CN105224715A - High wind three-dimensional fluctuating wind field comprehensive simulation method under the landforms of a kind of mountain area - Google Patents

Abstract

The invention discloses high wind three-dimensional fluctuating wind field comprehensive simulation method under the landforms of a kind of mountain area.Existing analogy method mostly only for the massif of simple shape, rule, more difficult be generalized to more complicated actual mountain area landforms under.The present invention is first for target mountain area landforms, the landform model of mountain area accurately of error requirements is met in conjunction with GoogleEarth and Imageware Software Create, then CFD technique computes is adopted to obtain the eddy stress of control point, and be introduced in random pulse wind field generation technique, the three-dimensional fluctuating wind speed field of target mountain area landforms is obtained with this.The present invention simulates in the process of three-dimensional fluctuating wind speed field, have employed pointwise simulation, do not need to decompose the huge cross-power spectrum matrix of exponent number, greatly reduce the requirement to computer software and hardware, can simulate for concrete, complicated mountain area landforms, the result drawn has more applicability, accuracy than simple, regular mountain area model simultaneously.

Description

High wind three-dimensional fluctuating wind field comprehensive simulation method under the landforms of a kind of mountain area

Technical field

The present invention relates to simulation of wind method, specifically high wind three-dimensional fluctuating wind field comprehensive simulation method under the landforms of a kind of mountain area.

Background technology

Wind can be regarded as and is made up of two parts: average wind and fluctuating wind.In recent years the research of mountain region wind field is got more and more, but most research is all for average wind field.And for being similar to the wind sensitive structure of Transmission Tower-line System, only consider that average wind field is comprehensive not, also must consider the power augmentation effect of wind sensitive structure under fluctuating wind field.Therefore, how obtain efficiently and accurately and specify the three-dimensional fluctuating wind field of high wind under the landforms of mountain area most important.

From current progress, obtaining the method for three-dimensional fluctuating wind field, to mainly contain two kinds: one be numerical computation method based on theory of random processes; Two is use Fluid Mechanics Computation (ComputationalFluidDynamics is called for short CFD) Method for Numerical Simulation velocity field.The former mainly comprises harmony superposition, wavelet method etc., wherein, harmonic and reactive detection method develops comparatively early and the fluctuating wind speed analogy method of comparative maturity, but there is following problem: on the one hand, to large scale structure or the larger situation of this spatial extent of mountain area landforms, need the feature wind speed point of investigation more, simulate assessing the cost too greatly of so many node Wind Velocity History, the restriction of computer hardware is also a problem; On the other hand, wind power spectrum used in harmonic and reactive detection method obtains after carrying out mathematic(al) manipulation and curve to ward actual measurement air speed data often, only can express the feature of energy in frequency domain of ward wind speed, therefore, under the complex landform being similar to mountain area landforms, its applicability remains to be discussed.

Along with the development of computer technology, computational fluid dynamics method can simulate the building periphery turbulent flow wind field obtaining having certain precision, and concrete grammar mainly contains the Reynolds method of average and Large eddy simulation method.The Reynolds equation of homogenizing when what the Reynolds method of average solved is, therefore has the feature that computing velocity is fast and counting yield is high; But the Reynolds method of average only can provide the statistical property of field of turbulent flow, the time-history analysis result of fluctuating wind speed cannot be provided.Although Large eddy simulation method can simulate instantaneous fluctuating wind field effectively, still need to solve a large amount of transient state N-S equation, calculate consuming time longer in its computation process, result of calculation is larger by the impact of grid simultaneously.Therefore, the more simple mountain region fluctuating wind field analogy method efficiently of one is needed.

In addition, existing most only for the massif of simple shape, rule to the wind field research under the landforms of mountain area, common as sine and cosine shape massif, Bel's shape, triangle etc., its result of calculation can reflect some basic wind field rules, but more difficult be generalized to more complicated actual mountain area landforms under.

Summary of the invention

Technical matters to be solved by this invention overcomes the defect that above-mentioned prior art exists, and provides a kind of more simple efficient and three-dimensional fluctuating wind field comprehensive simulation method that is high wind under being applicable to complex mountainous landforms.

For this reason, the present invention adopts following technical scheme: high wind three-dimensional fluctuating wind field comprehensive simulation method under the landforms of a kind of mountain area, and it comprises the following steps:

A. according to engineering demand, quantity, precision and scope that spot elevation gathers are set in the GoodyGIS system of GoogleEarth secondary development, derive longitude and latitude and height coordinate that target mountain area data extract point;

B. the longitude and latitude of said extracted point is converted and be met the planimetric coordinates of accuracy requirement, a final formation point cloud file;

C. the some cloud file that step B is formed is imported in Imageware software, use the function of " being gone to construct free form surface by point " to generate the landforms surface model in target mountain area;

D. the landforms surface model formed in step C is imported in the pre-processing software GAMBIT of Fluent, complete the modeling needed for CFD calculating, stress and strain model and definition boundary condition;

E. RSM model is used to calculate the eddy stress component value of each extraction point, wind power spectrum and some coherence function are revised, more true reasonably with the wind spectrum matrix that point is incorgruous under drawing mountain area landforms, revised wind spectrum matrix comprises auto-power spectrum and cross-power spectrum;

F. structure obtains extracting the incorgruous three-dimensional fluctuating wind field stimulation formula of the same point of a consideration at place, and introduces Fast Fourier Transform (FFT) technology.

The present invention is first for target mountain area landforms, the landform model of mountain area accurately of error requirements is met in conjunction with GoogleEarth and Imageware Software Create, then CFD technique computes is adopted to obtain the eddy stress of control point (namely extracting a little), and be introduced in random pulse wind field generation technique, the three-dimensional fluctuating wind speed field of target mountain area landforms is obtained with this.Because comprehensively employ two kinds of technology in solution procedure, therefore be called comprehensive simulation method.

As improving further and supplementing technique scheme, the present invention takes following technical measures:

In step e, when geomorphologic conditions change in mountain area is too complicated, when being difficult to generate the grid meeting RSM model computation requirement, using the k-ε turbulence model relatively low to mesh quality requirement, obtain the computing formula being solved each eddy stress component value by tubulence energy k.

The computing formula of three diagonal angle eddy stress component values is:

in formula, i is u, v or w;

Wherein,

${\mathrm{\α}}_v=\frac{{\mathrm{\α}}_v}{{\mathrm{\σ}}_u}=1-0.22{\cos }^4\left(\frac{\mathrm{\π}}{2}\frac{z}{H}\right)$

${\mathrm{\α}}_w=\frac{{\mathrm{\α}}_w}{{\mathrm{\σ}}_u}=1-0.45{\cos }^4\left(\frac{\mathrm{\π}}{2}\frac{z}{H}\right)$

α _u＝1，

In above formula, σ _u, σ _vand σ _wrepresent down wind, beam wind respectively to the root mean square with perpendicular wind direction fluctuating wind speed component, z is the height of three-dimensional fluctuating wind field simulated target point, and H is the gradient level of atmospheric boundary layer.

For off-diagonal eddy stress component value, due to down wind and beam wind to, beam wind to and perpendicular wind direction between almost there is no correlativity, τ _uvand τ _vwvalue minimum, ignore; Eddy stress component τ _uwestimation equation: wherein, u _*for wind friction velocity, u _*computing formula be: u _*=U ₁₀/ [2.5ln (10/z ₀)], U ₁₀for the mean wind speed of 10m At The Height, z ₀for Roughness Length.

In step e, auto-power spectrum is:

$s_{ii}^n=\frac{1}{2\mathrm{\π}}\·{\mathrm{\τ}}_{ii}\frac{L_i/U_z}{{\[1+70.8{({\mathrm{\ω}}_n{L}_i/2{\mathrm{\πU}}_z)}^2\]}^{5/6}},i=u,$

$S_{ii}^n=\frac{1}{2\mathrm{\π}}\·{\mathrm{\τ}}_{ii}\frac{L_i/{\stackrel{\‾}{U}}_z\[1+755.2{({\mathrm{\ω}}_n{L}_i/2\mathrm{\π}{\stackrel{\‾}{U}}_z)}^2\]}{{\[1+283.2{({\mathrm{\ω}}_n{L}_i/2\mathrm{\π}{\stackrel{\‾}{U}}_z)}^2\]}^{11/6}},i=v,w;$

Cross-power spectrum is:

$S_{ij}^n=\sqrt{S_{ii}^n{S}_{jj}^n}Coh\left({\mathrm{\ω}}_n\right),i\≠j,$

$Coh\left({\mathrm{\ω}}_n\right)=\frac{{\mathrm{\τ}}_{ij}}{\sqrt{{\mathrm{\τ}}_{ii}{\mathrm{\τ}}_{jj}}}\frac{1}{\sqrt{1+0.4{\[({\mathrm{\ω}}_n{L}_u/2\mathrm{\π}{\stackrel{\‾}{U}}_z)\]}^2}}.$

Wherein, wherein Δ ω=ω _u/ N, ω _ufor cut-off circular frequency (is selected according to data precision demand, generally can select 2 π or 4 π), N is that wind power spectrum is counted at the discrete of frequency domain, n is the n-th discrete point on wind power spectrum, i, j=u, v and w in cross-power spectrum formula represent respectively the down wind of fluctuating wind speed, beam wind to perpendicular wind direction, τ _ijfor eddy stress component value, L _ifor the turbulence integral length dimension in all directions, L _ufor the turbulence integral length dimension of down wind, U _zfor the wind speed at Z height place, for the mean wind speed at Z height place, Coh (ω _n) for using the some coherence function after Reynolds stress modifier.

The energy of the n-th Frequency point windward spectrum correspondence is adopt eigenwert eigendecomposition method to decompose the cross-power spectrum matrix on each Frequency point, obtain a stack features value and orthogonal tensor

In formula, for cross-power spectrum, ω _ufor cut-off circular frequency, with be orthogonal tensor, for eigenwert.

In step F, three-dimensional fluctuating wind field stimulation formula is as follows:

Wherein: ${\tilde{x}}_m=\frac{x_m}{L_m},{\tilde{d}}_k^n=\frac{d_k^n{c}^n}{{\mathrm{\η}}_{\left(k\right)}^n\sqrt{{\mathrm{\ω}}_u}},c^n=\sqrt{\frac{3}{2}{S}_{lm}^n{\mathrm{\ω}}_u\frac{d_l^n{d}_m^n}{d_k^n{d}_k^n}},$ for cross-power spectrum, $p_j^n={\mathrm{\ϵ}}_{jmi}{\mathrm{\ξ}}_m^n{d}_i^n,$ x _mfor extracting the volume coordinate of point, L _mfor turbulence integral scale, t represents the time, with for separate random number, and have $\begin{array}{ccc}{\mathrm{\ξ}}_m^n,{\mathrm{\ζ}}_m^n\∈N(0,1),& d_i^n\∈N(0,1/2),& d_k^n\∈N(0,1/2)\end{array},$ n (M, σ) represents that average is M, variance is the normal distribution of σ, ε _jmifor permutation tensor, and meet orthogonality condition: for orthogonal tensor, for eigenwert.

The present invention simulates in the process of three-dimensional fluctuating wind speed field, have employed pointwise simulation, do not need to decompose the huge cross-power spectrum matrix of exponent number, greatly reduce the requirement to computer software and hardware, can simulate for concrete, complicated mountain area landforms, the result drawn has more applicability, accuracy than simple, regular mountain area model simultaneously.

Accompanying drawing explanation

Fig. 1 is the true geomorphologic map in China island of application examples of the present invention.

Fig. 2 is the target island point cloud display figure of application examples of the present invention.

Fig. 3 is the target Landforms of Islands illustraton of model of application examples of the present invention.

Fig. 4-6 be the measuring point place, target island of application examples of the present invention three-dimensional fluctuating wind speed time-histories figure (Fig. 4 is down wind, Fig. 5 be beam wind to, Fig. 6 is perpendicular wind direction).

Fig. 7-9 be the measuring point place three-dimensional fluctuating wind auto-power spectrum of application examples of the present invention comparison diagram (Fig. 7 is down wind, Fig. 8 be beam wind to, Fig. 9 is perpendicular wind direction).

Embodiment

Below in conjunction with accompanying drawing, embodiments of the invention are further described.

One, mountain area model is set up

Complete the modeling work under the landforms of different mountain areas, be mainly divided into following step:

(1) acquisition of mountain area data

Need in this step to use the GoodyGIS system based on GoogleEarth secondary development, in the GoodyGIS system of professional version, need autonomous definition to intercept the number of the scope of three-dimensional coordinate point, precision and extraction point, longitude and latitude and height value that namely the parameter value satisfied the demands extracts target mountain area are set.

(2) arrangement of mountain area data converts

The mountain area coordinate obtained in previous step provides with the form of longitude and latitude, and conveniently modeling work below needs latitude and longitude coordinates formal transformation to become common planimetric coordinates form.When mountain area landforms are complicated, coordinate is counted numerous, uses following simple estimation equation to carry out coordinate conversion in the present invention:

Δx＝(x _i-x ₀)/0.00001

Δy＝1.1×(y _i-y ₀)/0.00001

Wherein x _iand y _irepresent that arbitrary data extract warp, the latitude of point respectively, and x ₀and y ₀be then warp, the latitude of true origin given in advance, the coordinate difference on the surface level calculated by above formula like this can obtain the three-dimensional planar coordinate figure of all coordinate points.

(3) formation of mountain area model

In the generative process on massif surface, need to use Imageware reverse engineering software, the function that this software can realize is: by given some cloud file, matching is met the three-dimension curved surface of requirement.Therefore, the three-dimensional point coordinate calculated is imported in Imageware software with a form for cloud file, then use the function of " constructing free form surface by a cloud " that the surface model of the mountain area landforms needed can be obtained in upper step.The parameter such as standard deviation, tension force, smoothness in matching face can also being set in this function, for obtaining the higher model of accuracy, standard deviation wherein need being set to minimum value.

Two, theory deduction

Theory deduction process of the present invention mainly comprises following 5 steps:

(1) reynolds stress model is adopted

In Fluid Mechanics Computation, often turbulent stress is assumed to the function of turbulent kinetic viscosity, therefore the key solving turbulence model is the turbulent kinetic viscosity how determining flow field, because this computation model does not do isotropic supposition to turbulence pulsation stress when why first-selected RSM model calculates, but directly set up the transport equation about eddy stress and solve, so the Renolds stress tensor of each point can be drawn, i.e. the covariance of each point fluctuating wind speed in flow field.The potential flows rotated for not having system, Reynolds stress transport equation has fairly simple form:

$\begin{array}{c}\frac{\∂\left({\mathrm{\ρ\τ}}_{ij}\right)}{\∂t}+\frac{\∂\left({\mathrm{\ρu}}_k{\mathrm{\τ}}_{ij}\right)}{\∂x_k}=\frac{\∂}{\∂x_k}(\frac{{\mathrm{\μ}}_t}{{\mathrm{\σ}}_k}\frac{\∂{\mathrm{\τ}}_{ij}}{\∂x_k}+\mathrm{\μ}\frac{\∂{\mathrm{\τ}}_{ij}}{\∂x_k})-\mathrm{\ρ}({\mathrm{\τ}}_{ik}\frac{\∂u_j}{\∂x_k}+{\mathrm{\τ}}_{jk}\frac{\∂u_i}{\∂x_k})\\ -C_1\mathrm{\ρ}\frac{\mathrm{\ϵ}}{k}({\mathrm{\τ}}_{ij}-\frac{2}{3}{\mathrm{k\δ}}_{ij})-C_2(P_{ij}-\frac{1}{3}{P}_{kk}{\mathrm{\δ}}_{ij})-\frac{2}{3}{\mathrm{\ρ\ϵ\δ}}_{ij}\end{array},i=u,v,w$

In formula: τ _ij, τ _ikand τ _jkfor Reynolds stress tensor, u _iand u _jbe respectively x _i, x _jthe mean wind speed in direction, μ _tfor turbulence viscosity, μ is kinematic viscosity, k and ε is respectively Turbulent Kinetic and dissipative shock wave, x _kfor durection component, u _kfor speed component, ρ is fluid density, P _ijfor the item that lateral fluid force produces, P _kkfor P _ijsum term, C ₁, C ₂and σ _kfor relevant coefficient, δ _ijfor Kronecker tensor.

(2) eddy stress method for solving is supplementary

Consider that RSM method is higher for the quality requirements of stress and strain model, when alpine terrain is comparatively complicated, the division difficulty of grid is comparatively large, especially near-wall grid, now, and the more difficult convergence of result of calculation.Which limits the scope of application of the present invention, for this reason, carry out supplementing to step (1) and revise.

Model the most frequently used engineering calculation is just become after k-ε turbulence model was suggested from 1972, this model have can calculate compared with high Reynolds number flow problem, calculated amount is little, convergence is fast, to mesh quality requirement far below advantages such as RSM models, correction requirement according to the invention.But k-ε turbulence model supposition turbulent kinetic viscosity is isotropic, cannot obtain the correlative character between 3 direction fluctuating wind speed in turbulent flow, namely directly cannot obtain concrete eddy stress component value.According to the statistics in ESDU, between the root mean square of all directions fluctuating wind speed, there is following proportionate relationship:

${\mathrm{\α}}_v=\frac{{\mathrm{\α}}_v}{{\mathrm{\σ}}_u}=1-0.22{\cos }^4\left(\frac{\mathrm{\π}}{2}\frac{z}{H}\right)$

${\mathrm{\α}}_w=\frac{{\mathrm{\α}}_w}{{\mathrm{\σ}}_u}=1-0.45{\cos }^4\left(\frac{\mathrm{\π}}{2}\frac{z}{H}\right)$

α _u＝1

Wherein: σ _u, σ _vand σ _wrepresent down wind, beam wind respectively to the root mean square with vertical fluctuating wind speed component, z is the height of three-dimensional fluctuating wind field simulated target point, and H is the gradient level of atmospheric boundary layer.And to a specified point, tubulence energy k equal Reynolds stress tensor three diagonal components and, when after the relational expression above having had, just can by following formula decomposition obtain three diagonal angle Reynolds components of stress:

in formula, i is u, v or w;

For off-diagonal Reynolds components of stress, the wind speed component covariance value also namely on different directions, according to existing achievement in research, down wind and beam wind to, beam wind to and perpendicular wind direction between almost there is no correlativity, so τ _uvand τ _vwvalue minimum, negligible; And to another Reynolds components of stress τ in ESDU _uwgive corresponding estimation equation: ${\mathrm{\τ}}_{uw}=-u_{*}^2{(1-z/H)}^2.$

Wherein u _*for wind friction velocity, computing formula is: u _*=U ₁₀/ [2.5ln (10/z ₀)], U ₁₀for the mean wind speed of 10m At The Height, z ₀for Roughness Length.

(3) correction of fluctuating wind speed cross-power spectrum and decomposition

Use the Renolds stress tensor drawn by preceding method, wind power spectrum and some coherence function are revised, thus more true reasonably with the incorgruous cross-power spectrum matrix of point under drawing mountain area landforms, revised wind spectrum matrix comprises auto-power spectrum and cross-power spectrum two parts:

Auto-power spectrum is:

$s_{ii}^n=\frac{1}{2\mathrm{\π}}\·{\mathrm{\τ}}_{ii}\frac{L_i/U_z}{{\[1+70.8{({\mathrm{\ω}}_n{L}_i/2{\mathrm{\πU}}_z)}^2\]}^{5/6}},i=u$

$S_{ii}^n=\frac{1}{2\mathrm{\π}}\·{\mathrm{\τ}}_{ii}\frac{L_i/{\stackrel{\‾}{U}}_z\[1+755.2{({\mathrm{\ω}}_n{L}_i/2\mathrm{\π}{\stackrel{\‾}{U}}_z)}^2\]}{{\[1+283.2{({\mathrm{\ω}}_n{L}_i/2\mathrm{\π}{\stackrel{\‾}{U}}_z)}^2\]}^{11/6}},i=v,w$

Cross-power spectrum is:

$S_{ij}^n=\sqrt{S_{ii}^n{S}_{jj}^n}Coh\left({\mathrm{\ω}}_n\right),i\≠j$

$Coh\left({\mathrm{\ω}}_n\right)=\frac{{\mathrm{\τ}}_{ij}}{\sqrt{{\mathrm{\τ}}_{ii}{\mathrm{\τ}}_{jj}}}\frac{1}{\sqrt{1+0.4{\[({\mathrm{\ω}}_n{L}_u/2\mathrm{\π}{\stackrel{\‾}{U}}_z)\]}^2}}.$

As can be seen from above-mentioned derivation, the non-diagonal composition of eddy stress have modified the cross-power spectrum part of traditional wind spectrum to a certain extent.

The energy of the n-th Frequency point windward spectrum correspondence is adopt eigenwert eigendecomposition method to decompose the cross-power spectrum matrix on each Frequency point, a stack features value can be obtained and orthogonal tensor for:

(4) random pulse wind field is generated

Cross-power spectrum is utilized to decompose the eigenwert of gained each Frequency point is zero to the average constructed, meets the three-dimensional fluctuating wind field of orthogonality condition and carry out scale transformation, can obtain:

$w_i(\stackrel{\→}{x},t)=\sqrt{\frac{2{\mathrm{\ω}}_u}{N}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}{v}_{\left(i\right)}^n{\mathrm{\η}}_{\left(i\right)}^n=\sqrt{2\mathrm{\Δ}\mathrm{\ω}}\underset{n=1}{\overset{N}{\mathrm{\Σ}}}{v}_{\left(i\right)}^n{\mathrm{\η}}_{\left(i\right)}^n$

To each trigonometric function ripple of three-dimensional fluctuating wind field, there is oneself a new coordinate system with initial coordinate x _ibetween transformational relation be: now with orthogonal tensor, orthogonal transformation is carried out to fluctuating wind field, the velocity field considering a coherence function can be obtained:

Above formula be the simulation formula of the fluctuating wind field under the consideration concrete mountain area landforms impact using method of the present invention to draw through deriving, and by using fast Fourier change (FFT) technology, improve the counting yield of the inventive method further.

Application examples

Should use-case with the mountain area landforms on certain island of China for research object, select this weather station, island 10m At The Height measuring point, use the step method in the present invention successively, obtain the three-dimensional fluctuating wind speed field at this measuring point place.

In GoogleEarth, observation obtains the true landforms situation on this island as shown in Figure 1, and the position of sign is implements point position of the present invention.As can be seen from the figure, the pole of figure on target island is irregular, mountain area landform shape is extremely complicated, and the wind field rule obtained under the mountain area landforms of simple rule accurately cannot reflect the wind field distribution characteristics of this area, and the method in the present invention can be used to complete the simulation of this area three-dimensional fluctuating wind field.

(1) Modling model

Use GoodyGIS system obtains the extraction point latitude and longitude coordinates information in the certain limit of target island.For reducing calculated amount, use irregular polygon to carry out frame choosing when selective extraction point coordinate scope, the outer along Dongtou Island is selected, and is ensureing, under the prerequisite included in target mountain area, to remove the inessential point of periphery.Utilize the method process data in the present invention, obtain the some cloud file that can import in Imageware, as shown in Figure 2.

On the basis of this cloud according to the present invention in the method that illustrates, can generate the true mountain area landform model on target island, net result is as shown in Figure 3.

The file layout that the model of generation can read with GAMBIT is preserved, as IGES form, is then read in GAMBIT and carries out the follow-up process of model, the grid model that Fluent calculates can be obtained.So far, the model work of setting up completes.

Due to should use-case Zhongshan District landforms extremely complicated, be difficult to mark off the grid of the quality that can meet RSM model calculation requirement, therefore, use the compensation process in the present invention---namely first use k-ε turbulence model to calculate, and then utilize decomposition formula to obtain each Reynolds components of stress.Editor UDF file, using category-B wind field as speed entrance, the mean wind speed under 10m height is taken as 10m/s.Adopt the coupling of SimpleC algorithm process pressure and speed, all discrete scheme all adopt Second-order Up-wind form, when being considered as calculating convergence during two convergence below meeting simultaneously: the iteration residual error that a. works as all variablees is reduced to 10 ^-5; B. the monitoring parameter calculation stability such as speed, pressure of control point, no longer changes.Theory deduction part in the present invention is all realized by MATLAB software programming, is updated in MATLAB program by the Reynolds components of stress value at the measuring point place obtained, and can draw the three-dimensional fluctuating wind speed of measuring point, result and checking are respectively as Figure 4-Figure 6.

As can be seen from three fluctuating wind speed time series figure, in three directions, the distribution of fluctuating wind speed is slightly different, this is because cause the distribution of energy in frequency of different wind speed direction windward spectrum is different.

Fig. 7-9 is that the fluctuating wind on three wind directions after normalization composes desired value and the comparison diagram utilizing the gained analogue value of the present invention, as can be seen from the figure, although simulate the wind speed that obtains and desired value has deviation slightly in low frequency part, at HFS, it is relatively good that both coincide.And in actual applications, in general, the HFS of fluctuating wind speed is more obvious to the effect of structure.Therefore, can think that simulating by the method in the present invention the three-dimensional fluctuating wind speed field obtained meets the requirement of engineer applied.

Claims (7)

1. a high wind pulsation dimensional wind comprehensive simulation method under the landforms of mountain area, it comprises the following steps:

A. according to engineering demand, quantity, precision and scope that spot elevation gathers are set in the GoodyGIS system of GoogleEarth secondary development, derive longitude and latitude and height coordinate that target mountain area data extract point;

B. the longitude and latitude of said extracted point is converted and be met the planimetric coordinates of accuracy requirement, a final formation point cloud file;

C. the some cloud file that step B is formed is imported in Imageware software, use the function of " being gone to construct free form surface by point " to generate the landforms surface model in target mountain area;

D. the landforms surface model formed in step C is imported in the pre-processing software GAMBIT of Fluent, complete the modeling needed for CFD calculating, stress and strain model and definition boundary condition;

E. RSM model is used to calculate the eddy stress component value of each extraction point, wind power spectrum and some coherence function are revised, more true reasonably with the wind spectrum matrix that point is incorgruous under drawing mountain area landforms, revised wind spectrum matrix comprises auto-power spectrum and cross-power spectrum;

F. structure obtains extracting the incorgruous three-dimensional fluctuating wind field stimulation formula of the same point of a consideration at place, and introduces Fast Fourier Transform (FFT) technology.

2. high wind pulsation dimensional wind comprehensive simulation method under the landforms of mountain area according to claim 1, it is characterized in that, in step e, when geomorphologic conditions change in mountain area is too complicated, when being difficult to generate the grid meeting RSM model computation requirement, use the k-ε turbulence model relatively low to mesh quality requirement, obtain the computing formula being solved each eddy stress component value by tubulence energy k.

3. high wind pulsation dimensional wind comprehensive simulation method under the landforms of mountain area according to claim 2, it is characterized in that, the computing formula of three diagonal angle eddy stress component values is:

in formula, i is u, v or w;

Wherein,

${\mathrm{\α}}_v=\frac{{\mathrm{\α}}_v}{{\mathrm{\σ}}_u}=1-0.22{\cos }^4\left(\frac{\mathrm{\π}}{2}\frac{z}{H}\right)$

${\mathrm{\α}}_w=\frac{{\mathrm{\α}}_w}{{\mathrm{\σ}}_u}=1-0.45{\cos }^4\left(\frac{\mathrm{\π}}{2}\frac{z}{H}\right)$

α _u＝1，

In above formula, σ _u, σ _vand σ _wrepresent down wind, beam wind respectively to the root mean square with perpendicular wind direction fluctuating wind speed component, z is the height of three-dimensional fluctuating wind field simulated target point, and H is the gradient level of atmospheric boundary layer.

4. high wind pulsation dimensional wind comprehensive simulation method under the landforms of mountain area according to claim 3, is characterized in that,

For off-diagonal eddy stress component value, due to down wind and beam wind to, beam wind to and perpendicular wind direction between almost there is no correlativity, τ _uvand τ _vwvalue minimum, ignore; Eddy stress component τ _uwestimation equation: wherein, u _*for wind friction velocity, u _*computing formula be: u _*=U ₁₀/ [2.5ln (10/z ₀)], U ₁₀for the mean wind speed of 10m At The Height, z ₀for Roughness Length.

5. high wind pulsation dimensional wind comprehensive simulation method under the landforms of mountain area according to claim 4, it is characterized in that, in step e, auto-power spectrum is:

$s_{ii}^n=\frac{1}{2\mathrm{\π}}\·{\mathrm{\τ}}_{ii}\frac{L_i/U_z}{{\[1+70.8{({\mathrm{\ω}}_n{L}_i/2{\mathrm{\πU}}_z)}^2\]}^{5/6}},i=u,$

$S_{ii}^n=\frac{1}{2\mathrm{\π}}\·{\mathrm{\τ}}_{ii}\frac{L_i/{\stackrel{\‾}{U}}_z\[1+755.2{({\mathrm{\ω}}_n{L}_i/2\mathrm{\π}{\stackrel{\‾}{U}}_z)}^2\]}{{\[1+283.2{({\mathrm{\ω}}_n{L}_i/2\mathrm{\π}{\stackrel{\‾}{U}}_z)}^2\]}^{11/6}},i=v,w;$

Cross-power spectrum is:

$S_{ij}^n=\sqrt{S_{ii}^n{S}_{jj}^n}Coh\left({\mathrm{\ω}}_n\right),i\≠j,$

$Coh\left({\mathrm{\ω}}_n\right)=\frac{{\mathrm{\τ}}_{ij}}{\sqrt{{\mathrm{\τ}}_{ii}{\mathrm{\τ}}_{jj}}}\frac{1}{\sqrt{1+0.4{\[({\mathrm{\ω}}_n{L}_u/2\mathrm{\π}{\stackrel{\‾}{U}}_z)\]}^2}}.$

Wherein, wherein Δ ω=ω _u/ N, ω _ufor cut-off circular frequency, N is that wind power spectrum is counted at the discrete of frequency domain, and n is the n-th discrete point on wind power spectrum, i, j=u, v and w in cross-power spectrum formula represent respectively the down wind of fluctuating wind speed, beam wind to perpendicular wind direction, τ _ijfor eddy stress component value, L _ifor the turbulence integral length dimension in all directions, L _ufor the turbulence integral length dimension of down wind, U _zfor the wind speed at Z height place, for the mean wind speed at Z height place, Coh (ω _n) for using the some coherence function after Reynolds stress modifier.

6. high wind pulsation dimensional wind comprehensive simulation method under the landforms of mountain area according to claim 5, is characterized in that, the energy of the n-th Frequency point windward spectrum correspondence is adopt eigenwert eigendecomposition method to decompose the cross-power spectrum matrix on each Frequency point, obtain a stack features value and orthogonal tensor

In formula, for cross-power spectrum, ω _ufor cut-off circular frequency, for orthogonal tensor, for eigenwert.

7. high wind pulsation dimensional wind comprehensive simulation method under the mountain area landforms according to claim 3 or 4, it is characterized in that, in step F, three-dimensional fluctuating wind field stimulation formula is as follows:

Wherein: ${\tilde{x}}_m=\frac{x_m}{L_m},{\tilde{d}}_k^n=\frac{d_k^n{c}^n}{{\mathrm{\η}}_{\left(k\right)}^n\sqrt{{\mathrm{\ω}}_u}},c^n=\sqrt{\frac{3}{2}{S}_{lm}^n{\mathrm{\ω}}_u\frac{d_l^n{d}_m^n}{d_k^n{d}_k^n}},$ for cross-power spectrum, $p_j^n={\mathrm{\ϵ}}_{jmi}{\mathrm{\ξ}}_m^n{d}_i^n,$ x _mfor extracting the volume coordinate of point, L _mfor turbulence integral scale, t represents the time, with for separate random number, and have $\begin{array}{ccc}{\mathrm{\ξ}}_m^n,{\mathrm{\ζ}}_m^n\∈N(0,1),& d_i^n\∈N(0,1/2),& d_k^n\∈N(0,1/2)\end{array},$ n (M, σ) represents that average is M, variance is the normal distribution of σ, ε _jmifor permutation tensor, and meet orthogonality condition: for orthogonal tensor, for eigenwert.