Local Aerodynamic Heat Optimization Method and Device Based on Discrete Adjoint

Abstract

The present disclosure provides a local aerodynamic heat optimization method and device based on discrete adjoint. The method includes: drawing a structural grid of the local component under the oncoming flow condition according to the initial shape of the local component to be optimized and the oncoming flow condition; obtaining a control box required for grid deformation of the structural grid, the interior of the control box contains grid boundary points to be optimized, and the grid boundary points form the physical surface of the local component; optimizing and updating the control box to obtain the optimized coordinates of the control box; optimizing the initial shape according to the optimized coordinates and the oncoming flow condition to obtain the target shape of the local component. In this embodiment, the shape of the local component can be optimized with the aerodynamic heat environment as the objective function, and a large heat reduction benefit can be obtained at the cost of a small shape deformation without affecting the overall aerodynamic performance of the aircraft.

Description

Technical Field

[0001] The present disclosure relates to the technical field of aircraft design, and particularly to a local aerodynamic heat optimization method and device based on discrete adjoint. Background Art

[0002] In the field of aircraft design, aerodynamic heat is usually a critical constraint. Taking hypersonic aircraft as an example, the heat protection requirements can be met by increasing the thickness of the heat protection layer of the aircraft.

[0003] With the continuous development of hypersonic aircraft towards high Mach numbers, long endurance, reusability, etc., the temperature of the aircraft gradually exceeds the material limit, and the optimization of the aerodynamic heat profile is extremely urgent.

[0004] In the existing aircraft shape design methods, the requirements for aerodynamic force, aerodynamic heat, and other aspects are designed separately, and each requirement is met one by one, and the design requirements are finally achieved through repeated iterations.

[0005] However, as a highly integrated complex system, hypersonic aircraft face a flight environment with strong nonlinear characteristics, making it difficult to effectively utilize traditional methods in the shape design of hypersonic aircraft. Summary of the Invention

[0006] The present disclosure provides a local aerodynamic heat optimization method and device based on discrete adjoint.

[0007] According to the first aspect of the present disclosure, a local aerodynamic heat optimization method based on discrete adjoint is provided. The method includes:

[0008] Drawing a structural grid of the local component under the oncoming flow condition according to the initial shape of the local component to be optimized and the oncoming flow condition;

[0009] Obtaining a control box required for grid deformation of the structural grid, where the interior of the control box contains grid boundary points to be optimized, and the grid boundary points form the surface of the local component;

[0010] Optimizing and updating the control box to obtain the optimized coordinates of the control box;

[0011] Optimizing the initial shape according to the optimized coordinates and the oncoming flow condition to obtain the target shape of the local component.

[0012] Optionally, optimizing and updating the control box to obtain the optimized coordinates of the control box includes:

[0013] Obtaining the optimization direction and optimization step length required for the control box;

[0014] Optimize the initial coordinates of the control box according to the optimization direction and the optimization step size to obtain the optimized coordinates of the control box.

[0015] Optionally, obtaining the optimization direction required by the control box includes:

[0016] Obtain the optimization objective function of the local component in the current optimization scenario;

[0017] Solve the flow field according to the oncoming flow condition to obtain the aerodynamic heat data under the initial shape;

[0018] Obtain the objective function according to the aerodynamic heat data;

[0019] Obtain the coefficient matrix according to the solved flow field and the automatic differentiation method;

[0020] Obtain the right-hand vector according to the objective function and the automatic differentiation method;

[0021] Establish an adjoint equation according to the coefficient matrix and the right-hand vector, and obtain the adjoint vector corresponding to the adjoint equation;

[0022] Obtain the objective function gradient according to the adjoint vector and the preset gradient calculation model;

[0023] Calculate the optimization direction according to the objective function gradient and the steepest descent method.

[0024] Optionally, obtaining the optimization objective function of the local component in the current optimization scenario includes:

[0025] When the optimization function in the current optimization scenario is the heat flux peak, determine the optimization objective function as the first objective function;

[0026] When the optimization function in the current optimization scenario is drag, determine the optimization objective function as the second objective function;

[0027] When the objective function in the current optimization scenario is drag without increasing the maximum heat flux, determine the optimization objective function as the third objective function;

[0028] When the objective function in the current optimization scenario is the maximum heat flux without increasing drag, determine the optimization objective function as the fourth objective function.

[0029] Optionally, obtaining the optimization step size required by the control box includes:

[0030] Use the flow field of the local component and the golden section method to iteratively calculate the step size that meets the preset conditions as the optimization step size.

[0031] Optionally, optimize the initial shape according to the optimized coordinates and the oncoming flow conditions to obtain the target shape of the local component, including:

[0032] Based on the optimized coordinates of the control box, complete the deformation of the local component shape and the corresponding aerodynamic heat environment calculation structure grid according to the FFD method to obtain a candidate shape of the local component;

[0033] Obtain the aerodynamic heat data under the candidate shape of the local component according to the optimized coordinates of the control box;

[0034] Obtain the objective function value of the optimization objective function according to the aerodynamic heat data;

[0035] When the objective function value of the candidate shape meets the optimization termination condition, determine the candidate shape as the target shape of the local component.

[0036] According to the second aspect of the present disclosure, there is provided a local aerodynamic heat optimization device based on discrete adjoint, and the device includes:

[0037] A structural grid drawing module, configured to draw a structural grid of the local component under the oncoming flow conditions according to the initial shape of the local component to be optimized and the oncoming flow conditions;

[0038] A control box acquisition module, configured to acquire a control box required for grid deformation of the structural grid, where the control box internally includes grid boundary points to be optimized, and the grid boundary points form the object surface of the local component;

[0039] An optimized coordinate acquisition module, configured to optimize and update the control box to obtain the optimized coordinates of the control box;

[0040] A target shape acquisition module, configured to optimize the initial shape according to the optimized coordinates and the oncoming flow conditions to obtain the target shape of the local component.

[0041] Optionally, the optimized coordinate acquisition module includes:

[0042] A direction step size acquisition sub-module, configured to acquire an optimization direction and an optimization step size required for the control box;

[0043] An optimized coordinate acquisition sub-module, configured to optimize the initial coordinates of the control box according to the optimization direction and the optimization step size to obtain the optimized coordinates of the control box.

[0044] Optionally, the direction step size acquisition sub-module includes:

[0045] An optimization function acquisition unit, configured to acquire an optimization objective function of the local component in the current optimization scenario;

[0046] The aerodynamic heat data acquisition unit is used to solve the flow field according to the oncoming flow conditions and obtain the aerodynamic heat data under the initial shape;

[0047] The objective function acquisition unit is used to obtain the objective function according to the aerodynamic heat data;

[0048] The coefficient matrix acquisition unit is used to obtain the coefficient matrix according to the flow field solution and the automatic differentiation method;

[0049] The right-hand vector acquisition unit is used to obtain the right-hand vector according to the objective function and the automatic differentiation method;

[0050] The adjoint vector acquisition unit is used to establish an adjoint equation according to the coefficient matrix and the right-hand vector, and obtain the adjoint vector corresponding to the adjoint equation;

[0051] The function gradient acquisition unit is used to obtain the objective function gradient according to the adjoint vector and the preset gradient calculation model;

[0052] The optimization direction acquisition unit is used to calculate the optimization direction according to the objective function gradient and the steepest descent method.

[0053] Optionally, the optimization function acquisition unit includes:

[0054] The first function determination subunit is used to determine the optimization objective function as the first objective function when the current optimization scenario is that the optimization function is the heat flux peak;

[0055] The second function determination subunit is used to determine the optimization objective function as the second objective function when the current optimization scenario is that the optimization function is the drag;

[0056] The third function determination subunit is used to determine the optimization objective function as the third objective function when the current optimization scenario is the drag under the condition that the maximum heat flux does not increase;

[0057] The fourth function determination subunit is used to determine the optimization objective function as the fourth objective function when the current optimization scenario is the maximum heat flux under the condition that the drag does not increase.

[0058] Optionally, the direction step size acquisition sub-module includes:

[0059] The optimization step size acquisition unit is used to iteratively calculate the step size that meets the preset conditions by using the flow field of the local component and the golden section method as the optimization step size.

[0060] Optionally, the target shape acquisition module includes:

[0061] A candidate shape acquisition sub-module, configured to perform deformation of a local component shape and a corresponding aerodynamic heat environment calculation structural grid according to the FFD method based on the optimized coordinates of the control box, and obtain a candidate shape of the local component;

[0062] An aerodynamic heat data acquisition sub-module, configured to acquire aerodynamic heat data under the candidate shape of the local component according to the optimized coordinates of the control box;

[0063] An objective function value acquisition sub-module, configured to acquire an objective function value of an optimization objective function according to the aerodynamic heat data;

[0064] An objective shape determination sub-module, configured to determine the candidate shape as the objective shape of the local component when the optimization objective function value of the candidate shape meets an optimization termination condition.

[0065] According to a third aspect of the present disclosure, there is provided an electronic device, including: a processor and a memory;

[0066] The memory is configured to store a computer program executable by the processor;

[0067] The processor is configured to execute the computer program in the memory to implement the method according to any one of the first aspects.

[0068] According to a fourth aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium, which can implement the method according to any one of the first aspects when an executable computer program in the storage medium is executed by a processor.

[0069] The technical solutions provided by the embodiments of the present disclosure may include the following beneficial effects:

[0070] In the solution provided in this embodiment, a structural grid of the local component under the oncoming flow condition can be drawn according to the initial shape of the local component to be optimized and the oncoming flow condition; then, a control box required for grid deformation of the structural grid is obtained, and the control box contains grid boundary points to be optimized inside, and the grid boundary points form the physical surface of the local component; then, the control box is optimized and updated to obtain the optimized coordinates of the control box; finally, the initial shape is optimized according to the optimized coordinates and the oncoming flow condition to obtain the objective shape of the local component. In this way, this embodiment can optimize the shape of the local component with the aerodynamic heat environment as the objective function, and can obtain a large heat reduction benefit at a small shape deformation cost without affecting the overall aerodynamic performance of the aircraft.

[0071] It should be understood that the above general description and subsequent detailed description are only exemplary and explanatory, and cannot limit the present disclosure. Description of the Drawings

[0072] Figure 1 Flow chart of a local aerodynamic heat optimization method based on discrete adjoint according to an embodiment of the present disclosure.

[0073] Figure 2 Schematic diagram of a structural grid of an initial shape of a local component according to an embodiment of the present disclosure.

[0074] Figure 3 Schematic diagram of drawing a control box on a structural grid according to an embodiment of the present disclosure.

[0075] Figure 4 Flow chart of obtaining optimized coordinates of a control box according to an embodiment of the present disclosure.

[0076] Figure 5 Flow chart of a local aerodynamic heat optimization method based on discrete adjoint according to an embodiment of the present disclosure.

[0077] Figure 6 Flow chart of another local aerodynamic heat optimization method based on discrete adjoint according to an embodiment of the present disclosure.

[0078] Figure 7 Flow field density cloud map of an initial shape of a local component according to an embodiment of the present disclosure.

[0079] Figure 8 Schematic diagram of the projection of the gradient of an objective function in the X direction according to an embodiment of the present disclosure.

[0080] Figure 9 Schematic diagram of the projection of the gradient of an objective function in the Y direction according to an embodiment of the present disclosure.

[0081] Figure 10 Schematic diagram of a computational grid of an aerodynamic heat environment obtained through six rounds of optimization deformation according to an embodiment of the present disclosure.

[0082] Figure 11 Flow field density cloud map corresponding to the final optimized shape obtained through six rounds of optimization according to an embodiment of the present disclosure.

[0083] Figure 12 Block diagram of a local aerodynamic heat optimization device based on discrete adjoint according to an embodiment of the present disclosure. Detailed implementation manners

[0084] Here, exemplary embodiments will be described in detail, and examples thereof are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present disclosure. On the contrary, they are merely examples of devices consistent with some aspects of the present disclosure as detailed in the appended claims.

[0085] Embodiments of the present disclosure provide a local aerodynamic heat optimization method, device, electronic device, and storage medium based on discrete adjoint. For a local aerodynamic heat optimization method based on discrete adjoint provided by the present disclosure, refer to Figure 1 , which includes steps 11 to 14.

[0086] In step 11, a structural grid of the local component under the oncoming flow condition is drawn according to the initial shape of the local component to be optimized and the oncoming flow condition.

[0087] In this step, the initial shape of the local component to be optimized in the aircraft can be obtained first. It can be understood that since the specific structure of the aircraft is known, the initial shape of its local component is also known.

[0088] In this step, the performance parameters of the aircraft can be obtained, and the oncoming flow condition of the aircraft can be determined according to the above performance parameters. When the oncoming flow condition of the aircraft is known, the oncoming flow condition of the local component is also known.

[0089] In this step, a structural grid required for calculating the aerodynamic heat environment can be drawn according to the initial shape of the local component and the oncoming flow working condition of the local component. The above structural grid needs to have good wall grid orthogonality, uniform and smooth transition, the wall normal growth ratio is controlled within 1.1, and the distance of the first layer of the wall meets the requirement that the grid Reynolds number is less than 8 (the grid Reynolds number is equal to the product of the wall distance and the Reynolds number per meter of the oncoming flow). Among them, the Reynolds number per meter of the oncoming flow is determined by the oncoming flow condition.

[0090] In one example, the automatic surface grid and automatic volume grid functions of the local component can be used by using Pointwise software; a structural grid can be established around the aircraft, and the generation of the grid can be controlled by setting parameters such as the maximum aspect ratio, mapping options, and subdivision. In another example, ANSYS ICEM CFD software is used to generate a structural grid. Those skilled in the art can select the structural grid drawing method according to the specific scenario, and the corresponding solution falls within the protection scope of the present disclosure.

[0091] Refer to Figure 2 , a structural grid 23 is sequentially drawn outward from the outer surface 22 of the local component 21, and the density of the structural grid 23 increases as the distance from the outer surface 22 increases.

[0092] In step 12, a control box required for grid deformation of the structural grid is obtained, and the interior of the control box contains grid boundary points to be optimized, and the grid boundary points form the physical surface of the local component.

[0093] In this step, a control box required for grid deformation of the structural grid is obtained. It can be understood that the control box can be drawn in the same way as the structural grid, except that the control box contains the grid boundary points to be optimized. Moreover, the size of the control box is larger than that of the structural grid, so as to reduce the calculation amount while ensuring the smoothness of the optimized outer surface of the object.

[0094] See Figure 3 , inside the control box 31, there are grid boundary points 32 to be optimized. It can be understood that the structural grid 23 surrounds the local component, so the outer surface of the local component 21 is in contact with the innermost grid edge boundary points of the structural grid, that is, the grid boundary points 32 form the outer surface of the local component 21. Then, when adjusting the position of the control box 31, the position of the grid boundary points 32 inside the control box can be adjusted synchronously; in other words, adjusting the position of the grid boundary points 32 is equivalent to adjusting the shape of the outer surface of the local component, that is, adjusting the outer shape of the local component 21.

[0095] In step 13, the control box is optimized and updated to obtain the optimized coordinates of the control box.

[0096] In this step, the control box is optimized and updated to obtain the optimized coordinates of the control box. See Figure 4 , which includes step 41 and step 42.

[0097] In step 41, the optimization direction and optimization step size required for the control box are obtained.

[0098] In an example, the optimization direction required for the control box can be obtained.

[0099] First, obtain the optimization objective function of the aircraft in the current optimization scenario, including:

[0100] (1) When the current optimization scenario is the peak heat flux, determine the optimization objective function as the first objective function. This first objective function is shown in Equation (1).

[0101] F = F max = q max (1)

[0102] In Equation (1), F max represents the first objective function for optimizing the peak heat flux, and q max represents the maximum heat flux value on the outer surface of the local component.

[0103] (2) When the current optimization scenario is drag, determine the optimization objective function as the second objective function. This second objective function is shown in Equations (2) - (4).

[0104] F = F drag = F p + Fμ (2)

[0105]

[0106] In Equations (2) to (4), F p represents piezoresistance, Ψ wall represents a grid cell containing the inclusion surface, P represents the surface pressure, P0 represents the oncoming flow pressure, S m represents the grid surface area, α represents the oncoming flow angle of attack, represents the unit normal vector of the wall surface, Re represents the Reynolds number, Δh represents the distance from the center point of the first layer of the wall grid to the wall surface, (N x , N y ) represents the unit center point normal vector of the wall surface, (u v) is the grid center velocity vector of the first grid of the wall surface.

[0107] (3) When the current optimization scenario is that the objective function is the drag force without increasing the maximum heat flux, determine that the optimization objective function is the third objective function. This third objective function is shown in Equation (5).

[0108]

[0109] In Equation (5), represents the drag force of the initial shape of the local component, F max represents the maximum heat flux value of the current shape, represents the maximum heat flux value of the initial shape, Large represents a larger coefficient (such as 1000, adjustable), and Tanh represents the hyperbolic tangent function.

[0110] (4) When the current optimization scenario is that the objective function is the maximum heat flux without increasing the drag force, determine that the optimization objective function is the fourth objective function. This fourth objective function is shown in Equation (6).

[0111]

[0112] Then, the optimization termination condition can be obtained. This optimization termination condition can be used to determine whether the optimization process terminates. In one example, the difference between the current objective function value and the previous round's objective function value satisfies: the ratio of this difference to the objective function value of the initial original shape is less than or equal to a preset ratio. The value range of this preset ratio is [0.5%, 5%]. In one example, the above preset ratio takes a value of 1%.

[0113] After that, the aerodynamic heat data under the initial shape of the local component can be obtained. For example, the hypersonic numerical simulation method can be used to solve the flow field of the above local component under the above inflow conditions, and aerodynamic heat data such as drag, heat flux distribution, and viscous drag can be obtained. Corresponding parameters can be selected as the aerodynamic heat data according to the specific scenario.

[0114] After that, the objective function value under the initial shape of the local component can be calculated according to the aerodynamic heat data and the above objective function.

[0115] Again, the adjoint equation can be established based on the objective function value, the coefficient matrix, and the right-hand vector, as shown in Equation (7).

[0116]

[0117] In Equation (7), R represents the residual of the flow field solution control equation (i.e., the NS equation (Navier-Stokes equations)), W represents the flow field variables, represents the coefficient matrix, F represents the gradient of the objective function selected in the above steps, Λ represents the adjoint variable to be solved, represents the right-hand vector (i.e., the constant term vector of the linear equation system).

[0118] It can be understood that the coefficient matrix and the right-hand vector in the above adjoint equation can be calculated using the automatic differentiation method. Then, the Block-LUSGS iterative method can be used to solve the above adjoint equation, so as to obtain the adjoint vector Λ in Equation (7). In this example, the solution efficiency of the adjoint equation can be improved.

[0119] Then, the objective function gradient can be calculated according to the adjoint vector and the preset gradient calculation model, and the calculation formulas are shown in Equations (8) to (10).

[0120]

[0121] In Equations (8) to (10), β k represents the coordinates of the kth control box, Δβ k represents the small change of the kth design variable. In one example, Δβ k takes 10 -5 .

[0122] When the objective function gradient is obtained, the optimization direction can be calculated. For example, taking the above objective function gradient as the input condition, the optimization direction can be calculated using the steepest descent method. Among them, the steepest descent method selects the direction that makes the objective function decrease fastest in each iteration step, and uses this direction as the optimization direction of the solution in this example, gradually approaching the minimum value of the objective function.

[0123] In one example, the optimized step size required for the control box can be obtained. For example, the process of calculating the optimized step size by the golden section method may include: obtaining the initial search interval [a0, b0], and calculating the relationship between the interval length and the golden ratio, where the golden ratio φ can be taken as 0.618. Then, select two trial points p and q within the above initial search region, where p = a0 + (1 - φ) * (b0 - a0); q = a0 + φ * (b0 - a0); the above two trial points p and q divide the initial search interval into three parts, and the length ratio of two of them is the golden ratio. After that, calculate the objective function values at the above two trial points p and q, that is, F(p) and F(q). When F(p) is less than F(q), the minimum value is within the interval [a0, q], and update the search interval to [a0, q]; when F(p) is greater than or equal to F(q), the minimum value is within the interval [p, b0], and update the search interval to [p, b0]; then reselect two trial points p and q within the updated search interval, and re-obtain the search interval until the preset stop condition is met (such as the search region is less than 10 -5 ), obtain the endpoints [a, b] of the smallest interval containing the minimum point, or output the optimized step size.

[0124] In step 42, optimize the initial coordinates of the control box according to the optimization direction and the optimized step size to obtain the optimized coordinates of the control box.

[0125] In this step, when the base optimization direction and the optimized step size are known, the initial coordinates of each control box can be moved according to the above optimization direction and the above optimized step size to obtain the optimized coordinates of the control box.

[0126] In step 14, optimize the initial shape according to the optimized coordinates and the oncoming flow conditions to obtain the target shape of the local component.

[0127] In this step, according to the latest coordinates of the control box, use the FFD free-form deformation method to complete the deformation of the local component shape and the corresponding computational structure grid of the aerodynamic heat environment, that is, after the local component shape changes, the structural grid of its flow field changes accordingly, and the updated structural grid still needs to meet the requirements of aerodynamic heat calculation (such as the wall normal grid and the wall orthogonality, the first layer distance meets the requirements, the transition is uniform, and the wall normal growth ratio is not greater than 1.1, etc.) to obtain the candidate shape of the local component.

[0128] It should be noted that the FFD free-form deformation is used for the mapping function between the component structural grid and the control box, and the coordinate position of the structural grid is updated by updating the coordinate position of the control box; when the coordinate position of the structural grid changes, the shape of the local component also changes.

[0129] Then, the aerodynamic heat data under the candidate shape of the local component can be obtained according to the optimized coordinates of the control box, that is, the flow field of the updated local component is solved by using the hypersonic numerical simulation method to obtain the aerodynamic heat data of the candidate shape such as drag and heat flux distribution (updated), and the objective function value of the objective function is calculated. When the optimized objective function value of the candidate shape meets the optimization termination condition, the candidate shape can be determined as the target shape of the local component.

[0130] So far, in the solution provided in this embodiment, the structural grid of the local component under the oncoming flow condition can be drawn according to the initial shape of the local component to be optimized and the oncoming flow condition; then, the control box required for grid deformation of the structural grid is obtained, and the interior of the control box contains the grid boundary points to be optimized, and the grid boundary points form the physical surface of the local component; after that, the control box is optimized and updated to obtain the optimized coordinates of the control box; finally, the initial shape is optimized according to the optimized coordinates and the oncoming flow condition to obtain the target shape of the local component. In this way, this embodiment can optimize the shape of the local component with the aerodynamic heat environment as the objective function, and can obtain a large heat reduction benefit at the cost of a small shape deformation without affecting the overall aerodynamic performance of the aircraft.

[0131] The following describes the local aerodynamic heat optimization method based on discrete adjoint with reference to examples, see Figure 5 , including steps 501 to step 514.

[0132] In step 501, the initial shape of the local configuration to be optimized and the oncoming flow condition are obtained, and the structural grid required for calculating the aerodynamic heat environment is drawn.

[0133] In step 502, the control box required for grid deformation is drawn. Among them, the physical surface boundary points inside the control box are the boundary points that need to be deformed, and the coordinates of the control box are the optimized design variables. By adjusting the coordinates of the control box, the boundary points of the physical surface inside the control box are adjusted, that is, the shape of the local component is adjusted.

[0134] In step 503, the optimization objective function is selected according to the current optimization scenario, such as the first objective function, the second objective function, the third objective function, and the fourth objective function respectively exemplified by formulas (1) to (6).

[0135] In step 504, the optimization termination condition is set, and this optimization termination condition can be used to judge whether the optimization process terminates. In one example, the difference between the current objective function value and the previous round of objective function values satisfies: the ratio of this difference to the objective function value of the initial original shape is less than or equal to the preset ratio. The value range of this preset ratio is [0.5%, 5%]. In one example, the above preset ratio is set to 1%.

[0136] In step 505, the hypersonic numerical simulation method is used to solve the flow field of the above-mentioned local component under the above-mentioned oncoming flow conditions, and aerothermal data such as drag, heat flux distribution, and viscous drag can be obtained; further, the objective function value of the objective function in step 503 is calculated.

[0137] In step 506, the automatic differentiation method is used to solve the coefficient matrix and the right-hand vector; and the adjoint equation as shown in Equation (7) is established.

[0138] In step 507, the Block-LUSGS iterative method is used to solve the above adjoint equation, so as to obtain the adjoint vector Λ in Equation (7).

[0139] In step 508, the objective function gradient is calculated using the objective function gradient calculation formulas as shown in Equations (8) to (10).

[0140] In step 509, the objective function gradient obtained in step 508 is used as the input condition, and the steepest descent method is used to calculate the optimization direction.

[0141] In step 510, the golden section method is used to calculate the optimization step size.

[0142] In step 511, using the optimization direction obtained in step 509 and the optimization step size obtained in step 510, the coordinates of the control box are optimized and updated to obtain the optimized coordinates.

[0143] In step 512, according to the optimized coordinates of the control box, the FFD method is used to complete the deformation of the structural grid of the local component's outer shape and the corresponding aerothermal environment calculation, and the candidate outer shape of the local component is obtained, realizing the grid deformation that can meet the requirements of aerothermal environment calculation and improving the effectiveness of the optimization result.

[0144] In step 513, the hypersonic numerical simulation method the same as that in step 505 is used to solve the flow field of the candidate outer shape of the local component, and the aerothermal data of the candidate outer shape such as drag and heat flux distribution are obtained, and further the objective function value of the objective function in step 503 is calculated.

[0145] In step 514, it is judged whether the objective function value of the candidate outer shape of the local component meets the optimization termination condition. If it does not meet, steps 506 to 513 are repeated; if it meets, the optimization terminates.

[0146] So far, by taking advantage of the property that the computational amount of the adjoint method gradient is independent of the number of design variables, an efficient solution to the gradient of the objective function based on the thermal environment has been achieved. Combining the calculation method of the steepest descent optimization direction and the golden section optimization step size calculation method, the local optimization of the aerodynamic thermal environment in the hypersonic field has been realized. The solution of this embodiment can solve the problem of large computational amount caused by a large number of design variables, can be applied to the field of hypersonic aerodynamic thermal environment optimization, and expands the application scope of the adjoint optimization method.

[0147] The adjustment process of a spherical head flat plate shape based on the solution of the above embodiment is as Figure 6 shown. Among them, the spherical head radius of the spherical head flat plate shape is 20 mm, the flat plate length is 500 mm, the flight conditions are 60 km, Mach number 10, angle of attack 0, and laminar flow.

[0148] First, draw the computational structural grid of the aerodynamic thermal environment according to the spherical head flat plate shape size and flight conditions, where the structural grid is as Figure 2 shown.

[0149] Second, set the optimization objective as the drag force without increasing the maximum heat flux, that is, select the third objective function.

[0150] Third, set the optimization termination condition as the difference between the objective function value of the current optimized shape and the objective function value of the previous round of optimized shape is less than or equal to 1% of the objective function value of the initial shape.

[0151] Fourth, draw the control frames required for the optimization of the spherical head flat plate shape, where the control frames are as Figure 3 shown. The number of each control frame is 2 * 10. The surface boundary points inside the control frame are the boundary points that need to be deformed, and the coordinates of the control frame are the optimized design variables. The control frame is as Figure 3 shown.

[0152] Fifth, use the hypersonic numerical simulation method to solve the flow field of the initial shape of the local component, obtain the maximum heat flux value and drag force value of the whole field, and after substituting them into the calculation formula of the third objective function, the objective function value of the initial shape can be calculated. Among them, the flow field density cloud map of the initial shape obtained by the flow field solution is as Figure 7 shown.

[0153] Sixth, use the automatic differentiation method to find the coefficient matrix and the right-hand vector and establish the adjoint equation as shown in Equation (7).

[0154] Seventh, use the Block-LUSGS iterative method to solve the adjoint equation and obtain the adjoint variables.

[0155] Eighth, calculate the gradient of the objective function using the objective function gradient calculation formula. The distribution nephogram of the projection of the objective function gradient in the X direction is as shown in Figure 8 shown, and the distribution nephogram of the projection in the Y direction is as shown in Figure 9 shown.

[0156] Ninth, calculate the optimization direction using the steepest descent method. The optimization direction is the opposite direction of the objective function gradient direction.

[0157] Tenth, calculate the optimization step size using the golden section method.

[0158] Eleventh, update the coordinates of the control box according to the optimization direction and the optimization step size.

[0159] Twelfth, according to the latest coordinates of the control box, use the FFD method to complete the deformation of the local component shape and the corresponding computational structure grid of the aerodynamic heat environment.

[0160] Thirteenth, use the hypersonic numerical simulation method to solve the flow field of the new shape and obtain the flow field of the new shape and the corresponding objective function value.

[0161] Fourteenth, perform an optimization termination judgment according to the objective function value of the new shape. When the optimization termination condition is not satisfied, repeat steps five to thirteen until the objective function value obtained from the optimized shape in the sixth round satisfies the optimization termination condition. Among them, the computational grid of the aerodynamic heat environment obtained through six rounds of optimization deformation is as shown in Figure 10 shown, and the flow field density nephogram corresponding to the final optimized shape obtained through six rounds of optimization is as shown in Figure 11 shown.

[0162] It can be understood that in the solution of this embodiment, the gradient of the objective function in the case of a large number of design variables can be obtained through the adjoint method, and the local shape optimization with the aerodynamic heat environment as the objective function can be realized by using the objective function gradient. Without affecting the overall aerodynamic performance of the aircraft, a large heat reduction benefit can be obtained at the cost of a very small shape deformation.

[0163] Based on a local aerodynamic heat optimization method based on discrete adjoint provided in an embodiment of the present disclosure, an embodiment of the present disclosure further provides a local aerodynamic heat optimization device based on discrete adjoint. Refer to Figure 12 and the device includes:

[0164] A structural grid drawing module 121, configured to draw a structural grid of the local component under the oncoming flow condition according to the initial shape of the local component to be optimized and the oncoming flow condition;

[0165] A control box acquisition module 122, configured to acquire a control box required for grid deformation of the structural grid. The control box contains grid boundary points to be optimized, and the grid boundary points form the surface of the local component;

[0166] The optimized coordinate acquisition module 123 is used to optimize and update the control box to obtain the optimized coordinates of the control box;

[0167] The target shape acquisition module 124 is used to optimize the initial shape according to the optimized coordinates and the oncoming flow condition to obtain the target shape of the local component.

[0168] In one embodiment, the optimized coordinate acquisition module includes:

[0169] The direction step size acquisition sub-module is used to obtain the optimization direction and optimization step size required by the control box;

[0170] The optimized coordinate acquisition sub-module is used to optimize the initial coordinates of the control box according to the optimization direction and the optimization step size to obtain the optimized coordinates of the control box.

[0171] In one embodiment, the direction step size acquisition sub-module includes:

[0172] The optimization function acquisition unit is used to obtain the optimization objective function of the local component in the current optimization scenario;

[0173] The aerodynamic heat data acquisition unit is used to solve the flow field according to the oncoming flow condition to obtain the aerodynamic heat data under the initial shape;

[0174] The objective function acquisition unit is used to obtain the objective function according to the aerodynamic heat data;

[0175] The coefficient matrix acquisition unit is used to obtain the coefficient matrix according to the flow field solution and the automatic differentiation method;

[0176] The right-hand vector acquisition unit is used to obtain the right-hand vector according to the objective function and the automatic differentiation method;

[0177] The adjoint vector acquisition unit is used to establish an adjoint equation according to the coefficient matrix and the right-hand vector to obtain the adjoint vector corresponding to the adjoint equation;

[0178] The function gradient acquisition unit is used to obtain the objective function gradient according to the adjoint vector and the preset gradient calculation model;

[0179] The optimization direction acquisition unit is used to calculate the optimization direction according to the objective function gradient and the steepest descent method.

[0180] In one embodiment, the optimization function acquisition unit includes:

[0181] The first function determination sub-unit is used to determine that the optimization objective function is the first objective function when the optimization function in the current optimization scenario is the heat flux peak;

[0182] The second function determination subunit is configured to determine that the optimization objective function is the second objective function when the current optimization scenario is that the optimization function is drag force.

[0183] The third function determination subunit is configured to determine that the optimization objective function is the third objective function when the current optimization scenario is drag force under the condition that the maximum heat flux does not increase.

[0184] The fourth function determination subunit is configured to determine that the optimization objective function is the fourth objective function when the current optimization scenario is the maximum heat flux under the condition that the drag force does not increase.

[0185] In one embodiment, the direction step size acquisition sub-module includes:

[0186] The optimization step size acquisition unit is configured to iteratively calculate a step size that meets a preset condition by using the flow field of the local component and the golden section method as the optimization step size.

[0187] In one embodiment, the target shape acquisition module includes:

[0188] The candidate shape acquisition sub-module is configured to complete the deformation of the local component shape and the corresponding aerodynamic heat environment calculation structural grid based on the optimized coordinates of the control box according to the FFD method to obtain the candidate shape of the local component.

[0189] The aerodynamic heat data acquisition sub-module is configured to acquire the aerodynamic heat data under the candidate shape of the local component according to the optimized coordinates of the control box.

[0190] The objective function value acquisition sub-module is configured to acquire the objective function value of the optimization objective function according to the aerodynamic heat data.

[0191] The target shape determination sub-module is configured to determine that the candidate shape is the target shape of the local component when the objective function value of the candidate shape meets the optimization termination condition.

[0192] It should be noted that the device shown in this embodiment matches the content of the method embodiment, and the content of the above method embodiment can be referred to and will not be elaborated here.

[0193] In some possible embodiments, an electronic device is provided, including: a processor and a memory;

[0194] The memory is used to store a computer program executable by the processor;

[0195] The processor is used to execute the computer program in the memory to implement the local aerodynamic heat optimization method based on discrete adjoint as described above.

[0196] In some possible embodiments, a non-transitory computer-readable storage medium is provided, and when the executable computer program in the storage medium is executed by a processor, the local aerothermal optimization method based on discrete adjoint as described above can be implemented.

[0197] Those skilled in the art will readily conceive of other embodiments of the present disclosure after considering the specification and practicing the disclosure herein. The present disclosure is intended to cover any variations, uses, or adaptations of the disclosure that follow the general principles of the disclosure and include known common knowledge or conventional technical means in the technical field not disclosed by the present disclosure. The specification and examples are only illustrative, and the true scope and spirit of the present disclosure are pointed out by the following claims.

[0198] It should be understood that the present disclosure is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present disclosure is only limited by the appended claims.

Claims

1. A local aerodynamic heat optimization method based on discrete adjoint, characterized in that The method includes: drawing a structural grid of the local component under the oncoming flow condition according to the initial shape of the local component to be optimized and the oncoming flow condition; obtaining a control box required for grid deformation of the structural grid, wherein the interior of the control box contains grid boundary points to be optimized, and the grid boundary points form the physical surface of the local component; optimizing and updating the control box to obtain optimized coordinates of the control box; optimizing the initial shape according to the optimized coordinates and the oncoming flow condition to obtain the target shape of the local component; optimizing and updating the control box to obtain optimized coordinates of the control box, including: obtaining an optimization direction and an optimization step size required for the control box; optimizing the initial coordinates of the control box according to the optimization direction and the optimization step size to obtain the optimized coordinates of the control box; obtaining the optimization direction required for the control box, including: obtaining an optimization objective function of the local component in the current optimization scenario; solving the flow field according to the oncoming flow condition to obtain aerodynamic heat data under the initial shape; obtaining the objective function according to the aerodynamic heat data; obtaining a coefficient matrix according to the solved flow field and the automatic differentiation method; obtaining a right-hand vector according to the objective function and the automatic differentiation method; establishing an adjoint equation according to the coefficient matrix and the right-hand vector to obtain an adjoint vector corresponding to the adjoint equation; obtaining a gradient of the objective function according to the adjoint vector and a preset gradient calculation model; calculating the optimization direction according to the gradient of the objective function and the steepest descent method.

2. The method according to claim 1, characterized in that, obtaining an optimization objective function of the local component in the current optimization scenario, including: when the optimization function in the current optimization scenario is the peak heat flux, determining the optimization objective function as the first objective function; when the optimization function in the current optimization scenario is the drag, determining the optimization objective function as the second objective function; when the objective function in the current optimization scenario is the drag under the condition that the maximum heat flux does not increase, determining the optimization objective function as the third objective function; when the objective function in the current optimization scenario is the maximum heat flux under the condition that the drag does not increase, determining the optimization objective function as the fourth objective function.

3. The method according to claim 1, characterized in that obtaining the optimization step size required for the control box, including: iteratively calculating a step size that meets a preset condition by using the flow field of the local component and the golden section method as the optimization step size.

4. The method according to claim 1, wherein optimizing the initial shape according to the optimized coordinates and the oncoming flow condition to obtain the target shape of the local component, including: based on the optimized coordinates of the control box, completing the deformation of the structural grid of the local component shape and the corresponding aerodynamic heat environment calculation according to the FFD method to obtain a candidate shape of the local component; obtaining aerodynamic heat data under the candidate shape of the local component according to the optimized coordinates of the control box; obtaining an objective function value of the optimization objective function according to the aerodynamic heat data; when the objective function value of the candidate shape meets the optimization termination condition, determining the candidate shape as the target shape of the local component.

5. A local aerodynamic heat optimization device based on discrete adjoint, characterized in that, The device includes: a structural grid drawing module, configured to draw a structural grid of the local component under the oncoming flow condition according to the initial shape of the local component to be optimized and the oncoming flow condition; A control box acquisition module, configured to acquire a control box required for grid deformation of the structured grid, wherein the control box contains grid boundary points to be optimized inside, and the grid boundary points form the physical surface of the local component; An optimized coordinate acquisition module, configured to optimize and update the control box to obtain the optimized coordinates of the control box; A target shape acquisition module, configured to optimize the initial shape according to the optimized coordinates and the oncoming flow condition to obtain the target shape of the local component; The optimized coordinate acquisition module includes: A direction step size acquisition sub-module, configured to acquire the optimization direction and the optimization step size required for the control box; An optimized coordinate acquisition sub-module, configured to optimize the initial coordinates of the control box according to the optimization direction and the optimization step size to obtain the optimized coordinates of the control box; The direction step size acquisition sub-module includes: An optimization function acquisition unit, configured to acquire the optimization objective function of the local component in the current optimization scenario; An aerodynamic heat data acquisition unit, configured to solve the flow field according to the oncoming flow condition to acquire the aerodynamic heat data under the initial shape; An objective function acquisition unit, configured to acquire the objective function according to the aerodynamic heat data; A coefficient matrix acquisition unit, configured to obtain a coefficient matrix according to the solved flow field and the automatic differentiation method; A right-hand side vector acquisition unit, configured to obtain a right-hand side vector according to the objective function and the automatic differentiation method; An adjoint vector acquisition unit, configured to establish an adjoint equation according to the coefficient matrix and the right-hand side vector to obtain the adjoint vector corresponding to the adjoint equation; A function gradient acquisition unit, configured to acquire the objective function gradient according to the adjoint vector and a preset gradient calculation model; An optimization direction acquisition unit, configured to calculate the optimization direction according to the objective function gradient and the steepest descent method.

6. The device according to claim 5, characterized in that, The target shape acquisition module includes: A candidate shape acquisition sub-module, configured to complete the deformation of the local component shape and the corresponding aerodynamic heat environment calculation structured grid based on the optimized coordinates of the control box according to the FFD method to obtain a candidate shape of the local component; An aerodynamic heat data acquisition sub-module, configured to acquire the aerodynamic heat data under the candidate shape of the local component according to the optimized coordinates of the control box; An objective function value acquisition sub-module, configured to acquire the objective function value of the optimization objective function according to the aerodynamic heat data; A target shape determination sub-module, configured to determine the candidate shape as the target shape of the local component when the objective function value of the candidate shape meets the optimization termination condition.

7. An electronic device, characterized in that, It includes: A processor and a memory; The memory is used to store the computer program executable by the processor; The processor is used to execute the computer program in the memory to implement the method according to any one of claims 1 to 4.

8. A non-transitory computer-readable storage medium, characterized in that, When the executable computer program in the storage medium is executed by the processor, it can implement the method according to any one of claims 1 to 4.

Images