A three-dimensional time-domain far-field noise hybrid prediction method

Abstract

The application provides a three-dimensional time-domain far-field noise mixing prediction method, comprising the following steps: calculating the sound source size through fluid velocity and pressure field, simulating the propagation process of sound waves in the sound source area by using the sound field wave equation, setting a closed integral surface in the sound source area, performing interpolation calculation according to the discrete points on the integral surface, obtaining the flow field variable and the sound field variable on the integral surface, calculating the noise contribution of each discrete point on the integral surface to the far-field observation point at a target time according to the flow field variable and the sound field variable, solving the noise contribution of the discrete points on the integral surface to the far-field observation point at all times based on the noise contribution at the target time, and performing interpolation summation according to the far-field time to obtain the time-domain signal of the far-field noise. The application avoids the solid sound coupling effect in the sound source area and the nonlinear influence of the fluid field on the sound field, improves the prediction accuracy of the far-field noise, and guarantees the calculation accuracy and efficiency of the far-field noise prediction.

Description

Technical Field

[0001] This invention relates to the field of noise prediction technology, and in particular to a three-dimensional time-domain far-field noise hybrid prediction method. Background Technology

[0002] Far-field noise refers to the pressure fluctuations (sound sources) generated by a non-uniform pressure field in a fluid. These fluctuations propagate rapidly over long distances within the fluid. When the propagation distance is several times or even hundreds of times greater than the source area, the sound pressure level of the distant sound wave increases. The magnitude of far-field noise can be obtained through far-field prediction.

[0003] Traditional far-field noise prediction methods are direct prediction methods, which directly calculate the magnitude of far-field noise in a stationary or uniformly moving fluid by integrating the fluid and acoustic variables on the closed surface of the sound source region. However, this method is only applicable to simple sound sources in stationary or uniformly moving fluids, neglecting the fluid-structure acoustic coupling effect and the nonlinear influence of the fluid field on the sound field, thus reducing the accuracy of far-field noise prediction. The fluid-structure acoustic coupling effect involves a non-uniform flow field containing non-uniform velocity and pressure fields. The non-uniform velocity field affects the propagation characteristics of sound waves, and the non-uniform pressure field can create new sound sources. Therefore, in the sound wave calculation process, it is necessary to consider the coupling effect of the flow field on the generation and propagation of sound waves.

[0004] To meet the demands of far-field noise prediction in large computational domains, far-field noise prediction methods based on acoustic analogy theory have been extensively studied. Among these, for far-field noise prediction in three-dimensional space, researchers have proposed a direct simulation method for far-field noise in the three-dimensional time domain. This method calculates the magnitude of far-field noise in stationary or uniformly moving fluids by integrating fluid and acoustic variables on the closed surface of the sound source region. However, it is only applicable to simple sound sources in stationary and uniformly moving fluids, neglecting the fluid-structure-acoustic coupling effect and the nonlinear influence of the fluid field on the sound field in the sound source region, thus reducing the accuracy of far-field noise prediction.

[0005] Therefore, there is an urgent need for a far-field noise prediction method that can reduce computational load and improve the accuracy of far-field noise prediction. Summary of the Invention

[0006] Therefore, it is necessary to provide a three-dimensional time-domain far-field noise mixing prediction method to address the aforementioned technical problems.

[0007] A three-dimensional time-domain far-field noise mixing prediction method includes the following steps: acquiring fluid velocity and pressure field; calculating the sound source size based on the fluid velocity and pressure field, and simulating the sound wave propagation process in the sound source region using the sound field wave equation; setting a closed integral surface in the sound source region, and performing interpolation calculations based on discrete points on the integral surface to obtain the flow field variables and sound field variables of the discrete points; calculating the noise contribution of each discrete point on the integral surface to the far-field observation point at the target time based on the flow field variables and sound field variables of the discrete points; solving for the noise contribution of the discrete points on the integral surface to the far-field observation point at all times based on the noise contribution at the target time, and interpolating and summing based on the far-field time to obtain the time-domain signal of the far-field noise.

[0008] In one embodiment, the step of acquiring the fluid velocity and pressure field, calculating the sound source size based on the fluid velocity and pressure field, and simulating the propagation process of sound waves in the sound source region using the sound field wave equation includes: acquiring the density field, and calculating the corresponding average flow field variables based on the fluid velocity, pressure field, and density field, respectively, using the following formula:

[0009]

[0010] in, All are average flow field variables, where τ0 and τ1 are the initial and final times for calculating the average flow field variables in the sound source region, respectively. The propagation of the sound wave in the sound source region is calculated using the sound field wave equation, as shown in the formula:

[0011]

[0012]

[0013] In the formula, u f (x,t) represents the fluid velocity, p f (x,t) represents the pressure field, ρ f (x,τ) represents the density field, S(x,τ) represents the magnitude of the sound source term, x represents the Eulerian grid node where the flow field variable is located, and τ represents the time variable.

[0014] In one embodiment, setting a closed integral surface in the sound source region and performing interpolation calculations based on discrete points on the integral surface to obtain the flow field variables and sound field variables of the discrete points includes: setting a cube with six surfaces centered on the sound source to form a closed integral surface; discretizing the integral surface into multiple discrete points using a Lagrange network; the grid area of ​​the multiple discrete points being ΔS; performing interpolation calculations on each discrete point to obtain the flow field variables and sound field variables on the integral surface; and uniformly labeling the flow field variables and sound field variables as Ψ. The interpolation calculation formula is:

[0015] Ψ(y,τ n )=ψ(x,τ n )δ(xX(y,τ n ))h 3

[0016]

[0017] Wherein, in the x-dimensional d = xX(y,τ) n ) / h, in the y-dimensional d=yY(y,τ n ) / h, in the z-dimensional d=zZ(y,τ) n ) / h.

[0018] In one embodiment, calculating the noise contribution of each discrete point on the integral surface to the far-field observation point at the target time based on the flow field variables and sound field variables of the discrete points includes: obtaining the target time τ when the sound wave is emitted and the time t when the sound wave is received at the far-field observation point; then, the propagation time of the sound wave from the integral surface in the sound source region to the far-field observation point is t-τ; and calculating the noise contribution of each discrete point on the integral surface at time τ to the far-field observation point at time t, using the following formula:

[0019] p a (x,y,t)=p a,T (x,y,t)+p a,L (x,y,t)

[0020]

[0021] In the formula:

[0022]

[0023] U a,n =U a,j ·n,U f,n =U f,j ·n

[0024] Where, p a (x,t) represents the sound pressure level at a far-field observation point at position x and time t, p a,T (x,t) and p a,L (x,t) represent the thick noise and load noise components in the far-field noise prediction, respectively, Q is the monopole source term caused by fluid mass fluctuations, and L... i Let R be the dipole sound source term caused by fluid momentum fluctuations, where c0 is the magnitude of the standard speed of sound in incompressible fluids. * R is the amplitude radius with respect to the observation point x and the integration point y, and R is the phase radius with respect to the observation point x and the integration point y. retThe relationship between the delay time R / c0 and the observation time t and the sound source time τ is t-τ=R / c0, Ma f,i Ma is the fluid Ma number at discrete points on the integral surface S. f In x i Components in direction; P ij The compressible stress tensor is calculated as (P-P0)δ. ij δ ij The symbol for Kronecker; Let n be the normal vector n of the discrete point y on the integral curve at x. j The component in the direction, the direction of the normal vector is pointing outwards from the integral surface S.

[0025] In one embodiment, it further includes: using Einstein's convention for summation, defining Ma... f,j r j =Ma f,1 r1+Ma f, 2r2+Ma f,3 r3, where Ma f,R =Ma f,i R i , as well as in According to Einstein's conventional summation definition, the spatial derivatives of the phase radius and magnitude radius are calculated using the following formulas:

[0026]

[0027] In the formula, r is the distance between the observation point x and the integration point y, and its calculation expression is |xy|; i Let the direction vector r = (xy) be in x i The component in the direction; γ is the influence parameter of the fluid Mach number, with a magnitude of 1 / (1-|Ma f | 2 ).

[0028] In one embodiment, the noise contribution based on the target time is used to solve for the noise contribution of discrete points on the integral surface to the far-field observation point at all times, and then interpolated and summed according to the far-field time to obtain the time-domain signal of the far-field noise, including: calculating the time t before and after the time interval Δt. n and t n+1 The time-domain signal of far-field noise is given by the following formula:

[0029] t=τ+R / c0,t n ≤t≤t n+1

[0030]

[0031] In the formula, p a (x, t) n ) for t n Far-field noise at time p a (x, t) n+1 ) for t n+1 Far-field noise at any given moment.

[0032] Compared with existing technologies, the advantages and beneficial effects of this invention are as follows: the size of the sound source is calculated by fluid velocity and pressure field, and the propagation process of sound waves in the sound source area is simulated by sound field wave equation. A closed integral surface is set in the sound source area, and interpolation calculation is performed based on discrete points on the integral surface to obtain the flow field variables and sound field variables on the integral surface. Based on the flow field variables and sound field variables, the noise contribution of each discrete point on the integral surface to the far-field observation point at the target time is calculated, thereby simplifying the calculation and improving the calculation efficiency. Based on the noise contribution at the target time, the noise contribution of discrete points on the integral surface to the far-field observation point at all times is solved, and interpolation and summation are performed based on the far-field time to obtain the time-domain signal of the far-field noise. This avoids the solid-sound coupling effect and the nonlinear influence of the fluid field on the sound field in the sound source area, improves the accuracy of far-field noise prediction, and ensures the calculation accuracy and efficiency of far-field noise prediction. Attached Figure Description

[0033] Figure 1 This is a flowchart illustrating a three-dimensional temporal far-field noise mixing prediction method in one embodiment.

[0034] Figure 2 This is a schematic diagram of an underwater cylindrical shell radiated noise experiment in one embodiment;

[0035] Figure 3 The image shows a comparison of the prediction results of the present invention and the direct far-field prediction method in one embodiment, wherein (1) is the prediction result of the present invention; and (2) is the prediction result of the direct far-field prediction method.

[0036] Figure 4 This is a schematic diagram showing the error comparison results between the present invention and the direct far-field prediction method in one embodiment. Detailed Implementation

[0037] Before describing the specific embodiments of the present invention, the overall concept of the present invention will be explained as follows:

[0038] This invention is primarily based on the development of a far-field noise prediction process within a large computational domain. Current traditional far-field noise prediction methods are direct prediction methods, applicable only to simple sound sources in stationary or uniformly moving fluids. They neglect the fluid-structure-acoustic coupling effect and the nonlinear influence of the fluid field on the sound field in the source region, thus reducing the accuracy of far-field noise prediction. Furthermore, existing three-dimensional time-domain direct simulation methods for far-field noise require calculating the changes in the fluid field and sound field at every spatial node at every moment, resulting in a massive computational load and low computational efficiency for far-field noise prediction.

[0039] Through analysis, the inventors discovered that the main reason for the aforementioned problems is the failure to consider the fluid-structure-acoustic coupling effect in the sound source region and the influence of non-uniform fluid fields on far-field noise prediction. By effectively combining direct source region simulation methods and direct far-field prediction methods, these problems can be avoided. Therefore, this invention proposes a three-dimensional time-domain hybrid far-field noise prediction method. By inputting the time-domain information of the flow field and sound field in three-dimensional space, the time-domain information of three-dimensional far-field noise can be obtained. Simultaneously, considering the fluid-structure-acoustic coupling effect in the sound source region and the influence of non-uniform fluid fields on far-field noise prediction, the computational accuracy and efficiency of far-field noise prediction are guaranteed.

[0040] Having introduced the overall concept of the present invention, to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below through specific embodiments in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0041] In one embodiment, such as Figure 1 As shown, a three-dimensional time-domain far-field noise mixing prediction method is provided, including the following steps:

[0042] Step S101: Obtain the fluid velocity and pressure field, calculate the size of the sound source based on the fluid velocity and pressure field, and simulate the propagation process of sound waves in the sound source area using the sound field wave equation.

[0043] Specifically, when sound waves propagate in a fluid medium under certain fluid velocity and pressure fields, the magnitude of the corresponding sound source can be calculated by obtaining the fluid velocity and pressure field of the fluid medium. Simultaneously, based on the known fluid velocity and pressure field, the velocity and pressure fields of the fluid in stable and fluctuating states are separated, and the magnitude of the sound source term and its sound pressure changes during propagation are calculated using the sound field wave equation. This numerical simulation of the fluid sound wave generation and propagation process in the entire computational domain facilitates the acquisition of relevant variable information during sound wave propagation and enables far-field noise prediction based on this variable information.

[0044] Step S101 includes: obtaining the density field, and calculating the corresponding average flow field variables based on the fluid velocity, pressure field, and density field, using the following formula:

[0045]

[0046] in, All are average flow field variables, where τ0 and τ1 are the initial and final times for calculating the average flow field variables in the sound source region, respectively. The propagation of the sound wave in the sound source region is calculated using the sound field wave equation, as shown in the formula:

[0047]

[0048] In the formula, u f (x,t) represents the fluid velocity, p f (x,t) represents the pressure field, ρ f (x,τ) represents the density field, S(x,τ) represents the magnitude of the sound source term, x represents the Eulerian grid node where the flow field variable is located, and τ represents the time variable.

[0049] Specifically, based on the fluid velocity, pressure field, and density field of the fluid medium, the corresponding average flow field variables are calculated respectively. Based on the average flow field variables, the sound field wave equation is used to obtain the propagation of the sound wave in the sound source area, so as to calculate the flow field variable information and sound field variable information during the propagation process.

[0050] Step S102: Set a closed integral surface in the sound source area, and perform interpolation calculations based on discrete points on the integral surface to obtain the flow field variables and sound field variables of the discrete points.

[0051] Specifically, a closed integral surface is set in the sound source region, and this integral surface is treated as a special boundary condition of the fluid medium to facilitate the calculation of relevant variables of the sound source propagating in the fluid medium. When setting the integral surface, the shape can be set according to needs or for calculation convenience. The set integral surface is discretized using a Lagrange net to obtain multiple discrete points. Interpolation calculations are performed on multiple discrete points to obtain the flow field variables and sound field variables on the integral surface.

[0052] Step S102 includes: setting a cube with six surfaces centered on the sound source to form a closed integral surface; discretizing the integral surface into multiple discrete points using a Lagrange network, with the grid area of ​​the multiple discrete points being ΔS; performing interpolation calculations on each discrete point to obtain the flow field variables and sound field variables on the integral surface; uniformly labeling the flow field variables and sound field variables as Ψ; and then using the interpolation calculation formula:

[0053] Ψ(y,τ n )=ψ(x,τ n)δ(xX(y,τ n ))h 3

[0054]

[0055] Wherein, in the x-dimensional d = xX(y,τ) n ) / h, in the y-dimensional d=yY(y,τ n ) / h, in the z-dimensional d=zZ(y,τ) n ) / h.

[0056] Specifically, when setting the integral surface, a cube with six surfaces is typically set around the sound source as a closed integral surface. This integral surface is discretized into a series of discrete points y(x,τ) using a Lagrange grid, with the grid area of ​​each discrete point being ΔS. By interpolating at each discrete point, the flow field variables and sound field variables on the integral surface are obtained, respectively U f (y,τ),P f (y,τ),ρ f (y,τ),U a (y,τ),P a (y,τ),ρ a (y,τ), denoted as Ψ, facilitates unified interpolation calculations. Furthermore, the flow field variables and sound field variables obtained based on discrete point calculations can be used to calculate variables in non-uniform flow fields. This allows for consideration of the coupling effect of the flow field on the generation and propagation of sound waves, thereby improving the accuracy of far-field noise prediction.

[0057] Step S103: Based on the flow field variables and sound field variables of the discrete points, calculate the noise contribution of each discrete point on the integral surface to the far-field observation point at the target time.

[0058] Specifically, based on the flow field variables and sound field variables at discrete points on the integral surface, the target time is determined, and the contribution of each discrete point on the integral surface to the far-field noise at the target time is calculated. This facilitates the calculation of the noise contribution at adjacent times, thereby simplifying the calculation and improving computational efficiency.

[0059] Step S103 includes: obtaining the target time τ of the sound wave emission and the time t of the sound wave reception at the far-field observation point; the propagation time of the sound wave from the integral surface of the sound source region to the far-field observation point is t-τ; and calculating the noise contribution of each discrete point on the integral surface at time τ to the far-field observation point at time t, using the following formula:

[0060] p a (x,y,t)=p a,T (x,y,t)+p a,L (x,y,t)

[0061]

[0062] In the formula:

[0063]

[0064] U a,n =U a,j ·n,U f,n =U f,j ·n

[0065] Where, p a (x,t) represents the sound pressure level at a far-field observation point at position x and time t, p a,T (x,t) and p a,L (x,t) represent the thick noise and load noise components in the far-field noise prediction, respectively, Q is the monopole source term caused by fluid mass fluctuations, and L... i Let R be the dipole sound source term caused by fluid momentum fluctuations, where c0 is the magnitude of the standard speed of sound in incompressible fluids. * R is the amplitude radius with respect to the observation point x and the integration point y, and R is the phase radius with respect to the observation point x and the integration point y. ret The relationship between the delay time R / c0 and the observation time t and the sound source time τ is t-τ=R / c0, Ma f,i Ma is the fluid Ma number at discrete points on the integral surface S. f In x i Components in direction; P ij The compressible stress tensor is calculated as (P-P0)δ. ij δ ij The symbol for Kronecker; Let n be the normal vector n of the discrete point y on the integral curve at x. j The component in the direction, the direction of the normal vector is pointing outwards from the integral surface S.

[0066] Specifically, since it takes time for sound waves to propagate from the integral surface in the sound source region to the far-field observation point, when the target time of sound wave emission in the sound source region is τ and the far-field time of sound wave transmission to the far-field observation point is t, the propagation time is t-τ. By calculating the noise contribution of each discrete point on the integral surface to the far-field observation point at time t, the fluid-structure-acoustic coupling effect and the nonlinear influence of the fluid field on the sound field in the sound source region can be avoided, thus improving the prediction accuracy of far-field noise.

[0067] Step S103 further includes: using Einstein's convention for summation, defining Ma... f,j r j =Ma f,1 r1+Ma f,2 r2+Ma f,3r3, where Ma f,R =Ma f,i R i , as well as in According to Einstein's conventional summation definition, the spatial derivatives of the phase radius and magnitude radius are calculated using the following formulas:

[0068]

[0069] In the formula, r is the distance between the observation point x and the integration point y, and its calculation expression is |xy|; i Let the direction vector r = (xy) be in x i The component in the direction; γ is the influence parameter of the fluid Mach number, with a magnitude of 1 / (1-|Ma f | 2 ).

[0070] Specifically, the spatial derivatives of the phase radius and amplitude radius are calculated so that the phase radius and amplitude radius can be measured in a way that changes with space, thereby enabling the calculation of the spatially varied acoustic waveform to facilitate far-field noise mixing prediction.

[0071] Step S104: Based on the noise contribution at the target time, solve for the noise contribution of discrete points on the integral surface to the far-field observation point at all times, and interpolate and sum according to the far-field time to obtain the time-domain signal of the far-field noise.

[0072] Specifically, after obtaining the noise contribution at the target time, the noise contribution of all points on the integral surface to the far-field noise can be solved. Then, by interpolating and summing according to the time of the far-field observation points, the time-domain signal of the far-field noise can be obtained, thereby enabling efficient prediction of far-field noise with high accuracy.

[0073] Step S104 includes: calculating the time intervals t before and after the time interval Δt. n and t n+1 The time-domain signal of far-field noise is given by the following formula:

[0074] t = τ + R / c0, t n ≤t≤t n+1

[0075]

[0076] In the formula, p a (x, t) n ) for t n The time-domain signal of far-field noise at time p a (x, t) n+1 ) for tn+1 The time-domain signal of far-field noise at time step.

[0077] Specifically, after obtaining the contribution of all points on the integral surface to the far-field noise, the interpolation and summation method is used to calculate the time-domain signal of the far-field noise at adjacent time points, thereby reducing the amount of computation, improving the far-field noise prediction efficiency, and achieving high accuracy in the prediction results.

[0078] In this embodiment, the size of the sound source is calculated using fluid velocity and pressure field, and the propagation process of sound waves in the sound source region is simulated using the sound field wave equation. A closed integral surface is set in the sound source region, and interpolation calculations are performed based on discrete points on the integral surface to obtain the flow field variables and sound field variables on the integral surface. Based on the flow field variables and sound field variables, the noise contribution of each discrete point on the integral surface to the far-field observation point at the target time is calculated, thereby simplifying the computation and improving the computational efficiency. Based on the noise contribution at the target time, the noise contribution of discrete points on the integral surface to the far-field observation point at all times is solved, and interpolation and summation are performed based on the far-field time to obtain the time-domain signal of the far-field noise. This avoids the solid-sound coupling effect and the nonlinear influence of the fluid field on the sound field in the sound source region, improves the accuracy of far-field noise prediction, and ensures the computational accuracy and efficiency of far-field noise prediction.

[0079] In one embodiment, the effect of the prediction method of the present invention is predicted by taking the far-field prediction of the radiated noise of an underwater cylindrical shell as an example.

[0080] Experimental setup such as Figure 2 As shown, the cylindrical shell is completely immersed in the anechoic pool to reduce the impact of reflection from the pool wall. A vibration excitation is given to the cylindrical shell, and the acceleration signal of the cylindrical shell surface is obtained through an accelerometer. The flow field variables of the sound source area are obtained through fluid-structure interaction calculation. Based on the three-dimensional time-domain far-field noise mixing prediction method of this patent, the noise level at the far-field observation point can be obtained.

[0081] In the numerical analysis, the mass density of the cylindrical shell is ρ. s =7850kg / m 3 The Young's modulus of the shell is E = 2.06 × 10⁻⁶. 11 Pa, the Poisson's ratio of the shell is υ = 0.3, and the shell thickness is h. s =0.008m. The fluid mass density is ρ. f =998kg / m 3 The speed of sound is c o =1483m / s, the uniform spatial grid size is h = 0.18m, and the computation time step is Δt = h / 2c0 = 6 × 10 -5s, and assume the fluid is in a static state. To facilitate comparison between the predicted results and the hydrophone test results, a commonly used variable in underwater acoustics is defined, namely the sound pressure level:

[0082] SPL water =20lg(p) water,e / p ref )

[0083] Where: SPL water is the underwater acoustic pressure level, measured in decibels (dB); lg is the logarithm base 10, i.e., log10; p water,e The effective sound pressure level predicted by hydrophone testing or prediction methods; p ref For reference sound pressure levels, a commonly used reference sound pressure level in underwater acoustics is 1 × 10⁻⁶. -6 Pa.

[0084] A comparison between the direct far-field prediction method and the method proposed in this invention yields the following results: Figure 3 As shown, the method of the present invention is more consistent with the experimental results and has higher calculation accuracy.

[0085] To further compare the calculation errors between the two, the error calculation formula is given as follows:

[0086]

[0087] In the formula: i is the hydrophone number; SPL p SPL represents the frequency amplitude of the sound pressure level at 20 hydrophone locations calculated using different prediction methods. e denoted as , where is the frequency amplitude of the sound pressure level at 20 hydrophone locations tested in the water tank experiment; n is the nth frequency component; and N is the last frequency component.

[0088] The calculations using the above formulas yield the error magnitudes of the two methods at different locations, as follows: Figure 4 As shown, 0°~60°, 150°~210°, and 300°~360° are the two sides of the anechoic pool, which have a certain reflection effect. Therefore, when these areas are not considered, the method of the present invention can improve the calculation accuracy by 5%.

[0089] Table 1 shows a comparison of the computational space and time required for the source region method, the direct far-field prediction method, and the far-field prediction method of this invention (programmed using MATLAB R2020b and run on a desktop computer with an Intel(R) Core(TM) i7-7700 CPU, 3.60GHz, and 24.0GB RAM). It is evident that the far-field prediction method of this invention has a similar computational time to the direct far-field prediction method, and both are significantly shorter than the computational time of the source region method, effectively demonstrating that the far-field prediction method of this invention can simultaneously guarantee the accuracy and efficiency of far-field noise prediction.

[0090] Table 1 Comparison of the calculation efficiency of the three methods

[0091] Method Sound source region method Direct far field prediction method Far field prediction method of the invention Computational space 41.3 MB 75.8 MB 84.9 MB Computational time 20267s 861.14s 1772.9s

[0092] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0093] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a computer storage medium (ROM / RAM, magnetic disk, optical disk) for execution by the computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Therefore, the present invention is not limited to any particular hardware and software combination.

[0094] The above description, in conjunction with specific embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such deductions or substitutions should be considered within the scope of protection of the present invention.

Claims

1. A three-dimensional time-domain far-field noise hybrid prediction method, characterized in that, The method comprises the following steps: obtaining a fluid velocity and pressure field, calculating a sound source size according to the fluid velocity and pressure field, and simulating a propagation process of sound waves in a sound source area by using a sound field wave equation; setting a closed integral surface in the sound source area, processing the integral surface as a special boundary condition of a liquid medium, discretizing by using a Lagrange network to obtain a plurality of discrete points, performing interpolation calculation according to the discrete points on the integral surface to obtain flow field variables and sound field variables of the discrete points; calculating a noise contribution of each discrete point of the integral surface to a far-field observation point at a target time according to the flow field variables and sound field variables of the discrete points; based on the noise contribution at the target time, solving noise contributions of the discrete points on the integral surface to the far-field observation point at all times, and performing interpolation summation according to a far-field time to obtain a time-domain signal of far-field noise.

2. The three-dimensional time-domain far-field noise hybrid prediction method according to claim 1, characterized in that, The method of obtaining a fluid velocity and pressure field, calculating a sound source size according to the fluid velocity and pressure field, and simulating a propagation process of sound waves in a sound source area by using a sound field wave equation comprises: obtaining a density field, and calculating corresponding average flow field variables according to the fluid velocity, pressure field and density field, with a formula being: ， wherein , , are average flow field variables, and are the initial time and the final time of the average flow field variable calculation in the sound source region, respectively, and the propagation of sound waves in the sound source region is calculated by the sound field wave equation, and the formula is: ， ， ， wherein is the fluid velocity, is the pressure field, is the density field, is the acoustic source term size, is the Euler grid node where the flow field variable is located, is the time variable.

3. The three-dimensional time-domain far-field noise hybrid prediction method according to claim 2, characterized in that, The method of setting a closed integral surface in the sound source area, processing the integral surface as a special boundary condition of a liquid medium, discretizing by using a Lagrange network to obtain a plurality of discrete points, and performing interpolation calculation according to the discrete points on the integral surface to obtain flow field variables and sound field variables of the discrete points comprises: A cube six surface is set with a sound source as a center to form a closed integral surface, the integral surface is discretized into a plurality of discrete points by a Lagrange network, and a grid area of the plurality of discrete points is ; interpolating each of the discrete points to obtain the flow field variable and the sound field variable on the integral surface, and uniformly marking the flow field variable and the sound field variable as The interpolation formula is: ， ， ， wherein, in dimension , in dimension , in dimension .

4. The three-dimensional time-domain far-field noise hybrid prediction method according to claim 3, characterized in that, The method of calculating a noise contribution of each discrete point of the integral surface to a far-field observation point at a target time according to the flow field variables and sound field variables of the discrete points comprises: The time instance when the sound wave is emitted and the time instance when the sound wave is received at the far field observation point The propagation time of the sound wave from the sound source region integral surface to the far field observation point is The calculation is The noise contribution of each integral surface discrete point to the far field observation point at the time instance is The noise contribution of each integral surface discrete point to the far field observation point at the time instance is ， ， ， In the formula: ， ， ， in, For the location of the far-field observation point Time is The sound pressure level at that time and These are the thick noise and load noise components in far-field noise prediction, respectively. This is the monopole sound source term caused by fluid mass fluctuations. The term is the dipole sound source term caused by fluid momentum fluctuations. This represents the standard speed of sound in incompressible fluids. It's about the observation point. and points of integration amplitude radius, It's about the observation point. and points of integration phase radius, The indicated delay time With observation time Harmony and sound source time The relationship is , To pass through the integral surface Fluid Mach number at discrete points exist Components in direction; The compressible stress tensor is calculated as follows: , The symbol for Kronecker; Discrete points on the integral curve normal vector exist The component in the direction, the direction of the normal vector is pointing towards the integral surface. external.

5. The three-dimensional time-domain far-field noise hybrid prediction method according to claim 4, characterized in that, The method further comprises: Using Einstein's summation convention, define , where , , , and where ; calculating spatial derivatives of the phase radius and amplitude radius according to an Einstein summation convention, with a formula being: ， where is the observation point and the integration point The distance between the observation point ; is the component of the direction vector in the direction; is the influence parameter for the fluid Mach number, with a magnitude of .

6. The three-dimensional time-domain far-field noise hybrid prediction method according to claim 5, characterized in that, The method of solving noise contributions of the discrete points on the integral surface to the far-field observation point at all times based on the noise contribution at the target time, and performing interpolation summation according to a far-field time to obtain a time-domain signal of far-field noise comprises: The computation time interval is the time before and after and the time-domain signal of the far-field noise, the formula is: ， ， ， wherein is far field noise at time is far field noise at time

Images