Method for acquiring density field of flow field and related device
By inputting the shadow image of the target flow field into the pre-trained physical information shadow network model, the problem of difficulty in efficiently obtaining quantitative data of the flow field density field in the prior art is solved, and efficient acquisition and precise evaluation of the flow field density field is achieved.
Patent Information
- Application Number
- CN202510329821.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-19
- Publication Date
- 2025-06-17
AI Technical Summary
The prior art is difficult to efficiently obtain quantitative data of the density field of the flow field, affecting the evaluation of the fluid motion state and performance.
Quantitative data of the density field of the flow field is generated by obtaining the shadow image of the target flow field and inputting it into the pre-trained physical information shadowing network model.
It realizes efficient acquisition of quantitative data of the density field of the flow field, and improves the evaluation accuracy of fluid motion state and performance.
Smart Images
Figure CN120163092A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of fluid mechanics, and particularly to a method and related device for obtaining the density field of a flow field. Background Art
[0002] The density field of a flow field is a scalar field in a three-dimensional space, and each point in the three-dimensional space has a corresponding density value. The change of the density field has an important impact on the motion state and performance of the fluid. For example, in the combustion process, the change of the density field affects the stability and efficiency of combustion; in the design of an aircraft, the density field distribution of the flow field is crucial for evaluating the aerodynamic performance of the aircraft; in environmental protection, the change of the atmospheric density field is of great significance for predicting weather and climate change. Therefore, how to efficiently obtain the quantitative data of the density of the flow field has become one of the technical problems to be solved urgently in the technical field of fluid mechanics. Summary of the Invention
[0003] Based on the above problems, the present application provides a method for obtaining the density field of a flow field to efficiently obtain the quantitative data of the density of the flow field.
[0004] The embodiments of the present application disclose the following technical solutions:
[0005] The first aspect of the present application discloses a method for obtaining the density field of a flow field, the method including:
[0006] Obtaining a shadow image of a target flow field; the shadow image is an image representing the light intensity distribution after parallel light penetrates the target flow field;
[0007] Obtaining the density field of the target flow field through a physical information shadow network model according to the shadow image; the physical information shadow network model is a pre-trained model for generating quantitative data of the density field of a flow field.
[0008] In an optional implementation manner, the training process of the physical information shadow network model includes:
[0009] Obtaining a sample shadow image of a sample flow field; the sample shadow image is an image representing the light intensity distribution after parallel light penetrates the sample flow field;
[0010] Inputting the sample shadow image into a pre-trained physical information neural network model to obtain feature data; the feature data indicates the light intensity distribution of each pixel point in the sample shadow image;
[0011] Obtaining a target residual based on the feature data and the refractive index distribution data of the sample flow field;
[0012] Based on the target residual and the preset residual threshold, train the pre-trained physics-informed neural network model to obtain the physics-informed shadow network model.
[0013] In an alternative implementation, the obtaining the density field of the target flow field by the physics-informed shadow network model according to the shadow image includes:
[0014] Determine a target physics-informed shadow network model from multiple physics-informed shadow network models based on the density range of the target flow field;
[0015] Input the shadow image into the target physics-informed shadow network model to obtain the density field of the target flow field.
[0016] In an alternative implementation, after obtaining the shadow image of the target flow field, the method further includes:
[0017] Perform normalization processing on the shadow image to obtain a target shadow image;
[0018] The inputting the shadow image into the target physics-informed shadow network model to obtain the density field of the target flow field is specifically:
[0019] Input the target shadow image into the target physics-informed shadow network model to obtain the density field of the target flow field.
[0020] In an alternative implementation, the obtaining the shadow image of the target flow field includes:
[0021] Obtain a shadow test device; the shadow test device includes a light source, a slit, a first concave mirror, a fluid test area, a second concave mirror, and an optical camera; place the light source, the slit, the first concave mirror, the fluid test area, the second concave mirror, and the optical camera in sequence along the light transmission direction; the light source is located at the focal point of the first concave mirror; the optical camera is located at the focal point of the second concave mirror; the first concave mirror reflects the light emitted by the light source; the second concave mirror reflects the light onto the optical camera;
[0022] Process the target flow field through the shadow test device to obtain the shadow image; the target flow field is located within the fluid test area.
[0023] In an alternative implementation, the step of obtaining the refractive index distribution data of the sample flow field includes:
[0024] Obtain initial light intensity distribution data; the initial light intensity distribution data is the light intensity distribution data obtained when no flow field is loaded in the fluid test area;
[0025] Obtain the test light intensity distribution data; the test light intensity distribution data is the light intensity distribution data obtained after loading the sample flow field in the fluid test area.
[0026] Obtain the first distance and the second distance; the first distance is the distance between the first boundary line and the second boundary line; the first boundary line is the boundary line of the fluid test area close to the first concave mirror; the second boundary line is the boundary line of the fluid test area close to the second concave mirror; the second distance is the distance from the boundary line of the fluid test area close to the second concave mirror to the imaging plane of the optical camera.
[0027] Substitute the initial light intensity distribution data, the test light intensity distribution data, the first distance, and the second distance into the Poisson equation to obtain the refractive index distribution data.
[0028] The second aspect of the present application discloses a device for obtaining the density field of a flow field, and the device includes:
[0029] A shadow image acquisition module, configured to acquire a shadow image of the target flow field; the shadow image is an image representing the light intensity distribution after parallel light penetrates the target flow field.
[0030] A density field acquisition module, configured to obtain the density field of the target flow field according to the shadow image through a physical information shadow network model; the physical information shadow network model is a pre-trained model for generating quantitative data of the density field of a flow field.
[0031] In an optional implementation manner, the device further includes:
[0032] A sample shadow image acquisition module, configured to acquire a sample shadow image of the sample flow field; the sample shadow image is an image representing the light intensity distribution after parallel light penetrates the sample flow field.
[0033] A feature data acquisition module, configured to input the sample shadow image into a pre-trained physical information neural network model to obtain feature data; the feature data indicates the light intensity distribution of each pixel point in the sample shadow image.
[0034] A target residual acquisition module, configured to obtain a target residual based on the feature data and the refractive index distribution data of the sample flow field.
[0035] A model training module, configured to train the pre-trained physical information neural network model based on the difference between the target residual and the preset residual threshold to obtain the physical information shadow network model.
[0036] In a third aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the steps of the method described in any implementation manner of the first aspect are implemented.
[0037] In a fourth aspect of the present application, an electronic device is provided, including:
[0038] A memory on which a computer program is stored;
[0039] A processor configured to execute the computer program in the memory to implement the steps of the method described in any implementation manner of the first aspect.
[0040] Compared with the prior art, the present application has the following beneficial effects:
[0041] A method for obtaining a density field of a flow field is disclosed in the present application, including: obtaining a shadow image of a target flow field; inputting the shadow image into a pre-trained physics-informed shadow network model to obtain the density field of the target flow field. Since the physics-informed shadow network model in the present application is a pre-trained model for generating quantitative data of the density field of a flow field; after inputting the shadow image of the target flow field into this model, quantitative data of the density field corresponding to the target flow field can be efficiently obtained. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0043] Figure 1 It is a flowchart of a method for obtaining a density field of a flow field provided by an embodiment of the present application;
[0044] Figure 2 It is a schematic structural diagram of a shadow test device provided by an embodiment of the present application;
[0045] Figure 3 It is a flowchart of a method for obtaining a physics-informed shadow network model provided by an embodiment of the present application;
[0046] Figure 4 It is a schematic structural diagram of another shadow test device provided by an embodiment of the present application;
[0047] Figure 5 It is a schematic structural diagram of a device for obtaining a density field of a flow field provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0048] As mentioned above, the density field of the flow field is a scalar field in three-dimensional space, and each point in the three-dimensional space has a corresponding density value. Obtaining the density distribution in the flow field can study key parameters such as fluid movement and flow field structure, thereby improving the understanding of flow behavior, which plays an important role in aerodynamics, meteorology, combustion, energy and industry.
[0049] Currently, the density field of the flow field can be obtained through optical testing methods, background-guided schlieren methods, and fluid mechanics numerical calculation methods. However, the optical testing method has the problem of being unable to obtain quantitative data of the density field; the background-guided schlieren method is highly sensitive to spatial resolution and can only obtain good calculation accuracy when the density gradient is large; during the calculation process of the fluid mechanics numerical calculation method, different ideal boundary conditions need to be set for different types of fluids, and the calculation is relatively cumbersome.
[0050] How to efficiently obtain quantitative data on the density of a flow field has become one of the technical problems that need to be urgently solved in the field of fluid mechanics technology.
[0051] To solve the above problems, the present application discloses a method for obtaining the density field of a flow field, including: obtaining a shadow image of a target flow field; inputting the shadow image into a pre-trained physical information shadow network model to obtain the density field of the target flow field. Since the physical information shadow network model in the present application is a pre-trained model for generating quantitative data of the density field of a flow field; after inputting the shadow image of the target flow field into the model, the quantitative data of the density field corresponding to the target flow field can be efficiently obtained.
[0052] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0053] Figure 1 A flow chart of a method for obtaining a density field of a flow field provided in an embodiment of the present application.
[0054] Combination Figure 1 As shown, the method for obtaining the density field of the flow field disclosed in the present application includes:
[0055] S101, obtaining a shadow image of a target flow field.
[0056] The flow field refers to the distribution state of a fluid with a certain velocity and direction in space. The flow direction can be classified according to different criteria. For example, according to the variation law of velocity and direction in the flow field, the flow field can be divided into one-dimensional flow field, two-dimensional flow field and three-dimensional flow field; according to the nature of velocity distribution, the flow field can be divided into rotational flow field and irrotational flow field; according to whether the physical quantities in the flow field change with time, the flow field can be divided into steady field and unsteady field, etc. There are various classification methods for the flow field, and different classification methods help to understand and study the characteristics of the flow field from different perspectives.
[0057] The flow field in this application is distinguished according to density. Specifically, in this application, the flow field with a density of 0 - 1.2 kg / m 3 is regarded as the first type of flow field; the flow field with a density of 1.2 - 3 kg / m 3 is regarded as the second type of flow field. The target flow field in this application is one of the first type of flow field or the second type of flow field.
[0058] The shadow image in this application is an image that characterizes the light intensity distribution after parallel light penetrates the target flow field. The target flow field can be processed by a shadow test device to obtain the shadow image.
[0059] Figure 2 is a schematic structural diagram of a shadow test device provided by an embodiment of this application. As shown in combination with Figure 2 the shadow test device in this application includes: a light source, a slit, a first concave mirror (concave mirror 1), a fluid test area, a second concave mirror (concave mirror 2) and an optical camera ( Figure 2 industrial camera in).
[0060] Among them, the light source, the slit, the first concave mirror (concave mirror 1), the fluid test area, the second concave mirror (concave mirror 2) and the optical camera (industrial camera) are arranged in sequence along the light transmission direction. The light source is located at the focal point of the first concave mirror; the optical camera is located at the focal point of the second concave mirror. The fluid to be tested is placed in the fluid test area, that is, in the test section. Figure 2 The fluid test area of is located at the center of the optical path.
[0061] Figure 2 After the light emitted by the light source in passes through the slit, it is incident on the first concave mirror (concave mirror 1), and the first concave mirror converts the divergent light incident on this mirror surface into parallel light; these parallel lights pass through the fluid to be tested located in the test section and then are incident on the second concave mirror; the second concave mirror collects the incident parallel light and focuses it on the imaging plane of the optical camera ( Figure 2 industrial camera in), so as to ensure that Figure 2 the change in light intensity caused by the change in the refractive index of the fluid to be tested inside the silver fluid test area can be captured by the industrial camera with high fidelity.
[0062] The shadow image of the target flow field in this application, that is, after placing the target flow field in the Figure 2 flow field test area in, start Figure 2 the shadow test device in, Figure 2 the image captured by the optical camera in.
[0063] In this application, the parameters of each component in the shadow test device are not limited, and those skilled in the art can select the parameters of each component in the shadow test device based on actual needs.
[0064] S102, through the physical information shadow network model, according to the shadow image, obtain the density field of the target flow field.
[0065] The physical information shadow network model in this application is a pre-trained model for generating quantitative data of the density field of the flow field. The specific training process of the physical information shadow network model will be introduced in detail in the subsequent embodiments of this application.
[0066] The flow field in this application is distinguished according to density. The density is in the range of 0 - 1.2 kg / m 3 is the first type of flow field; the density is in the range of 1.2 - 3 kg / m 3 is the second type of flow field.
[0067] The model parameters of the physical information shadow network model corresponding to the first type of flow field in this application are not the same as the model parameters of the physical information shadow network model corresponding to the second type of flow field.
[0068] In an optional implementation manner, after determining the type of the target flow field, that is, determining the range where the density of the target flow field is located, the target physical information shadow network model can be determined from multiple physical information shadow network models (two physical information shadow network models in this application).
[0069] For example, after determining the density range of the target flow field, the model parameters corresponding to the physical information shadow network model can be selected; then the model parameters are loaded into the physical information shadow network model to obtain the target physical information shadow network model.
[0070] For example, after determining the type of the target flow field, that is, determining the range where the density of the target flow field is located, the target physical information shadow network model that matches the type of the target flow field can be directly selected from multiple models of the physical information shadow network model.
[0071] After obtaining the target physical information shadow network model, the shadow image of the target flow field can be directly input into the target physical information shadow network model to obtain the density field of the target flow field.
[0072] However, in order to further improve the accuracy of the density field of the obtained target flow field, after obtaining the shadow image of the target flow field, the light intensity data in the shadow image can be extracted, and the light intensity data can be normalized, such as removing outliers, removing noise, and filtering operations; and the normalized data can be organized into the form of a tensor for storage; then the stored tensor-form data is input into the target physical information shadow network model to obtain a density field of the target flow field with higher accuracy.
[0073] It should be noted that since the normalization process of the light intensity data is well-known to those skilled in the art, the specific content of the normalization process will not be described in detail in this application.
[0074] This application discloses a method for obtaining the density field of a flow field, including: obtaining a shadow image of a target flow field; inputting the shadow image into a pre-trained physical information shadow network model to obtain the density field of the target flow field. Since the physical information shadow network model in this application is a pre-trained model for generating quantitative data of the density field of the flow field; therefore, after inputting the shadow image of the target flow field into this model, quantitative data of the density field corresponding to the target flow field can be efficiently obtained.
[0075] Figure 3 It is a flowchart of a method for obtaining a physical information shadow network model provided by an embodiment of this application. Combining Figure 3 As shown, the method for obtaining a physical information shadow network model disclosed in this application includes:
[0076] S301, obtaining a sample shadow image of a sample flow field.
[0077] The sample flow field in this application includes the first type of flow field and the second type of flow field divided according to density mentioned in the foregoing embodiments.
[0078] It can be understood that in order to enable the trained physical information shadow network model to generate both the density field of the first type of flow field and the density field of the second type of flow field; the first type of flow field is used as the sample flow field to train the model to be trained, and the second type of flow field is used as the sample flow field to train the model to be trained, so as to obtain the physical information shadow network model corresponding to each type of flow field in this application.
[0079] Since the process of using the first type of flow field as the sample flow field to train the model to be trained and using the second type of flow field as the sample flow field to train the model to be trained to obtain the physical information shadow network model corresponding to each type of flow field in this application is exactly the same, therefore, in this application, any one of the first type of flow field or the second type of flow field is taken as an example to detail the process of obtaining the physical information shadow network model corresponding to this type of flow field.
[0080] Refer to the method for obtaining the shadow image of the target flow field introduced in the foregoing embodiments to obtain the sample shadow image of the sample flow field. This application will not elaborate on this content further.
[0081] S302. Input the sample shadow image into the pre-trained physics-informed neural network model to obtain feature data.
[0082] In this application, the physics-informed neural network model (Physics-Informed Neural Networks, PINN) is used as the neural network model to be trained. The physics-informed neural network model is a machine learning model that combines deep learning and physics knowledge.
[0083] The working principle of the PINN model is to integrate physical knowledge into deep learning and use physical laws to guide the model, thereby improving the generalization ability of the model. This is mainly reflected in adding a physical information term, that is, the physical law followed, to the loss function of the PINN. During the training process, the PINN not only needs to minimize the data error but also minimize the physical information error to ensure that the prediction results conform to the physical laws.
[0084] In this application, after obtaining the pre-trained physics-informed neural network model, the sample shadow image is input into the pre-trained physics-informed neural network model to obtain feature data. The manifestation form of the feature data can be a one-dimensional array or multi-dimensional data. This application does not limit the specific form of the feature data.
[0085] The feature data output by the pre-trained physics-informed neural network model is data indicating the light intensity distribution of each pixel point in the sample shadow image.
[0086] S303. Based on the feature data and the refractive index distribution data of the sample flow field, obtain the target residual.
[0087] In an optional implementation manner, the following method can be used to obtain the refractive index distribution data of the sample flow field, specifically:
[0088] (1) Obtain the initial light intensity distribution data.
[0089] The initial light intensity distribution data is the light intensity distribution data obtained when the light of the light source is not affected by any disturbance.
[0090] In Figure 2 When the fluid test area in the shadow test device shown is not loaded with the fluid to be measured, start the Figure 2 shadow test device shown. Through the optical camera in the shadow test device, take the initial shadow image obtained after the light of the light source penetrates the empty fluid test area, and then perform data processing on the initial shadow image to obtain the initial light intensity distribution data.
[0091] (2) Obtain the test light intensity distribution data.
[0092] The test light intensity distribution data is the light intensity distribution data obtained when the light of the light source is affected by the flow field to be measured.
[0093] In Figure 2 When the sample flow field is loaded in the fluid test area of the shadow test device shown, start Figure 2 the shadow test device shown. Take a test shadow image of the light of the light source passing through the sample flow field through the optical camera in the shadow test device, and then perform data processing on the test shadow image to obtain the test light intensity distribution data.
[0094] (3) Obtain the first distance and the second distance.
[0095] In this application, the first distance is the distance between the first boundary line and the second boundary line; the first boundary line is the boundary line of the fluid test area close to the first concave mirror; the second boundary line is the boundary line of the fluid test area close to the second concave mirror; the second distance is the distance from the boundary line of the fluid test area close to the second concave mirror to the imaging plane of the optical camera.
[0096] Figure 4 It is a schematic structural diagram of another shadow test device provided by an embodiment of this application. Figure 4 The shadow test device in Figure 2 is an abstract expression of the shadow test device in Figure 4 The parallel light in Figure 2 is the light reflected by the concave mirror 1 in Figure 4 The imaging plane in Figure 2 is the imaging plane of the industrial camera in Figure 4 As shown in combination with Figure 4 the first distance in this application is Figure 4 D in
[0097] Substitute the initial light intensity distribution data, the test light intensity distribution data, the first distance, and the second distance into the Poisson equation to obtain the refractive index distribution data.
[0098] The expression of the Poisson equation is as shown in formula (1):
[0099]
[0100] In formula (1), I0 is the initial light intensity distribution data, (x i , y i ) is the coordinate point on the first boundary line of the fluid test area, I0(x i , y i) is the light intensity at the coordinate of the first boundary line (x i , y i ) of the fluid where the light of the light source is incident; I s is the test light intensity distribution data, (x s , y s ) is the coordinate point on the imaging screen of the optical camera ( Figure 4 the imaging plane in); I s (x s , y s ) is the light intensity at the coordinate of (x s , y s ) where the light of the light source is incident on the imaging screen of the optical camera; the light intensity at the point (x s , y s ) on the screen is the result of several light beams moving from the position (x i , y i ) and mapping to the point x s and y s on the screen; D is the first distance; L is the second distance; n(x, y) is the average refractive index of the fluid with a length of D in the light propagation direction (z direction) of the light of the light source.
[0101] After obtaining the characteristic data and the refractive index distribution data of the flow field, the target residual can be obtained based on formula (2). Formula (2) is specifically:
[0102]
[0103] I in formula (2) s,norm is the characteristic data; n(x, y) is the average refractive index of the fluid with a length of D in the light propagation direction (z direction) of the light of the light source.
[0104] S304. Based on the difference between the target residual and the preset residual threshold, train the pre-trained physics-informed neural network model to obtain the physics-informed shadow network model.
[0105] In this application, a physics neural network model is used as the pre-trained model. The physics neural network model includes various residuals, including but not limited to: numerical calculation residuals, gradient residuals, and boundary condition residuals, etc. On the basis of the above traditional residuals, this application adds physics-informed residuals, that is, the target residuals calculated in S303 include physics-informed residuals and traditional data residuals.
[0106] Compare the sum value of multiple residuals in the physical information residual physical network neural model in this application with a preset residual threshold; if the residual sum value is greater than the preset residual threshold, adjust the parameters of the pre-trained physical information neural network model, calculate the sum value of the new multiple residuals, and stop training until the sum value of the calculated multiple residuals is less than the preset residual threshold, and obtain the trained physical information shadow network model.
[0107] Based on the method for obtaining the density field of the flow field disclosed in the foregoing embodiments, this application further discloses an apparatus for obtaining the density field of the flow field. Figure 5 It is a schematic structural diagram of an apparatus for obtaining the density field of a flow field provided in an embodiment of this application. Combining Figure 5 As shown, the density field acquisition apparatus 500 disclosed in this application includes:
[0108] A shadow image acquisition module 501, configured to acquire a shadow image of a target flow field; the shadow image is an image representing the light intensity distribution after parallel light penetrates the target flow field;
[0109] A density field acquisition module 502, configured to obtain the density field of the target flow field according to the shadow image through a physical information shadow network model; the physical information shadow network model is a pre-trained model for generating quantitative data of the density field of the flow field.
[0110] In an optional implementation manner, the density field acquisition apparatus 500 further includes:
[0111] A sample shadow image acquisition module 503, configured to acquire a sample shadow image of a sample flow field; the sample shadow image is an image representing the light intensity distribution after parallel light penetrates the sample flow field;
[0112] A feature data acquisition module 504, configured to input the sample shadow image into a pre-trained physical information neural network model to obtain feature data; the feature data indicates the light intensity distribution of each pixel point in the sample shadow image;
[0113] A target residual acquisition module 505, configured to obtain a target residual based on the feature data and the refractive index distribution data of the sample flow field;
[0114] A model training module 506, configured to train the pre-trained physical information neural network model based on the difference between the target residual and the preset residual threshold to obtain the physical information shadow network model.
[0115] In an optional implementation manner, the density field acquisition module 502 includes:
[0116] A target model acquisition unit, configured to determine a target physical information shadow network model from multiple physical information shadow network models based on the density range of the target flow field;
[0117] A density field acquisition unit, configured to input the shadow image into the target physical information shadow network model to obtain the density field of the target flow field.
[0118] In an optional implementation manner, the density field acquisition unit includes:
[0119] A density field acquisition subunit, configured to input the target shadow image into the target physical information shadow network model to obtain the density field of the target flow field.
[0120] In an optional implementation manner, the density field acquisition device 500 further includes:
[0121] A test device acquisition module, configured to acquire a shadow test device; the shadow test device includes a light source, a slit, a first concave mirror, a fluid test area, a second concave mirror, and an optical camera; the light source, the slit, the first concave mirror, the fluid test area, the second concave mirror, and the optical camera are arranged in sequence along the light transmission direction; the light source is located at the focal point of the first concave mirror; the optical camera is located at the focal point of the second concave mirror;
[0122] A shadow image acquisition module, configured to process the target flow field through the shadow test device to obtain the shadow image; the target flow field is located within the fluid test area.
[0123] In an optional implementation manner, the target residual acquisition module 505 includes:
[0124] An initial light intensity acquisition unit, configured to acquire initial light intensity distribution data; the initial light intensity distribution data is the light intensity distribution data obtained when no flow field is loaded in the fluid test area;
[0125] A test light intensity acquisition unit, configured to acquire test light intensity distribution data; the test light intensity distribution data is the light intensity distribution data obtained after the sample flow field is loaded in the fluid test area;
[0126] A distance data acquisition unit, configured to acquire a first distance and a second distance; the first distance is the distance between a first boundary line and a second boundary line; the first boundary line is the boundary line of the fluid test area close to the first concave mirror; the second boundary line is the boundary line of the fluid test area close to the second concave mirror; the second distance is the distance from the boundary line of the fluid test area close to the second concave mirror to the imaging plane of the optical camera;
[0127] The refractive index data acquisition unit is configured to substitute the initial light intensity distribution data, the test light intensity distribution data, the first distance, and the second distance into the Poisson equation to obtain the refractive index distribution data.
[0128] Based on the apparatus and method for obtaining the density field of a flow field provided in the foregoing embodiments, correspondingly, the present application further provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, some or all of the steps in the foregoing method for obtaining the density field of a flow field are implemented.
[0129] Based on the apparatus and method for obtaining the density field of a flow field provided in the foregoing embodiments, the present application further provides an electronic device, including:
[0130] A memory, on which a computer program is stored;
[0131] A processor, configured to execute the computer program in the memory to implement some or all of the steps in the foregoing method for obtaining the density field of a flow field provided in the foregoing embodiments.
[0132] It should be noted that the embodiments in this specification are all described in a progressive manner. Similar parts among the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the apparatus embodiments, since they are basically similar to the method embodiments, they are described relatively simply. The relevant parts can be referred to the partial descriptions of the method embodiments. The apparatus embodiments described above are only illustrative. The units described as separate components may or may not be physically separated, and the components indicated as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative efforts.
[0133] The above is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.
Claims
1. A method for obtaining a density field of a flow field, characterized in that: The method comprises: Acquire a shadow image of the target flow field; the shadow image is an image representing the light intensity distribution after the parallel light penetrates the target flow field; The density field of the target flow field is obtained according to the shadow image through a physical information shadow network model; the physical information shadow network model is a pre-trained model for generating quantitative data of the density field of the flow field.
2. The method according to claim 1, characterized in that The training process of the physical information shadow network model includes: Acquire a sample shadow image of the sample flow field; the sample shadow image is an image representing the light intensity distribution after the parallel light penetrates the sample flow field; Inputting the sample shadow image into a pre-trained physical information neural network model to obtain feature data; the feature data indicates the light intensity distribution of each pixel in the sample shadow image; Obtaining a target residual based on the refractive index distribution data of the sample flow field of the characteristic data; Based on the target residual and the preset residual threshold, the pre-trained physical information neural network model is trained to obtain the physical information shadow network model.
3. The method according to claim 1, characterized in that The method of obtaining the density field of the target flow field according to the shadow image through the physical information shadow network model includes: Determining a target physical information shadow network model from a plurality of physical information shadow network models based on a density range of the target flow field; The shadow image is input into the target physical information shadow network model to obtain the density field of the target flow field.
4. The method according to claim 3, characterized in that: After obtaining the shadow image of the target flow field, the method further includes: Normalizing the shadow image to obtain a target shadow image; The shadow image is input into the target physical information shadow network model to obtain the density field of the target flow field, specifically: The target shadow image is input into the target physical information shadow network model to obtain the density field of the target flow field.
5. The method according to claim 2, characterized in that: The step of obtaining a shadow image of a target flow field comprises: Obtain a shadow testing device; the shadow testing device comprises a light source, a slit, a first concave mirror, a fluid testing area, a second concave mirror and an optical camera; the light source, the slit, the first concave mirror, the fluid testing area, the second concave mirror and the optical camera are sequentially placed along the transmission direction of light; the light source is located at the focus of the first concave mirror; the optical camera is located at the focus of the second concave mirror; the first concave mirror reflects the light emitted by the light source; the second concave mirror reflects the light onto the optical camera; The target flow field is processed by the shadow testing device to obtain the shadow image; the target flow field is located in the fluid testing area.
6. The method according to claim 5, characterized in that The step of acquiring the refractive index distribution data of the sample flow field comprises: Acquiring initial light intensity distribution data; the initial light intensity distribution data is light intensity distribution data acquired when no flow field is loaded in the fluid test area; Acquire test light intensity distribution data; the test light intensity distribution data is light intensity distribution data acquired after the sample flow field is loaded in the fluid test area; Acquire a first distance and a second distance; the first distance is the distance between the first boundary line and the second boundary line; the first boundary line is the boundary line of the fluid testing area close to the first concave mirror; the second boundary line is the boundary line of the fluid testing area close to the second concave mirror; the second distance is the distance from the boundary line of the fluid testing area close to the second concave mirror to the imaging plane of the optical camera; Substituting the initial light intensity distribution data, the test light intensity distribution data, the first distance and the second distance into the Poisson equation, the refractive index distribution data is obtained.
7. A device for obtaining a density field of a flow field, characterized in that: The device comprises: A shadow image acquisition module, used to acquire a shadow image of a target flow field; the shadow image is an image representing the light intensity distribution after parallel light penetrates the target flow field; The density field acquisition module is used to obtain the density field of the target flow field according to the shadow image through a physical information shadow network model; the physical information shadow network model is a pre-trained model for generating quantitative data of the density field of the flow field.
8. The device according to claim 7, characterized in that The device also includes: A sample shadow image acquisition module is used to acquire a sample shadow image of a sample flow field; the sample shadow image is an image representing the light intensity distribution after parallel light penetrates the sample flow field; A feature data acquisition module, used to input the sample shadow image into a pre-trained physical information neural network model to obtain feature data; the feature data indicates the light intensity distribution of each pixel in the sample shadow image; A target residual acquisition module, used to obtain a target residual based on the characteristic data and the refractive index distribution data of the sample flow field; The model training module is used to train the pre-trained physical information neural network model based on the difference between the target residual and the preset residual threshold to obtain the physical information shadow network model.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 6 are implemented.
10. An electronic device, characterized in that: include: a memory having a computer program stored thereon; A processor, configured to execute the computer program in the memory to implement the steps of the method according to any one of claims 1 to 6.