine are taken as a sample set; wherein each sample set comprises fuel components, engine parameters and current working conditions as input, and structured flow field data as output; at least two sample sets are used to train the prediction model; the training of the prediction model using at least two sample sets comprises: for each sample set, the following steps are performed: inputting the fuel components, engine parameters and current working conditions in the sample set into the prediction model to output predicted structured flow field data; and obtaining the real structured flow field data corresponding to the sample set; frequency domain feature extraction is performed on the predicted structured flow field data and the real structured flow field data to obtain predicted frequency domain features and real frequency domain features; comparing the predicted frequency domain features and the real frequency domain features to obtain a first loss value; comparing the predicted structured flow field data and the real structured flow field data to obtain a second loss value; extracting the predicted structured flow field data from all sample sets to input into the characteristic equation for evaluation to obtain an evaluation value; according to the first loss value, the second loss value and the evaluation value, a total loss value is calculated; if the total loss value is less than a preset loss threshold, the pre-trained prediction model is obtained; if the total loss value is not less than the preset loss threshold, the total loss value is transmitted through back propagation to adjust the parameters of the prediction model for further training; the characteristic equation comprises: where p denotes density; u denotes velocity field; denotes the divergence operator symbol; p denotes pressure field; denotes the Laplacian operator symbol; m denotes dynamic viscosity; g denotes bulk viscosity; f denotes external force; f denotes internal energy; k denotes thermal conductivity; T denotes temperature; denotes viscous dissipation; N denotes the number of tail gas components; W k denotes the molar mass of the kth tail gas component, R is the gas constant.

2. The method of claim 1, wherein, the engine parameters comprise combustion chamber pressure, combustion chamber temperature and nozzle configuration; and the current working conditions comprise environmental pressure, environmental temperature, flight Mach number and environmental air humidity.

3. The method according to claim 1 or 2, characterized in that, the evaluation of the predicted plume flow field distribution according to the characteristic equation to obtain an evaluation result comprises: obtaining the velocity, temperature and pressure at each spatial node of the predicted plume flow field distribution; inputting the velocity, temperature and pressure at each spatial node into the characteristic equation, and determining that the evaluation result meets the preset condition when the characteristic equation is established.

4. An engine plume flow field generating device based on characteristic equation guidance, characterized by, for implementing the method of any one of claims 1 to 3, comprising: The prediction generation module is configured to input the obtained fuel component, engine parameter and current working condition of the target engine into a pre-trained prediction model, and output a predicted plume flow field distribution; The characteristic acquisition module is configured to acquire physical information of the target engine during operation; wherein the physical information comprises a plurality of characteristic equations; The evaluation module is configured to evaluate the predicted plume flow field distribution according to the characteristic equations to obtain an evaluation result; and when the evaluation result meets a preset condition, determine that the predicted plume flow field distribution is a target plume flow field. 5.A computing device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the method of any one of claims 1-3. 6.A computer readable storage medium, having stored thereon a computer program, which when executed in a computer, causes the computer to perform the method of any one of claims 1-3.

7. A computer program product, characterised in that, comprising computer instructions, which when executed by a processor, implement the steps of the method of any one of claims 1-3.

Images